Herman Grooten

Chess Endgames for Club Players

The Essential Skills for a Forceful Finale

New In Chess 2022

© 2022 New In Chess

Published by New In Chess, Alkmaar, The Netherlands

www.newinchess.com

All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in
any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior
written permission from the publisher.

Cover design: rouwhorst + van roon

Translation and supervision: Peter Boel
Proofreading: René Olthof, Mariska de Mie
Production: Sandra Keetman, Anton Schermer

Have you found any errors in this book?

Please send your remarks to editors@newinchess.com. We will collect all relevant corrections on the
Errata page of our website www.newinchess.com and implement them in a possible next edition.

ISBN: 978-94-93257-49-8
Contents
Foreword by Karsten Müller
Preface

Part I Fear of the endgame?

Chapter 1 Introduction
Chapter 2 Ingrained fear
Chapter 3 Liquidation
Chapter 4 Knowledge and skill
Chapter 5 Fortresses in practice
Chapter 6 Various theoretical draws
Chapter 7 Exercises
Chapter 8 Solutions

Part II Elementary endgames

Chapter 9 Introduction
Chapter 10 Queen versus pawn
Chapter 11 Rook versus pawn
Chapter 12 Bishop versus pawn(s)
Chapter 13 Knight versus pawn
Chapter 14 Exercises
Chapter 15 Solutions

Part III Various techniques

Chapter 16 Introduction
Chapter 17 Pawn endings
Chapter 18 Knight endings
Chapter 19 Bishop endings
Chapter 20 Rook endings
Chapter 21 Queen endings

Part IV Curious endgames

Chapter 22 Introduction
Chapter 23 Thinking in schemes
Chapter 24 Analysis of practical endgames
Chapter 25 Exercises
Chapter 26 Solutions

Index of names
Explanation of symbols
Bibliography
Foreword: mastering the endgame
There are already many endgame books, so what is the point of another one, I hear you ask. But most of
them deal with theoretical positions and concepts and are sorted by material. Grooten does this as well of
course. But in most books the important motifs and rules of thumb are underrepresented.
Here comes Herman Grooten’s real point. He deals with them in deep detail. Of course he fully
understands that all guidelines also have exceptions and so he also sharpens the reader’s eye for their
applications.
He has added exercises, which I think is a very valuable didactical concept and also graphical rules like
the winning zones with queen against bishop’s and rook’s pawn on its seventh rank. I like these graphical
rules a lot as they are easy to see and to remember.
Of course Grooten also deals deeply with the very frequent and important rook endings. I would advise
you to study these sections several times to make sure that you really remember the relevant theoretical
positions and the rules of thumb.

I hope that Herman Grooten’s book gives you as much pleasure as it has given me.

GM Dr Karsten Müller,
Hamburg, September 2022
Preface
When I saw a chess endgame for the first time in my life, I was surprised by the simplicity and the beauty of
the theme. It must have been at my club – someone was showing the final phase of an endgame study. Like
most club players I preferred to concentrate on studying openings and solving tactical exercises. I was
always told that to become better, you especially needed to work hard on your openings. Although I never
studied openings in the way most people do, and I was more interested in the middlegame from an early
age, through the years I gradually started noticing that my treatment of endgames was lagging behind
considering the level I had reached at a certain point. This changed, however, due to a quite special
occurrence.
In 1980, despite being a complete outsider, I managed to qualify for the Dutch Championship that was
going to be held in Leeuwarden in 1981. I qualified in the semi-finals, among others because I created a
positional zugzwang situation against one of the favourites, the late IM Leon Pliester, in the position given
below.

1
Leon Pliester
Herman Grooten
Eindhoven ch-NED sf 1980 (2)
XIIIIIIIIY
9-+-+-+-+0
9zp-+-+p+k0
9q+-zp-+-zp0
9+-zpPzp-+-0
9P+n+P+-+0
9+-+Q+P+l0
9-tr-+-+pzP0
9tR-mK-vL-tR-0
xiiiiiiiiy
Black to move
Black has sacrificed an exchange to obtain the initiative, but here it looks like White has full control again.
The black rook on b2 is more or less hanging in the air (after Bc3), and it’s not clear how Black can create
threats against the white king. But with the next move, which I regard as one of the most aesthetically
pleasing moves I ever played, I managed to force my strong opponent to his knees:
31...Bc8!!
I have to protect my queen since I want to make a move with the c4-knight in some variations. While I was
considering this bishop move, it took me a while to realize that White finds himself in a kind of positional
zugzwang. Each move with a piece will have great drawbacks, so all White is left with are pawn moves, but
these cannot avoid the breakdown.
32.f4
Due to this pawn move, things go rapidly amiss for White.
A) Bad was 32.Rb1 in view of 32...Ra2 when the a4-pawn falls, after which there is no defence;
B) On 32.Qc3 I had prepared 32...Re2!, with the point that 33.Qd3 is met with the pretty 33...Ne3!, e.g.
34.Qxa6 Bxa6 and Black threatens 35...Rc2 followed by ...Bd3. Now, 35.Bd2 is met with the quite artistic
35...Nf1! 36.Ba5 Bd3 37.Ra3 c4 and there is nothing sensible to be found against the threat ...Nxh2-f3;
C) I was very proud of the variation I found after 32.Bc3: 32...Ne3!!. After this move all Black’s pieces are
hanging, but it all fits like clockwork!
C1) 33.Qxa6 runs into the zwischenzug 33...Rc2+ 34.Kb1 Bxa6–+
 XIIIIIIIIY
9-+-+-+-+0
9zp-+-+p+k0
9l+-zp-+-zp0
9+-zpPzp-+-0
9P+-+P+-+0
9+-vL-snP+-0
9-+r+-+pzP0
9tRK+-+-tR-0
xiiiiiiiiy
analysis diagram
and again the black pieces are cooperating fantastically well. There is nothing White can do against this
overwhelming force, e.g. 35.Ba5 Bd3 36.Ra3 Rf2+ 37.Rxd3 Rf1+ 38.Ka2 Rxg1 39.Rxe3 Rh1 and the pawn
queens;
C2) 33.Qxe3 Re2 sort of catches the queen. Material is still even after 34.Qd2 Rxd2 35.Kxd2 but after
35...Bh3 Black will decide the game on the light squares. An illustrative sample variation runs 36.Rge1 f5
37.Re2 Qc4–+.
32...Bg4
Now that White has weakened his light squares, the bishop can enter the position from the other side.
33.f5 Be2 34.Qg3
XIIIIIIIIY
9-+-+-+-+0
9zp-+-+p+k0
9q+-zp-+-zp0
9+-zpPzpP+-0
9P+n+P+-+0
9+-+-+-wQ-0
9-tr-+l+pzP0
9tR-mK-vL-tR-0
xiiiiiiiiy
White wants to confuse the issue with f5-f6, but Black’s attack is in full swing now.
34...Qb7!
The final hammer blow.
35.Qc3
The only move to prevent an immediate mate, but now White is treated to a ‘light-square massage’.
After 35.f6, White would have been checkmated first with 35...Rc2+! 36.Kxc2 Qb2#.
35...Rb3 36.Qc2 Bd3 0-1
Here Leon shook my hand, completely dizzy from the cyclone that had raged over him. During the post-
mortem he was very friendly, telling me that he had rarely been beaten in such an elegant way.
A possible continuation is 36...Bd3 37.Qa2 Rb2 and the curtain falls.
Some time after this, I received a phone call from the then national coach of the Netherlands, IM Cor van
Wijgerden. He told me that young players who had qualified for the Dutch Championship were entitled to
trainings by him. However, since I was already 22 years old at the time, I wasn’t eligible for this since the
chess federation ‘didn’t train elderly people’. This was meant as a joke, and the above game must have
helped, as became clear to me when he offered me to visit him on three different days to get trainings.
As I had never received any ‘official’ training before, I thankfully accepted the proposal. Van Wijgerden
quickly added that he ‘didn’t do openings’ and asked me if I had an idea for a training subject. Since I knew
that he had just revised an endgame book by Euwe (Volume 3 of the Theory of the Endgame series: Rook
Endings), I suggested that we study solely rook endings. As Van Wijgerden was a great expert on this
subject, he opened an entirely new world for me at the time. Not only did he manage to awaken an interest
in the endgame in me, but he also gave me some (difficult) exercises to solve at home.
One time, for example, he had been sent an adjourned position from a club match for arbitration. The
players from both teams could file a claim – win or draw – and substantiate it with variations. This was the
position in question:

2
Adjourned position
Cor van Wijgerden
XIIIIIIIIY
9-+-+k+-+0
9+R+-+p+p0
9-zP-+-+p+0
9+-+-zp-zP-0
9-+-+-+KzP0
9+r+-zP-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move

The White player claimed a win, and this claim looks justified. But the Black player appealed, among others
since a member of his club had invented an insane defence. They tried to substantiate their assessment
with variations.
1...Kf8 2.Kf3 Rb4 3.e4! h6!!
An incredible move that had been found by the analysing team of the Black player. Black wins a crucial
tempo by offering the pawn at this moment. The normal move is 3...Kg7 after which the game could
continue as follows: 4.Ke3 h6 (first Black has to solve the problem of the potential weakness of his h7-
pawn, and then he has to return to e8 with his king) 5.Kd3 hxg5 6.hxg5 Kf8 7.Kc3 Rb1 8.Kc4 Ke8. But
White is winning after 9.Kc5. The fact that Black had to spend two tempi to trade the h7-pawn for the g5-
pawn proved essential.
4.Ke3
The pawn capture also leads to a draw: 4.gxh6 Kg8.
4...hxg5 5.hxg5 Ke8 6.Kd3 Kd8
And White cannot win.
Van Wijgerden told me he had spent many hours trying to refute this unusual move, but hadn’t managed to
find a win for White. In his book, he wrote: ‘The following analyses prove that Black can achieve a draw by
the skin of his teeth, in spite of the fact that White has indisputable advantages in the diagram position.’
Many years later, I took another look at this position, and I even managed to trace a few new ideas, but I
also failed to find a conclusive path to a win for White. Another few years later, I presented the position to
my then pupil Loek van Wely (see the anecdote in Chapter 20) and although he managed to pull out even
more ideas than Cor and me had come up with together, he wasn’t able to find a win for White either – even
though he had found a brilliant move somewhere in one line which received a different assessment as a
consequence.
It would take us much too far to discuss this position here due to its high degree of complexity. But with
this example, the foundation was laid for my interest in endgames – also because it made me see that in
endgames the difference between a half and a full point sometimes depends on an apparently minuscule
detail. And isn’t that what we are always doing – playing for a half or a full point? In the endgame, the
margin between success and ruin is quite small...
Analysis of endgames
After the Dutch Championship, which was a school of hard knocks for me, my results were a little worse for
a while at first, until suddenly I made a giant leap. This was mainly caused by the fact that my endgame
knowledge had increased significantly, and my endgame technique had improved considerably.
Nevertheless, in practice, playing endgames was still an arduous business. Sometimes you think you know
something, but it always turns out to be a lot more difficult to put your knowledge into practice. That can be
a hard pill to swallow at times, but I had come to enjoy analysing endgames and studying books on the
subject, even consciously liquidating into an endgame now and then to enhance my skills level, and I had
one important success in 1984.
In the English spa of Ramsgate, during a two-hour break, I analysed an adjourned position better than
my opponent, and also played it better after the resumption, managing to convert a drawn rook ending into
a win. This was against no less a player than Julian Hodgson, who went on to become a strong grandmaster
and a four-time champion of England. That was something to be proud of, and something to build on.
Young players of today can hardly imagine it, but in our days there was no Internet, no laptops, and there
were neither strong engines nor Tablebases. A chess player would set off with a large suitcase filled with
chess books and only a toothbrush. Games were adjourned (as recalled above) and you could analyse them
before the resumption. On such occasions, precisely because it was your own game and you wanted to
achieve an optimal result, you dived deeply into the secrets of a position. This demanded energy and could
give you sleepless nights, but it strongly improved your understanding of the game, especially the endgame.
Adjourned games could turn into adventures, emotional roller-coasters in which you were hovering
between success and failure.

Training/coaching
There was yet another aspect in my career that caused me to focus increasingly on endgames. In my
student years, I had applied for two trainer courses, enabling me to acquire the necessary documents,
ratified by the federation, to go through life as a qualified chess trainer. It hasn’t done me any harm; my
entire life I’ve been working as a trainer/coach with great pleasure. And during my work as a trainer, I
found out that a lot of profit can be gained in the endgame. This is true not only for junior players, but also
for club players – for anybody, in fact! If you have a certain knowledge of the endgame, you will also play
better in the middlegame – this is a known fact. It’s just that many players do not seem to have this interest
in the endgame – and this, in turn, has to do with knowledge. Without knowledge you are groping in the
dark... it’s the well-known chicken-and-egg problem.
Sometimes you need to have someone who can help you remove the deadlock – and that someone can be
your trainer/coach. I’ve always regarded it as my job to convey my love for the endgame to my chess
students. I try to do this as lively as possible, by presenting the ‘tedious theory’ to them in constantly
varying ways. Thus, I hoped to awaken an interest in my chess students by giving them a couple of ‘handles’
in the endgame. If you increase the skills of your pupils, they will be more successful. Also, success is often
the key to more pleasure. One thing I have also noticed during my career as an active player: if you’re
having fun, the rest follows automatically. And the pleasure I have always experienced in the study of
endgames has brought me immense benefit. I hope that with this book I can convey, besides the theoretical
stock-in-trade that every club player should have, especially the pleasure that comes with analysing
endgames. If, for example, during a training you can invent an expressive analogy that clarifies the Vancura
principle (by introducing the notion of the ‘bus stop’, see Chapter 20), and more than twenty years later
you meet one of your former students and he’s still talking about it, then you know you’ve fixed something
in his mind at least!

Endgame studies
The reader will notice that this book contains a substantial number of endgame studies. This has been done
with a clear idea in mind. In my career as a trainer I gradually found out that good training involves having
good examples at your disposal. To illustrate the theme at hand, an example for instruction should, in my
opinion, be clear and transparent – it should contain as little ‘noise’ as possible. With ‘noise’ I mean factors
that distract the student from the things that you really want to convey.
 Of course, it is also important to show examples from tournament practice, and this is what I have also
done in this book – preferably games by my pupils. But in my experience, while you are discussing
tournament games, you often get ‘in-between’ questions like ‘Why can’t I play this?’, ‘Why can’t Black
defend in this way?’, etcetera. Critical pupils often had a point with such questions, but this meant that we
were soon distracted from the central theme. Side variations would start playing first fiddle all of a sudden.
When at the end of the lesson I asked ‘What have we learned today?’, often there was no-one who could
give a good answer...
What I find important in endgame studies is that they contain play, counterplay and hidden defences, but
above all a clear path to the win (or to the draw) with a highly aesthetic character that not only shows the
theme in its full glory, but also shows, in my view, how beautiful chess can be. This, too, is an important goal
of this book: to demonstrate the pure beauty that is often hidden in an endgame.
At this point, I would like to offer my apologies to various study composers. In many cases I have
removed the complicated and sometimes laboured ‘foreplay’ phase in order to put the spotlight on the
point, the basic idea immediately. This is often regarded as a ‘deadly sin’. Such ‘operative interventions’
may cut to the heart of some composers, but my aim as an instructor is entirely different from theirs. To
bring to the fore the main themes that fit into the didactic frame of this book, I sometimes have to resort to
such rigorous measures.
Naturally, I haven’t omitted including many games from tournament practice but also from club players,
in which you can see the trial-and-error experiences that are so familiar to every chess player. That is why I
have created separate sections everywhere that I have called ‘This endgame in practice’.
The reader will also notice that I have used many endgames by Magnus Carlsen – the reason is that, in
my opinion, the Norwegian World Champion often plays the endgame almost perfectly. And not only that –
he also pursues crystal-clear strategies to obtain victory. So clear, sometimes, that playing through these
endings is a feast for the eye and for the mind.
Ultimately, the purpose of working through the examples in this book is that the main themes become
‘anchored’ in your memory. By demonstrating the themes as clearly as possible, I hope to achieve that the
reader will be actually successful with them in practice.

Endgame skills
What I wrote above brings me to another aspect of chess training which has perhaps been chronically
underestimated. Many trainers/coaches put the emphasis on increasing knowledge, trusting that more
knowledge will automatically lead to more points. In my opinion, this is a misconception, and it often leads
to disappointment. In chess, every move is a decision. Then why does an International Master often make
better decisions than a club player? And why, in turn, does the grandmaster frequently do a better job than
the IM? And why do elite players outsmart most other players in the world? Because our game is so
complex and the situation can sometimes change on the next move. This demands flexibility, orientation,
knowledge, insight, intuition, and often also quite a lot of skill, for example, to convert an advantageous
position into a win. In fact, knowledge and skill are your two main guides on the road to success. A chess
player needs both! This is precisely what determines differences in playing strength.
If we apply this principle to the endgame, then we can say that even if we know that a certain ending is a
win, this is not yet a guarantee that we can actually convert the position at hand to a win. That is why in this
book I have offered an impetus to the training of a couple of endgame skills – the usefulness of which,
naturally, depends on which type of ending there is on the board. We can classify many endgame skills by
formulating different techniques in the endgame. To make this more explicit, I have split the material into:

Rules of thumb
Techniques
With the help of a rule of thumb (what are the clues we can find in a position?) we can force a win or a draw
with a certain technique (what should we pay attention to? which plan should we choose?). Isn’t that what
every chess player wants?
Improving your skills is not something you do by reading a book. You have to do some active work. For this
reason, I have made up a number of exercises for you to work on. The American top player Hikaru
Nakamura has become so strong by playing blitz games all day, alternated with solving chess puzzles on the
Internet. I once talked to a coach who had worked with him; he was amazed by what Hikaru didn’t know... If
you would like to see how Nakamura can convert a winning position into a full point lightning-fast and with
almost perfect play, take a look at Hikaru’s Twitch channel (https://www.twitch.tv/gmhikaru).

To practice at home, therefore, you have to play out certain positions. This can be done in different ways:
• against a sparring partner
- physically, on a chessboard with a clock
- online
• against your own trainer/coach, individually or in a simul
• against a chess computer

If this book can serve as an important tool in this way, then this means that one of my main aims has been
achieved.

Conclusions
I would like to conclude with a small anecdote. When in the 1990s I worked as a ‘top sport coordinator’ on
behalf of the Dutch Chess Federation to assemble the training programme for talented junior players, it
seemed like a good idea to make use of the know-how we had available in our country. I succeeded in
trapping GM Jan Timman into giving a number of guest lectures at a central location. When he went outside
after a certain training session and I asked him how things had gone, he replied that there was a lot of
talent, but that he was amazed by their lack of ‘elementary knowledge’ of rook endings. When I asked him
what he meant by this, Timman said he was talking about the ‘standard rook endings’ with 3 vs 3 pawns on
the kingside and one pawn on the other side of the board (see diagram). ‘On my question which is the most
favourable pawn for White, they couldn’t give the right answer,’ Timman said. The diagram shows what the
right answer is...
3
XIIIIIIIIY
9-+-+-+-+0
9+R+-+pmk-0
9-+-+-+p+0
9+-+-+-+p0
9-zP-+-+-zP0
9+-+-+KzP-0
9-tr-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
Portisch-Petrosian, Palma de Mallorca ct m 1974

What is ‘standard’ for Timman, usually isn’t for ordinary humans. This type of ending can be so complex
that we will not pay attention to it in this book. Perhaps in a future edition? The point of departure for this
book is that it first offers the basics of most ending types. I have tried to present the required basic
techniques in every case, to show the student the best way to treat those endings. Practising these endings
is also stimulated by focused exercises. Above all, I want to emphasize that the endgame shows us so many
beautiful things that, even if our results may not be satisfying at a certain moment, enjoyment of the beauty
of the game should prevail over our craving for success. For if there is one area in chess in which simplicity
and beauty go hand in hand, it is the endgame!

I am very honoured that one of the greatest endgame specialists in the world, GM Karsten Müller, has been
found prepared to write a foreword. I also want to express my gratitude and appreciation for the club
players and endgame lovers Michel Hoetmer and Armin Kohlrausch (the latter unfortunately passed away
highly unexpectedly on 25 March 2022), who have not only taken care of the correction work for the
original manuscript, but have also provided me with a lot of advice, ideas and beautiful examples. Finally, I
would like to thank several of my students, among them Sascha Kurt, Andrew Imbens and again Armin
Kohlrausch, for patiently studying (online) various subjects I showed them for this book. Their reactions
sometimes provided me with new insights, stimulating me to rewrite certain passages.

Herman Grooten,
Eindhoven,
September 2022
 Part I
Fear of the endgame?
Chapter 1
Introduction
The only Dutch World Champion Max Euwe wrote a long time ago that most
of the mistakes in chess are made in the endgame. Probably it has never
been investigated whether that is really true. It is a fact, however, that
players often treat the endgame wrongly due to a lack of knowledge. Many
chess players never get around to playing an endgame in their games, and
that is why they haven’t studied the tricks of the trade. And why should we
blame them? Where should you start? Which book should you read? Which
endgames should you study first, to benefit from it in your play? What is the
practical value of studying difficult endgames that you will probably never
get on the board?
These are all legitimate questions to which it is not easy to give a
conclusive answer. If we visit a specialized bookseller and we look on the
‘endgame’ shelf, we will see various books that might provide answers to the
above questions.
However, closer investigation teaches us that these answers are still not
easy to obtain for the average club player. In the more dated endgame books, the author often expected a
high basic knowledge level of the reader, and unquestioningly assumed that most of the ‘handles’ were
already known to him. When as a junior player I tried to pick up the thread and wanted to study the
endgame more deeply, I soon found out that in a substantial number of books the contents had been
organized around specific material balances; for example, a chapter on rook endings would start with rook
+ pawn vs rook, the following section would discuss rook + two pawns vs rook, then rook + two pawns vs
rook + one pawn, etc.
Long after this, after I had been giving chess trainings for years, the idea came to me to classify endgames
according to rules of thumb and techniques. Of course, you will always take a specific material balance as a
starting-point, but in my view the reader who plays though the material will mainly benefit from having
guidelines given to them, and having explained how they should assess a position, and, especially, what
factors they should pay attention to if they want to bring such an endgame to a successful end in a practical
game. We can’t get around discussing the basic principles first in each type of ending, purely to know which
position is winning and which isn’t. These are endings that should be the stock-in-trade of every player.
I hope that I am offering this theory, which is sometimes ‘tedious’, in a slightly more variegated way than
it is usually done. With the help of my examples, I hope that the principles of a certain endgame will
automatically come to the foreground. In this book, I will try to make the ‘implicit’ rules of thumb and
techniques more ‘explicit’ by bringing them to the forefront. Hopefully, this will give you better grip of the
subject matter, and you will be able to internalize the deeper meaning of certain manoeuvres and ideas.

As said, we can’t get around knowing which endgame position is winning and which one is drawn. Here, the
devil is sometimes in the apparently small details. It is precisely this knowledge that will enhance your
understanding of the middlegame, and will tell you when and how you can liquidate into a winning position.
Unfortunately, knowledge alone is not sufficient. You also need the skills to convert a theoretically winning
position into victory. But more on that later.
Chapter 2
Ingrained fear
Many people will remember those incidents from their childhood years, when as a child they ran into
something new and unknown. Imagine a mother saying to her eight-year-old child, ‘Here you have some
money, go to the baker’s shop and buy a few loaves of bread.’ Of course, the child has often been to the
baker’s with its mother. But now the tiny tot is suddenly on its own. It has to manage by itself – quite
exciting though, isn’t it?
Club players also get such an uneasy feeling when they land in an endgame. Many chess players display a
kind of ingrained fear of the endgame. To them, a liquidation into an ending is a leap in the dark. Many
prefer not to take that leap, but sometimes there is no alternative, and they land in an endgame ‘by
accident’.
Immediately, they are seized by an unsteady feeling, since there are few pieces and pawns left on the
board, while the possibilities seem endless... how empty the board is now! Where should I put my pieces?
There are so many squares... a rook gives check to my king. Where should it go? To the left? Or to the right?
Who can tell? The ‘steady hand’ they had in the opening and part of the middlegame has suddenly gone. An
intense feeling of uncertainty, or rather incapacity, takes hold of them. The right orientation is lost, the right
knowledge is lacking, to say nothing of the skill that is necessary to convert a winning position or steer a
theoretically drawn position into the safe haven of a draw.
Where does this ingrained fear come from? As a chess trainer, I have worked a lot with talented young
players, but also with many adult club players who are definitely able to fend for themselves in many
situations. When I looked at games sent to me by my students, gradually it became clear to me that the
endgame is hard to fathom for many players. I see players selling themselves short in promising
middlegames by avoiding a logical queen trade, banishing their queen to faraway places while the trade
might have resulted in a large or even decisive advantage.
A poignant example occurred in one of my games with a talented player from the same province, a
Candidate Master who is known for his tactical skill and is regarded as a dangerous attacking player. See,
for example, this rapid game in which he managed to overcome Anish Giri, who was still young at the time
but was already the much stronger player on paper.

4
Remco Sprangers 2257
Anish Giri 2523
Roosendaal 2009 (4)
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nc6 5.Nc3 Qc7 6.Be2 a6 7.Be3 Nf6 8.0-0 Be7 9.f4 d6 10.Kh1 0-0
11.Qe1 Nxd4 12.Bxd4 b5 13.Qg3 Bb7 14.a3 Rad8 15.Rad1 Rd7 16.Bf3 Qd8 17.Bxf6 Bxf6 18.e5 Be7
19.Bxb7 Rxb7 20.exd6 Bxd6 21.Ne4 Rd7
Better was 21...Rb6.
22.Qd3 Qc7?!
XIIIIIIIIY
9-+-+-trk+0
9+-wqr+pzpp0
9p+-vlp+-+0
9+p+-+-+-0
9-+-+NzP-+0
9zP-+Q+-+-0
9-zPP+-+PzP0
9+-+R+R+K0
xiiiiiiiiy
After 22...Qe7!, there would have been no trouble for Black at all: 23.f5 Rfd8=. However, he gets problems
after the text.
23.f5! exf5
After 23...e5, White would still have had a tremendous position, but the damage would have been limited.
24.Rxf5
 XIIIIIIIIY
9-+-+-trk+0
9+-wqr+pzpp0
9p+-vl-+-+0
9+p+-+R+-0
9-+-+N+-+0
9zP-+Q+-+-0
9-zPP+-+PzP0
9+-+R+-+K0
xiiiiiiiiy
24...Re8?
The decisive mistake. Black had an ingenious way to keep on fighting here: after 24...Be7 25.Qxd7 Rd8
26.Qxd8+ Bxd8 27.Rc5 Qb8 28.Nd6, White has a substantial advantage but he would still have had a lot to
prove.
25.Rd5! Rde7 26.Rxd6 Kf8
Obviously, 26...Rxe4 was not possible in view of 27.Qxe4!+–.
27.Ng3 Re3 28.Qd2 h5 29.Rd7 Re1+ 30.Rxe1 Rxe1+ 31.Qxe1 Qxd7 32.Nxh5 Qf5 33.Qd1 1-0
33.Qe8+ Kxe8 34.Nxg7+ would have been nice as well.
When I faced Remco in a tournament, I emerged from the opening with a slightly more pleasant position.
But just when I was starting to ask myself whether I had any advantage at all, my opponent opted for a
highly peculiar move:

5
Herman Grooten 2370
Remco Sprangers 2191
Vlissingen 2007 (7)
XIIIIIIIIY
9r+l+r+k+0
9zppvl-wqpzp-0
9-snp+-+P+0
9+-+-+-+p0
9-+-zPp+-+0
9zP-sN-zP-+-0
9LzPQ+-zPP+0
9tR-vL-tR-mK-0
xiiiiiiiiy
Black to move
I had just played 20.g6 and expected the logical reply 20...Be6. However, to my great surprise my opponent
played
20...Bf5??.
That put me in shock for a brief moment. This move costs a rook; is this serious? But then I realized that
Remco was a dangerous attacking player, and that this continuation was a conscious attempt to play for an
attack. As Remco told me after the game, he preferred not to go for 20...Be6 since then after 21.Qxe4 Bxa2
22.Qxe7 Rxe7 23.Rxa2, the queens would have been exchanged, and he hadn’t looked forward to playing an
endgame (against me). Apparently, my reputation as an endgame expert had preceded me...
However, in the game the first thing for me to do was to ward off his attack, which always requires special
care against such creative tacticians.
By the way, after 23...f5!
 XIIIIIIIIY
9r+-+-+k+0
9zppvl-tr-zp-0
9-snp+-+P+0
9+-+-+p+p0
9-+-zP-+-+0
9zP-sN-zP-+-0
9RzP-+-zPP+0
9+-vL-tR-mK-0
xiiiiiiiiy
analysis diagram
there are hardly any problems for Black, and White isn’t in any way better: the g6-pawn will be lost, and
then White will have to find a way to make progress.

20...Bf5?? 21.gxf7+ Kh8 22.fxe8=Q+ Rxe8

Now Black is ready for either ...Qh4 or ...Qd6.
23.Ne2!?
This is the human way to defend the white position.
I immediately rejected lines like 23.f4 exf3 24.Qxf5 Qh4 because I did not fancy getting caught in a
labyrinth of variations. The engine doesn’t agree, and suggests the dry 25.gxf3 Qxe1+ 26.Kg2 Qg3+ 27.Kf1,
flatly claiming that White is totally winning! For example, after 27...Qh2 with the intention 28...Bg3, White
has the strong 28.Bf7, threatening to trade the queens (!), while the rook on e8 is also hanging. There is still
a lot we can learn from the silicon monster!
23...Qd6 24.Ng3
I could also have played 24.Nf4 here, e.g. 24...g5 25.Qc5+–.
XIIIIIIIIY
9-+-+r+-mk0
9zppvl-+-zp-0
9-snpwq-+-+0
9+-+-+l+p0
9-+-zPp+-+0
9zP-+-zP-sN-0
9LzPQ+-zPP+0
9tR-vL-tR-mK-0
xiiiiiiiiy
24...Bg4
We both had calculated that 24...h4 didn’t work on account of 25.Nxf5 Qh2+ 26.Kf1, which brings Black no
benefits, not even after 26...Qh1+ (26...h3 27.gxh3 Qh1+ 28.Ke2 Qf3+ 29.Kd2 Qxf5 30.f4!) 27.Ke2 Qxg2
28.Nxh4 Qg4+ 29.f3 Qxh4 30.fxe4 Rxe4 31.Bd2, and Black has nothing.
25.Qc5
If the opponent doesn’t want to trade queens, I usually know what to do!
25...Qg6 26.Bb1
This is more than sufficient to win. Perhaps 26.Rf1 was even more accurate, with the possible continuation
26...h4 27.Nxe4 Qxe4 (27...Rxe4 28.Qf8+ Kh7 29.Qg8+ Kh6 30.f3+–) 28.f3+–.
26...h4
 XIIIIIIIIY
9-+-+r+-mk0
9zppvl-+-zp-0
9-snp+-+q+0
9+-wQ-+-+-0
9-+-zPp+lzp0
9zP-+-zP-sN-0
9-zP-+-zPP+0
9tRLvL-tR-mK-0
xiiiiiiiiy
27.Nxe4!
After this, Remco threw in the towel. Indeed, Black is completely busted now, since after 27...Rxe4 White
liquidates with 28.f3! Bxf3 29.Qf8+ Kh7 30.Qxf3+–, or after 27...Bf5 with 28.Nd6! h3 29.g3 Bxb1 30.Nxe8
h2+ 31.Kxh2 Be4 32.Qf8+ Kh7 33.Qxg7+ and the queens will be exchanged, after which Black’s attack
comes to a definitive halt.
This type of pertinently wrong decision occurs much more often than many people think. It is high time for
us to discuss this phenomenon of ‘ingrained fear’, and acquaint chess players with some basic knowledge.
Perhaps even more importantly, if we hand you some rules of thumb and techniques that apply in many
endgames, hopefully your uncertainties in this ‘barren’ field will disappear like snow in summer!
Chapter 3
Liquidation
Just like the transition from the opening to the middlegame, the transition from the middlegame to the
endgame harbours a lot of secrets for many club players. Which pieces should they trade off? Which pieces
should they keep on the board? Which pawn(s) will be essential later on? These are all difficult questions
when you do not have any basic knowledge of the endgame.
The loathing some club players have for the endgame is founded on their first impression of endgames:
they are supposed to be boring. I have already indicated that this conviction is reinforced by the fact that
many books on the endgame are boring as well. In conversations with club players, I noticed that many of
them once started to work through such a book in good spirits. But after reading a mere few pages, their
spirits sagged and they lay the book aside. With that, they also subconsciously developed a tendency to
simply avoid the endgame in their games. Their aversion to the endgame induces them, for instance, to
avoid exchanging queens in the late middlegame. Without queens on the board, the chance to get an attack
on the enemy king is gone, and the hated endgame is coming closer and closer...
But shouldn’t we be turning this around? Shouldn’t we start with the idea that we know some of the laws
of the endgame – and that in a certain type of ending, maybe even by coincidence, we know what we are
doing? Wouldn’t it be child’s play to just trade the queens, and other pieces, to enter an ending we have
previously studied?
In my book Attacking Chess for Club Players, I showed that it is sometimes wise during an attack on the
king to focus on the final aim – the mate motif, for instance. If you have a good picture of this in your mind,
then you know what you have to strive for. This works in exactly the same way in the endgame. If you know
exactly into which endgame you have to liquidate, you can work on that with full confidence. Someone once
said to me that if you know more about the endgame, you will also automatically play better in the
middlegame.
Let’s see how former World Champion Garry Kasparov once worked from the opening towards an ending
in which a trivial trick – which he had studied beforehand – is the underlying cause of his victory. We can
assume that Kasparov had the two basic ideas of this ending somewhere in the back of his head throughout
the game. In the variation of the Caro-Kann in which Black plays 4...Bf5 and later 13...0-0-0, often a
structure arises in which Black exchanges his c-pawn for White’s d-pawn by means of the ...c6-c5 push.
6
Help position 1
XIIIIIIIIY
9-+-+-+-+0
9zpp+-+pzp-0
9-+p+p+-zp0
9+-+-+-+P0
9-+-zP-+-+0
9+-+-+-zP-0
9PzPP+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Let’s assume we have a middlegame here with almost all the pieces still on the board. Stereotypically,
White mostly continues

1.c4
after which Black tries to free his game with
1...c5.
If then the white d-pawn is traded off for the black c-pawn, it is to be expected that the rooks and maybe
even the queens will be exchanged on the d-file as well. Various minor pieces will also disappear from the
board, and so both players should know what can happen in the endgame that remains.

2.d4xc5
and a black piece takes back on c5:
2...Q/R/B/Nxc5
An endgame is looming in which the pawn structure looks as follows:
XIIIIIIIIY
9-+-+-+-+0
9zpp+-+pzp-0
9-+-+p+-zp0
9+-+-+-+P0
9-+P+-+-+0
9+-+-+-zP-0
9PzP-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
7
Help position 2
 XIIIIIIIIY
9-+k+-+-+0
9+p+-+pzp-0
9p+-+psn-zp0
9+-+-+-+P0
9-+P+-+P+0
9+-vL-+P+-0
9PzP-+-+-+0
9+K+-+-+-0
xiiiiiiiiy
White to move
In this position, White wants to liquidate into a winning pawn ending. So how does he do this?
1.Bxf6 gxf6 2.f4
The idea is clear: he is threatening to force a breakthrough with g4-g5. Black’s king should hurry to the
kingside to anticipate this:
2...Kd7 3.g5 fxg5 4.fxg5 Ke7
XIIIIIIIIY
9-+-+-+-+0
9+p+-mkp+-0
9p+-+p+-zp0
9+-+-+-zPP0
9-+P+-+-+0
9+-+-+-+-0
9PzP-+-+-+0
9+K+-+-+-0
xiiiiiiiiy
Just in time, the king enters the square of White’s h-pawn. But now a new catastrophe awaits him: White
has a pawn majority on the queenside, and by setting this majority in motion, he is able to create a second
passed pawn. The black king can’t be everywhere at once, and so this second passed pawn will decide the
game:
5.gxh6 Kf8 6.b4 Kg8 7.c5 Kh7 8.a4 Kxh6 9.b5 axb5 10.axb5 Kxh5 11.c6 bxc6 12.bxc6
And the pawn is unstoppable. Please note that Black’s apparently beautiful pawns on e6 and f7 were of no
use to him.

8
Help position 3
 XIIIIIIIIY
9-+k+n+-+0
9+p+-+pzp-0
9p+-+p+-zp0
9+-+-+-zPP0
9-+P+-zP-+0
9+-vL-+-+-0
9PzP-+-+-+0
9+K+-+-+-0
xiiiiiiiiy
White to move
Before this position arose, White had pawns on f3, g4 and h5. Logically, the knight wanted to avoid the
exchange on f6 and retreated to e8. But now suddenly a dark cloud appears over Black’s position: White
crashes through!
1.Bxg7!!
This beautiful shot, creating a passed h-pawn, decides the game in White’s favour. After
1...Nxg7 2.gxh6
the pawn queens. Please note that if it had been Black’s move in the initial position, he could have avoided a
lot of damage with ...f7-f6.
So now it is high time to look at this extraordinary game:

9 – Caro-Kann Defence
Garry Kasparov 2595
Milan Vukic 2460
Skara Ech-tt 1980 (7)
1.e4 c6 2.d4 d5 3.Nd2 dxe4 4.Nxe4 Bf5 5.Ng3 Bg6 6.h4 h6 7.Nf3 Nd7 8.h5 Bh7 9.Bd3 Bxd3 10.Qxd3
e6 11.Bd2 Qc7 12.0-0-0 Ngf6 13.Ne4 0-0-0 14.g3 Nxe4 15.Qxe4 Be7 16.Kb1 Rhe8 17.Qe2 Bd6
18.Rhe1 Re7 19.c4
XIIIIIIIIY
9-+ktr-+-+0
9zppwqntrpzp-0
9-+pvlp+-zp0
9+-+-+-+P0
9-+PzP-+-+0
9+-+-+NzP-0
9PzP-vLQzP-+0
9+K+RtR-+-0
xiiiiiiiiy
19...c5
Black does not allow c4-c5, since then White would get control of a nice diagonal with Bf4. Besides, Black
wants to free his own game.
20.Bc3 Nf6 21.Ne5
Another idea was 21.dxc5 Bxc5 22.Ne5, with advantage to White.
21...cxd4 22.Rxd4
 XIIIIIIIIY
9-+ktr-+-+0
9zppwq-trpzp-0
9-+-vlpsn-zp0
9+-+-sN-+P0
9-+PtR-+-+0
9+-vL-+-zP-0
9PzP-+QzP-+0
9+K+-tR-+-0
xiiiiiiiiy
22...Bxe5
Black thinks he can obtain equality by trading off as many pieces as possible.
23.Rxd8+ Qxd8 24.Bxe5 Rd7 25.Bc3 Qb6 26.g4
The pawn on h5 can become vulnerable in the endgame, and therefore Kasparov strives for the pretty
formation with pawns on f3, g4 and h5.
26...Qd6 27.f3 a6 28.a4 Qd3+
XIIIIIIIIY
9-+k+-+-+0
9+p+r+pzp-0
9p+-+psn-zp0
9+-+-+-+P0
9P+P+-+P+0
9+-vLq+P+-0
9-zP-+Q+-+0
9+K+-tR-+-0
xiiiiiiiiy
Black is still thinking that the more pieces he trades off, the better his chances of a draw will be. But he will
be rudely awakened.
29.Kc1
29.Qxd3 Rxd3 30.Rf1 was to be preferred, since Black’s rook cannot be maintained on d3 – it will be driven
off by the white king.
29...Kc7
It was high time to withdraw the knight from the exchange against the bishop with 29...Ne8, so as to create
a block against that same bishop with ...f7-f6, as we observed above.
30.Qxd3
Anyway!
30...Rxd3 31.Rf1 Kc6 32.Kc2 Rd7
 XIIIIIIIIY
9-+-+-+-+0
9+p+r+pzp-0
9p+k+psn-zp0
9+-+-+-+P0
9P+P+-+P+0
9+-vL-+P+-0
9-zPK+-+-+0
9+-+-+R+-0
xiiiiiiiiy
33.a5!?
With this pawn move, White would like to ‘fix’ the queenside structure, but there was a stronger possibility.
Presumably, 33.b4 would have been much more useful. Black has various ways to defend, but many of his
attempts are doomed to failure:
A) Interesting seems to be 33...b5?, but then follows 34.Bxf6 gxf6 35.Rd1 Rc7 (not 35...Rxd1 36.axb5+ axb5
37.Kxd1, when the pawn ending is an easy win for White; the same also follows after 35...bxa4 36.b5+ axb5
37.cxb5+ Kc7 38.b6+ Kc6 39.Rxd7 Kxd7 40.f4, and White will create a second passed pawn with g4-g5; if
Black withdraws his rook from the trade with 35...Re7, White plays 36.cxb5+ axb5 37.a5±, and his winning
chances are huge with this protected passed pawn), and the endgame with a protected passed pawn
resulting after 36.cxb5+ axb5 37.a5 has to be winning for White. For example, 37...Kb7+ 38.Kb3 and since
Black is lacking counterplay, White’s prospects are very good;
B) 33...Rd8 may be the most tenacious defence;
C) After 33...a5, White’s strongest reply is 34.bxa5! Kc5 35.Rb1 Kxc4 36.Rb4+ Kc5 37.Rb5+ Kc4...
XIIIIIIIIY
9-+-+-+-+0
9+p+r+pzp-0
9-+-+psn-zp0
9zPR+-+-+P0
9P+k+-+P+0
9+-vL-+P+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
after which 38.g5! gives White great chances.
33...Ne8
Black withdraws his knight from the trade.
34.Re1
Preventing 34...f6.
34...Rd6
Black defends e6, hoping to re-introduce the option of ...f7-f6.
35.f4 Nf6?
A serious mistake that immediately costs Black the game.
If 35...Kd7 36.g5 Rc6 37.Rd1+ Kc8
 XIIIIIIIIY
9-+k+n+-+0
9+p+-+pzp-0
9p+r+p+-zp0
9zP-+-+-zPP0
9-+P+-zP-+0
9+-vL-+-+-0
9-zPK+-+-+0
9+-+R+-+-0
xiiiiiiiiy
analysis diagram
38.Bxg7!+–.
However, now it becomes clear that 35...f6? is still impossible in view of 36.Bb4 Rd4 37.Rxe6+ Kd7 38.Re7+
Kd8 39.Kb3 Rxf4 40.Rxb7, and White wins.
Black couldn’t have saved himself either with 35...Kc5
XIIIIIIIIY
9-+-+n+-+0
9+p+-+pzp-0
9p+-trp+-zp0
9zP-mk-+-+P0
9-+P+-zPP+0
9+-vL-+-+-0
9-zPK+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
analysis diagram
(to exclude the possibility of Bb4), since after 36.g5!, slowly but surely the tactical shot Bxg7 starts to
emerge – which, by the way, isn’t possible right away: 36...Kc6 37.b3! (not 37.Bxg7? in view of 37...Nxg7
38.gxh6 Nf5, and the h-pawn is kept under control by the black rook: 39.h7 Rd8, and Black wins).
XIIIIIIIIY
9-+-+n+-+0
9+p+-+pzp-0
9p+ktrp+-zp0
9zP-+-+-zPP0
9-+P+-zP-+0
9+PvL-+-+-0
9-+K+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
analysis diagram
Now Black is slowly but surely counted out, since here 37...Rd8 is met by the winning pawn break 38.g6!.
 XIIIIIIIIY
9-+-+-+-+0
9+p+-+pzp-0
9p+ktrpsn-zp0
9zP-+-+-+P0
9-+P+-zPP+0
9+-vL-+-+-0
9-zPK+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
36.Bxf6!
Kasparov liquidates to a winning pawn ending.
36...gxf6 37.Rd1!
The rooks have to be exchanged immediately, since otherwise Black would be able to use his rook to stop
one of the breakthrough pawns. After the text move, Vukic realized that his situation was hopeless and
resigned.
XIIIIIIIIY
9-+-+-+-+0
9+p+-+p+-0
9p+ktrpzp-zp0
9zP-+-+-+P0
9-+P+-zPP+0
9+-+-+-+-0
9-zPK+-+-+0
9+-+R+-+-0
xiiiiiiiiy
If Black trades with 37...Rxd1 38.Kxd1 Kc5 (38...Kd7 also fails after 39.g5 fxg5 40.fxg5 Ke7 41.gxh6 Kf8
42.b4 f5 43.b5 f4 44.bxa6 bxa6 45.c5, and Black cannot do anything against these two distant passed
pawns), White has the breakthrough 39.g5 fxg5 40.fxg5 hxg5 41.h6 and wins.
So, this game was quite striking since the contours of the later endgame treatment became visible in the
opening. We don’t need to doubt that Kasparov was steering towards this liquidation, and it must have
been in the back of his mind all the time.
Let’s have a look at another striking example I came across in a book by Alexander Koblenz.
Again, before we show this example, some foreknowledge is important. In this case, we are talking about
the principle of ‘rook behind the passed pawn’. In endgame theory, this is one of the best known themes. I
am happy to show this with the help of a practical game fragment.

10
José Castro 2313
I.Alahakoon 2120
Tromsø ol 2014 (5)
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+kzpp0
9-+-+-+-+0
9+-+-+r+-0
9P+-+R+-+0
9+-+-+-+P0
9-+-+-+P+0
9+K+-+-+-0
xiiiiiiiiy
White to move
White has a healthy extra pawn and now he has to find a way to exploit it. One of the ‘standard ideas’ in
rook endings is to place the (own) rook behind the passed pawn. There, the rook as it were ‘drives the
pawn forward’. This strategy works especially well if the opponent is forced to put his rook in front of the
passed pawn. Preferably, the defender places his king in front of the pawn. We will see why this is so in this
example.
40.Re2!
The rook is striving for a2; in passing, it also defends the second rank.
40...h5 41.Kb2
First White improves the position of his king, which he will need to do in any case.
Putting the rook behind the pawn immediately was also possible, since then Black has nothing better than
putting his rook in front of the pawn: 41.Ra2 Ra5 (useless is 41...Rf1+ 42.Kc2 Rf2+ 43.Kb3; also, the
passage of the black king to the queenside does not make a lot of difference, as becomes clear after 41...Ke6
42.a5 Kd7 43.a6 Rf8 44.a7 Ra8 45.Kc2, and the white king makes its entry on the kingside), and White’s
king will be able to conquer the black pawns, or else walk to the queenside anyway to menace the black
rook.
41...g5 42.Kb3 Kg6
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+rzpp0
9P+-+-+-+0
9+K+-+-+P0
9-+-+R+P+0
9+-+-+-+-0
xiiiiiiiiy
The black king remains cut off along the e-file.
43.Ra2
Now the rook gets behind the pawn, and White threatens to push it.
43...Ra5 44.Kb4 Ra8 45.a5 Kf5
Soon, the black rook will be forced into complete passivity, while on the other hand the white rook is quite
active behind its own pawn.
46.a6
 XIIIIIIIIY
9r+-+-+-+0
9+-+-+-+-0
9P+-+-+-+0
9+-+-+kzpp0
9-mK-+-+-+0
9+-+-+-+P0
9R+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
46...Ke4
This is useless.
46...Kf4 didn’t help either, but at least it would have enabled Black to try something. White has to watch out
here, because the black king is moving in the direction of the white pawns. The cleverest continuation here
is 47.Ra3!, to prevent the black king from approaching his pawns. Incidentally, 47.a7 is okay too. After
47...Kg3, White still has to make an effort, as after 48.Kc5...
XIIIIIIIIY
9r+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-mK-+-zpp0
9-+-+-+-+0
9+-+-+-mkP0
9R+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Black can still sacrifice the rook with 48...Rxa7, so as after 49.Rxa7 Kxg2 (49...g4 50.hxg4 hxg4 51.Ra2) to
conquer the pawn on g2. The rook vs pawn ending, however, is lost for him since White’s king is close
enough. For example, 50.Rg7 Kxh3 51.Rxg5 h4, and White wins because his king is approaching quickly:
52.Kd4 Kh2 53.Ke3 h3 54.Kf2 Kh1 55.Ra5 Kh2 56.Ra3, with mate on the next move.
47.a7
This is the clearest illustration of the power of the rook behind the passed pawn.
XIIIIIIIIY
9r+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-zpp0
9-mK-+k+-+0
9+-+-+-+P0
9R+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
Black’s rook is doomed to complete passivity, while White’s rook can move freely along the a-file. Now the
white king walks to b7 to win the black rook:
47...Ke3 48.Kc5 g4 49.hxg4 hxg4 50.Kb6
And Black gave up the fight.
Once we have seen this, we immediately realize that the following position could very well be winning for
White as well:

11
Liquidation example 1
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-zpp0
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
9-+r+-+PzP0
9+-+-tR-+K0
xiiiiiiiiy
White to move
This ending can be won if White plays the following important move:
9.Rb1! 1-0
According to the motto: ‘The rook belongs behind the passed pawn’! As we have seen, the white rook takes
up an ideal position here, forcing Black’s rook to stand in front of the pawn later.
With this scheme in mind, we can now look at a liquidation from the middlegame, from the book Schach
Positionell – Training mit Alexander Koblenz.

12
Liquidation example 2
XIIIIIIIIY
9-+-trr+-mk0
9+-+l+-zpp0
9-wqp+-zp-+0
9+-sN-sn-+-0
9-zP-sNR+-+0
9+-wQ-+-+-0
9-+P+-+PzP0
9+-+-tR-+K0
xiiiiiiiiy
White to move
Imagine we are in a middlegame with all the major pieces still on the board, and with two minor pieces on
each side. If we only look at the pawn structure, we see that White has a 2 vs 1 majority on the queenside,
while Black has 3 vs 2 on the kingside. Also notable is the quite dominant position of White’s knights on c5
and d4, and, superficially speaking, Black is struggling with a not-very-pretty-looking bishop on d7 that is
tied to the defence of the c6-pawn.
So, at first sight, White has a slight positional advantage, but can he get something concrete out of it? White
starts investigating if he can gain something concrete by trading on d7.
1.Nxd7
Giving his beautiful knight for the not-so-pretty bishop. However, it is also clear that now the protection of
the c6-pawn becomes a problem, and this carries more weight.
1...Rxd7 2.Qxc6!?
A pretty shot, capturing a pawn. Suddenly all the tricks are based on Black’s vulnerable back rank.
Thanks to this motif, 2.Nxc6 was also possible here. Funnily enough, it even turns out to work tactically:
2...Rc7 seems to win a piece, but White has the surprising defence 3.Rc4, and now 3...Rec8 is not exactly a
solution for Black in view of 4.Nxe5!, and White wins. Also, 2...Rc8 fails to the surprising 3.Nxe5 Rxc3
4.Nxd7, and since the black queen on b6 is hanging, and mate is threatened at the same time, Black has to
give up material. There are no other options for Black to exploit the pin on the white knight. In short, White
could have won a pawn in this way.
2...Qd8
Trying to stay in the game. The recapture of a pawn with 2...Qxb4? would have cost Black dearly after
3.Qxd7! Nxd7 4.Rxe8+ Nf8 5.c3! (to deflect the queen from the attack on the Re1) 5...Qa3 6.Ne6, winning.
Neither was 2...Qc7 enough to save himself after, for example, 3.h3, solving any possible problems on
White’s own first rank, resulting in a highly favourable ending for White.
3.Nf3
Black’s back-rank problems continue.
3...Rc7
Obviously, 3...Nxc6?? was not possible due to 4.Rxe8+ Qxe8 5.Rxe8#. But now White comes up with a new
surprise:
XIIIIIIIIY
9-+-wqr+-mk0
9+-tr-+-zpp0
9-+Q+-zp-+0
9+-+-sn-+-0
9-zP-+R+-+0
9+-+-+N+-0
9-+P+-+PzP0
9+-+-tR-+K0
xiiiiiiiiy
4.Nxe5!?
Liquidating into a rook ending, in the conviction that this will result in a winning position for him.
An alternative was 4.Qa4, after which Black could continue to complicate matters with 4...Rc3.
4...fxe5
Forced – but now, again, White has a nice shot up his sleeve.
After 4...Rxc6 5.Nf7+ Kg8 6.Nxd8 Rxd8, White continues with 7.c4
XIIIIIIIIY
9-+-tr-+k+0
9+-+-+-zpp0
9-+r+-zp-+0
9+-+-+-+-0
9-zPP+R+-+0
9+-+-+-+-0
9-+-+-+PzP0
9+-+-tR-+K0
xiiiiiiiiy
analysis diagram
after which the connected passed pawns will bring him an easy win.
5.Qxe8+! Qxe8 6.Rxe5
Forcing the exchange of even more pieces.
6...Qg8 7.Re8 Rxc2 8.Rxg8+ Kxg8
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-zpp0
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
9-+r+-+PzP0
9+-+-tR-+K0
xiiiiiiiiy
White had calculated until here. Now he has to find one more important move, connected to the entire
transaction that has led to this rook ending:
9.Rb1!
‘Rook behind the passed pawn’! Since we already know this position from the ‘Liquidation example 1’
diagram, we know that we can go for the liquidation! That the conversion is not at all trivial from here is
not important. The knowledge that this position is winning for him, along with the confidence that he has
the skill to bring this to a successful finish, provides a player with an important extra weapon in the
endgame! Further on in this book, we will devote some more attention to the characteristic conversion of
this ending.
9.h3?, for example, would have led to a drawish ending after 9...Rb2!. Now it is the black rook that has
conquered a spot behind the white passed pawn, and here it fulfils a double role: it prevents the march of
the pawn and also aims its arrows to the white pawns on the kingside.
9...Kf7
Black could also have tried to block the pawn as quickly as possible with 9...Rc6. After 10.g3 Kf7 11.Kg2
Ke6 12.b5 Rb6 13.Kf3, White has to react in the right way to each defending method.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpp0
9-tr-+k+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+KzP-0
9-+-+-+-zP0
9+R+-+-+-0
xiiiiiiiiy
analysis diagram
Black has two basic defences:
A) He brings his king to the front of White’s passed pawn (by moving to the b7-square), freeing his rook
from its passive placement;
B) The king remains in the centre to hinder the white king’s invasion of the kingside.
For each of these ideas, I have selected a plausible series of moves that show how the winning process
works for White:
A) 13...Kd7 14.Rb2 Kc7 15.Rb4 Rh6 16.h4 Rd6 17.Re4 Rd3+ 18.Kf4 Rd7 19.Re6 Kb7 20.Ke5 Kc7 21.h5
Kb7 22.g4 Kc7 23.Ra6 Kb7 24.Rd6 Re7+ 25.Kd5 Kc8 26.Re6 Rf7 27.b6 Kb7 28.Kd6.
 XIIIIIIIIY
9-+-+-+-+0
9+k+-+rzpp0
9-zP-mKR+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Since White can now liquidate into a winning pawn ending, it will be clear that here Black’s resistance is
broken;
B) 13...Kf5 14.Rb4 Ke5 15.h4 Kf5 16.Ke3 Ke5 17.Kd3 Kd5 18.Rg4, and now:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpp0
9-tr-+-+-+0
9+P+k+-+-0
9-+-+-+RzP0
9+-+K+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
B1) Certainly not 18...Rxb5?, since the ensuing rook ending is a simple win for White after 19.Rg5+ Kc6
20.Rxg7 (20.Rxb5?? Kxb5 21.Kd4 Kc6 22.Ke5 Kd7=) 20...Kd6 (if 20...h5 21.Ke4 followed by Rg5) 21.Ke4
(not 21.Rxh7?, as this loses the g3-pawn, while the h-pawn is an unfortunate pawn in most rook endings:
21...Rb3+ 22.Ke4 Rxg3, and a draw is the most probable result) 21...Rb3 22.Kf4.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-tRp0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-mK-zP0
9+r+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
The king takes over the protection of g3, after which the h-pawn is lost: 22...h5 23.Rg5, and after White has
won the h5-pawn, the rest is elementary;
B2) 18...Rb7 19.Rg5+. The rook has switched positions, which serves to drive the black king further back
and so bring his own king into the vicinity of the passed pawn and support it: 19...Kd6 20.Kc4 Rc7+
21.Kb4 g6 22.Rg4 Kd5 23.Rf4 Re7 24.Rc4.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-tr-+p0
9-+-+-+p+0
9+P+k+-+-0
9-mKR+-+-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Thus, the black king is cut off from the white passed pawn, allowing this pawn, with the support of its king,
to march freely.
A sample line is 24...Rb7 25.g4 Kd6 26.Rc6+ Kd5 27.g5 Rf7 28.Rc5+ Kd6 29.b6, and White wins.
10.b5
The pawn marches on as swiftly as possible, to force the black rook to stand in front of it.
10...Ke6 11.b6 Rc8
XIIIIIIIIY
9-+r+-+-+0
9+-+-+-zpp0
9-zP-+k+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+PzP0
9+R+-+-+K0
xiiiiiiiiy
12.h3!?
There are various roads leading to Rome here. Either way, White will have to start using his king, and it will
eventually try to invade on the kingside. At the same time, White may also want to push his pawns forward,
enabling him to attack Black’s pawn structure there.
In principle, there wasn’t much wrong with 12.h4 either, since after 12...Kf5 13.Kh2 Kg4, Black cannot
exploit the weakening of the g4-square. With 14.b7 Rb8 15.Rb4+, everything turns out right for White after
all; the black king will be driven off.
12...Rb8 13.Kh2
Obviously, at this point the white king has to start taking part in the struggle.
13...Kd5
 XIIIIIIIIY
9-tr-+-+-+0
9+-+-+-zpp0
9-zP-+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+PmK0
9+R+-+-+-0
xiiiiiiiiy
In almost all cases, Black’s strategy is to try to place his king in front of White’s passed pawn, thus freeing
his rook to enable it to engage in the struggle with the white king.
14.Kg3 Kc6 15.Kg4
It is always important to investigate pawn endings that may arise. However, 15.b7? spoils half a point here,
since after 15...Rxb7 16.Rxb7 Kxb7 17.Kf4 Kc6 18.Ke5 Kd7, White can make no progress: draw.
15...Kb7
The king has managed to settle in front of the pawn. With this, the necessity to keep the rook behind the
passed pawn disappears for White, and he immediately uses the opportunity to move the rook to the side,
where it will be much more active.
16.Re1!
XIIIIIIIIY
9-tr-+-+-+0
9+k+-+-zpp0
9-zP-+-+-+0
9+-+-+-+-0
9-+-+-+K+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-tR-+-0
xiiiiiiiiy
Making use of the fact that the pawn ending is winning for White.
16...Rg8
16...Kxb6 17.Rb1+ Kc7 18.Rxb8 Kxb8, and now that the white king is closer, it can walk in unhindered:
19.Kf5 Kc7 20.Ke6...
XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-zpp0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
and game over.
17.Re6
Black has landed in a passive position. Any rook move will cost material, and pawn moves also have their
drawbacks, so he is limited to king moves.
17...Ka6
Weakening the pawn structure with 17...h6 hands squares over to the white king: 18.Kh5, and the king
walks in.
XIIIIIIIIY
9-+-+-+r+0
9+k+-+-zp-0
9-zP-+R+-zp0
9+-+-+-+K0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
After 17...g6, White plays 18.Re7+ Kxb6 19.Rxh7 followed by Kg5 and Rh6, and the g6-pawn falls.
18.Kg5
18.Kf5 would have been unwise in view of 18...Rf8+ 19.Kg5 Rf2 20.g4 Rf7, and the black rook is better
placed here than on g8.
XIIIIIIIIY
9-+-+-+r+0
9+-+-+-zpp0
9kzP-+R+-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
The king moves forward to approach the black pawns. White avoids the f-file, as then the black rook can
give check and then protect the pawns from the side.
18...Kb7
Black persists in ‘doing nothing’. Perhaps he could have ventured a breakout with the rook by 18...Rc8, even
if this amounts to nothing with good play: 19.Re7 Rc2 20.g4 Rc3 21.b7 Ka7 22.h4 Rc5+ 23.Kf4 Rc4+.
 XIIIIIIIIY
9-+-+-+-+0
9mkP+-tR-zpp0
9-+-+-+-+0
9+-+-+-+-0
9-+r+-mKPzP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Now it is important to manoeuvre to the right squares with the king: 24.Kf3! Rc3+ 25.Kf2 Rc2+ 26.Ke3
Rc3+ 27.Kd4 Rh3 28.h5 g6
XIIIIIIIIY
9-+-+-+-+0
9mkP+-tR-+p0
9-+-+-+p+0
9+-+-+-+P0
9-+-mK-+P+0
9+-+-+-+r0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
29.Rxh7 gxh5, and here White has various ways to haul in the loot. The most convincing is 30.g5, although
the ‘riskless’ 30.Rxh5 Ra3 31.Ke5 is not a problem either: 31...Kxb7 32.g5 Kc7
XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-mK-zPR0
9-+-+-+-+0
9tr-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
33.Rh7+! (driving the king away again, since if it moves to the back rank, White would win simply by
pushing his pawn), and now:

A) After 33...Kd8, the black king is cut off along the ‘absolute seventh rank’: 34.g6 Rg3 (if 34...Ke8, White
wins with 35.Rh8+, taking control of the promotion square, after which the pawn moves to the other side
unhindered: 35...Ke7 36.g7) 35.Kf6 Rf3+ 36.Ke6 Rg3 37.g7, and Rh8+ followed by g8=Q cannot be
prevented;
B) 33...Kc6 34.g6, and since the black king is now cut off horizontally, the win is child’s play.
19.h4!
White needs a passed pawn on the kingside, so he first pushes this pawn and will see later how he can
make a breach in Black’s pawn formation.
19...Ka6 20.h5 Kb7 21.g4 Ka6
XIIIIIIIIY
9-+-+-+r+0
9+-+-+-zpp0
9kzP-+R+-+0
9+-+-+-mKP0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
22.Kh4
White has to keep manoeuvring carefully, since 22.h6? gxh6+ 23.Kh5 would throw away the win, as after
23...Rg5+ 24.Kh4 Rg6, White cannot make any progress.

22...Kb7 23.h6!
XIIIIIIIIY
9-+-+-+r+0
9+k+-+-zpp0
9-zP-+R+-zP0
9+-+-+-+-0
9-+-+-+PmK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, yes!
23...gxh6
For the defender, it is always important to exchange as many pawns as possible, which is why 23...g5+
would have been bad due to 24.Kh5.
24.Rxh6
He could also have won the hindmost of the two pawns, but then we land in a variation we already saw
above: 24.Re7+? Kxb6 25.Rxh7 Rg6 26.Kh5 Rg5+ 27.Kh4 Rg6, and White can’t make progress.
24...Rg7 25.Kh5
25.g5 would have amounted to the same thing. After 25...Ka6 26.Kh5 Kb7 27.Re6 followed by Kh6, the
pawn on h7 is due for capture.
25...Ka6
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-trp0
9kzP-+-+-tR0
9+-+-+-+K0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
26.Rc6!
The most convenient continuation, since White can use the point of support on c7 for a clever liquidation.
26...Re7
The best chance to create complications. It is clear that the pawn ending after 26...Kb7 27.Rc7+ Rxc7
28.bxc7 Kxc7 29.Kh6 Kd7 30.Kxh7 Ke7 31.g5 is winning for White.
27.Rc7 Re5+
XIIIIIIIIY
9-+-+-+-+0
9+-tR-+-+p0
9kzP-+-+-+0
9+-+-tr-+K0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
28.g5
White could still have botched it by continuing 28.Kh6 Re4 29.b7 Ka7, and now he would have to find
30.Rg7!. After that, he can calmly play Kxh7, reaching a winning position. But not 30.g5??, in view of
30...Rh4+! 31.Kg7 Rh5 32.Kf6 Kb8 33.Rf7 Ka7, and it is not to be seen how White can make progress.
28...Kxb6 29.Rxh7
XIIIIIIIIY
9-+-+-+{+0
9+-+-+-+R0
9-mk-+-+-+0
9+-+-tr-zPK0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The final black pawn has fallen, and, very important: Black’s king is cut off horizontally. Thus, it will never
be able to reach the promotion square of the pawn (without help). This technique will be addressed
extensively in Part III of this book.
White now finishes off as follows:
29...Kc6 30.Kh6 Kd6 31.g6 Re1 32.Rf7!
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+R+-0
9-+-mk-+PmK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tr-+-0
xiiiiiiiiy
Creating both a horizontal and a vertical cut-off!
32...Ke6 33.Rf2 Ra1
33...Rh1+ 34.Kg7 Rg1 35.Kh7 Rh1+ 36.Kg8 Rg1 37.g7 Rh1 38.Re2+ Kd7 39.Re4.
34.g7 Rh1+ 35.Kg6 Rg1+ 36.Kh7 Rh1+ 37.Kg8 Ke7
XIIIIIIIIY
9-+-+-+K+0
9+-+-mk-zP-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-tR-+0
9+-+-+-+r0
xiiiiiiiiy
This is known as the ‘Lucena Position’, and the winning method is called ‘building a bridge’. This endgame
technique will be brought to your attention subsequently in a systematic way. At this point, we will just
show you the way White wins from this position. For now, we can use the knowledge that this is a winning
position as a point of departure:
38.Re2+ Kd7
38...Kf6 39.Kf8; 38...Kd6 39.Re4 Kd5 40.Re8 Rh2 41.Kf7 Rf2+ 42.Ke7 Re2+ 43.Kd7.
39.Re4 Rh2 40.Kf7 Rf2+ 41.Kg6 Rg2+ 42.Kf6 Rf2+ 43.Kg5 Rg2+ 44.Rg4 1-0
In the previous example, White, calculating accurately, exchanged almost all the pieces with the aim to be
left with a ‘rook behind the passed pawn’ ending. Later in this book we will give more thought to the
characteristic examples of this type of ending.
During the discussion of this rook ending, the reader encountered a number of themes like ‘horizontal cut-
off’, ‘king in front of the pawn’, switching position’, ‘absolute seventh rank’ and the like, which may raise
some questions. These notions are part of a wide arsenal of techniques that will be reviewed in this book.
Armed with these basic techniques, each club player will be able to find his way through the most difficult
technical phases of an endgame.
We are gradually finding out that, apparently, strong players already see the contours of a favourable
ending during the opening in certain cases. We also see this in the following game, taken from the book
Fischer and his Predecessors in the World Chess Championship by Max Euwe, who analysed his defeat at the
hands of Fischer. I have thankfully made use of his comments at several points.

13
Bobby Fischer
Max Euwe
Leipzig ol 1960 (7)
 XIIIIIIIIY
9r+-+-vl-tr0
9zp-+qmkpzpp0
9-+-+p+-+0
9+Q+-+-+-0
9-+-+-+-+0
9+-zP-+P+-0
9P+-+-zP-zP0
9tR-vL-mK-+R0
xiiiiiiiiy
White to move
In the well-known Panov Variation of the Caro-Kann, we quickly move from the opening and the
middlegame to the endgame. In his book, Euwe wrote that at the time this game was played, ‘the theory
concludes that Black is at least equal. However, Fischer had analysed this position thoroughly and had
come to the conclusion that White’s position is clearly to be preferred. He manages to substantiate this
assessment convincingly.’
15.Rb1!
Euwe: ‘This is the key move, originally suggested by grandmaster Pal Benko and gratefully adopted by
Fischer. The purpose is clear: to make the completion of Black’s development as difficult as possible. In a
few moves, it will become clear that Black can only complete his development at the cost of a pawn. In most
cases, the black a-pawn will be the victim, which is even more dangerous since White then disposes of a
passed pawn.’
15...Rd8?!
Euwe wanted to try to make a stand on the seventh rank, but this unsettles the development of his kingside.
It seems that Fischer recommended 15...Qxb5, with the possible continuation 16.Rxb5 Kd6 17.Rb7 f6
18.Ke2 Kc6 19.Rf7, and now, I think, Black’s best option is to play 19...a5 (not 19...g5?, as indicated by
Fischer, since White simply wins two pawns with 20.Rxf6 Re8 21.Bxg5), although White still has the better
prospects after 20.Rd1.
16.Be3 Qxb5 17.Rxb5 Rd7
XIIIIIIIIY
9-+-+-vl-tr0
9zp-+rmkpzpp0
9-+-+p+-+0
9+R+-+-+-0
9-+-+-+-+0
9+-zP-vLP+-0
9P+-+-zP-zP0
9+-+-mK-+R0
xiiiiiiiiy
18.Ke2!
Euwe: ‘White is not in a hurry. He can capture the black a-pawn at any moment. This is one of the strong
sides of Fischer’s style: postponing material gain to the most favourable point in time.’
The immediate 18.Ra5 would have provided Black with some counterplay after 18...Kf6 (Euwe gives
18...g6?! 19.Rxa7?! Rxa7 20.Bxa7 Bg7=, but this is much less strong since White has the better move
19.Ke2!). Yet, White retains the upper hand if he goes about things accurately: 19.Ke2! (19.Rxa7?! Rxa7
20.Bxa7 Be7, and White has nothing after 21.Be3 Ra8) 19...Bd6 20.Bg5+ Kg6 21.Rg1 Rb8 22.Be7+ Kh6
23.Bxd6 Rxd6 24.Rxa7...
 XIIIIIIIIY
9-tr-+-+-+0
9tR-+-+pzpp0
9-+-trp+-mk0
9+-+-+-+-0
9-+-+-+-+0
9+-zP-+P+-0
9P+-+KzP-zP0
9+-+-+-tR-0
xiiiiiiiiy
analysis diagram
and White has a nice extra pawn at any rate.

18...f6
Now, 18...g6 would have been met by 19.Bc5+ Ke8 20.Rb8+ Rd8 21.Rhb1, and Black ends up in stormy
weather.
XIIIIIIIIY
9-+-+-vl-tr0
9zp-+rmk-zpp0
9-+-+pzp-+0
9+R+-+-+-0
9-+-+-+-+0
9+-zP-vLP+-0
9P+-+KzP-zP0
9+-+-+-+R0
xiiiiiiiiy
19.Rd1!?
A typical Fischer move. He trades off Black’s only active piece to be able to use his own pieces optimally
afterwards.
19.Rhb1 would have led to the same goal.
19...Rxd1 20.Kxd1 Kd7 21.Rb8
Euwe: ‘Again, admirable self-control. The black a-pawn is not running away, and White now threatens
22.Bc5.’

21...Kc6 22.Bxa7
Finally, he takes the pawn.

22...g5
Black has to try to untangle by preparing ...Bg7.

23.a4 Bg7 24.Rb6+ Kd5 25.Rb7 Bf8 26.Rb8 Bg7 27.Rb5+ Kc6 28.Rb6+ Kd5
 XIIIIIIIIY
9-+-+-+-tr0
9vL-+-+-vlp0
9-tR-+pzp-+0
9+-+k+-zp-0
9P+-+-+-+0
9+-zP-+P+-0
9-+-+-zP-zP0
9+-+K+-+-0
xiiiiiiiiy
Euwe: ‘This is the same position as four moves ago. We see such move repetitions (not three times, but two
times the same position!) more often in Fischer’s games. In this way, he gives his opponent the opportunity
to make a wrong choice, and moreover the notion is drummed into the opponent that he has to dance to
Fischer’s piping entirely.’

29.a5!
Euwe: ‘Remarkable straightforwardness of White’s strategy. His a-pawn will have to do all the work. This is
also what is going to happen – in the most efficient way possible.’

29...f5
The alternative 29...Rc8 also leads to a very precarious position for Black, e.g. 30.a6 Rxc3 31.Rb7 Bf8...
XIIIIIIIIY
9-+-+-vl-+0
9vLR+-+-+p0
9P+-+pzp-+0
9+-+k+-zp-0
9-+-+-+-+0
9+-tr-+P+-0
9-+-+-zP-zP0
9+-+K+-+-0
xiiiiiiiiy
analysis diagram
and now 32.Rd7+! is immediately winning (Euwe analyses 32.Be3 Ra3 33.a7 followed by Rb8, which
indeed also wins) 32...Kc6 33.Rd8, after which Black has to give up material. White threatens both 34.Rxf8
and 34.Rc8+. Here, 33...Bc5 obviously fails to 34.Bxc5 Rxc5 35.Rc8+, when the pawn ending wins for White.

30.Bb8
Of course, Fischer puts the bishop on this square. If now he is also allowed to push a5-a6-a7, the game is
over.

30...Rc8?
This loses without a fight. The best defence was 30...Rd8, since then a5-a6?? fails to ...Kc5. But also the
move indicated by Euwe, 30...Be5, misses the mark due to a nasty little trick: 31.c4+! (not 31.Rb5+ as
indicated by Euwe, in view of 31...Kc4 32.Rxe5 (32.Bxe5 Rd8+!) 32...Rxb8 ‘and White’s task is not so simple’)
31...Kd4 32.Bxe5+ Kxe5 33.Kc2, and the ensuing rook ending with two white passed pawns is not hard to
win.

31.a6 Rxc3
 XIIIIIIIIY
9-vL-+-+-+0
9+-+-+-vlp0
9PtR-+p+-+0
9+-+k+pzp-0
9-+-+-+-+0
9+-tr-+P+-0
9-+-+-zP-zP0
9+-+K+-+-0
xiiiiiiiiy
32.Rb5+?!
This moment of carelessness by Fischer passes by unnoticed in the analysis. The American involuntarily
makes things difficult for himself with this move.
32...Kc4
Euwe gives himself a question mark for this move. He wrote that he should have played 32...Kc6 so as after
33.Ra5 to continue with 33...Bd4 ‘after which the obvious 34.Be5 only leads to a draw due to 34...Rc5.’ This
variation is entirely correct; Black’s position is even slightly favourable. But the Dutch former World
Champion did not recognize that after 34.Ke2 Rc2+!? (Euwe exclusively analysed 34...Rb3, which loses to
35.a7 Rb2+ 36.Kd3 Bxa7 37.Bxa7, and the ‘obstacles’ White still has to overcome – as Euwe wrote – do not
look all that insurmountable) Black starts a liquidation that leads to complications White had not catered
for, e.g. 35.Kd3 Bb6 36.a7
XIIIIIIIIY
9-vL-+-+-+0
9zP-+-+-+p0
9-vlk+p+-+0
9tR-+-+pzp-0
9-+-+-+-+0
9+-+K+P+-0
9-+r+-zP-zP0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
36...Rd2+!. Incidentally, White retains ample winning chances with 37.Kc4!.
33.Rb7 Bd4 34.Rc7+ Kd3 35.Rxc3+ Kxc3
XIIIIIIIIY
9-vL-+-+-+0
9+-+-+-+p0
9P+-+p+-+0
9+-+-+pzp-0
9-+-vl-+-+0
9+-mk-+P+-0
9-+-+-zP-zP0
9+-+K+-+-0
xiiiiiiiiy
36.Be5!
A beautiful final move that ends all resistance. Black resigned.
One remark in Euwe’s argument kept echoing in my mind: ‘His a-pawn will have to do all the work. This is
also what is going to happen – in the most efficient way possible.’
I thought this was so aptly put that I used this idea in many of my own games to put a passed pawn to work.
Here is an example:

14
Ton Ellenbroek 2078
Herman Grooten 2328
Hoogeveen 2009 (1)
XIIIIIIIIY
9r+-+-+-+0
9+-+-tR-zpp0
9-+-+-zp-+0
9zpp+k+-+-0
9-+-zP-+-+0
9+PmK-+-zP-0
9-+-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Black has managed to place his rook behind a potential passed pawn. In the meantime, White has posted
his rook on the seventh rank, hoping to create sufficient counterplay. But this will backfire:
34...a4
Here, I decided to ‘let the a-pawn do the work’. The engine thinks 34...b4+ is also a good move, if not a
better one. White’s best reply then is 35.Kd2 (an ‘ordinary human being’ would opt for 35.Kd3 here, but
then 35...Rc8! is very strong, with the possible continuation 36.Rd7+ Ke6 37.Ra7 Rc3+ 38.Kd2 Rxb3
39.Rxa5 Rb2+ 40.Kd3 Rxh2, and Black is winning), but Black still has plenty of winning chances after
35...g5! 36.Rxh7 Kxd4.

34...a4 35.bxa4 bxa4 36.Kb2

Forced – otherwise 36...a3 is immediately decisive already. As we know, the king has to move in front of the
enemy passed pawn if there is an enemy rook behind it.
36...g5!
It is very important to keep ‘pawn pairs’ intact, especially in rook endings. That’s why I opted for this move.
37.Rxh7 Kxd4
XIIIIIIIIY
9r+-+-+-+0
9+-+-+-+R0
9-+-+-zp-+0
9+-+-+-zp-0
9p+-mk-+-+0
9+-+-+-zP-0
9-mK-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
Black has an extra pawn, but he still has to do some work to convert it. It requires some endgame
knowledge.
38.h4
White tries to create counterplay in some way or other. He could also have defended with 38.g4, when
Black has to take some time to determine how to convert his extra pawn. Best seems to be 38...Re8 39.Ka3
Re4 40.h3 Ke3 41.Rf7 Rf4 42.Ra7 Kf3 43.Rh7 Kg2 44.Rh6
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-zp-tR0
9+-+-+-zp-0
9p+-+-trP+0
9mK-+-+-+P0
9-+-+-+k+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
44...Re4!, and White has no useful move. 45.Kb2 is strongly met by 45...Re3, cutting off the enemy king
along the third rank, after which Black can follow up with a winning liquidation: 46.Rxf6 Kxh3, and the rest
is simple.
38...Ke4
The king has to move in the direction of the kingside as quickly as possible, not only to join its own pawns,
but also to be able to attack the enemy pawns.
39.hxg5
He tries to exchange as many pawns as possible, which is a good defensive strategy in principle.
Here, 39.h5 was also hopeless as after, for example, 39...Kf3 40.h6 Kxg3 41.Rf7 Rh8 42.Rxf6 g4 43.Rc6 Kh3,
Black’s g-pawn decides the issue.
39...fxg5 40.Rf7
XIIIIIIIIY
9r+-+-+-+0
9+-+-+R+-0
9-+-+-+-+0
9+-+-+-zp-0
9p+-+k+-+0
9+-+-+-zP-0
9-mK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This way White keeps the black king at bay for a while. This is the signal for Black to activate his rook from
behind the a-pawn:
40...Rb8+!
Black opts for a liquidation to an ending that is ‘theoretically winning’ for him. Thus, he exchanges one
advantage for another, on the basis of the required knowledge. It isn’t very hard to see that Black is winning.
41.Ka3 Rb3+ 42.Kxa4 Rxg3 43.Kb4
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+R+-0
9-+-+-+-+0
9+-+-+-zp-0
9-mK-+k+-+0
9+-+-+-tr-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
43...Re3
Here, White surrendered. The only remaining pawn is supported by both the king and the rook, while
White’s king is not only standing at a large distance, but is horizontally cut off as well. The win is no longer
difficult, and might go something like this: 44.Kc4 g4 45.Rg7 g3 46.Rg8 Kf3 47.Kd4 Kf2 48.Ra8 g2 49.Ra2+
Re2–+.
Chapter 4
Knowledge and skill
We have already briefly mentioned that there is a difference between knowledge and skill – i.e., the
difference between ‘know’ and ‘can’. As the famous football player Johan Cruijff once said, you can ‘know’
something, but that doesn’t mean that you ‘can’ work with this knowledge. There are many chess players
who live according to the adage ‘Knowledge is power’, and in a certain sense this works in chess. But
without certain skills, there will be situations when you will not be able to apply this knowledge in the right
way.
Let’s take a look at a lesson in which I tested the endgame skills of a group of my students. For this
purpose, I constructed a number of positions in a rook ending. I played out these positions in a
simultaneous setting against ten or so players in the age range of 8-11 years. They were allowed to take
back their moves. The players wrote down their moves on ‘simul sheets’ we had developed for the occasion
(see the screenshot below). The idea was that when they had the feeling they had done something wrong,
they could go back to a position in which they thought they could do better. They could write down their
new move in the next column.

As soon as I thought that a player had handled a certain position successfully, he could move to the next
position, in which the win was a little more complicated.
 XIIIIIIIIY
9-+-+-tr-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+P+0
9+-+-+P+-0
9-+-+-+K+0
9+-+R+-+-0
xiiiiiiiiy
Level I
XIIIIIIIIY
9-+-tr-+-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+P+0
9+-mK-+P+-0
9-+-+-+-+0
9+-+-+-+R0
xiiiiiiiiy
Level II
XIIIIIIIIY
9-+-tr-+-+0
9+-+-+-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+P+0
9+-mK-+P+-0
9-+-+R+-+0
9+-+-+-+-0
xiiiiiiiiy
Level III
Remarkably, all the players realized they had a winning position, but more than half of them had no idea
how to bring it to a favourable conclusion. Therefore, I would stop the play now and then to formulate a
few guidelines together with the pupils. These are the ‘rules of thumb’ that are so important for an
awareness of the essence of a winning process. For example, we can formulate the following rules of thumb
for this endgame:

Rules of thumb
• The player who has the two pawns has to push those pawns together.
• Preferably, the pawns should be supported by their own king.
• The stronger side has to prevent the pawns from being blocked.
Here are a few typical flaws in the play of the students:
1. Inexperienced young players often have the tendency to push only one of the pawns, and then losing it
somewhere along the line.
2. Or they don’t realize yet that it is good to use their king to help pushing the pawns.
3. Finally, it is important to use their own rook well to drive the enemy king back and prevent it from
blocking the pawns in a favourable way.
With the help of the above rules of thumb, we tracked down an essential technique for cases where their
own king was cut off from the pawns by the enemy rook (see Level II). To experienced players, it is clear
that White first has to remove the blockade by manoeuvring the rook, for example, to d2 (with Rh1-h2-d2),
to enable the king to support its own pawns before they are pushed forward. This is possible because the
pawn ending would be an easy win. We also made this technique explicit by writing it down on the
whiteboard:

Technique
In a rook ending with two extra passed pawns where the king has ‘drifted off’, the stronger side
can remove the blockade of the king by threatening a liquidation into a pawn ending. This
forces the opponent to give up the blockade, and then the king can move closer to its pawns.
In this way, we put the rules of thumb and one technique into words, and this brought light into the
darkness! By verbalizing a certain problem every time, we also found the solution. The beautiful thing was
that the students started to play this endgame better and better as we went along – some of them even
played it to a win in a perfect way! This approach is precisely what helps club players to internalize the
problems as well as the skills.
Of course, a trainer can go on endlessly doing such exercises with his pupils. But since the contact hours
in training sessions are generally expensive and therefore scarce, I came up with another way to train this.
Playing out a position against a chess engine on the computer is a good way to master such skills. A trainer
can give a number of ‘pre-cooked’ positions to ambitious players who can work on them at home. It is
recommended to arrange the order of the positions according to their degree of complexity.
The following examples show how a student can dabble with them during training or at home. In these
exercise positions, the stronger side has to overcome a couple of obstacles to force the win.
XIIIIIIIIY
9-+-tr-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+k+p0
9-+K+-+-zP0
9+-+-+P+-0
9-+-+R+P+0
9+-+-+-+-0
xiiiiiiiiy
Level IV
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpk0
9-+-+-+-+0
9+-+R+-+P0
9-+-+-+-+0
9+-+-+P+-0
9r+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
Level V

XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
9-tR-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
Level VI
The conclusion is that without the right skills, a player may allow certain winning positions to peter out to a
draw.
Who better than Magnus Carlsen can show us what it takes to convert a winning endgame against a strong
opponent? In the following rook ending, he has two extra pawns, but these are not connected passed pawns.
This makes things a lot more difficult, but it is a joy to see how easily Magnus performs this technical task:

15
Magnus Carlsen 2813
Teimour Radjabov 2740
Bazna 2010 (7)
Comments are by Magnus Carlsen himself.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mk-+-0
9R+P+-+-+0
9+-+-+-+-0
9-+r+-zPK+0
9+-+-+-+-0
xiiiiiiiiy
White to move
50.Ra5+!
‘My only problem was the passive rook on a4. Now it will be activated either in front of, or on the other side
of the c4-pawn.
50...Ke6
50...Kf4 51.Rc5+–; 50...Kd4 51.Rd5+! Kxc4 52.Rd8 cuts off the king and wins easily: 52...Ra2 53.Kg3 Ra7
54.f4 Rg7+ 55.Kf3 Rf7 56.Kg4 Rg7+ 57.Kf5+–.
51.Rc5 Kd6 52.Rd5+ Ke6 53.Rd4 Ke5 54.Rh4 Rc3 55.c5
Exploiting the opportunity to advance the pawn. The pawn endings are always won.
55...Kf5 56.Rh8!
The most accurate.
56...Kf4 57.Rc8 Ke5 58.c6 Kd6 59.f3
Preparing Kg3 and Kg4 followed by f4 and so forth.
59...Rc5 60.Kg3 Rg5+ 61.Kh4 Rg1 62.f4 Ke7 63.c7 Rc1
Sets a final trap.
63...Kd7 64.Rd8+ Kxc7 65.Rd2 wins easily.
64.Kg5 1-0
64.Rh8?? or 64.Rg8 was not to be recommended!’

Tablebase endgames
That such skills cannot be learned from one day to the next will be demonstrated with the help of two quite
extreme examples. These skills can sometimes be of such a high level – as in the following endings – that a
human player is forced to pull out. The Tablebases know better! What matters is that a club player has to
realize that knowledge isn’t power as long as you don’t master certain skills.
The most extreme example may be winning with two bishops against a lone knight. The Tablebases tell
us that this is winning for the stronger side, but the manoeuvres of the two-bishops side are so intricate
that it is infeasible for most human players to develop the skill to convert such an ending to a win with a
steady hand. Incidentally, in this case, the same applies to the defender.
It has been clear for several decades that the entry of the computer has proved its great worth for chess.
What I find intriguing is that these days we can use the computer to establish with a large degree of
certainty whether a game has been played well or not. With the Tablebases, this probability has even been
turned into 100% certainty. The numbers indicate how many moves it takes to win or draw from a certain
position in the fastest way. It will be clear that in many cases, a human player is not able to internalize these
‘clean’ optimal moves during a game under pressure from the clock. Nevertheless, I am sometimes amazed
by what the strongest grandmasters in the world can manage. With pleasure, I present the following
example, connected with an anecdote, that I once published on the Dutch-language website schaaksite.nl.
If two endgame specialists face each other, you can bet your boots that an endgame will come on the board.
One especially interesting game was played between the Dutchman Jan Timman and the Brit Jonathan
Speelman in Linares 1992. In that game, Speelman came up with a totally unexpected rook sacrifice that
turned out to be winning, But Speelman didn’t continue correctly, and he ended up a piece down in an
endgame. He defended extraordinary well, after which the curious ending of two bishops vs knight arose. It
used to be assumed that this endgame was a draw, but with the modern Tablebases it can be conclusively
established that the two bishops always beat the knight – even in less than 50 moves.
The story goes that from Linares Timman called the well-known Dutch professor Jaap van den Herik to
discuss his adjourned game. Apparently, Van den Herik set a research assistant to work to make a specific
database on this ending type. They sent Timman a large number of A4 sheets by fax with rough data with
which he tried in his hotel room to discover certain regularities. For a human, this seems a well-nigh
impossible job. I can see Timman sitting there with a chessboard and a pile of paper, trying to find his way
through the endless stream of variations...
The only thing I had ever read about this ending was that the defender had to try to get the knight on the
‘ugly’ square g2 (or g7, b7 or b2), in the vicinity of its king. It seems that it is hard to drive the knight away
from the king in that case. However, it is also known that the bishops side can drive the knight away from
this square, and then the knight side has to try to get the knight to one of the other ‘ugly’ squares, and let
his own king run after it.
I have consulted the current Tablebases again, and have indicated after every move how many moves the
optimal path to the win takes. If a player manages to do this within a margin of, say, five moves of the
optimal conversion, I think that is very impressive. Two endgame virtuosos facing each other should be
capable of a top performance. When you play through the following fragment, you can also see where the
defender misses a much more tenacious defence.

16
Jan Timman 2620
Jonathan Speelman 2630
Linares 1992 (10)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-vL0
9+L+K+-+-0
9-+-+-+-+0
9+-+-+ksn-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
In my opinion, Timman played this ending very strongly – for a human player. Also, the final position, in
which Speelman resigned, is remarkable. The conversion would still have taken quite a lot of moves, which
are moreover not very obvious to the layman, but Speelman assumed that a player of Timman’s calibre
would find the path to the win. The numbers probably look a bit sterile here, but that doesn’t make
Timman’s performance any less impressive!
After White’s previous move (66.Bxb5), the last pawn has been removed from the board, and from here on
the 50-move rule is valid. Speelman finds the best defence, after which it is mate in 60 moves according to
the Tablebases.
Let me make a slight distinction here: mate in 60 moves doesn’t necessarily mean that White cannot win
due to the 50-move rule. If, for instance, he can win the knight on move 48 during the winning process, and
he needs 12 more moves to give mate, he has remained within the 50-move margin in both cases.
66...Nf5 67.Bf8 Mate in 64.
67.Bg5 is mate in 60.
67...Ne3+ Mate in 59.
67...Kf4 was the most tenacious (mate in 63).
68.Kd4 Mate in 60.
68.Ke5 Mate in 59.
68...Ng2 Mate in 58.
Speelman clearly knows the defence with the knight on the ‘ugly’ square g2.
68...Nf5+ Mate in 59.
69.Kd3 Mate in 66.
69.Bc6+ was mate in 58.
69...Nf4+ Mate in 65.
70.Kd2 Mate in 69.
70.Kc4 was mate in 65.
70...Ng2 Mate in 56.
70...Ke4 was mate in 68.
71.Bc6+ Mate in 56.
71...Kf2 Mate in 55.
72.Bd6 Mate in 57.
72.Bb7 Mate in 55.
72...Nh4 Mate in 56.
73.Bc5+ Mate in 56.
73...Kg3 Mate in 55.
74.Bb6 Ng2 Mate in 52.
74...Kf4 was mate in 60.
75.Bd5 Mate in 54.
75.Bc7+ was mate in 52.
75...Nh4 Mate in 49.
75...Nf4 was mate in 53.
76.Ke2 Mate in 49.
76...Kf4 Mate in 48.
77.Bb3 Mate in 48.
77...Nf5 Mate in 45.
77...Ng6 was mate in 47.
78.Bc7+ Mate in 45.
78...Kg5 Mate in 44.
79.Be5 Mate in 44.
79...Kg4 Mate in 43.
80.Bc2 Mate in 44.
80.Kf2 was mate in 43.
80...Ng3+ Mate in 41.
80...Ne7 was mate in 43.
81.Kf2 Mate in 41.
81...Nf5 Mate in 40.
82.Bd1+ Mate in 40.
82...Kg5 Mate in 39.
83.Kf3 Mate in 39.
83...Nh4+ Mate in 38.
84.Ke4 Mate in 38.
84...Nf5 Mate in 37.
85.Ba4 Mate in 39.
85.Bc2 was mate in 37.
85...Ne7 Mate in 35.
85...Kg6 was mate in 38.
86.Bd6 Mate in 40.
86.Bc3 was mate in 35.
86...Ng8 Mate in 28.
86...Nf5 was mate in 39.
87.Bf4+ Mate in 37.
87.Ke5 was mate in 28.
87...Kg6 Mate in 28.
87...Kf6 was mate in 36.
88.Ke5 Mate in 28.
88...Nf6 Mate in 26.
88...Kf7 was mate in 27.
89.Bb5 Mate in 30.
89.Ke6 was mate in 26.
89...Kf7 Mate in 29.
90.Bc4+ Mate in 29.
90...Kg6 Mate in 28.
91.Ke6 Mate in 28.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+Ksnk+0
9+-+-+-+-0
9-+L+-vL-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Here, Speelman resigned, even though the win will still take quite a few moves. A sympathetic gesture,
showing a lot of respect for Timman’s endgame skills.
After 91...Ng4 92.Be2 Nh6 93.Bd2 Ng8 94.Bd3+ Kg7 95.Bc3+ Kf8 96.Bb2 Nh6 97.Bg6 Ng8 98.Bh5 Nh6
99.Kf6 Ng8+ 100.Kg6 Ke8 (100...Ne7+ 101.Kh7 Nd5 (101...Ng6 is objectively the most tenacious, but
then the rest is quite clear) 102.Ba3+ Ne7 103.Kh6 Kg8 104.Bxe7101.Kg7+ Ke7 102.Kxg8, it is now mate
in 16 moves.
Late 2021 saw the World Championship Match between the title defender Magnus Carlsen and his
challenger Ian Nepomniachtchi. After the first five games had ended in draws after some tough fighting, an
epic struggle took place on Friday 3 December. The World Champion had determined for himself that on
this day he was going to take a risk. During the middlegame, the evaluation of the position fluctuated
several times, and eventually a highly curious ending arose on the board in which White had a rook plus a
knight and two pawns against Black’s queen. The players had been confined to the board for many hours
already. In passing, they broke the record of the longest-ever game in a World Championship match. That
record had been held by Viktor Kortchnoi and Anatoly Karpov, who had played a game of 124 moves in the
fifth round of their match in Baguio City 1978.
The whole world could see that this ending between Carlsen and Nepomniachtchi could be found in the
Tablebases. While I was following the game live, I also tried to use a Tablebase site online. Of course the site
was down, very probably because there were a few more people who had the same idea! However, I saw on
the Chess.com livestream that the commentators had managed to check the ending in the Tablebases. It
was a theoretical draw, but this was impossible for flesh-and-blood players to figure out during a game.
We enter the game at the moment when ‘Nepo’ diverges from his defence, which had been excellent until
that point, and founders.

17
Magnus Carlsen 2855
Ian Nepomniachtchi 2782
Dubai Wch m 2021 (6)
XIIIIIIIIY
9-+-+k+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-zPR+-0
9-+-+-zP-+0
9+-+-+-sNK0
9q+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
In this position, the Tablebases indicate that there are only two moves that draw, i.e. 130...Qb1 and
130...Qc2, working from a distance with the queen. For a human player, it is impossible to determine
during a game which move draws and which one doesn’t. The move Black plays is losing:
130...Qe6? 131.Kh4 Qh6+ 132.Nh5 Qh7
XIIIIIIIIY
9-+-+k+-+0
9+-+-+-+q0
9-+-+-+-+0
9+-+-zPR+N0
9-+-+-zP-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
133.e6!
It has to be admitted that the World Champion – who must have been tired at the end of this long game –
finds a perfect way to liquidate to the win. This is a fantastic move since, among other things, the rook is
taboo.
133...Qg6
At this point, the Tablebases say that there are only two winning moves for White. Magnus played one of
them – and quite quickly too!
Of course, 133...Qxf5 would have been refuted by 134.Ng7+.
134.Rf7! Kd8
The trick is that the pawn ending is winning after 134...Qxe6 135.Ng7+ Kxf7 136.Nxe6 Kxe6 137.Kg5 Kf7
138.Kf5!. That is some liquidation!
135.f5 Qg1
 XIIIIIIIIY
9-+-mk-+-+0
9+-+-+R+-0
9-+-+P+-+0
9+-+-+P+N0
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-wq-0
xiiiiiiiiy
136.Ng7!
With this powerful move, White ends all resistance. The pawn march e6-e7-e8=Q cannot be prevented.
After 136...Qh2+ 137.Kg5 Qg3+ 138.Kh6 Qh3+ 139.Kg6 Qg4+ 140.Kh7 Qh3+ 141.Kg8, the king has
escaped from the checks.Nepomniachtchi surrendered, and thus Carlsen had dealt an important blow in
this match that kept him on the throne for two more years.
Why did I show you the above two fragments? Should a club player work with these Tablebases? The
answer is: no! I showed the fragments to make it clear that there is a broad division line between theory
and practice. In other words, a player has to master certain (elementary) skills to be able to convert a
winning position. For top grandmasters, the bar is high, and in some cases apparently too high, as the
Tablebases show.
Can we blame Nepomniachtchi for not finding the narrow path to the draw? I think not. Not only was he
struggling with the clock, but Carlsen could do what he wanted in this ending and also because of this,
Nepo’s defensive task was a well-nigh impossible one.
The advice for club players is that they should acquire those skills that can reasonably be learned. It is
precisely these skills that can be the key to success. And in this book I hope to tell you what these basic
skills are.
Chapter 5
Fortresses in practice
Analysing with strong(-er) players will always provide you with new insights. I have done this during my
entire career as a chess player, and I know that these new insights have made me stronger. But what can
you do with this as a tournament player? That is the big question. However, it is clear to me by now that if
you are interested in something – or, even better, intrigued by it – then you will store an image of it
somewhere in your memory. It has happened to me fairly frequently that I was able to use such an image to
my advantage later. A kind of ‘hatch’ is opened in your memory, enabling you to get access to the
knowledge you need to determine, for instance, whether a chess position is a win or a draw. I have the
same situation with pop music. When I hear a certain song, I get spontaneous associations with situations
that occurred earlier in my life.
The notion of a ‘fortress’ was a huge eye-opener when I first acquainted myself with it.

18
Fortress R+§ vs Q
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+kwq-+-+0
9+R+-+-+-0
9-+P+-+-+0
9+K+-+-+-0
xiiiiiiiiy
White to move
It is possible to hold a draw with rook + pawn vs queen in certain positions.

The conditions are quite easy to understand:

- the side with the rook must always have tempo moves at its disposal;
- the enemy king should not be allowed to enter;
- the pawn should be on the second rank, so that the enemy queen cannot enter the position from the rear.
This means that pawns on your third rank are no good.

In this position, White has tempo moves along the third rank with the rook, where it cannot be won with a
check. Preferably, White moves the rook to and fro between the protected squares b3 and d3.
1.Rd3 Qe4 2.Rb3 Qe1+ 3.Kb2 Qd2 4.Rd3 etc... ½-½

19
Fortress R+§ vs Q
 XIIIIIIIIY
9-wQ-+-+-+0
9+-+pmk-+-0
9-+-+r+-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Philidor already proved in 1777 that the side with the rook can also draw with a central pawn. He always
has two moves with the rook at his disposal: ...Rc6 and ...Re6 – provided that it keeps occupying the sixth
rank to prevent the white king from crossing it. If the white king could approach the black king on the
seventh rank, e.g., on g7, the position would be lost.
So these positions in which a queen cannot beat a rook + pawn are intriguing. There is a whole arsenal of
positions here that you can find – should you be interested – in chess literature or on the Internet. As a
chess player, you will always be curious about the ‘limits’ of the game, and these are amply expressed in the
notion of a ‘fortress’.
I have always been interested to know where these limits lie, and there is no way to learn this better than
by hearing it from absolute top players – hence the following anecdote.
Once I was in the Swiss lakeside resort of Lugano, where a strong Open was played annually. In one of those
years, former World Champion Boris Spassky and World Championship candidate Viktor Kortchnoi played
there. Apparently, both gentlemen were not at war with each other at the time, as they were later during
one of their matches. In Lugano, they were getting on quite well.

I was intrigued by a game Kortchnoi was playing against someone – I think it was the American Jay
Whitehead. The latter had been lost with White for half of the game, but the Swiss GM hesitated a few times
to deal the decisive blow. Several times, he did not go for a certain sharp breakthrough in the centre
because he saw something that was ‘not clear’. As it turned out later, the American had hardly been aware
of all this. But in the meantime, Kortchnoi had made a bad mistake, and he had a losing position at the
adjournment.
When, during the post mortem, Whitehead naively asked why Kortchnoi hadn’t played a certain move, the
latter played a long variation on the board, shouting, ‘What do you mean, didn’t I see this variation? Yes, of
course I calculated this variation. But after this, this and this I win. But why didn’t I play it? Because of this,
this, that, this and then this!’ For a moment, the American was speechless under this torrent of violence, but
he didn’t allow himself to be overwhelmed, and a little later he tried another move he had considered.
Seated next to Kortchnoi, Spassky was following the analysis, and when Kortchnoi failed to produce an
answer at some point, this time the American was instantly given a ready-made answer by the former
World Champion. These moves too were played on the board at a furious pace and with great cogency, with
a triumphant look in Spassky’s eyes, saying something like ‘what makes you think that we world top
players wouldn’t see something like that...’

As said, Kortchnoi had blundered just before the time control, and after 40 moves the adjourned position
was totally losing for him. The game would be resumed two hours later. While his opponent went for a bite
to eat, Kortchnoi remained seated at the board to analyse. I wanted to leave too, but I was also very curious
how this would end. I followed the scene from some distance.

Kortchnoi got help from another grandmaster – the Romanian Florin Gheorghiu, if I remember correctly.
Kortchnoi asked Gheorghiu to try to beat him from that winning position. And here it became clear how
strong – more importantly, how tenacious Kortchnoi was in the endgame. The Romanian hardly managed
to prove a win for White, even though all of those who had gathered around them were completely
convinced that there was one.
The game continued, and when I returned later in the evening, they were still playing. As it turned out,
Kortchnoi had been forced to give up material, but he was building a fortress now. When Whitehead didn’t
play optimally, Kortchnoi got what he had aimed for: an impregnable fortress.
Of course, the question is: what is a fortress? Let’s start with the type of fortress Kortchnoi got.

20
Fortress R+N+§ vs Q+2§
After Whitehead-Kortchnoi
XIIIIIIIIY
9-+-+-+-+0
9+-+nmk-+-0
9-+-+r+p+0
9+P+Q+-zP-0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The type of fortress that came on the board between Whitehead and Kortchnoi had the form of this position,
constructed by me. Black’s rook and knight work together to control the b6-square. Both pieces are
protected by the black king, and there is no way for the white king to cross the e-file. The fortress
characteristic that the weaker side should not be put into zugzwang applies here as well. The knight can
shuttle endlessly between b6 and d7. There is no way through, even though modern engines indicate scores
of around +3 for White.
Back to the struggle of queen vs rook. An analogous case from grandmaster practice appeared on the board
in the famous Wijk aan Zee tournament, also between two endgame virtuosos.

21
Jan Timman 2655
John Nunn 2590
Wijk aan Zee 1982 (2)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-wq-+-+-0
9p+-+k+-+0
9+-+R+-+-0
9-mKP+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
In this position, Black even has an extra a-pawn, but funnily enough this doesn’t make a difference; it is still
a draw.
63.Ra3
White being allowed to play the rook to this square (of course, the b3-square is not available) is one of the
reasons why this position is a draw: White has this square at his disposal since the rook is also protected
here.
63...Qb4+ 64.Ka2 Kd5 65.Rd3+ Kc5 66.Ra3 Qc4+ 67.Kb2 Kb4
Slowly but surely, Black has put his pieces where he wants them, but the position is still a draw.
68.Rd3 Qe4!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9pmk-+q+-+0
9+-+R+-+-0
9-mKP+-+-+0
9+-+-+-+-0
xiiiiiiiiy
A nasty move, though there is still no problem for White. But now there is only one good move, and White
has to find it:
69.Ka2??
Wrong! And losing after all.
The only good move was 69.Ra3! after which it is still a draw – also, of course, because 69...Qxc2+ 70.Kxc2
Kxa3 does not win.
69...a3!
Now White loses his pawn. Timman resigned immediately; the queen vs rook ending is winning. How it is
to be played is another story – we will work this out further on in this book.
Another possible variation is 70.Rb3+ (70.Kb1 Qe1+ 71.Ka2 Qc1, and the c2-pawn is again lost) 70...Kc4
71.Rd3 (71.Rxa3 Qxc2+ 72.Ka1 is a theoretical loss) 71...Qe2 72.Kb1 a2+, and Black wins.
Accidents can easily happen...
Sometimes, a picture tells more than a thousand words. Therefore, I will now present to you a collection of
positions that are useful to peruse and store in your memory. Experience teaches us that there will come a
moment when you will be able to use one of these cases in your own practice – perhaps to make a draw or
otherwise to prevent it from appearing on the board.
Obviously, you will never get most of these cases on the board. But precisely the knowledge of such
positions may help you in your decision to steer clear from such obscure cases, or, on the other hand, to
steer for them when you are in trouble – like Kortchnoi did!

22
Fortress B+N vs Q
 XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-vl-0
9-+-+-+-+0
9+-+-sn-+-0
9-+-+-+-+0
9+-+-+-+-0
9-mKQ+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Black holds the draw with two minor pieces vs a queen with this set-up. The knight covers important
squares, preventing the white king from coming nearer. The knight is protected by the bishop, which, in
turn, is protected by the king. Since the king will always have sufficient moves at its disposal, Black can
never be put in zugzwang.

23
Fortress N+N vs Q
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-snn+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-mKQ+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Two knights are also able to keep a queen at bay, provided that they are neatly placed next to each other,
close to their own king, so they can control a whole series of squares. White can do no better than draw.
This elicited a joke from the Dutch grandmaster Hein Donner, who said, ‘these days a queen is worth
nothing. Here you can take if off the board and put it back again – the position is a draw either way!’

24
R+2§ vs B+2§
XIIIIIIIIY
9-+-+-+k+0
9+R+-+-+-0
9-+-+-vlp+0
9+-+K+-+p0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
White to move
A position in which White is a ‘full exchange’ up (rook vs bishop) can quite easily be a draw, even with two
pawns on both sides. I once managed to escape with a draw in such a situation. It is essential here that the
pawns are placed on light squares (g6/h5) and that the bishop can stay on the long diagonal.
1.Ke6 Bd4 2.g4 hxg4 3.hxg4
XIIIIIIIIY
9-+-+-+k+0
9+R+-+-+-0
9-+-+K+p+0
9+-+-+-+-0
9-+-vl-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black can continue making moves with the bishop along the a1-h8 diagonal.
3...Bc3
Of course not 3...Bh8??, when Black’s pieces are so unfortunately placed that White wins anyway: 4.Rb8+
Kh7 5.Rxh8+! Kxh8 6.g5, and White wins the remaining pawn ending.
There is still a big snake in the grass if White tries to slip inside with his king via g5. He can try this as
follows:
4.Kd5 Bf6 5.Ke4 Bc3 6.Kf4
Black has to prevent Kf4-g5 under any circumstances. He can do this in two ways, with both ...Bd2+ as well
as the main line (maintaining the bishop on the long diagonal):
6...Bf6 7.Rc7
Here, Black has to take great care: there are only two moves that hold the draw.
7...Kf8
And White cannot make progress.
7...Bh4 also saves Black. However, this is a very strange square for the bishop, and therefore I wouldn’t
advise playing such a move during a game: 8.g5
XIIIIIIIIY
9-+-+-+k+0
9+-tR-+-+-0
9-+-+-+p+0
9+-+-+-zP-0
9-+-+-mK-vl0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
8...Be1! (absolutely the only move already, which makes you think...) 9.Rd7 (now White has ‘bad luck’, since
the king cannot move to e5 in view of 9...Bg3+) 9...Bc3 10.Ke4 Bb2, and White doesn’t get any further.
But after 7...Kh8?, Black would lose after all!: 8.Rf7! Bh4 9.Ke5 Bg5 10.Ke6 Kg8 11.Rb7 Bh4 12.Rb5 Kg7
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+K+p+0
9+R+-+-+-0
9-+-+-+Pvl0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
13.g5! and after 13...Be1 14.Rb7+ Kg8 15.Kf6, White gets the f6-square for his king and wins.
The following two remarkable endgame studies, in which the white player has a seemingly unbridgeable
material disadvantage of only one piece vs a queen (with pawns on both sides), show that in chess,
wonders never cease.

25
End of a study by Ernest Pogosyants
Bulletin Central Chess Club 1965
XIIIIIIIIY
9-+-+-+-+0
9+-+-sN-zpk0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-+-+0
9mK-+-zp-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White to play and draw
In this position, Black has a dangerous passed pawn that cannot be stopped by the white king. He will have
to stop it with the knight, but that is also impossible at the moment. Therefore, he seeks salvation in the
only forcing move available:
1.g6+! Kh8
It is clear that after 1...Kh6, the e-pawn is lost to 2.Nf5+.
Now the knight cannot stop the black pawn either. So, should White resign?
2.Kb4!
No! White can make a draw by creating a quite curious type of fortress.
2...e2 3.Kc5 e1=Q 4.Kd6
 XIIIIIIIIY
9-+-+-+-mk0
9+-+-sN-zp-0
9-+-mK-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-wq-+-0
xiiiiiiiiy
Just in time, the king manages to join the knight. It is essential that the knight stays on e7, keeping the black
king permanently locked up in its ‘cage’. Since Black has no other material besides his queen to undertake
something, he has to settle for a draw. A remarkable fortress!

26
B vs Q
Study (source unknown)
XIIIIIIIIY
9-+-+-mk-+0
9+-+p+p+q0
9-+-zP-vLp+0
9+-+-+-zPp0
9-+-+-+-mK0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White to play and draw
It seems improbable, but White can hold the draw in this position by keeping the black queen (and king)
locked up in the corner. Black has a clear plan: king to e8, queen to f8, attacking the pawn on d6. Then Be7
fails to ...Qxe7 and ... Kxe7, after which Black wins the pawn ending. What can White actually do about this?
He embarks on a remarkable king march:
1.Kg3! h4+
Or 1...Ke8 2.h4 Qg8 3.Kf4 Qf8 4.Ke5.
Now it seems like the queen can also escape from her chains via h5.
2.Kf2 Qh5 3.Ke1!
 XIIIIIIIIY
9-+-+-mk-+0
9+-+p+p+-0
9-+-zP-vLp+0
9+-+-+-zPq0
9-+-+-+-zp0
9+-+-+-+P0
9-+-+-+P+0
9+-+-mK-+-0
xiiiiiiiiy
The only move to keep the queen locked in her ‘cage’. Note how beautifully White’s pawns control the
escape squares of the queen.
3...Ke8
Black already vacates the f8-square for the queen, since it will land on that square after three giant steps,
securing the win. Or not...?
4.Kd2
The king has to stay in the vicinity of the d1- and e2-squares. But is White also doing anything constructive
with this move?
4...Qh7 5.Kd3! Qg8 6.Kd4! Qf8 7.Kd5
XIIIIIIIIY
9-+-+kwq-+0
9+-+p+p+-0
9-+-zP-vLp+0
9+-+K+-zP-0
9-+-+-+-zp0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
Just in time, the king reaches the centre to protect the d6-pawn, making Be7 unnecessary. It is a pure
optical illusion: from d2, the white king also needed only three moves to get to the d6-pawn.
7...Qg8
The queen is on her way to h5 again. So, the white king has to return in a hurry, but even if this still looks
like a race between a hare (the black queen!) and a turtle (the white king), they both reach the finish at the
same time!
8.Kd4 Qh7 9.Kd3 Qh5 10.Kd2 ½-½
Chapter 6
Various theoretical draws
Chess can be unfathomable sometimes, and there are things you cannot conceive behind the board in a
practical game. In such cases, knowledge is useful. Therefore I will now present a collection of well-known
and perhaps lesser-known drawn positions.

Bishop of the wrong colour

27
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-mK-0
9-+-+L+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This situation may be known to you. It illustrates the theme of the ‘bishop of the wrong colour’. With a
dark-squared bishop, this position would be winning for White. With the light-squared bishop, White can
make no progress, since he cannot drive the enemy king away from the corner square h8 to promote his
pawn.
You could even add a couple more white h-pawns and light-squared bishops, and the position remains a
draw!

Knight + rook’s pawn

The struggle of knight versus rook’s pawn is a different story, as we will see further on in this book. Here,
we present two bizarre situations in which one side even has a knight plus a rook’s pawn, and still cannot
win.
In the position on the right, the black king refuses to leave the corner square a8, and without giving
stalemate, White can make no progress.

28
XIIIIIIIIY
9-+-+-+-+0
9zPk+-+-+-0
9-+-+-+-+0
9+NmK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
29
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-sn-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+K+-zp0
9+-+-+-+k0
xiiiiiiiiy
White to move
Again, Black has an extra knight plus a rook’s pawn, but if White now plays the right move, it is also a dead
draw. He has to choose between 1.Kf2 and 1.Kf1, so there is a 50% chance to save half a point. Which one
would you choose?
1.Kf2!
With 1.Kf1??, White throws away the draw: 1...Nf5 2.Kf2 Ne3!, and the white king cannot move to f1, and
so the black king will emerge from the corner on the next move to enable the pawn to promote.

1.Kf2! Nh5 2.Kf1 Ng3+ 3.Kf2 Ne4+ 4.Kf1

And now the white king cannot be driven away from the f1- and f2-squares.

An extra bishop
The material advantage of a bishop and a pawn usually means a simple win. But also here, there are a
couple of notable exceptions that a chess player has to have seen at least once in his life. Fancy making your
escape to a draw in this way...

30
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mkL0
9-+-+-+P+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White is a bishop plus a pawn up – not even a rook’s pawn, and still the position is a draw! Indeed, the black
king cannot be driven away from g7 and h8 without being stalemated.

31
 XIIIIIIIIY
9-+-+-+k+0
9+-+-+-+p0
9-+-mK-+pzP0
9+-+-+-zP-0
9-+-+-+-+0
9+-vL-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White has a full extra bishop, of the right colour even, and also far advanced pawns, but nevertheless this
position too is a dead draw. There is no way he can chase the black king out of the corner. Even more
frustrating, if we add the same amount of white and black pawns on the respective diagonals c1-h6 and b1-
h7, the position still cannot be won. White does have one way to give Black the opportunity to go wrong:
1.Ke6 Kf8 2.Kf6 Kg8 3.Bb2
A waiting move!
3...Kf8!
And White doesn’t get any further.
After 3...Kh8?? 4.Kf7, Black would suddenly be mated!
Once during a team match, I saw one of my teammates resigning in the last diagram position... clearly he
wasn’t hampered by any knowledge of this remarkable draw position.

Extra exchange + pawn

32
XIIIIIIIIY
9-+-+-mk-+0
9tR-+-+-+-0
9-+-+-zP-+0
9+-+-mK-+-0
9-+-+-+-+0
9+l+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This is a curious case. White has an extra exchange plus an extra pawn, but he cannot win with correct play.
Even worse: if White had the chance to move the pawn back one square, the position would be winning for
him. But in this case, the bishop remains on the diagonal to prevent the f6-f7 push. If the white king walks
to g6 to try to carry through f6-f7, it gets a check from the rear on the b1-h7 diagonal. So, Black has to see to
it that he is able to give that check at the right moment, and that the checking square on that diagonal is not
‘coincidentally’ covered by the white rook:
1.Rc7
Of course, White can start with 1.f7, but then Black replies 1...Kg7! – certainly not 1...Bxf7??, on account of
2.Kf6 Bh5 (2...Be8 3.Ra8+–) 3.Ra8+ Be8 4.Rb8, and the bishop is lost.
1.Rc7 Ba2 2.Rb7 Bc4 3.Kf5 Bd3+ 4.Kg5 Bc4
XIIIIIIIIY
9-+-+-mk-+0
9+R+-+-+-0
9-+-+-zP-+0
9+-+-+-mK-0
9-+l+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
5.Kg6
5.Rd7 Bb3 6.Kg6 Bc2+.
5...Bd3+
And the white king has to retreat – draw.
Here too, we can say that it is quite improbable that the bishop side can hold the draw. Black has to find
some very clever manoeuvres, but if he manages to do this, the split of the point is guaranteed!
In this chapter, we have identified (sometimes implicitly) a number of techniques and rules of thumb that
we will mention explicitly one more time:

Techniques
• A passed pawn has to do the work – preferably in the most efficient way.
• A fortress is a position in which the side that is substantially down on material can still hold
the draw.
• In a rook ending in which the stronger side has two extra connected passed pawns and a
‘drifted-off’ king, he can break the blockade of his king by threatening to liquidate into a pawn
ending. Then, the opponent has to abandon the blockade, and the stronger side’s king can join
its pawns.

Rules of thumb
• The rook belongs behind the passed pawn.
• In many endgames (especially rook endings), it is important to keep pawn pairs intact.
• In a rook ending, the side that has two connected passed pawns has to push these pawns
together.
• Pawns should preferably be supported by their own king.
• The stronger side has to prevent his pawns from being blocked.
In the following chapters, we will highlight many techniques and rules of thumb in the same way, to make
the club player acquainted with various ‘weapons’ and ‘tools’ that he can employ in the endgame.
Chapter 7
Exercises
Exercise 1
XIIIIIIIIY
9Q+-+-+-+0
9+-+-+-mk-0
9-+-+Pzp-+0
9+-+-+P+-0
9p+-+L+K+0
9+l+-+-+-0
9-+-+-+-wq0
9+-+-+-+-0
xiiiiiiiiy
Black to move
In this sharp position, in which both sides are attacking, does Black have a forced win?

Exercise 2
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-zPl+0
9+-mk-mK-+-0
xiiiiiiiiy
White to move
How will this end?

Exercise 3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
9-+K+-mk-+0
9+-+-+-+-0
9-vl-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
White to move
With two pawns for a bishop, only White has winning chances. But Black can control the b-pawn with ...Be5,
and the g-pawn will be a prey for the black king. So can White achieve anything here?

Exercise 4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+R+-+-0
9K+-+-+-+0
9+-+-+-+-0
9Pzp-+l+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The situation looks hopeless for White. The rook is hanging, and at the same time Black threatens to
promote with 1...b1=Q. But there is a very pretty hidden salvation. What is it?

Exercise 5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9K+-+-+-+0
9+-+-+-+-0
9-+-+-+-sN0
9+-+-+-+p0
9-+-+-+-mk0
9+-+-+-+-0
xiiiiiiiiy
White to move
Can White draw in this position? Analyse a few options.

Exercise 6
XIIIIIIIIY
9-+-+-+-+0
9+k+-+-+-0
9nzP-+-+-+0
9zPK+-+-+-0
9-+-+-+-+0
9+-+-vL-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
My opponent thought he could win here. Can he? Or can Black make a draw in one way or other?
Chapter 8
Solutions
1
Constructed position
1997
Black has to give check, as otherwise he will be checked, starting with 2.Qb7+.
1...Bd1+
The advantage of this check is that White has only one answer:
2.Bf3 Qg2+!
And so this move is also logical.
3.Kf4
Other moves lose as well: 3.Kh5 Qg5#; or 3.Kh4 Qf2+ 4.Kg4 (4.Kh5 Bxf3+–+; 4.Kh3 Bxf3–+)
4...Bxf3+–+.
3...Qxf3+
Black liquidates into a winning pawn ending.
4.Qxf3 Bxf3 5.Kxf3 a3 0-1
This pawn cannot be stopped.

2
End of a study by Pietro Rossi
WCCT 2004
White wins with:
1.f4!
The idea of this move is that the two passed pawns cannot be both stopped by the bishop – and
certainly not by the black king.
1...gxf3
If 1...g3, 2.f5 Bd5 3.f6 g2 4.Kf2, and at least one of White’s pawns will get through.
2.Kf2!
Of course, the f3-pawn has to be blocked, as otherwise Black would be able to open the diagonal with
2...f2 check. Now Black cannot prevent 3.a8=Q.
1-0

3
Study by Henri Rinck
Basler Nachrichten 1937
The answer is: yes! With
1.Kd5!
White attacks the e5-square that is needed for the bishop to stop the b-pawn. The problem for Black is
that his king has to support the e5-square.
1...Be5
Or 1...Kf5 2.b7 Be5 3.g4+, and White wins in the same way as in the main line.
2.g3+! Kf5 3.g4+
The g-pawn marches on with gain of tempo, exploiting the fact that the black king has to keep
protecting the bishop.
3...Kf6
Also after 3...Kf4 4.g5 Kf5, the black king has to stay in the square of the pawn, but that won’t help him:
5.g6 Kf6 6.g7, and the pawn breaks through.
4.g5+ Kf5 5.g6 Kf6 6.g7 1-0
 Such a pawn walking straight to the other side of the board is known in the world of endgame studies
as the ‘Excelsior theme’.

4
Study by V.Yakhontov
Shakhmatnoye Obozrenye 1982
Only with
1.Rd1!!
can White achieve something:
1...Bxd1+ 2.Ka3! b1=N+
Promotion to a heavy piece (queen or rook) leads to a draw: 2...b1=Q and 2...b1=R are both stalemate!
3.Kb2
Attacking the knight.
3...Nd2 4.Kc1!
And this double attack guarantees the draw.

5
Constructed position
2008
White’s best chance is:
1.Nf5
But if Black finds the right reply, he still wins:
1...Kg1!
The only winning move. 1...Kg2 is met by 2.Ne3+, and White holds the draw by circling around with
his knight on the e3-, f1-, h2- and g4-squares. The knight cannot be forced to abandon these squares, as will
be discussed in Part II: 2...Kg3 3.Nf1+ Kf2 4.Nh2 Kg2 5.Ng4 Kg3 6.Ne3 etc.
2.Nd4
Now Black has to watch out!
2...Kf2! 0-1
The pawn cannot be stopped. 2...Kg2 also suffices, since the knight cannot come near the pawn;
however, after 2...h2?? 3.Nf3+, White holds the draw.

6
Dennis de Vreugt
Herman Grooten
Hilversum ch-NED 2007 (1)
82...Nc7+!
The only move, but Dennis hadn’t seen it coming. The knight can always exploit the fact that White
cannot afford to lose his b-pawn.
83.Kc4?
Now it is a forced draw.
He could still have won with 83.Kc5! Na6+ 84.Kd6. Now the bishop has to try to dominate the knight.
This can be done in a way that we can learn from the Tablebases: 84...Nb4 85.Bf2 Na6 86.Bg3 Nb4 87.Kc5
Nc2 (after 87...Na6+ 88.Kb5, the knight has no more good moves) 88.Bf2 Ka6 89.Kc6! Nb4+ 90.Kc7 Nd5+
(90...Kxa5 91.Be1) 91.Kb8 Kxa5 92.b7 Ka6 93.Bc5 Nf6 94.Kc7 Nd7 95.Be3+–.
 XIIIIIIIIY
9-+-+-+-+0
9+PmKn+-+-0
9k+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-vL-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram

And Black has ended up in zugzwang.

83...Kc6!
Exactly this move forces the draw. The king controls a number of important light squares from here.
After any other move, the knight might have got stuck, but after the text move, the half point is almost in the
pocket.
84.Bf4
Now the draw is certain, but there was already no win available. After 84.Bd4 Nd5 85.Bf2 Nc7 86.Bg3
Nd5, White cannot make any progress because Black is always threatening ...Nxb6, e.g. 87.Bf2 Nc7.
84...Nd5
Suddenly, the knight has managed to reach this square, threatening not only 85... Nxf4 but also
85...Nxb6+.
85.b7 Kxb7 86.Kxd5 Ka8 ½-½
 Part II
Elementary endgames
Chapter 9
Introduction
When as a young player (14) I’d got to playing ‘normal games’, there was no way around it: sometimes I
would get an endgame on the board. I have no memory of the way I handled those endgames, but I cannot
imagine I was very skilful in them. Neither did I have any knowledge of endgames, but fortunately there
was a boy at school who not only was a better chess player than me, but also had a number of chess books
in his possession. Every now and then we would open such a book, whenever one of us had suffered
another painful defeat in the endgame. We would try to look up the type of ending in question, to see if we
could discover any principles.
For example, we had Yury Averbakh’s famous ‘manuals’ (translated into German as Lehrbuch der
Schachendspiele, ‘Band’ 1 and 2), which should have been capable of helping us to find our way in
theoretically known positions. However, this material was not only very dull, but much of it was also way
over our heads. Nevertheless, we acquitted ourselves as well as we could of the task of improving our
endgame knowledge, with the aim of playing the endgame much better than we had been doing. Alas, we
were disappointed. There is a big difference between knowledge and skill. You may know something, but to
be able to use this knowledge is another thing. This tends to go wrong often, and we learned this the hard
way.
However, determined as we were (also determined to learn German, which was a required subject in
secondary schools), we tried to analyse our way through various sections. Later, my father, who also played
chess at my club, gave me another book by Averbakh (published in 1963) that had been translated from
Russian into Dutch and presented endgames in a more elementary form. The title was What Every Chess
Player Should Know About the Endgame. I now found out that I had already learned quite a bit in the
meantime. In his final chapter (titled ‘Practical endgames’) Averbakh discussed various types of endgames.
He showed endgames with multiple pieces, corresponding squares in pawn endings, opposite-coloured
bishops endings and, for example, the eternal battle between knight and bishop. A new world opened up for
me – a large piece of virgin territory...
There was one big difference between those old days and the times we know now. Games were
adjourned, and could be analysed before the resumption. It was the pre-engine age! I have noticed that
analysing adjourned games was an ideal method to discover the secrets of a certain ending. After all, if you
have an adjourned position, you want to win it if your position is better, and you want to make it as hard as
you can for your opponent if your position is worse or even losing. Both in the competition of the local
chess club and in matches against other clubs, games could be adjourned. This could even happen in open
tournaments, and then you would continue the game in the evening, after a two-hour break. If the game
wasn’t finished by then, there would be a new session in the next morning, and if then another
adjournment was necessary, the adjourned game would be counted as half a point for the pairing of the
next round. Those were real marathon sessions, since everything had to be done between the ‘normal’
round games. In my experience, analysing with various chess friends, but also with passers-by who were
prepared to roll up their sleeves, was always very instructive and useful. We would reap the fruits of this
labour at a much later stage – we will get back to this below.

For the sake of completeness, we can’t avoid presenting the elementary endgames to you in this book. And
neither can we avoid charting the most common principles in the various types of piece vs pawn struggles.
Many endings can, and do, lead to such ‘bare’ endings of piece vs pawn. And our game is so difficult that
miracles can happen even with such scarce material! Only if you start realizing this, can you start to
appreciate the endgame more – at least, this is my experience. How great it is for a practical player to know
whether the ending you are about to liquidate into is a win or not!
Sometimes, the margin between a win and a draw depends on a single tempo, or an apparently
minuscule difference. So, we are playing a long and tough game, only to find out that the difference between
a win and a draw is there for the taking in the far endgame! In my career as a trainer, I have often seen half
or even full points being thrown away in games, due to a chronic lack of knowledge or sometimes a lack of
insight. This may sound harsh, but it is the reason why it pays to study endgames, since liquidating from the
middlegame becomes a lot easier when you know which pieces you want to keep on the board in the
ending. And there is one thing we should never forget: endgames can be very interesting and, above all, fun.
It is the same as with all things: the more you know about them, the more enjoyable it gets!

So, let’s say that we will start this part by determining the goal in the ending to which we want to liquidate
from the middlegame. In such cases, knowledge is power! I hope that my explanations of what we call the
‘elementary endgames’ will be a little more lively than you are used to from most other manuals. As a chess
trainer, I like to see my pupils taking an active learning attitude, as this is the best way to systematically
store knowledge in memory.
Therefore, you will find a few special features in this part, but also in the rest of this book:

• Rules of thumb
Making various principles explicit is very important. For the reader, ‘verbalizing’ such laws in a position is
very useful to ‘internalize’ the concept in question. Please pay attention! The term ‘rule of thumb’ already
suggests that the rule will apply often, but not always. Of course, everything depends on the specific
situation on the board.

• Visual support
It doesn’t always work, but I like to work with visual aids. Sometimes, marks, lines, arrows and other
symbols have been added to a diagram to illustrate the issues.

• Link to practice
With all the theoretical reflections, we cannot neglect practical play in my opinion. Wherever I could, I have
traced and analysed examples from games by my students, club players, and also sometimes from my own
practice as a player, but also from games by strong players, so that I could link them to their theoretical
background. Naturally, we will see that strong players too can miscalculate, precisely because their
knowledge is ‘in the way’!

• Various working methods:

1. Analysis assignments
You could think of an analysis session with a chess friend. It is always nice to work together with someone,
and these days this can be very well done online.
2. Quizzes
By reproducing a position on a chessboard, covering the text below with a piece of cardboard paper and
slowly sliding it downwards, you can actively try to find a certain sequence of moves. You will be given the
solution directly after each question, but it is often better to be forced to think independently to memorize
something.
3. Positions for playing out
A game situation can be very well imitated during a training session. Therefore, I have constructed a few
nice positions for chess students to try to play out to the end. Ideally, this is done at the board, with a clock,
and against a flesh-and-blood player of similar strength. If that isn’t possible, it can also be done online. And
even if that isn’t an option, there are always plenty of chess programs on the computer (or on the internet)
that can be set on different levels to play out these positions. You may take back a move, and the computer
won’t complain! Such exercises are, in fact, essential to learn to apply certain principles.
Chapter 10
Queen versus pawn
In this chapter. I would like to present the queen vs pawn ending. At first sight, this is not a very interesting
type of ending, as in most cases the queen will easily beat the enemy pawn. Only if the pawn has reached
the penultimate rank, do (technical) problems appear. Let’s have a look at the first examples.

33
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9wQ-+-+-+-0
9-+-+-+-+0
9+-+-mk-+-0
9-+-zp-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
If the queen can get in front of the enemy pawn, the position is a simple win:

1.Qh5! Kf2
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+Q0
9-+-+-+-+0
9+-+-+-+-0
9-+-zp-mk-+0
9+-+-+-+-0
xiiiiiiiiy
2.Qd1 1-0
Thus, the queen prevents the king from reaching the promotion rank. The rest is simple. After 2...Ke3, the
white king will approach quickly and will drive the black king away from the protection of the pawn.

34
 XIIIIIIIIY
9-+-wQ-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9-+kzp-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
If the white king is close to the enemy pawn, it is simple:

1.Ke2 1-0
The king prevents the promotion of the pawn, which will be lost on the next move.
With the help of these examples, we can establish a few rules of thumb:

Rules of thumb
• Queen on the promotion square always wins.
• King close to the enemy pawn wins most of the time.

Now, if our king is very far away, can we then still force the win? The answer is: yes, we can. There is a
technique for it.

35
XIIIIIIIIY
9Q+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-zpk+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This position is winning for White. The technique is to force the black king in front of its own pawn by
giving checks with the queen, after which the white king can move closer. This process is repeated many
times, until the white king is able to help with the elimination of the enemy pawn, or with delivering
checkmate:
1.Qe4+ Kf2 2.Qd3
 XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+Q+-+-0
9-+-zp-mk-+0
9+-+-+-+-0
xiiiiiiiiy
2...Ke1 3.Qe3+ Kd1
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-wQ-+-0
9-+-zp-+-+0
9+-+k+-+-0
xiiiiiiiiy
The first stage is completed, and the white king can take a step.
4.Kg7 Kc2 5.Qe2 Kc1 6.Qc4+ Kb2 7.Qd3 Kc1 8.Qc3+ Kd1 9.Kf6 Ke2

Stage 2.
10.Qc2 Ke1 11.Qe4+ Kf2 12.Qd3 Ke1 13.Qe3+ Kd1 14.Ke5 Kc2

Stage 3.
15.Qe2 Kc1 16.Qc4+ Kb2 17.Qd3 Kc1 18.Qc3+ Kd1 19.Ke4 Ke2

Stage 4.
20.Qe3+ Kd1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-wQ-+-0
9-+-zp-+-+0
9+-+k+-+-0
xiiiiiiiiy
21.Kd3 Kc1 22.Qxd2+ 1-0

The pawn has been captured, and checkmate follows.

With the help of this example, we can establish the following basic technique:
Technique
By manoeuvring in the right way, the queen forces the enemy king to stand in front of its own
pawn to enable the attacking king to take a step closer. This process is repeated until the king
of the queen’s side can help to capture the pawn or even give checkmate.

In the previous positions, we looked at the ending with a black d-pawn. What happens if Black has a pawn
on the c-file?

36
XIIIIIIIIY
9-+-+-wQ-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-mkp+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Qb4+ Ka2 2.Qc3 Kb1 3.Qb3+
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+Q+-+-+-0
9-+p+-+-+0
9+k+-+-+-0
xiiiiiiiiy
Normally, the defending king has to move in front of its own pawn, but this is not necessary here:
3...Ka1!
As after
4.Qxc2
Black is stalemated. So, White can make no progress here: it is a draw.
We have seen that White can make no progress against a c-pawn and has to settle for a draw. Of course, this
also applies to an f-pawn, and in fact the same applies also for both rook pawns. In these four cases, the
defending side holds the draw. In other words, common theory states that both rook’s and bishop’s pawns
guarantee a draw to the weaker side, while the stronger side wins with all the other pawns. Unfortunately,
of course, there are also exceptions to this rule. If the king of the stronger side is close enough, White can
still force a win in some of these ‘draw situations’. He can do this by making use of a special motif:
domination.
The following example will make many things clear:
37
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+Qzpk+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Kg4
The white king is approaching quickly.
1...Kg1 2.Kg3!
This way of ‘nailing the enemy king to the edge’ is called domination.
2...f1=Q
2...f1=N+ is relatively better, but of course this also loses immediately to 3.Kh3.
3.Qh2#
The examples discussed above may be (partly) known to you. But I think there is no harm in repeating old
subject matter, to refresh your knowledge. Contrary to what I found in many endgame books, we can
formulate a simple but very adequate rule of thumb:

Rule of thumb
The queen side always wins, unless it is stalemate!
In the final example above, we came across a new concept – that of domination. With the help of this idea,
we can define the following technique:

Technique
Domination means that the pieces of the stronger side, as it were, ‘push the enemy pieces to the
edge of the board’, thereby creating mating patterns. It is important here that the opponent is
unable to save himself by giving a check.
With the above rule of thumb (‘The queen side always wins, unless it is stalemate’), the only thing we have
to remember is that the side with the pawn can sometimes make a draw if it has any of two rook’s pawns or
bishop’s pawns. The given rule avoids exceptions like the above Queen vs pawn fragment No. 36.
This leads us to a curious phenomenon: apparently, a win is possible if the king can move within a
certain range of the bishop’s or rook’s pawn. Fortunately, endgame theoreticians have investigated this for
us, and we can check this if we know where the limits are.

Critical area
Apparently, if the white king can reach the vicinity of the black pawn, there are quite a few cases where the
stronger side can win after all. The queen’s side allows the opponent to promote his pawn at the right
moment, and then checkmates him. This brings us to the so-called ‘critical area of the bishop’s pawn’.
In all the cases we will discuss from here on, we will see that the stronger side’s king achieves the
following in one or several moves:
• Either it succeeds in creating a mating pattern so that the opponent cannot promote his pawn;
• Or the king forces a ‘domination position’ after the promotion of the pawn.
For example, in the following case:
38
Domination
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+K+0
9+-+-+-+-0
9-+-wQ-+-+0
9+-+-+qmk-0
xiiiiiiiiy
White to move
With 1.Kg3!, White completes the domination over the black pieces. Both Black’s king and queen are
pushed to the edge of the board, reducing their activity to a minimum. Please note that White plays a ‘quiet
move’ here, and Black is unable to give check due to the passive placement of his queen. The mate threat on
h2 cannot be parried. This domination theme occurs quite often in endings.
It would go too far to have all the different positions of the king pass the review one by one, but I have
selected a few typical set-ups to illustrate the above-mentioned laws. I hope these examples will also make
clear why certain squares do not belong to the critical area, because the king is quite simply too late to
achieve either of its two aims.

A) with the bishop’s pawn, defending king at the corner side

39
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-wQ-+-+-0
9{}{+-+-+0
9+-+{mK-+-0
9-mkp}{+-+0
9+-+-}-+-0
xiiiiiiiiy
White to move
In this diagram, we see that White wins if his king is on one of the squares a4, b4, c4, d3, e3, e2 or e1. It is
simple with the king on e3:

1.Kd2
And the pawn will be lost:
1-0

40
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+K+-+-+0
9+-+-+-+-0
9-mkp+Q+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
With the king on c4, White can win as well. But he has to pay attention:

1...Ka1!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9{}K+-+-+0
9+-+{}-+-0
9-+p}Q+-+0
9mk-+-}-+-0
xiiiiiiiiy
Black gladly uses the stalemate trick, but:

2.Qd2! Kb2
2...Kb1 3.Kb3 c1=Q 4.Qa2#.

3.Kd3
And White wins the pawn, after which it is game over.

41
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
9-mkp+Q+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Let’s check first that with the white king on d4, there is no win. Indeed, after
1...Ka1!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9{}{mK-+-+0
9+-+{}-+-0
9-+p}Q+-+0
9mk-+-}-+-0
xiiiiiiiiy
Black holds the draw. Taking the pawn leads to stalemate, and there is no way for White to bring his own
king closer.
2.Qe1+
2.Qd2 doesn’t win in view of 2...Kb1.
2...Kb2
Forced; but not 2...Ka2?? in view of 3.Qc1!, and White wins.
3.Qd2 Kb1 ½-½
Also here, the white king cannot intervene.
By the way, not 3...Ka1??, as then 4.Qc1+ wins the pawn with check!

B) with the bishop’s pawn, defending king at the centre side

42
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-wQ-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+pmk-+-+0
9+-+-+-mK-0
xiiiiiiiiy
White to move
With his king inside the critical area, White can force the win. Here, we have a significantly larger amount of
squares than we saw with the black king at the other side of the pawn. This is logical, because here Black
cannot work with the stalemate trick. The win from the various squares is quite ingenious. With the king on
g1 (though in your mind you can also put it on g4), White’s first task is to put his queen on the second rank
– obviously, without giving Black the chance to promote his pawn in the meantime:
1.Qb4+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9}{}{+-+-0
9{wQ{}{}{+0
9}{}{}{}-0
9{}pmk-}{+0
9}-+-+{mK-0
xiiiiiiiiy
1...Kd1
1...Kc1?? would be a big blunder due to 2.Kf2 Kd1, since in that case White wins immediately with 3.Qe1#.
The same goes for 1...Kd3?? 2.Qe1!, and the queen reaches the promotion square of the pawn.
2.Qb3 Kd2 3.Qb2 Kd1
Here we see White’s idea.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9}{}{+-+-0
9{}{}{}{+0
9}{}{}{}-0
9{wQp+-}{+0
9}-+k+{mK-0
xiiiiiiiiy
The queen indirectly controls the e2-square, and thus White can create a mating pattern:
4.Kf2!
This is the trick. Black cannot promote to queen on pain of checkmate on the e2-square. This could also
have happened if the white king had to come from g4.
4...Kd2
4...c1=Q 5.Qe2#.
5.Qd4+ Kc1
Black may think he can escape to the corner with his king now, but White uses a clever little ploy to end all
resistance at once:
6.Qb4! Kd1 7.Qe1#

43
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9}{}K+-+-0
9{}{wQ{}{+0
9}{}{}{}-0
9{}pmk-}{+0
9}-+-+{}-0
xiiiiiiiiy
White to move
Now, we will check if the position with, for instance, the king on d5 is indeed winning for White. The
simplest path to the win is by digging out the well-known principle of forcing the defending king to move in
front of its own pawn.
This can be done as follows:
1...Ke2
The flight to the corner also doesn’t work for Black: 1...Kc1 2.Kc4 Kb1, and now White has a wide choice of
wins. I will give one here: 3.Kb3 c1=Q 4.Qd3+ Ka1 5.Qa6+ Kb1 6.Qa2#.

1...Ke2 2.Qc3 Kd1 3.Qd3+ Kc1

XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+K+-+-0
9-+-+-+-+0
9+-+Q+-+-0
9-+p+-+-+0
9+-mk-+-+-0
xiiiiiiiiy
Now, the white king has to be directed to b3 as quickly as possible. This also explains why White cannot
win if his king is further away (squares e5, f5, etc.).
4.Kc4 Kb2 5.Qd2 Kb1 6.Kb3 c1=Q 7.Qa2#

C) Critical area – rook’s pawn

If his king is inside the marked area, White can win. The following two examples with the king from a
certain square will make things clear.

44
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-wQ-+-zp0
9+-+-+-mk-0
xiiiiiiiiy
White to move
The rook’s pawn is often an unfavourable pawn for the defender. In this scheme, we see that the white king
is relatively far removed from the pawn, but White can still force the win. To achieve this, we dig out the
‘domination’ theme again:
1.Kf4!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-}{}{0
9-+-}{mK{}0
9+-+{}{}{0
9-+-wQ{+-zp0
9+-+{}-mk-0
xiiiiiiiiy
1...h1=Q 2.Kg3! 1-0
The black king and queen are dominated at the same time.
With, for instance, the white king on d3, there is another mating pattern that plays a role in the winning
process.
45
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+K+-+-0
9-+-wQ-+-zp0
9+-+-+-mk-0
xiiiiiiiiy
White to move
After
1.Ke3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-}{}{0
9-+-}{}{}0
9+-+{mK{}{0
9-+-wQ{+-zp0
9+-+{}-mk-0
xiiiiiiiiy
Black cannot promote to a queen with
1...h1=Q
in view of
2.Qf2#.

Exceptional wins
Sometimes the king is not in one of the areas indicated above. Still, there are two such cases in which White
can win.

Case 1: Tempo gain

In the following study by Lolli, the king is too far away, but there is a way to gain time.

46
Study by Giambattista Lolli
Osservazioni Teorico-Pratiche 1763
XIIIIIIIIY
9-+-+-+-+0
9+QmK-+-+-0
9-+-+-+-+0
9}{}{+-+-0
9-+-+{+-+0
9+-+-}-+-0
9p+-+{+-+0
9mk-+-}-+-0
xiiiiiiiiy
White to play and win
At first sight, there is nothing to be gained here for White; his king is too far away to play a significant role.
But there is something White can do about this!
1.Kb6!!
A marvellous move, lifting the stalemate and thus enabling the king to move closer with tempo gain.
 XIIIIIIIIY
9-+-+-+-+0
9+Q+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9p+-+-+-+0
9mk-+-+-+-0
xiiiiiiiiy
1...Kb2 2.Kc5+
And now the king has already entered the area we have come to know.
2.Ka5+ comes down to the same thing, e.g., 2...Kc2 3.Qg2+ Kb1 4.Kb4 a1=Q 5.Kb3.
2...Kc2 3.Qe4+ Kb2 4.Qe2+ Kb1
4...Kb3 is, again, met by the standard move 5.Qe5, after which the queen can occupy the promotion square.
5.Kb4 a1=Q 6.Kb3! 1-0
Domination!

Case 2: A second enemy pawn functions as ‘harmful material’

47
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-wQ-+0
9+-+-+-+-0
9-+-+-+kzp0
9+-+-+-+-0
xiiiiiiiiy
White to move
Black not only has a pawn on h2, but one on h7 as well. And it is precisely this second pawn that will
become fatal for him!
First, Black’s king has to be driven into the corner.
1.Qg4+
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+Q+0
9+-+-+-+-0
9-+-+-+kzp0
9+-+-+-+-0
xiiiiiiiiy
1...Kf2 2.Qh3 Kg1 3.Qg3+ Kh1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-wQ-0
9-+-+-+-zp0
9+-+-+-+k0
xiiiiiiiiy
The first phase has been completed: Black’s king is stalemated. But this is of no use to us unless we can
remove the stalemate. White does this with an, at first sight, bizarre move:
4.Kg7!
The utterly wrong direction, it seems! But with this move, White gives free passage to the h7-pawn, in the
meantime working on an instructive mating pattern!
Also possible is 4.Kh5 h6 5.Kh4 h5 6.Kh3 h4 7.Qe1# (7.Qg2#).
4...h5 5.Qf2
A little trick worth remembering.
5...h4 6.Qf1#

Technique
If the stalemate can be removed, sometimes a win is possible.

Quiz
It is always useful to test whether you can put the knowledge you have just acquired into practice.
Have a look at the following endgame study, and answer the three quiz questions.

48
Study by Nikolay Grigoriev
Source unknown, 1932
 XIIIIIIIIY
9Q+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+k+-+-0
9-+p+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This might be the most complex problem in existence in this type of queen vs pawn ending. Any player who
is able to find the following win during a game can boast that he passed the exam with distinction! In other
words, he could claim that he understands the finesses of this endgame fully!

Question 1: With which move should White start? Choose between:

A) 1.Qa1;
B) 1.Qc6;
C) 1.Qd5+;
D) 1.Qe4+.
1.Qd5+!
This is the only move that wins.
A) If 1.Qa1 Kd2 2.Qa2, 2...Kc3! is the only defence: the white king has to be kept back. White can make no
progress;

B) 1.Qc6 serves to stop the pawn, but after 1...Kd2, 2.Qd5+ Ke1 won’t make any progress either. By the
way, not 2...Ke2?? in view of 3.Qa2, a variation we will also see elsewhere.

Also attractive is D) 1.Qe4+, but then Black draws with 1...Kd2! (not 1...Kc3? in view of 2.Qd4+ Kb3 3.Qa1
– this is a theme, where the queen gets in front of the pawn, which we will see more often).

1.Qd5+! Ke3
1...Kc3? is strongly met by 2.Qd4+! Kb3 3.Qa1!, and 1...Ke2 is met by 2.Qa2! Kd1 (after 2...Kd3 3.Qb2 Kd2
4.Ke4, Black is also pushed to the edge: 4...Kd1 5.Kd3 c1=Q 6.Qe2#) 3.Kd4 c1=Q 4.Kd3, and White wins
by domination.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+QmK-+-0
9-+-+-+-+0
9+-+-mk-+-0
9-+p+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black plays the critical move, and now White has to come up with a brilliant continuation to win.
Question 2: What are the things White has to reckon with? Translate this into the correct continuation.
2.Qg2!!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-mk-+-0
9-+p}{+Q+0
9+-+-+-+-0
xiiiiiiiiy
A fantastic move with multiple points. Firstly, it prevents the black king from moving to d2, and at the same
time it enables the white queen to manoeuvre in front of the pawn, and the black king cannot prevent this.
2...Kd3
2...c1=Q is not possible in view of 3.Qg5+, winning the newborn queen.
3.Qg5!
And again, the queen gets in front of the pawn, after which the position is winning.
3...Ke2 4.Qc1 and White wins.

This ending in practice

For some time, thanks to sponsoring from the Max Euwe Centre in Amsterdam (initiated by Hans Böhm), I
was the trainer of the young Bianca Muhren, who was a strong player from her early youth, and later ended
up in the Dutch Olympiad Women’s team. She remained active in the chess world, was one of the chess
players who laid the foundation for the ‘ChessQueens’ foundation, and became the chairwoman of the Royal
Dutch Chess Federation in late 2020.
Once, Bianca played a game against Tjapko Struik, who was a difficult opponent for her. The following
fragment from that game turns out to be highly instructive – as was the phase before the queen vs pawn
ending.
I will give only the part with that ending; the preceding phase will be treated in Chapter 12.
What was it again – the area in which the white king has to stand to win?

49
Tjapko Struik
Bianca Muhren
Netherlands tt 2002/03 (1)
XIIIIIIIIY
9Q+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+kzp-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Let’s refresh our memory, and project the earlier markings on the current position.
After a tough fight, the notorious ending of queen versus pawn on the penultimate rank has arisen. Having
in mind the rule of thumb that the queen’s side always wins unless the defender can make use of a
stalemate trick, Black has just played her pawn from f3 to f2. Of course, Bianca knew the classics, but she
will experience a few problems here as there are two unfavourable factors for Black: her king is on the
wrong side of the pawn, and the white king is relatively close.
60.Qe4+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-}{}{0
9-}{}Q}{}0
9+{}{}{}{0
9-}{+kzp{}0
9+{}-+-+{0
xiiiiiiiiy
What was the normal mechanism for White? The queen drives the black king in front of the pawn, after
which the white king can hurry closer. 60.Qg2 was an important alternative which, however, does not get
White anywhere in this case. It would win if his king were on b4.

(60.Qg2) Here, we see that such an apparently minuscule detail can make all the difference! Because in that
case, White would suddenly have a win thanks to a deeply hidden mate trick. After 60...Ke1 follows 61.Kc3!!
f1=Q 62.Qd2 mate! Obviously, 61...f1=Q is not forced, but also after 61...Ke2 White hauls in the point
because his king is now close enough: 62.Kd4 Ke1 63.Ke3 f1=N+ 64.Kd3.

After (60.Qg2) 60...Ke1 61.Qg3 Ke2 62.Qh2, 62...Kf3! is absolutely the only move, but it immediately
forces the draw (certainly not 62...Ke1?? in view of 63.Kd4 f1=Q 64.Ke3, and the domination theme does
the job again!).
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+k+-0
9-+-+-zp-wQ0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
And now it turns out that the white king is just outside the winning zone. So it is a draw, but it is a narrow
path in a practical game.

60.Qe4+ Kf1
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+Q+-+0
9+-+-+-+-0
9-+-+-zp-+0
9+-+-+k+-0
xiiiiiiiiy
What did theory say again about this position? If we assume that the black king will be on g1 on the next
move, then what is the area in which the white king has to stand to win the game? This turns out to be the
d1-d4-h4-h1 square, from which (unfortunately for Tjapko, and fortunately for Bianca) the e4-square has
been ‘cut off’. So, it is still a draw if the white king is on d4 or e4, but if the king is, for instance, on f4, White
can just win since his king can reach the g3-square.
61.Qh1+
Whatever it takes, White tries to chase the black king back to the centre.
After 61.Qe3 Kg2 62.Qd2 Kg1, now White would be able to force the win with the king from f4 (inside the
aforementioned square).
61.Qf4 Kg2 62.Qg4+ Kh2 63.Qf3 Kg1 64.Qg3+ Kh1.
61...Ke2 62.Qh2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+kzp-wQ0
9+-+-+-+-0
xiiiiiiiiy
This is the right idea, since now 62...Ke1 is met by 63.Kd4!.
62...Kf3!
Bianca is alert! She plays the only move that draws, as after 62...Ke1? 63.Kd4 f1=Q 64.Ke3, we again see
the domination motif, leading to the black king being checkmated.
A move like 62...Ke3? would also have lost, for example: 63.Qg2! Ke2 64.Kd4 Ke1 65.Ke3 f1=Q 66.Qd2#.
63.Qxf2+
White has seen enough – there is nothing to be gained, e.g., 63.Qh1+ Ke2.
63...Kxf2 ½-½
A heroic draw!
Chapter 11
Rook versus pawn
Even though a rook is much stronger than a pawn, it turns out that there are still many rook vs pawn
positions that are drawn, if the stronger side’s king cannot join the struggle in time. A race ensues that is
sometimes decided by a single tempo. The result of the game is hanging by a thread, and the stronger side
has to play accurately. A study of the ‘border cases’ will provide insight into the precepts of this endgame
type.
We will start with two rules of thumb, where we formulate a few principles that may come in handy:

Rules of thumb
• The stronger side cannot win without help of the king.
• In principle, the best position for the rook is behind the passed pawn.
In present-day football, but especially in rugby and American Football, shoulder budges are frequent. In
chess, there are various endgames in which a fight arises between the two kings, who try to make their way
through the woods of their own and enemy pawns. Inevitably, it can happen that they are very much in
each other’s way at some point. Keeping off or ‘outflanking’ the enemy king with your own king is also
called ‘shouldering’.

Technique
The weaker side’s king should try to keep the enemy king at bay by ‘shouldering’.

50
XIIIIIIIIY
9-+-+-+R+0
9+-+-+-+-0
9K+-+-+-+0
9+-+-+-+-0
9-+-+-+pmk0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Kb5!
The king hurries closer.
1...g3
Black can make an attempt to ‘shoulder off’ the enemy king with 1...Kg3 2.Kc4 (2.Rf8 Kh2 3.Kc4 g3 4.Kd3
g2) 2...Kf3 3.Kd3 g3. It seems Black’s mission has been accomplished. However, now that the kings are
facing each other, Black’s king gets driven to the fatal side anyway: 4.Rf8+ Kg2 5.Ke2 Kh2 6.Rg8 (the rook
takes up its ideal position – behind the pawn) 6...Kg2 (after 6...g2 7.Kf2, it is immediately over)
 XIIIIIIIIY
9-+-+-+R+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-zp-0
9-+-+K+k+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
7.Rg7!. A tempo move, after which Black has to concede ground again: 7...Kh2 8.Kf3, and White wins.
2.Kc4 Kh3 3.Kd3 g2 4.Ke2 Kh2 5.Kf2
Just in time, White takes sufficient control of the promotion square.
5...Kh1 6.Rh8#
Certainly not 6.Rxg2??, with stalemate.

51
XIIIIIIIIY
9-+-+-+R+0
9+-+-+-+-0
9K+-+-+-+0
9+-+-+-+-0
9-+-+-mkp+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
With the black king on the other side, the half-point is secured. White’s king does not arrive in time to
‘shoulder off’ its opposite number:

1.Kb5 g3 2.Kc4 Kf3 3.Kd3 g2

Also 3...Kf2 is sufficient for the draw, but then Black has to continue shouldering: 4.Rf8+ Ke1! (not 4...Kg1
5.Ke2 g2 6.Rg8 Kh2 7.Kf2 Kh1 8.Rh8# – and certainly not 8.Rxg2?? stalemate!) 5.Ke3 g2 6.Rg8 Kf1, with a
draw.
 XIIIIIIIIY
9-+-+-+R+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+K+k+-0
9-+-+-+p+0
9+-+-+-+-0
xiiiiiiiiy
Everything hangs on one tempo here!

4.Rf8+ Kg3 5.Rg8+ Kf2

With a draw. The pawn cannot be kept from promoting.

The shoulder-budge

The following examples show how important it can be to shoulder off the enemy king.

52
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zp-+K+-0
9-+-mk-+-tR0
9+-+-+-+-0
xiiiiiiiiy
Black to move
1...Kd3!
Black saves the half point by keeping off (shouldering!) the enemy king.
1...Kd1? loses to 2.Ke3, and now 2...c2?? fails to 3.Rh1#, but after 2...Kc1 3.Kd3, White also wins as the
pawn falls.

1...Kc1? loses as well since after 2.Ke3 c2 3.Kd3, the pawn will be lost.
1...Kd3! 2.Rh8 c2 3.Rd8+
Thus, White drives off the black king, but the pawn has advanced far enough to force the draw.
3...Kc3
And there is nothing more to be gained for White:
½-½

53
XIIIIIIIIY
9-+-tR-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-zpk+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
We have already established that the white rook is ideally placed behind the pawn. However the white king
is still at a great distance here. Yet, White wins by ‘crossing over’ to the other side of the pawn with the king,
winning by one tempo:
1.Kd6!
Not 1.Ke6? as then the black king can keep its opposite number at bay for one more move by staying put
and moving the pawn one square closer to the back rank: 1...d3 2.Kd6 and now White is just one tempo
short: 2...Ke3 3.Kc5 d2 4.Kc4 Ke2 5.Kc3 d1=Q with a draw.

1.Kd6! d3 2.Kc5
XIIIIIIIIY
9-+-tR-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+k+-+0
9+-+p+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The king arrives on the other side. Since Black’s plan of marching with the pawn with support by the king is
quite slow, White gets there just in time:
2...Ke3 3.Kc4 d2 4.Kc3 1-0

54
 XIIIIIIIIY
9-tR-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-mkp+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Black’s pawn has advanced quite far, and the black king is keeping the enemy king at bay, but still Black
draws the short end of the stick here. Since the two kings are standing opposite each other, White can drive
the enemy monarch from its dominant position with checks. We have seen this procedure before:
1.Re8+ Kf2
Moving behind the pawn doesn’t make much sense now: White wins simply with 1...Kf4 2.Kd2 Kg3 3.Ke1
Kg2 4.Rg8+ Kh2 5.Kf2, and the pawn falls.
2.Kd2 Kg1
If Black tries to hold off the white king with 2...Kf1, he will still be too late after 3.Rf8 Kf2 4.Rf7 Kg2 5.Ke3.
XIIIIIIIIY
9-+-+R+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+p+-0
9-+-mK-+-+0
9+-+-+-mk-0
xiiiiiiiiy
3.Rf8 f2
The picture we saw in the previous variation arises after 3...Kf2 4.Rf7 Kg2 5.Ke3, and White wins.
4.Ke2
And White swallows up the pawn.

55
XIIIIIIIIY
9R+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-mkp+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Rf8!
Rook behind the pawn! White cannot force the win by moving the rook to the edge of the board: 1.Ra5+
Ke4 2.Ke6?!. Following the black king with his own king isn’t going to work. At the end, White gets fooled:
2...f4 3.Ra4+ Ke3 4.Ke5 (now, it would be too late to ‘cross over’ to the other side: 4.Kf5 f3 5.Kg4 f2 6.Rf4
Ke2, and Black is right on time) 4...f3 5.Ra3+ Ke2 6.Ke4 f2 7.Ra2+ Ke1 8.Ke3
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9R+-+-zp-+0
9+-+-mk-+-0
xiiiiiiiiy
analysis diagram
8...f1=N+. Underpromotion to a knight! As a rule, the rook cannot beat the knight, even if the defending king
is nailed to the edge of the board. The weaker side only loses if the knight is in the corner. White’s attempts
lead to nothing; for instance, 9.Kd3 Ng3 10.Rb2 Nf1 11.Re2+ Kd1 12.Ra2 Ke1, with a draw. We should
note that the knight can be caught if it moves away from its own king, but White cannot force this in this
position.
What White can do is implement the usual strategy, walking to the other side of the pawn rather than the
black king, e.g. 2.Kf6 f4 3.Kg5 f3 4.Kg4 f2 5.Rf5.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
9-+-+k+K+0
9+-+-+-+-0
9-+-+-zp-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Finally, White has placed the rook behind the pawn, and now the win is certain: 5...Ke3 6.Kg3.

1.Rf8! f4 2.Rf7!
A tempo move, forcing the black king to give ground.
In this instance, White gets nowhere with king moves like 2.Kf7 as Black will shoulder off White’s king with
2...Kf5! 3.Kg7+
 XIIIIIIIIY
9-+-+-tR-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+k+-0
9-+-+-zp-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
3...Kg5! (definitely not 3...Kg4? in view of 4.Kf6 f3 5.Ke5 Kg3 6.Ke4 f2 7.Ke3) 4.Rf7!. Now White can pick
up the thread of the main line.
2...Ke4 3.Kf6
And this gives White the opportunity to move to the g-file.
3...f3 4.Kg5 Ke3 5.Kg4 f2 6.Kg3
We have seen this position before: White wins.
As for that underpromotion – when I was fifteen years old or so, I refused to believe that positions with
rook versus knight could be a draw. One of my games in the competition of my club at the time was
adjourned just before I would get such an ending on the board.
At our club, an adjourned game had to be resumed at the home of one of the players. So, on one evening in
that same week I set forth to the home of my opponent, a friendly gentleman who was patiently waiting for
me to see that my winning attempts wouldn’t yield anything.
Thinking back on this, I feel a little ashamed about keeping him confined to the board in a theoretically
drawn position. What did I know...?

56
XIIIIIIIIY
9-+-+-+K+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-zp-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
White to move
Sometimes, the rook doesn’t manage to get behind the pawn. Standing in front of the pawn is unfavourable
in principle, because the rook is more vulnerable there, sometimes enabling the opponent to gain a crucial
tempo to achieve the draw.
1.Rd1+!
If we try to bring the king closer with 1.Kf7, we see that White is just too late: 1...e4 2.Kf6 Kd3 3.Kf5 and
now:
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-+p+-+0
9+-+k+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
analysis diagram
3...Kd2! (this is the important tempo Black has to win to hold the draw by one tempo; or 3...e3 4.Kf4 e2
5.Kf3 Kd2=) 4.Ra1 e3 5.Kf4 e2, and White’s king is too late to stop the pawn.

1.Rd1+! Kc3
This king move is the most important try in most cases. The black king moves in the direction of the rook,
trying at the same time to get to the promotion square as quickly as possible.
The alternative 1...Ke3 is not to be recommended. Apart from the fact that the king is now standing in front
of its own pawn, it is of not much use to him after White’s reply 2.Re1+, when after 2...Kf4, the king reaches
the other side of the pawn, but in turn the white king now hurries to the other side of the e-file: 3.Kf7 e4
4.Ke6 Kf3 5.Ke5 e3 6.Kd4 e2 7.Kd3, winning.
2.Re1!
This is the true intention behind the check: White improves the position of his rook. He is attacking the
pawn, and so in fact he has been allowed to make an extra move (moving the rook from c1 to e1), and after
Black protects the pawn, it’s his move again. This makes a crucial difference, as we will soon see.
2...Kd4
XIIIIIIIIY
9-+-+-+K+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-zp-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
White now takes aim at the pawn, and this is just enough to win the game.
3.Kf7 e4 4.Ke6 e3
Here is the difference: 4...Kd3 5.Kf5!. White attacks the pawn twice, so 5...e3 is forced, and after 6.Kf4, the
pawn is hanging again, as a result of which Black has no tempo gain on the rook with ...Kd2: 6...e2 7.Kf3,
and the pawn falls.
5.Kf5 Kd3 6.Kf4 e2 7.Kf3
And White wins.
We can derive the following rule of thumb from this winning method:

Rule of thumb
By giving check, the stronger side can improve the position of the rook with gain of tempo, to
attack the pawn.
Here is a variation on the same theme.

57
XIIIIIIIIY
9-mK-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+p+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
White to move
At first sight, Black seems to have good drawing chances here. White’s rook isn’t ideally posted, and neither
is his king.
Still, there is a way to conquer the pawn. White has to go about things systematically:
1.Rc1
First, the rook attacks the pawn, so that Black has to lay his cards on the table:
1...Kc5
Black can also try 1...Kd5. Now White has to take care. If he reacts in the wrong way, he will lose half a
point:
2.Kb7!. White starts a little tempo play, in the meantime bringing his king closer to the promotion rank (if
2.Kc7?, 2...Kc5!; the black king shoulders off its opposite number, and so we have a reciprocal zugzwang
situation – draw!, for example: 3.Kb7 (also after 3.Kd7, 3...Kd5! draws) 3...Kb5): 2...Kc5 (2...Kd4 3.Kc6 c3
4.Kb5 Kd3 5.Kb4 c2 6.Kb3)
XIIIIIIIIY
9-+-+-+-+0
9+K+-+-+-0
9-+-+-+-+0
9+-mk-+-+-0
9-+p+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
analysis diagram
3.Kc7!. Now, it will be clear that Black is in zugzwang, as in the main line below.
 XIIIIIIIIY
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mk-+-+-0
9-+p+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
2.Kc7!
This basic position is very important to memorize. This is a reciprocal zugzwang position. If it’s Black’s
move, he has to concede ground and loses. With White to move, Black can hold the ‘opposition’: he can
meet Kb7 with ...Kb5! and Kd7, of course, with ...Kd5!, with a draw. So in this position, White runs off with
the full point as follows:
2...Kd4
2...Kb4 3.Kc6 c3 4.Kd5 Kb3 5.Kd4 c2 6.Kd3 is identical to the main line, but mirrored on the c-file.
3.Kc6 c3 4.Kb5
Now, White’s king finds itself on the other side of the pawn, giving him just enough time to win:
4...Kd3 5.Kb4 c2 6.Kb3 1-0

58
XIIIIIIIIY
9K+-+-+-+0
9+-+-+-+-0
9-+-+-mk-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
White to move
With White’s king far away and his rook not optimally placed, the draw seems to be within reach for Black.
But there is one factor that does bring White the win:
1.Rc5!
 XIIIIIIIIY
9K+-+-+-+0
9+-+-+-+-0
9-+-+-mk-+0
9+-tR-+-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This is called ‘a horizontal cut-off on the fifth rank’. Black’s king cannot support its pawn, and so White can
leisurely move his king to within the reach of the pawn.

Not 1.Rf1+? Ke5 2.Rg1 Kf4, and Black holds the draw easily.

1.Rc5! Kg6
After 1...g3 2.Rc3 g2 3.Rg3, the pawn is lost. Here, it turns out that the division line is on the fifth rank, as it
is from this rank that White’s rook can catch the pawn just in time.
2.Kb7 Kf6 3.Kb6 Kg6 4.Kb5 Kf6
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mk-+0
9+KtR-+-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White’s king is moving closer, but it has to be careful not to get under the feet of the rook:
5.Rd5!
5.Kc4 would be inaccurate in view of 5...g3, when the rook cannot go after the pawn. Incidentally, White
wins anyway with 6.Kd4 g2 7.Rc1 Kf5 8.Ke3, since also here the black king is unable to help out.
5...Kg6 6.Kc4 Kf6 7.Kd3 1-0
The rest is simple.
Now we can formulate the following rule of thumb:

Rule of thumb
If the rook can cut off the enemy king from its pawn on the fifth rank, the rook’s side wins
simply.

Quiz
59
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-tR-0
9-+k+-+-+0
9+p+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
It seems improbable that White can win from this position, but...

Question 1: Which is the best move for White to start with?

1.Rg5!
This forces Black’s king to move in front of its pawn via the ‘wrong side’, and because of this, White will be
just in time.
1...Kb6
1...b4 would produce the winning position with the ‘fifth-rank cut’ we have seen above.
2.Kg7 Ka5

Question 2: What to do now?

3.Kf6 Kb4
The king is ‘creeping along the front of the pawn’ to keep off the enemy king.
3...Ka4 would offer no chance at all in view of 4.Ke5 b4 5.Kd4 b3 6.Kc3.
4.Ke5 Kc4
Black has managed to ‘shoulder’ the white king, but it won’t help him much. 4...Kc3 isn’t any better on
account of 5.Kd5 b4 6.Rg3+ Kc2 7.Kc4, and the pawn will be lost.
5.Rg4+ Kc3

Question 3: What is the most clever move for White now?

6.Kd5!
Thus, White conquers the pawn.
6...b4 7.Rc4+
Now, 7.Rg3+ amounts to the same: 7...Kc2 8.Kc4, and the pawn falls.
7...Kb3
Question 4: What was the intention behind White’s previous move?
8.Kc5 1-0

This ending in practice

Not so rarely, knowledge can be ‘in the way’ of a player. This is because certain subtleties can sometimes
make a world of difference. One important factor is that endgames only appear on the board in the final
hour of the game. If the previous part of the game has been tough going, fatigue may strike. And this can
cause a player who is normally able to find his way in such a situation, to make a mistake, perhaps also
under pressure from the clock.
Here are two special cases in which one of the players used his ‘knowledge’ in the wrong way:

60
David van Kerkhof 2287
Frank Erwich 2359
Netherlands tt 2012/13 (3)
XIIIIIIIIY
9-+-+-+-+0
9mK-+-+-+-0
9-+-+-tR-+0
9+-+-+p+-0
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
This position arose after a long game. Frank Erwich, who is known to have a good knowledge of endgames,
now has to choose between two moves: 49...f4 or 49...Kf4. What would you play?
49...Kf4?
The wrong choice! This is one of those examples in which we see that sometimes knowledge can be ‘in a
player’s way’. Erwich understood that his king had to be on the other side of the pawn, to keep out the
enemy king, so he decided to move it there at once. If he had calculated accurately, there is no doubt that he
would have seen that this manoeuvre costs a crucial tempo, leading to the loss of the game.
With immediately 49...f4! 50.Kb6 and now moving the king to the other side with 50...Kf3!, Black could
have forced the draw, e.g., 51.Kc5 Ke3!=. It essential that after, for example, 52.Re6+ (52.Kc4 f3 53.Kc3 f2
(53...Ke2) 54.Re6+ Kf3)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+R+-+0
9+-mK-+-+-0
9-+-+-zp-+0
9+-+-mk-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
52...Kd3!, Black can keep the white king at bay: 53.Rf6 Ke3 54.Kc4 f3 55.Kc3. The thing is that in this
position it should be Black to move and not White (as in the game), when 55...f2 draws.

49...Kf4? 50.Kb6 Ke4 51.Kc5 f4 52.Kc4 Ke3 53.Kc3 f3

 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-tR-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-mkp+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
We have reached the same position as in the above comment, only with the difference that it is White’s
move! And this means that the black king can be driven to the other side. The game is over, and Erwich
resigned immediately after:
54.Re6+
There could have followed: 54...Kf2 55.Kd2 Kg2 56.Rg6+ Kf1 57.Rf6 Kf2 58.Rf8 Kg2 59.Ke3, and the
pawn is lost.

61
Herman Grooten
Dirk Floor
Netherlands tt 2001/02 (1)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-mk-+-+KzP0
9+-zp-+-tR-0
9-+-+-+-+0
9+-+-+-+r0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
After a long, tough battle, in which Black had defended excellently, Floor had the draw for the taking. This
was the final game in a club match in the second KNSB league between De Stukkenjagers (Tilburg) and Paul
Keres (Utrecht). The match score was 4½-4½ – as tense as it gets! In this position, Black has to take a
momentous decision. Should he continue 71...Kb5 or 71...c4 ?
71...Kb5?
After the game, we found out that only 71...c4! would have held the draw. Again, we see how knowledge can
be in the way of a player. Didn’t we just see several examples of the ‘fifth-rank cut’? After the pawn move,
the black king and pawn ‘lose contact’, and this leads to a loss in some cases. This is what Black must have
estimated to be the case, and that is why he crossed the fifth rank with his king before advancing the pawn.
However, in this position, the ‘cut’ turns out to be irrelevant. After 71...c4, Black draws: 72.h7 (building a
bridge with 72.Rh5 doesn’t win, as the pawns reach the other side on the same move: 72...Rxh5 73.Kxh5 c3
74.h7 c2 75.h8=Q c1=Q, with a draw) 72...c3 73.Kg7 (stopping the pawn with 73.Rg2 brings no benefit
either: 73...Kc5 74.Kg7 Kd4 75.h8=Q Rxh8 76.Kxh8 Kd3, and it’s a draw; 73.Rh5 Rxh5 74.Kxh5 c2
75.h8=Q c1=Q 76.Qb8+ Ka6) 73...c2 74.Rg1 Kb5 75.Rc1.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mKP0
9-+-+-+-+0
9+k+-+-+-0
9-+-+-+-+0
9+-+-+-+r0
9-+p+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
analysis diagram
Thus, White can eliminate the black pawn. However, Black draws with 75...Rg3+ 76.Kf7 (White also gains
nothing by 76.Kh8 Rg2) 76...Rh3 77.Kg7 Rg3+ etc.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+KzP0
9+kzp-+-tR-0
9-+-+-+-+0
9+-+-+-+r0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
71...Kb5? 72.Rh5!
Of course, now White does make use of the ‘bridge’ on h5.
Not 72.h7 Rxh7 73.Kxh7 Kc4, and Black draws; nor the nonsensical 72.Rg4 in view of 72...c4.
72...Rd3
After 72...Rg3+, we reach an important moment. How can White force a technical win?
With 73.Kf7!, the king prevents the black rook reaching the back rank, e.g. 73...Rf3+ 74.Ke7 Re3+ 75.Kd7
Rd3+ 76.Kc7 Rf3 (the only way to stop the pawn) 77.h7 Rf8 (77...Rf7+ 78.Kd6 Rxh7 79.Rxh7 c4 80.Rc7
Kb4 81.Kd5 c3 82.Kd4) 78.h8=Q Rxh8 79.Rxh8 c4.
XIIIIIIIIY
9-+-+-+-tR0
9+-mK-+-+-0
9-+-+-+-+0
9+k+-+-+-0
9-+p+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
And White wins by one tempo: 80.Kd6! (preventing 80...Kc5) 80...c3 81.Rc8 Kb4 82.Kd5 Kb3 83.Kd4 c2
84.Kd3, and the pawn is lost.
After the exchange, White comes first: 72...Rxh5 73.Kxh5 c4 74.h7 c3 75.h8=Q, and wins.
73.h7 Rd8 74.h8=Q Rxh8 75.Rxh8
XIIIIIIIIY
9-+-+-+-tR0
9+-+-+-+-0
9-+-+-+K+0
9+kzp-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This position turns out to be winning in more than one way.
75...Kc4
Black opts for playing the king to the other side of the pawn, to keep the enemy king at bay. As it turns out,
this doesn’t work, but there was no defence anyway.
After 75...Kb4, hauling in the full point is a piece of cake for White: 76.Kf5 c4 77.Ke4 Kb3 78.Rc8, and the
pawn will be lost.
76.Rd8!
As a trainer, obviously you should opt for the thematic way...
I hadn’t calculated that 76.Kf5 ‘unfortunately’ wins as well: 76...Kd3 77.Ke5 c4 78.Rh3+ Kd2 79.Kd4, and
also here the pawn goes.
76...Kc3 77.Kf5 c4 78.Ke4 Kb2 79.Kd4 c3 80.Rb8+
And Black surrendered, which meant we won the match!
In the post-game analysis, we signalled the interesting theme of ‘alternating attacks’.

Rule of thumb
With the rook behind the pawn, sometimes the win can be achieved by alternating checks with
attacks on the enemy pawn.
Chapter 12
Bishop versus pawn(s)
In this chapter, we will pay attention to the struggle of a bishop against several pawns. The bishop is a
strong piece, but it has one big drawback: it can only play on squares of one colour. Often, it has no trouble
stopping an enemy pawn since it can exercise control from a great distance.
Have a look at the first diagram position.

62
XIIIIIIIIY
9-+K+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-vl0
9+-+-+-+-0
xiiiiiiiiy
Not so interesting – this is a draw. Black’s bishop on h2 easily stops the pawn, for as soon as White
promotes to queen with 1.b8=Q, Black gives his bishop for the queen.
The struggle of bishop versus several pawns is more interesting. In that case, it needs the help of its own
king to stop the enemy passed pawns.
Let’s put the following diagram under the microscope. Fen:

63
XIIIIIIIIY
9-+K+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-vl-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+-mk-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
In this position, Black’s bishop stops the b-pawn, while the king has to stop the h-pawn. As we learn in
pawn endings, the king has to enter the square of the pawn to be able to stop it. Here, the square of White’s
h-pawn is defined as the area marked by the squares h3-h8-c8-c3. With Black to move, he is well able to
pull off this division of tasks:
1...Ke3 2.h4 Kf4 3.h5 Kg5 ½-½
In the following diagram, the bishop is less fortunately placed, but this is not a problem since the king can
approach the enemy pawn by moving along the light squares:
64
XIIIIIIIIY
9-+K+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+k+-+P0
9-+-+-+-vl0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.h4 Ke4 2.h5
Black’s king just walks through:
2...Kf5 3.h6 Kg6
The division of tasks is clear: the bishop keeps the b-pawn under control, while the king has to stop the h-
pawn: ½-½
Just a slight difference in the situation may change a draw into a loss. The previous diagram and the
following one show how hard it is to distinguish between such situations. The difference is so small that
there doesn’t seem to be any problem in the position below. But still, it turns out to be lost:

65
XIIIIIIIIY
9-+K+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mk-+-+P0
9-+-+-+-vl0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.h4 Kd4 2.h5
 XIIIIIIIIY
9-+K+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-+-vl0
9+-+-+-+-0
xiiiiiiiiy
Now, Black’s king cannot enter the square via e5, since then it interrupts the range of its bishop. The bishop
diagonal has to remain open, as otherwise White queens with 3.b8. This motif is called ‘interference’ in the
world of endgame studies.
2...Bf4
2...Ke4 doesn’t help either.
3.h6
XIIIIIIIIY
9-+K+{+-}0
9+P+-+-+-0
9-+-+-+-zP0
9+-+-]-+{0
9-+-mk-vl-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The black king is too late. White wins.
This interruption of the bishop’s range by causing the king and bishop to stand in each other’s way is a
rewarding subject. It’s time to look at some tactical shots that are made possible by this motif. Have a look
at the next diagram position.

66
End of a study by Henry Otten
New York Sunday World 1891
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-mk-vl0
9+-+K+-+-0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
At first sight, White doesn’t seem to have a win here. Black’s bishop can stop the white b-pawn on the b8-h2
diagonal, and the king will keep White’s g-pawn in check.
1.b7 is met by 1...Bf4, and Black stops the pawn.

If 1.Ke4, to prevent 1...Bf4, Black replies 1...Bf8 and then after 2.b7, he gets to the important diagonal with
2...Bd6.
However, White has a nice combination here:

1.g5+!
With this pawn sacrifice, White disrupts the cooperation between Black’s king and bishop.
1...Bxg5
1...Kxg5 is met by the immediate 2.b7, and the pawn gets through.
2.Ke4!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-mk-+0
9+-+-+-vl-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Since Black’s king is in the bishop’s way on f6, the bishop is unable to reach the b8-h2 diagonal in time:
2...Bh4 3.Kf3 1-0

Quiz
Now for a more complicated case.

67
Study by Alexey Selezniev
Deutsches Wochenschach 1917
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+L+-+0
9+-+-mK-+-0
9-zp-+-+-zp0
9+-+-mk-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White has to combat two dangerous passed pawns. An obstacle is formed by the black king. The question is,
which pawn will he stop with the bishop, and which one with the king?
Question 1: How does White make a draw? The idea that the bishop has to stop the b-pawn and the king,
the h-pawn, is correct. But the solution is highly paradoxical:
1.Kd6!!
As it turns out, this is absolutely the only move. 1.Kf6 looks logical, to walk towards the h-pawn with the
king, but the path is cut off by Black’s king with 1...Kf4!. After 2.Kg6 (2.Ke7 Kg3 3.Kd6 h3 4.Kc5 h2 5.Bd5
b3) 2...Kg3 3.Kf5, White has to change tack. The h-pawn is unreachable for the king, and now it turns out
that after 3...h3 4.Ke4 h2, the king is also in the way of its own bishop, leading to an inevitable loss. Here,
we have another example of interference: the king has to step inside the square of the b-pawn, but by doing
this, it prevents the bishop from stopping the h-pawn.

1.Kd6!! Ke4!
The composer assigns an exclamation mark to this move, because it is a clever trick to thwart White’s plans.
1...Kd4 draws as well, but it’s not as evident as it seems. The bishop has to remain on e6, and the king has
to stay in the vicinity of the d5-square, as we will see below. So, White needs another brilliant move here to
make a draw.

Question 2: Which move would that be?

2.Kc6!!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+K+L+-+0
9+-+-+-+-0
9-zp-mk-+-zp0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Again, it’s a matter of dividing tasks. You would expect the king to go after the b-pawn, while the bishop
takes care of the h-pawn. However, the opposite is true. White makes a kind of waiting move, staying in the
vicinity of the d5-square, so as to be able to quickly reach the square of the h-pawn later via d5, e4 and f3.
The bishop will just be able to stop the pawn on b2 via a2: 2...Kc3 3.Kd5 b3 4.Ke4 b2 5.Ba2, with a draw.
2.Kd7!
A look at the Tablebases shows that there are no less than five moves that lead to a draw here.
2.Kc5? loses to 2...Ke5!, expelling the bishop from its ideal position: 3.Bd5 h3 4.Bb7 b3. This is also the
reason why the other king moves do not lead to the desired result: 2.Kc6= 2...Ke5 3.Kd7 Kd4 4.Kd6 Ke3
5.Kc5; 2.Kc7=; 2.Ke7=; 2.Bd5+=.
 XIIIIIIIIY
9-+-+-+-+0
9+-+K+-+-0
9-+-+L+-+0
9+-+-+-+-0
9-zp-+k+-zp0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
At first sight, the text move is bizarre, but it is perfectly well possible to explain the motivation behind it.
2...Kd4
2...Ke5 3.Ke7.
3.Kc6 Kc3
The crux is that 3.Kc6 Ke5 can be answered by 4.Kd7!, so that the bishop can remain in place! Once you
get this trick, the whole thing is not so difficult...
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+K+L+-+0
9+-+-+-+-0
9-zp-+-+-zp0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, White has to find the only right idea – which we have seen already.

Question 3: Which idea is that?

4.Kd5!
The king crosses over!
4...b3 5.Ke4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+L+-+0
9+-+-+-+-0
9-+-+K+-zp0
9+pmk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
... and arrives in the square of the h-pawn just in time.
5...b2 6.Ba2
The bishop takes care of the b-pawn. The division of tasks between the king and the bishop has worked!
6...h3 7.Kf3 with a draw.
This endgame in practice
We have seen the final phase of the following game in Chapter 10. But what went before that was at least as
interesting.

68
Tjapko Struik
Bianca Muhren
Netherlands tt 2002/03 (1)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9p+-+-+-+0
9+-mK-+p+-0
9P+-zpk+-+0
9+-+-+-+-0
9-+-+L+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
In this position, White has a bishop for two pawns, but Black’s king is nicely placed in the centre,
supporting both of its passed pawns on d4 and f5. How would you start with White?
54.Bxa6?
Who wouldn’t take this pawn? But in reality, this move costs White half a point.
White could have won in study-like fashion with 54.Kc4!, which is the only move that leads to the win. The
next moves are more or less forced: 54...f4 (54...a5 is no solution either in view of 55.Bd3+ Ke5 56.Kb5 f4
57.Kxa5 f3 58.Kb4, and White wins) 55.a5! (a good moment to fix the black pawn on the colour of the
bishop) 55...Ke3 (after 55...f3 56.Bd3+ Ke3, we get an analogous position) 56.Bd3! (not 56.Bf1 Kf2
57.Kxd4 Kxf1 58.Ke4 Ke2 59.Kxf4 Kd3=) 56...f3.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9p+-+-+-+0
9zP-+-+-+-0
9-+Kzp-+-+0
9+-+Lmkp+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This is a crucial position that is very hard to understand. There is only one winning concept for White here:
A) If we try the logical move 57.Bf1?, it turns out that Black has a miraculous escape: 57...Kf2 58.Bh3
(58.Bd3 Ke3).
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9p+-+-+-+0
9zP-+-+-+-0
9-+Kzp-+-+0
9+-+-+p+L0
9-+-+-mk-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Here, Black has a brilliant idea to save the half point: 58...d3!! (absolutely the only move! After, for example,
58...Kg3 59.Bf5 Kf4 60.Bd3, White triumphs by harking back to the concept from the main line) 59.Kxd3
(this is actually the square that was destined for the bishop) 59...Kg3 60.Bc8 f2 61.Ke2 Kg2 62.Bh3+ Kxh3
63.Kxf2 Kg4, with a draw. What a save!;
B) 57.Bf5 is one of several moves on the b1-h7 diagonal that are all winning for White – but why? 57...Kf4
58.Bh3 Kg3 59.Bf1 Kf2 60.Bd3 Ke3, and now we go back to the winning idea;
C) 57.Bg6!! (or also 57.Bb1 or 57.Bh7).
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9p+-+-+L+0
9zP-+-+-+-0
9-+Kzp-+-+0
9+-+-mkp+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This is a move that is extremely hard to find if you have the position on the board, let alone foreseeing it in
your calculations. The bishop is temporarily tucked away on a square where it cannot be attacked with
tempo gain. After this beautiful move, Black is in zugzwang. Let’s see:
C1) 57...Kf4 58.Kxd4, and White wins;
C2) 57...f2 58.Bd3, with zugzwang, e.g. 58...Kd2 59.Kxd4 Ke1 60.Kc5 (60.Bxa6);
C3) 57...Ke2 58.Kxd4 f2 59.Bd3+ and White wins, e.g. 59...Ke1 60.Bxa6.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9L+-+-+-+0
9+-mK-+p+-0
9P+-zpk+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
54.Bxa6? 54...d3!
Very well done by Bianca – this is the move that draws!
54...f4? would have lost.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9L+-+-+-+0
9+-mK-+-+-0
9P+-zpkzp-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
But again, there is only one winning idea for White:
A) 55.Kb4!. White’s king has to run along with the d-pawn, while the bishop keeps an eye on the f-pawn –
again, a matter of the right division of tasks: 55...f3 (55...d3 56.Kc3 d2 57.Kxd2, and White wins) 56.a5 f2
57.Bf1! (now, the bishop has to stop the f2-pawn) 57...Ke3 58.Kc4!, and White is winning;
B) The logical 55.a5? doesn’t win: 55...d3 (55...f3? 56.Bb7+ Ke3 57.Bxf3 Kxf3 58.Kxd4). Now, if the king
runs along with 56.Kb4?, it is a draw (56.Bc4 Ke3 57.a6 f3 58.a7 f2 59.Bxd3 (59.a8=Q f1=Q is also a draw)
59...Kxd3 60.a8=Q Ke2 leads to the position that came on the board in the game, ending in a draw): 56...d2
57.Be2 f3 58.Kc3 fxe2 59.Kxd2 Kd5, and the black king has reached the square.
55.Bxd3+
This is not strictly necessary, but it looks logical.
White could have waited with this for one more move, keeping marginal winning chances in the ending
after 55.a5 d2 56.Be2 Ke3 57.Bd1 f4 58.a6 f3 59.a7 f2 60.a8=Q f1=Q=.
 XIIIIIIIIY
9Q+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-mk-+-0
9-+-zp-+-+0
9+-+L+q+-0
xiiiiiiiiy
analysis diagram
In principle, this queen + bishop vs queen ending (in this case, even with an extra pawn) is a theoretical
draw, but White could always give it a try. You never know, and quite a lot of endgame studies have been
made in which the stronger side still managed to win in an artistic way.
55...Kxd3 56.a5
Now, suddenly we have a race in a pawn ending.
56...f4 57.a6 f3 58.a7
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+k+p+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
58...Ke2
Of course, White had the following in mind: 58...f2 59.a8=Q, and now certainly not 59...f1=Q?? in view of
60.Qa6+, winning the black queen with an X-ray check, but 59...Ke2!, and Black ends up in the same
variation as in the game.
59.a8=Q f2
XIIIIIIIIY
9Q+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+kzp-+0
9+-+-+-+-0
xiiiiiiiiy
And here we have the starting position from Chapter 10 on the board (No. 49).
Chapter 13
Knight versus pawn
With endings of queen vs pawn and rook vs pawn, respectively, we saw that the pawn became a more
formidable weapon as it came closer to the promotion square. The further the pawn has advanced, the
harder it gets for the stronger side to stop it. It is not different in knight versus pawn endings. In general,
the knight is quite clumsy when it has to stop a pawn, and as we will see, especially a rook pawn can make
life difficult for the knight.
Clearly, the maximum result that the side with the knight can achieve is a draw – with the exception of
one bizarre case with the rook’s pawn. We will discuss this case right away, so we can leave it out of
consideration as we continue with this subject.

69
Exception: knight gives mate
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+p0
9-+-+N+-mk0
9+-+-+-+-0
xiiiiiiiiy
White to move
In principle, the knight’s side can never win if it has no pawns left of its own. But this is an exception!
1.Kf3!
The only move that wins.
1.Kf2?? is only a draw. As we will see below, it is a matter of zugzwang, and therefore it is of great
importance to put the king on the right square immediately. The reason why the king move to f2 doesn’t
win is that White has no good continuation after 1...Kh1. He has to manoeuvre his knight, but also he has to
be able to play Ng3# immediately as soon as Black plays ...h3-h2. This would only be possible in case of
2.Kf1 h2, but obviously Black would reply 2...Kh2, and if now the knight goes on a stroll, Black’s king will
escape from the corner. And mate in the corner is impossible without re-routing the knight.

1.Kf3! Kh1 2.Kf2 Kh2

It is clear that 2...h2 3.Ng3 would be checkmate.
3.Nd4
Now the knight takes a walk around the block, intending to end up on the f1-square, after which ...h3-h2
will be forced at the right moment.
Or 3.Nc3 Kh1 4.Ne4 Kh2 5.Nd2 Kh1 6.Nf1 h2 7.Ng3#.
3...Kh1 4.Nf5 Kh2
White has to be able to meet 4...h2 with 5.Ng3# at any time.
5.Ne3 Kh1 6.Nf1
Now, Black’s king cannot move to h2, and so he has to allow the ‘helpmate’:
6...h2 7.Ng3#
The following fragment, taken from a video on Hikaru Nakamura’s Twitch livestream
(https://www.youtube.com/watch?v=RQn_YImo2mo), shows that even experienced top grandmasters can
go wrong in such endings with limited time. Dutch junior player Sascha Kurt, who gets a lot of inspiration
from following Hikaru on his channel, pointed this video fragment out to me.

70
Wesley So
Hikaru Nakamura
Titled Tuesday blitz 2021
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+nmk-mK0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White has given up his knight, but of course the position is a dead draw. There followed:
1.Kxh7 Nf4 2.h5
Actually, it doesn’t matter what White plays – except what he played in the game...
A move like 2.Kg8 was okay.
2...Kf7
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+K0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-sn-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
3.h6??
Of course, this was a blitz game with little time on the clock, but this is a horrible blunder. Here, Hikaru
cried: ‘Wesley oh my gosh oh Wesley just blundered, this oh my gosh, let’s check’.
3...Ne6 4.Kh8 Nf8 5.h7 Ng6#
This goes to show that even a top grandmaster can botch up pretty badly!
In most situations, the knight’s side will be able to secure the half point, but the knight is an especially poor
piece when it comes to stopping certain pawns.
We will take a look at a number of situations in which the knight has to wriggle and squirm to stop an
enemy pawn. Before we do that, it will be useful to formulate a number of rules of thumb for this ending
type:

Rules of thumb
• In almost all cases, a knight standing in front of the pawn will be able to hold the draw on its
own, i.e., without help from its king.
• The rook’s pawn is the natural enemy of the knight.
• The knight’s side can sometimes save itself by using a knight fork.
• The side with the pawn should preferably put its king diagonally opposed to the enemy
knight (or horizontally, with two squares in between), to try to keep the knight at a distance
from the pawn.
Obviously, these rules of thumb need some more explanation, and we will provide this by showing a few
examples.

The pawn cannot be stopped

71
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9n+-+-+K+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The pawn can just move to the promotion square because the knight is too far away, and the black king is
outside the square.
1.c6
XIIIIIIIIY
9-+{+{+-+0
9+-+-+-mk-0
9-+P+{+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9n+-+-+K+0
9+-+-+-+-0
xiiiiiiiiy
1...Nb4
1...Nc3 2.c7 – please note that if the white king were on, for example, g3, the knight would be able to stop
the pawn by first giving a check on e4.
2.c7 Nd5 3.c8=Q 1-0

The knight jumps up and down

72
XIIIIIIIIY
9-+-+-+n+0
9+-+-+KzP-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
1...Nh6+
The knight keeps jumping up and down. This diagram position is an illustration of a theme we have already
mentioned: the knight controls the square in front of the pawn and has exactly enough room to move to
avoid being driven off.
2.Kg6 Ng8 3.Kf7 Nh6+ 4.Kg6 Ng8 ½-½

Not enough squares

If we move this entire position one file to the right, things are completely different. Here, we see what is
meant by the comment we gave under the previous position. On the edge (and in the corner), the knight
loses a great deal of its activity.

73
 XIIIIIIIIY
9-+-+-+-sn0
9+-+-+-+P0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The attentive reader will have spotted that this position is lost for Black. By moving the position one file to
the right, the knight has been driven to the most horrible place on the board for it: the corner. This
underlines the statement that the rook’s pawn is the natural enemy of the knight!
1.Kg7 Ke3 2.Kxh8 Kf4 3.Kg7 1-0

Giving up the knight for the pawn

74
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+K0
9-+-+-+-+0
9+-mk-+-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+n+-0
xiiiiiiiiy
Black to move
Black manages to capture the white pawn and thus hold the draw:
1...Ng3
White to move would win with 2.h6 Ne3 3.Kg6 Ng4 4.h7 Ne5+ 5.Kg7.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mKP0
9-+-+-+-+0
9+-mk-sn-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
1...Ng3 2.h6 Nf5! 3.Kg6 Nxh6 ½-½

Locking up the enemy king

Sometimes, it doesn’t have to be fatal if the knight is caught in the corner, as long as its own king is close.

75
XIIIIIIIIY
9-+-+-+-sn0
9+-+-+-+P0
9-+-+K+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The situation looks hopeless for Black – the knight will be lost, and then what?
1.Kf6 Kd5
But the black king hurries in pursuit of the white king...
2.Kg7 Ke6 3.Kxh8 Kf7
... and arrives just in time to prevent the white king’s escape from the corner. Stalemate!
Now that we are talking about lock-ups/imprisonment, we can’t omit to look at the following situation,
even if it doesn’t have the material balance that is the subject of this chapter:

76
XIIIIIIIIY
9-+-+-+-mK0
9+-+-mk-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-sN0
9+-+-+-+-0
xiiiiiiiiy
Black to move
White even has an extra knight and pawn, but still Black can force a draw. He has to lock up the white king
in the corner, but which square should his king move to – f7 or g8?
1...Kf8!!
To the colour square that the enemy knight is on at this moment! Now, the white knight won’t manage to
chase the black king away from the f7/f8 squares.
1...Kf7?? loses! After 2.Ng4 Kf8 3.Ne5, the black king is expelled from the f8-square, and then the white
king escapes from the corner.

1...Kf8!! 2.Ng4 Kf7 3.Ne5+ Kf8 4.Ng6+ Kf7 5.Ne7 Kf8

Obviously, Black persists in playing his king to and fro between these two squares. It’s a draw, since White
cannot ‘lose a tempo’ with his knight. In a pawn ending, a king can also ‘lose a tempo’ to reach the same
position with the other player to move, and then it is called ‘triangulation’. But in such cases, the king
moves to a square of the same colour on two consecutive moves. A knight cannot do this – it always jumps
from a dark square to a light one. It can never jump to the same colour square in two consecutive moves.

Knight carousel
Now that we have seen the correct path to the win in one of the earlier diagram positions, drawing a
conclusion in the next diagram position doesn’t seem difficult. However, this conclusion turns out to be
premature:

77
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mKn0
9-+-+-+-zP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
If the knight can stop the pawn just before it reaches the seventh rank, it is still a draw! We call this the
‘knight carousel’.
1...Ng5 2.Kg6
XIIIIIIIIY
9-+-+-}-+0
9+-+-+-+{0
9-+-+{+KzP0
9+-+-+-sn-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
xiiiiiiiiy
2...Ne6! 3.Kf6
Or 3.h7 Nf8+ and draw.
3...Nf8 4.Kg7
3...Nf8 4.Kf7 Nh7=.
4...Ne6+ ½-½
The ‘revolving movement’ of the knight on the squares h7-g5-e6-f8 guarantees Black the half point. This is
an important motif to keep in mind. So, a pawn on the sixth rank can be stopped if the knight is able to
reach one of the squares in the carousel. The most remarkable thing here is that the knight can even
manage without the help of its own king.

Rule of thumb
The knight’s side can hold the draw against a rook’s pawn on the third/sixth rank if it can
reach one of the four indicated squares in the diagram. We call this the ‘knight carousel’.

Knight fork
78
XIIIIIIIIY
9-+-+-mK-+0
9+-+P+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+n+-+-0
9-+-+-+-+0
9+-mk-+-+-0
xiiiiiiiiy
Black to move
At first sight, it looks as if Black has to resign here. However, he can hold the draw in the nick of time with a
tactical trick:
1...Nc5!
But not 1...Nf4?? 2.Kf7+–.

1...Nc5! 2.d8=Q Ne6+ ½-½

King and knight are in each other’s way

79
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-sn-0
9-+-+k+-+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
If the knight and king are in the vicinity of the enemy pawn, the pawn won’t be able to achieve anything on
its own – unless the knight is highly unfortunately placed!
1.h6 Kf7
1...Nf5 2.h7+–.
2.h7+–
The king can’t reach the pawn!

Keeping the knight at bay

Not unimportantly, the king has a way to keep the enemy knight at bay. This can be done vertically with
one square in between, or horizontally, with, preferably, two squares in between. In the rules of thumb, we
touched upon this method already.
Schematically, we can illustrate this theme as follows:

80
Schematic diagram
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-sN-+n+0
9+{+-+-+-0
9-mk{+-}-}0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
In this schematic diagram, the squares are indicated which are inaccessile for the knights due to the
positioning of the respective kings.

81
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+P+-+nmk0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Kd6!
White prevents the knight from stopping the pawn by moving to e7. Now, Black has to take a detour, and he
will be too late. We call this ‘keeping the knight at bay’.
1.c7 Ne7 2.Kd6 Nc8+ 3.Kd7 Nb6+ 4.Kc6 Nc8 5.Kd7 Na7.

1.Kd6! Nf4 2.c7 1-0

Locking in the knight

There is one extreme type of situation in which the knight is so unfortunately placed that it can be locked in:

82
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+K+-+-0
9-zP-+-+k+0
9+n+-+-+-0
xiiiiiiiiy
White to move
White can lock in the knight and remains with a winning pawn ending:

1.Kc2 Kf3
1...Na3+ 2.bxa3 Kf3 3.Kd3 Kf4 4.Kd4 Kf5 5.Kd5 Kf6 6.Kd6 Kf7 7.Kd7 Kf6 8.a4 Ke5 9.a5 Kd5 10.a6 and
the pawn queens.
XIIIIIIIIY
9-+-+-+-+0
9+-+K+-+-0
9P+{+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
2.Kxb1 Ke4 3.Kc2 Kd4
And now:

4.Kb3!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9{}{mk-+-+0
9+K+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Thus, White controls the key squares in this pawn ending (see also Chapter 17).
4...Kc5 5.Ka4 Kb6 6.Kb4 1-0

Knight barrier
83
XIIIIIIIIY
9-mk-+-+-+0
9+-+-+-+P0
9-zPK+-+n+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
In this position, composed by the author, White has to determine the best way to move towards the black
knight with his king. This seems to be a piece of cake, but in practice it turns out to be not so easy:

With 1.Kd6? d5e4eWhite loses a crucial tempo, and as a result he is not fast enough in the race towards the
corner. The scheme (diagram on the right) makes clear what is going on here.
XIIIIIIIIY
9-mk-+-}-}0
9+-+[}-+P0
9-zPK+[+n+0
9+-+-},+-0
9-+-+,}-}0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The squares marked with an ‘{’ are directly covered by the knight, and cannot be accessed by the king.
The squares indicated with a ‘[’ can in fact not be accessed by the king either since then Black has ...Nf8+,
winning the pawn. In fact, Black has erected a kind of palisade with the knight, not allowing the white king
to pass so easily.
However, the visual image indicates that there are ‘holes in the fence’: on e8 (unreachable) and e4/f5
(indicated with a ‘,’), where the king can creep through to attack the knight. For example: 1...Kb7 2.Kd5
Kxb6 3.Ke4 Kc6 4.Kf5 Nh8 5.Kf6 Kd7 6.Kg7 Ke7 7.Kxh8 Kf7, and Black obtains the draw in the nick of
time.

1.Kd7? immediately loses the essential pawn by 1...Nf8+.

1.Kd5!
d5e4e4f5Only the text move wins. In the second diagram, we can immediately see where the barrier is, and
how the white king can sneak through to chase away the knight.
1...Kb7
Black has to eliminate the b6-pawn first, and this takes him one move to many to save the game.
2.Ke4
Again, not 2.Ke6?? in view of 2...Nf8+.
2...Kxb6 3.Kf5 Nh8 4.Kf6 Kc6 5.Kg7 Kd6 6.Kxh8 Ke7 7.Kg7
And White hauls in the loot.
Quiz
It’s time for a real brain-teaser. In the following beautiful, instructive and entertaining study, the knight
goes off on a veritable peregrination to be able to sacrifice itself for the black pawn in the nick of time.

84
Study by Nikolay Grigoriev
Source unknown, 1938
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+N+k+-+-0
9-zp-+-+-+0
9+-+-+-+-0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
White to move
White to play and draw
1.Nc7+ Kc4
Thus, Black keeps the knight away from the important b5-square, and the battle seems lost for White.
After any other move, e.g. 1...Kd4, it is important that the pawn can never be pushed forward, since the
knight can be moved in front of the pawn from the ‘side’, e.g. 2.Kg2 b3 3.Nb5+ Kc4 4.Na3+ or 1...Ke5
2.Na6 b3 3.Nc5 b2 4.Nd3+.

Question 1: Where should the knight go to be able to approach the pawn in one way or another?
2.Ne8!
Absolutely the only move, even though it looks like the knight is straying here.
To the strange move 2.Na8?, Black has only one correct reply: 2...Kc5! (not 2...b3?, after which White draws
with 3.Nb6+ Kb5 4.Nd5 b2 5.Nc3+, and the knight has reached the promotion square), and White can
forget the draw, e.g. 3.Nc7 b3 4.Na6+ Kc4.
After 2.Ne6?, the knight is very unfortunately placed since it cannot stop the black b-pawn.
Now there is one way for Black to avoid an immediate draw:
XIIIIIIIIY
9-+-+N+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-zpk+-+-+0
9+-+-+-+-0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
2...Kc5!
Black’s king takes away the d6-square from the knight.
A logical attempt is 2...b3, but after 3.Nd6+ Kb4 (or 3...Kd3 4.Nb5 b2 5.Na3) 4.Ne4 b2 5.Nd2, the knight
will reach the promotion square either from the right or from the left, securing the draw again.
3.Nf6
White cannot hesitate, and has to move the knight.
3.Ng7? moves the knight too far away; after 3...b3 4.Nf5 (4.Ne6+) 4...b2, there is no way to prevent
promotion.
The same goes for 3.Nc7? b3 4.Ne6+ Kc4 – again, the king is diagonally opposed to the knight, and it’s
game over.
3...Kd4
We know from earlier examples that a diagonal opposition between the king and the knight poses great
problems to the knight.

Question 2: Now what? Is there a route for the knight along which it can stop the pawn in time?
4.Ne8!
Again, brilliant! The knight has to move to the edge of the board again – completely in the wrong direction,
you would think – to get on the right track towards the promotion square of Black’s pawn. It can reach the
promotion square just in time via either of two routes: e8-c7(or d6)-b5-a3 or e8-d6-e4-d2.
XIIIIIIIIY
9-+-+N+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-zp-mk-+-+0
9+-+-+-+-0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
4...Ke5!
Using king manoeuvres to make life as difficult as possible for the opponent.
For example: 4...b3 5.Nd6 Kc3 (thus, Black tries to cut the knight short; if 5...b2 6.Nb5+ Kd3 7.Na3, White
is right on time!) 6.Ne4+! (now, the knight has to take the other route, via d6-e4-d2; if 6.Nb5+ Kb4!,
winning!) 6...Kc2 7.Nd6! (quick – back again) 7...b2 8.Nc4, and again White is exactly in time to force the
draw.
XIIIIIIIIY
9-+-+N+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mk-+-0
9-zp-+-+-+0
9+-+-+-+-0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
Question 3: Again, White faces the problem of how to get to the pawn from such a large distance. Which
move should he play here?
5.Nc7! Kd6!
Black’s king keeps harassing the knight, putting White on the spot.
Clearly, pushing the pawn with 5...b3 would be met by 6.Nb5 b2 7.Nc3, and again White has made it!
 XIIIIIIIIY
9-+-+-+-+0
9+-sN-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-zp-+-+-+0
9+-+-+-+-0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
Question 4: What is the only correct continuation for White?
6.Ne8+!
The only salvation, paradoxical as it may seem.
Moving the knight in the other direction doesn’t work; after 6.Nb5+?, Black wins with 6...Kc5 7.Nc7 b3
8.Ne6+ Kc4!.
6...Kc5
Not 6...Ke7? in view of 7.Nc7, when the knight ends up in front of the pawn after 7...b3 (with 7...Kd6, Black
could end up in the same position as before after 8.Ne8+) 8.Nd5+ Kd6 9.Nc3, with a draw.
7.Nf6 Kd4
We know this position; we had it on the board after the third move.
8.Ne8 b3
Black’s options are exhausted, and so he plays his trump card.
9.Nd6 Kc3
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-sN-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+pmk-+-+-0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
The alternative has been discussed above.
10.Ne4+!
10.Nb5+ is met by 10...Kb4 11.Nd4 b2, and the pawn queens.
10...Kc2!
Again, the knight is screened off diagonally. After 10...Kd4 11.Nd2, the knight gets in front of the pawn.
Question 5: Now, the knight is again unfortunately placed, but White can save himself with another
manoeuvre. Which one?
11.Nd6! b2
After 11...Kd3, the knight again crawls towards the pawn’s promotion square along the edge: 12.Nb5! b2
13.Na3.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-sN-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-zpk+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
Question 6: And now, the finishing touch!
12.Nc4!
The only way to save himself, but it works!
12...b1=Q
And now the saving knight fork:
13.Na3+ ½-½
A heroic duel between the black king and the white knight!

This endgame in practice

85
Zeynab Mamedjarova 2318
Herman Grooten 2387
Hoogeveen 2006 (5)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9p+-zp-+-+0
9zPn+P+kzP-0
9-+K+-vL-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Endgame study composer Nikolay Grigoriev
This game was played in the Hoogeveen Open. The tournament hall was almost empty at this point. We are
in the blitzing-out phase of the game, in a position where White has a healthy extra pawn.
While I was thinking about my next move, suddenly I saw the brother of my opponent, the famous
grandmaster Shakhriyar Mamedyarov (who was playing in the Crown Group) entering the hall.
Immediately, he struck up a conversation with his sister, and there was no arbiter in sight... I thought to
myself, ‘Now I’ll have to face 2750+ Elo too with this lousy ending’...
However, one should never think like that, and so in the second instance I assumed that they were talking
about the weather, not the game. Actually, my opponent didn’t see how she could make progress by normal
means, and therefore she sacrificed the bishop to be able to penetrate into my position with her king. With
little time on the clock, I had to make several difficult decisions now.
58.Bxd6!? Nxd6+ 59.Kc5 Ne4+ 60.Kb6
This is the first moment when Black has to determine the best way to defend this position.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9pmK-+-+-+0
9zP-+P+kzP-0
9-+-+n+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Afterwards, it turned out that there was more than one way to hold the draw here, but during a game a
player has to make choices. Dutch grandmaster Jan Timman once said that you have to choose the one
variation you trust the most, and try to calculate it to the end. If this variation leads to the desired goal, you
should play it, after checking it first. Here, I opted for:
60...Ke5
My idea was to remove the potentially dangerous passed d5-pawn from the board first. Then, Black has to
divide the tasks between his king and his knight. There are two ideas:
1. If the white king takes on a6 quickly, the black king takes the d5-pawn and then moves his king towards
the white g-pawn, since it is still in the square of this pawn. The knight can then control the white a-pawn
by making use of the famous ‘carousel’.
2. If White tries to run off with the g-pawn right away, the black king still takes the d5-pawn and then
directs its attention to the white a-pawn. It is important that if the white king takes on a6, it can be nailed to
the edge of the board by the black king immediately. In these cases, the knight can stop the g-pawn. I knew
that even if the white king escaped from the edge later and tried to chase off the knight, the draw could be
secured by jumping up and down with the knight. So, the variations were connected with the schemes I had
in my head.
With hindsight, it turns out that 60...Kxg5 was also a quite simple way to achieve the draw: 61.Kxa6 Kf6
62.Kb6 Nd6 63.a6 Ke7, and the knight carousel does the job again – 63...Ke5 could also be played, as long
as the d5-pawn can be captured in time.
61.Kxa6
Of course, for the sake of consistency we also have to investigate 61.g6!?.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9pmK-+-+P+0
9zP-+Pmk-+-0
9-+-+n+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Now, 61...Kxd5! is the only move! (the immediate 61...Nf6?? would be a terrible blunder in view of 62.Kxa6
Kxd5 – 62...Nxd5 also loses in view of 63.g7 Ne7 64.Kb7, and here too, the a-pawn decides – and Black is
too late after 63.Kb7!, when the white king cannot be kept on the edge, and so the a-pawn can march
straight to the other side) 62.Kxa6 Kc6 63.g7 Nf6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zP-0
9K+k+-sn-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
With a draw; for example, 64.Ka7 Kc7 65.a6 Kc8 66.Kb6 Kb8 67.Kc5 Ka7 68.Kd4 Kxa6 69.Ke5 Ng8
70.Ke6 Kb6 71.Kf7 Nh6+ 72.Kg6 Ng8 73.Kf7 Nh6+ 74.Kg6 etc.
61...Kxd5 62.Kb6
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-mK-+-+-+0
9zP-+k+-zP-0
9-+-+n+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Once more, an important moment, for now the first defensive scheme works. The division of tasks is thus:
the knight has to stop the a-pawn, and the king focuses on the g-pawn.
62...Nd6 63.a6
63.g6 Ke6.
63...Ke6 64.Kc5
She tries a different path. The carousel works in case of 64.Kc6 Nc8 65.Kc7 Na7 66.Kb7 Nb5 67.Kb6 Nd6
etc.
64...Nc8 65.Kc6 Kf5 66.Kc7 Na7
XIIIIIIIIY
9-+-+-+-+0
9sn-mK-+-+-0
9P+-+-+-+0
9+-+-+kzP-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Again, the knight does a good job with the carousel!
67.Kb6 Nc8+ 68.Kb7 Nd6+ 69.Kc7 Nb5+ 70.Kb6 Nd6 71.a7 Nc8+ 72.Ka6 Nxa7 73.Kxa7 Kxg5 ½-
½
A struggle until the bitter end.
Chapter 14
Exercises
Chess is learnt by thinking independently and trying to find solutions to problems that come up during a
game. Therefore, I have tried to compose a nice exercise for each endgame type that you can try to solve –
perhaps together with a chess friend.

1 – Queen vs pawn
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9p+-+-zp-+0
9zP-+-+-+P0
9-+k+-+-+0
9+-+-+-+K0
xiiiiiiiiy
White to move
Play out this position (possibly online) against a sparring partner or a computer program that can be set,
for instance, on ‘club player’ strength. You will be playing an ‘educative game’, which means that you can
take back any move (on either side), not only to try to convert this position into a win, but also, especially,
to acquaint yourself with various endgame principles through practice.

2 – Rook vs pawn
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-zp-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-+0
9+-+-+-+-0
9R+-+-+-+0
9+-+-+-mk-0
xiiiiiiiiy
White to move
Play out this position (possibly online) against a sparring partner or a computer program that can be set,
for instance, on ‘club player’ strength.

3 – Bishop vs pawn
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+-+p0
9-+K+-+-+0
9+L+-+-+-0
xiiiiiiiiy
White to move
Try to solve this endgame study: White to play and draw.

4 – Knight vs pawn
XIIIIIIIIY
9-+-+-+-sN0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zp0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Exercise A: Use the two help diagrams given below and indicate the knight barrier by drawing crosses on
the squares that cannot be used by the black king without being in check or losing the pawn by a knight
fork.
Exercise B: Now solve this endgame study: White to play and win.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-sN-0
9-+-+-+-zp0
9+-+-+-+-0
xiiiiiiiiy
help diagram 1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-sN-zp0
9+-+-+-+-0
xiiiiiiiiy
help diagram 2
Chapter 15
Solutions

1 – Queen vs pawn

I have regularly played out this position against a number of chess pupils who were under my tuition.
Sometimes, I would consciously play a move that was not the best, to find out if the pupil knew exactly
what he was doing!
The winning method is given below.
1.Kg2!
A deceptive mistake that was regularly made is the following, which superficially looks like an
attractive option: 1.h4?. But now, suddenly the black king is in time to support its own f-pawn, even
enabling Black to win: 1...Kd2! 2.Kg2 (2.h5 f3 3.h6 f2, and after 4.h7, Black queens with check) 2...Ke2 3.h5
f3+ 4.Kg3 f2.

1.Kg1, a move that was played by some ‘whiz-kids’, is not the best. The black king enters the square of
the white h-pawn with 1...Kd3, thereby forcing the draw. White will win the race for the a-pawns, but his
king won’t be in time to keep its opposite number from reaching the corner.
Here is a sample line to illustrate this: 2.Kf2 Ke4 3.h4 Kf5 4.Kf3 Ke5 5.h5 Kf5 6.h6 Kg6 7.Kxf4 Kxh6
8.Ke5 Kg6 9.Kd5 Kf6 10.Kc5 Ke6 11.Kb5 Kd7 12.Kxa4 Kc7 13.Kb5 Kb7, and Black is easily on time.

1.Kg2!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9p+-+-zp-+0
9zP-+-+-+P0
9-+k+-+K+0
9+-+-+-+-0
xiiiiiiiiy
1...f3+!
The best and most devious move, forcing White to play very inventively. Mostly, I don’t play this move
as my main variation – first, I force my pupil to find out how he has to win in the alternative variation.
We could call 1...Kb3 the ‘most straightforward’ variation. It leads to the following liquidation: 2.h4
Kxa3 3.h5 Ka2 (the king has to move in front of its own pawn, so as to keep the promotion square a1
protected later on; obviously, Black can also play 3...Kb2 here, even though White is also on time in that
case, since the king will be in check from h8; certainly not 3...Kb3?, because White then promotes first and
even has time to put his queen in front of the black pawn, after which the win is simple: 4.h6 a3 5.h7 a2
6.h8=Q Kc2 7.Qa1+–) 4.h6 a3 5.h7 Kb1 6.h8=Q a2.
 XIIIIIIIIY
9-+-+-+-wQ0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-zp-+0
9+-+-+-+-0
9p+-+-+K+0
9+k+-+-+-0
xiiiiiiiiy
analysis diagram

Normally this is a draw, you would think – and actually that is what most people think. But since Black
has an extra pawn, the whole idea of stalemate that the defender usually has in these positions is gone.
White can ‘abuse’ the black f-pawn. Things could continue as follows: 7.Qb8+ Kc2 8.Qa7 (the queen is
zigzagging closer; in a simul, I always accelerate this zigzagging mechanism somewhat by, for instance,
putting the queen on b4 immediately) 8...Kb2 9.Qb6+ Kc2 10.Qc5+ Kb2 11.Qb4+ Kc2 12.Qa3 Kb1
13.Qb3+ Ka1.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-zp-+0
9+Q+-+-+-0
9p+-+-+K+0
9mk-+-+-+-0
xiiiiiiiiy
analysis diagram

The black king is now stalemated, but here White has a beautiful mating idea we have already seen in
the instruction part: 14.Qc2! f3+ 15.Kf1 f2 16.Qc1#.
Once we have found this, we return to the position after 1.Kg2 and I play the nasty move 1...f3+!. As
the students have wisened up a little by this stage, they will be looking with suspicion at the capture of the
pawn!
2.Kf2!
Due to the above stalemate idea, the king would prefer to move to f1 here, but then Black again
escapes to a draw by going after the white h-pawn: 2.Kf1? Kd3!. The black king enters the square, and
again, White won’t win the race to the queenside: 3.Kf2 Ke4 4.h4 Kf4 5.h5 Kg5 6.Kxf3 Kxh5 7.Ke4 Kg6
8.Kd5 Kf7 9.Kc6 Ke7 10.Kb5 Kd7 11.Kxa4 Kc8.
2...Kb3
Black needs to collect White’s a-pawn, but as we already know, White will get to the other side first.
If 2...Kd3, now the white king cuts off the black king with 3.Kxf3, winning. The king keeps ‘shouldering
off’ its opposite number: 3...Kd4 4.Kf4 Kd5 5.Kf5 etc.
3.h4 Kxa3 4.h5 Kb2 5.h6 a3 6.h7 a2 7.h8=Q+ Kb1
 XIIIIIIIIY
9-+-+-+-wQ0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+p+-0
9p+-+-mK-+0
9+k+-+-+-0
xiiiiiiiiy
Here, we have almost the same situation as in the above-mentioned line, but this time the white king
also obstructs the push of Black’s f-pawn. This causes an extra complication which, however, can be solved
in an elegant way.
8.Qb8+ Kc2 9.Qc7+
Again, we are zigzagging towards the third rank.
9...Kb2 10.Qb6+ Kc2 11.Qc5+ Kb2 12.Qb4+ Kc2 13.Qa3 Kb1 14.Qb3+ Ka1
Again, the black king is stalemated. What was it we knew about theoretically winning positions? We
have to consult theory here. In section A in Chapter 10, ‘Queen versus pawn’, we defined the critical area in
which the stronger side’s king has to stand. Obviously, we have to mirror that position in our mind’s eye,
and then we come to the following scheme:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9}{}{+-+-0
9{}{}{+-+0
9+Q}{}p+-0
9p+{}{mK-+0
9mk-}{}-+-0
xiiiiiiiiy
15.Ke3!
From the scheme, we can conclude that White’s king is able to enter the critical area, after which the
position is winning.
15...f2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+Q+-mK-+-0
9p+-+-zp-+0
9mk-+-+-+-0
xiiiiiiiiy
The previous move by White’s king was possible because Black wasn’t stalemated yet, and had to push
his f-pawn. But now that the white king has entered the critical area, it is important that we find a clever
way to get rid of the f-pawn, which has now turned into a ‘nuisance’. Absolutely the wisest solution here is
to think prophylactically. We know that Black’s next move will be ...f2-f1=Q, so we can anticipate that by
keeping the black king imprisoned in its ‘cage’ and in the meantime controlling the f1-square with the
queen, as follows:
16.Qb5!
Incidentally, there are various other winning methods that take more time, but are of course not
wrong.
16...f1=Q 17.Qxf1+ Kb2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9pmk-+-+-+0
9+-+-+Q+-0
xiiiiiiiiy
We know the mechanism: queen to the second rank, king to the third rank, and if Black then queens,
we give mate on c2. So, it makes sense to take a good look at which square is useful for White to give check
on:
18.Qf2+!
This square is to be preferred by far over, for instance, e2, since after 18.Qe2+, Black can annoy White
a little longer with 18...Kb3. Now, White would like to move the queen to the long a1-h8 diagonal so it can
jump on the promotion square next. But White’s own king is in the way! So, therefore – 18.Qf2!.
18...Kb1
Forced. Now, 18...Kb3 wouldn’t exactly help in view of 19.Qf6! Kc2 20.Qa1.
19.Kd3 a1=N
A last spasm, bit it doesn’t save the game for Black.
After 19...a1=Q, we have achieved our mating pattern: 20.Qc2#.
20.Kc3 Nb3 21.Qb2#

2 – Rook vs pawn
Study by Josef Moravec, La Stratégie 1912

The only winning move here is an incredible one:

1.Kh7!!
How many players wouldn’t just gulp down the ‘tasty morsel’ on g7? But after 1.Kxg7, White can no
longer win! Following 1...h4 2.Kg6 h3 3.Kg5 h2 4.Kg4 h1=Q, White doesn’t have the desired idea of
domination with 5.Kg3? (5.Ra1+! Kh2 6.Rxh1+) due to the hidden defence 5...Qh8!, when the a1-square is
protected, and White even loses! So, he shouldn’t allow Black to queen his pawn, but rather he should
eliminate the pawn, after which it is a draw.

1.Kh7!! h4
Black can also try 1...g5, but White wins too after 2.Kg6 h4 3.Kxg5 h3 4.Kg4 h2 5.Kg3 h1=N+
(5...h1=Q 6.Ra1#) 6.Kf3, and the knight is lost.
2.Kg6 h3 3.Kg5 h2 4.Kg4
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zp-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+K+0
9+-+-+-+-0
9R+-+-+-zp0
9+-+-+-mk-0
xiiiiiiiiy
4...h1=N
The only move to keep on resisting, since 4...h1=Q would now lead to domination by the white king
after 5.Kg3!. We see here that the g7-pawn is ‘harmful material’ for White, since he no longer has the move
Qh8 to cover the a1-square. What an apotheosis!
5.Kf3 g5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-zp-0
9-+-+-+-+0
9+-+-+K+-0
9R+-+-+-+0
9+-+-+-mkn0
xiiiiiiiiy
This pawn might throw a spanner in the works, so White has to play carefully not to allow the knight
to escape from the corner. Here again, we apply the technique of ‘prophylaxis’. We know Black’s next move
will have to be 6...g4+. What is the best way to anticipate this?
6.Ra4!
The knight remains caught in the corner, and Black’s king is also unable to make a decent move.
Black has to give up his pawn:
6...Kh2
6...g4+ 7.Rxg4+ Kf1 (if 7...Kh2, a waiting move with the rook on the g-file wins the knight, e.g. 8.Rg8)
8.Rg2, and the knight is lost.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-zp-0
9R+-+-+-+0
9+-+-+K+-0
9-+-+-+-mk0
9+-+-+-+n0
xiiiiiiiiy
7.Rg4 Kh3 8.Rxg5 Kh2
After 8...Kh4 9.Rg8, Black will lose his knight as well.
9.Rg8
After this waiting move, it is definitively over for Black.

3 – Bishop vs pawn
Study by Andrey Selivanov,
Shakhmatnaya Kompozitsia 1993

The question, of course, is how White can stop Black’s h-pawn, and as it turns out there is only one
good move!
1.Kd3!
Curiously, White has to make the most paradoxical move on the board to save the game. Any
discovered check loses, but by keeping his king on the bishop’s diagonal White is able to save his skin. The
idea is to control the e4-square for the bishop, to keep Black’s pawn under control.
1.Ba2? is met by 1...Ke5!, when the pawn cannot be stopped.

The same happens after 1.Kc3+? Ke5!, and here too, the pawn will queen.

The final alternative, 1.Kd2+?, also doesn’t work: 1...Kf4! 2.Ba2 Ke5!, and again Black’s pawn will
queen.

1.Kd3! Ke5
The king already moves out of the check, but Black cannot prevent the coming defence.
Here, 1...Kf4 amounts to the same as the main line: 2.Kd4 h2 3.Be4. If Black tries to queen right away
with 1...h2, White has 2.Ke3+ Ke5 3.Be4.
2.Ke3 h2 3.Be4
Mission accomplished! ½-½

4 – Knight vs pawn
Study by Nikolay Grigoriev, 64 1932

White is facing the difficult task of stopping the black h-pawn. His problem is that his knight and king
are rather unfortunately placed. White just manages to save himself by making ingenious use of a knight
barrier:
1.Nf7!
1.Ng6? loses: 1...h3 2.Nf4 h2 3.Ne2+.
XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mk-+-+-0
9-+-+N+-zp0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram

Since the knight has to move to g3 next, Black can anticipate this with his king. He can bypass the
knight barrier with 3...Kd2! 4.Ng3 Ke1. Now the king is closing in on the knight at great speed: 5.Kd6 Kf2
6.Nh1+ Kg2 – the knight is lost, and White’s king is too far away to lock the black king in the corner.
1.Nf7! h3 2.Ng5 h2 3.Ne4+!
XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+N+-+0
9+-mk-+-+-0
9-+-+-+-zp0
9+-+-+-+-0
xiiiiiiiiy
This is the difference with the first variation. From this square, White has a choice between putting the
knight on f2 or on g3 to stop the pawn. The choice depends on Black’s next king move.
We have used help diagrams before to look closer at two different scenarios. We will do the same here.
In the first scenario, we put the knight on g3 and see what the barrier looks like by drawing ‘{’ and ‘[’ on
the ‘forbidden squares’:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+{+{0
9-+-+{+-+0
9+-+-]-sN-0
9-+-]{+-zp0
9+-+-+{+{0
xiiiiiiiiy
help diagram 1

In the second scenario, we put the knight on f2 and draw crosses again:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-]-]0
9+-+-]-+-0
9-+-+{+{+0
9+-+{]-+{0
9-+-+-sN-zp0
9+-+{+-+{0
xiiiiiiiiy
help diagram 2

The crosses and slashes indicate the squares where the enemy king cannot go, which is why this is
called the ‘knight barrier’. Actually this is a kind of palisade around the knight that shuts out the enemy
king. Now that we know this, it is easy to solve the study:
3...Kd3
3...Kd4 4.Nf2!.
 XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-sN-zp0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram

Now the black king has to make a lengthy detour, as follows: 4...Kc3 (if 4...Kd5, White cuts off the king
with 5.Kd7) 5.Kd6 Kd2 6.Ke5 Ke2 7.Nh1 Kf3 8.Kd4 Kg2 9.Ke3 Kxh1 10.Kf2 with a draw.
4.Ng3!
XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+k+-sN-0
9-+-+-+-zp0
9+-+-+-+-0
xiiiiiiiiy
And now the detour Black’s king has to make is too long, allowing White’s king to get to the knight on
time:
4...Kc2 5.Kd6 Kd1 6.Ke5 Ke1 7.Kf4 Kf2 ½-½
 Part III
Various techniques
Chapter 16
Introduction
If in the middlegame many pieces are exchanged, a player can start considering what the endgame will look
like for him. The great masters in chess history knew better than other players which pieces they wanted to
exchange and which pieces they want to keep on the board. Once there was an opponent of former World
Champion Tigran Petrosian who asked rhetorically how it could be that whenever he played against
Petrosian, he would always end up with the wrong pieces...
This is what the middlegame is often about: you have to have an open eye for the type of ending you
want to liquidate to. To offer you something to go on, in this part I have decided to present a number of
endings in which the material is the same on both sides. The arrangement is as follows:
Chapter 17 – Pawn endings
Chapter 18 – Knight endings
Chapter 19 – Bishop endings (divided into endings with same-coloured bishops and endings with opposite-
coloured bishops)
Chapter 20 – Rook endings
Chapter 21 – Queen endings
In each of these chapters, I will endeavour to ‘catch’ many specific aspects and describe them as Rules of
thumb and Techniques which I will then illustrate with the help of sample games. Apart from the text, I
hope that the diagrams will clarify things, by illustrating the theme at hand with a ‘picture’. If all goes well,
such a visual image of a concept will be a ‘memory anchor’ for the reader, and will emerge at the right
moment when the theme occurs in a game.

86
Massimiliano Negra 1955
Herman Grooten 2325
Caorle 1988 (7)
XIIIIIIIIY
9-+-+-+-+0
9zpp+-+-+-0
9-+-+-+-+0
9mKP+k+-+-0
9-+-+n+-+0
9+-+-+N+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
I had been trying for quite some time to win this ending, even though White will most probably be able to
draw.
66...Kc5 67.Ne5 Kd5
 XIIIIIIIIY
9-+-+-+-+0
9zpp+-+-+-0
9-+-+-+-+0
9mKP+ksN-+-0
9-+-+n+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Of course, White could move back with the knight, and then I would have had to look for something else.
But White was a bit fed up with it all and wanted to force the draw, which caused him to make a mistake...
68.Nc6??
My opponent thought that this move would enable him to liquidate all of Black’s pawns, but he was rudely
awakened because here I remembered a certain mechanism just in time.
If 68.Nf3, 68...Nd6 doesn’t bring Black any benefits after 69.Ng5 Nc4+ (69...Kc5 70.Ne6+) 70.Kb4.
68...bxc6 69.bxc6 Nd6!
The knight is just able to protect Black’s last pawn. Of course, not 69...Kxc6?, the move White had counted
on: after 70.Ka6, the last pawn leaves the board.
70.Ka6 Nc8!
XIIIIIIIIY
9-+n+-+-+0
9zp-+-+-+-0
9K+P+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The right position for the knight, as someone had once showed me. The knight has to move to the rear of
the pawn to protect it. If the enemy king then wants to drive away the knight, it will move outside the
square of the pawn, leading to a loss.
Perhaps he had calculated 70...Nb5, after which White holds the draw after 71.c7 Nxc7+ 72.Kxa7.
71.Kb7 a5! 72.c7
Of course, it is essential that 72.Kxc8 can be met by 72...Kxc6.
72...Ne7
Black saves his knight, and since White cannot stop the a-pawn now, he had to resign here.
72...Nd6+ was also possible.
In the following chapters, you will be required to find the right link between the themes and the diagrams,
and store those links in your memory as ‘anchors’. The purpose is, of course, for you to expand your arsenal
of weapons drastically.
Chapter 17
Pawn endings
Pawn endings form the basis of all other endings. It is important in every endgame to make the right
assessment of the ending that remains after all the pieces have been exchanged. If it is a win, this is the
signal to start a liquidation. On the other hand, a player who is worse in an ending may try to liquidate into
a pawn ending that cannot be won by the opponent.
Many times already, we have claimed that winning a won position requires not only the right knowledge
but also the right skills. Club players who have taken the trouble to study the various endgame ‘handles’
will be able to look back with satisfaction on occasions when they succeeded to score full points by using
those ‘handles’. Therefore, in this chapter, we will show a number of aspects of pawn endings that are
useful, and sometimes even essential, for making the right decisions. We will discuss a number of common
motifs and techniques, and will sometimes show how these can emerge from various other ending types. As
we will see, quite regularly the subjects presented below are related to each other. For example, the pawn
race often occurs in pawn endings. If such a race leads to promotion to a queen by both sides, it is useful to
foresee what will happen in the queen ending after that.

§ 17.1 Key squares and opposition

Many manuals introduce the term ‘opposition’ to the reader. This involves two kings standing opposite
each other, when the obligation to move determines which of the two kings has to stand aside.

87
XIIIIIIIIY
9-+-+-mk-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Every chess player has seen this position at least once in his life. The two kings are standing opposite each
other, which means that the player who is to move draws the short end of the stick. At least, that is what
every manual teaches us. But there is an exception to this rule, as we see in this position. It does not matter
whose move it is here: White always wins by taking control of the promotion square with his king.
With White to move:
1.Ke6 Ke8
Black has the opposition but still loses!
2.f6 Kf8 3.f7
Now, the black king has to leave the promotion square, and Black cannot prevent White from queening.
3...Kg7 4.Ke7
Followed by 5.f8=Q.
1-0

And with Black to move:

1...Kg8
Already on the first move, Black has to cede the promotion square to his opponent.
2.Ke7 Kg7 3.f6+
And the pawn walks on unhindered.
So, what is this ‘opposition’ all about? We can see this in the following fragments:

88
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
If we move the pieces down one rank, the rule does apply. If White is to move, Black holds the draw with
the right defence:
1.Ke5 Ke7 2.f5 Kf7 3.f6 Kf8!
This move is essential – the other two moves lose for Black. A terrible blunder would be, e.g., 3...Ke8??,
since after 4.Ke6 Kf8 5.f7, Black’s king is driven from the promotion square and White wins: 5...Kg7 6.Ke7.
4.Ke6
And now Black does not have to concede ground:
4...Ke8 5.f7+ Kf8 6.Kf6
Stalemate:
½-½

89
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
If it is Black’s move, he has to concede ground, i.e., abandon the opposition:
1...Ke7
1...Kg7 2.Ke6 amounts to the same thing. White profits immediately by making a ‘passing move’, as we
know from football or rugby:
1...Ke7 2.Kg6 Kf8 3.Kf6
This is the principally correct method to force the win. Here, 3.f5 wins as well, for example, 3...Kg8 4.f6
Kh8 5.Kf7 (5.f7?? is stalemate) 5...Kh7 6.Ke7 Kg6 7.f7.
3...Kg8 4.Ke7
Again this ‘passing move’, taking control of the promotion square and thus nursing the pawn to the eighth
rank.
4...Kg7 5.f5 Kg8 6.f6 etc:
1-0
Opposition simply means that the king of the side whose move it is has to make way. But personally, I think
that the term ‘opposition’ is a little misleading. The suggestion is created that the side who has to ‘abandon
the opposition’ has a problem. But the fact that the stronger side’s king has to make way does not
necessarily mean that he cannot win, as we have seen in the first example above. There, it was White’s
move, allowing Black to hold the opposition, but still White managed to win. The ‘opposition’ maxim does
apply in examples No. 88 and 89.
The Steps Method (https://www.stappenmethode.nl) – developed in the Netherlands by teacher and
International Chess Master Cor van Wijgerden and psychologist/pedagogue Rob Brunia – uses the term
‘key squares’. Working from the viewpoint of the stronger side, they use the following definition: ‘If the king
can occupy a key square, this means that it will ultimately be able to reach the sixth rank in front of the
pawn’. So, what it comes down to is that if the stronger side’s king can occupy a key square with his king,
this side will always win with correct play.
So, what are these ‘key squares’? Pictures often tell us more than words:

90
XIIIIIIIIY
9-+-+-+-+0
9+-+-+{}{0
9-+{}{}{}0
9}{}-+-zP-0
9-+-zP-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The markings show the key squares for each of the pawns: a5, b5, c5 for the b3-pawn; c6, d6, e6 for the d4-
pawn, and the pawn on g5 has no less than six key squares: f6, g6, h6 and f7, g7, h7. This means that if the
white king controls one of these squares, it doesn’t matter where the black king is – White always wins with
correct play.
We have to add a marginal note here. This rule doesn’t apply in case of rook’s pawns, and we have to make
a small reservation as well: a pawn can choose whether it moves one or two squares forward on its first
move. There is also something special with a pawn on the fifth rank: it has no less than six key squares.
Now that we have a clear picture of where these key squares are situated, we can take a look at an example
from practice.

91
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-mK{}-+-+0
9+-+-+-+-0
9-+P+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White’s king is already standing on one of the key squares. This means that he wins, no matter whose move
it is.
If it is White’s move, he plays:
1.Kc4!
This move is important: thus, the king controls the three key squares, and Black has to concede decisive
ground.
Certainly not 1.c3?:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+k+-+-+0
9+{}{+-+-0
9-mK-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This move would spoil the win. There is a simple explanation for this with the help of the ‘key squares’
theory. By moving the pawn forward, White transfers the key squares to b5/c5/d5 (see diagram). Black
retains control of the key squares with 1...Kb6 and now there is nothing to be gained for White.

1.Kc4! Kb6 2.Kd5 Kc7 3.Kc5!

Making trouble for the opponent once again in the same way, as is often the best method.
3...Kb7 4.c3
 XIIIIIIIIY
9-+-+-+-+0
9+k+-+-+-0
9-+-+-+-+0
9+{mK{+-+-0
9-+-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Thus, White moves the key squares up by one rank (b5/c5/d5), and now the game of conquering ground
repeats itself:
4...Kc7 5.c4
XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-+-0
9-}{}-+-+0
9+-mK-+-+-0
9-+P+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Once again, the key squares move up by one rank! And White’s king keeps them under control since it’s
Black’s move.
5...Kb7 6.Kd6 Kc8 7.Kc6 Kd8 8.Kb7 Kd7 9.c5
And White wins.

92
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+{}{+0
9+-+-+-+-0
9-+-+KzP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
If the white king can occupy one of the key squares, this means in this case it can also reach the sixth rank.
1.Ke5!
The right move. White threatens to play 2.Ke6, occupying a key square. Black has to prevent that, but he
will still lose control of the key squares. We have seen in the previous example how White can manoeuvre
to take control of the key squares, and here we can practise it once more.
Certainly not 1.Kf5? Kf7 2.Ke5 Ke7 3.f5 Kf7;

and 1.f5? is also a blunder, due to 1...Kf6 2.Kf4 Kf7 3.Kg5 Kg7 4.f6+ Kf7 5.Kf5 Kf8 6.Kg6 Kg8 7.f7+ Kf8
8.Kf6 stalemate.

1.Ke5! Kf7
If 1...Kf8, 2.Kf6 Ke8 3.Kg7 Ke7 4.f5 Kd7 5.f6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
2.Kf5
Thus, the white king controls the three key squares, and Black has to concede decisive ground.
2...Kf8
2...Ke7 is met by 3.Kg6 Kf8 4.Kf6 Kg8 5.Ke7 Kg7 6.f5 Kg8 7.f6 Kh7 8.f7 Kg6 9.f8=Q.
3.Kf6 Ke8 4.Kg7 Ke7 5.f5 Ke8 6.f6 Kd7 7.f7 1-0

93
Study by Jan Drtina, Ceskoslovensky Sach 1908
XIIIIIIIIY
9-+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+K+-+-0
xiiiiiiiiy
White to move
Question 1: What are the key squares for this pawn?
Question 2: With which move should White start to gain control of these three key squares later on?
1.Kc2!!
1.Kd2 looks very logical, but doesn’t lead to the goal: 1...Ke7 2.Kd3.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-mk-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zPK+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Black has only one good reply here, but it draws right away!: 2...Kd7! (2...Kd6? loses to 3.Kd4 Kc6 4.Kc4).
Here, the problem for White is that whatever move he makes with the king, Black can always ‘mirror’ it
with his reply:
A) If 3.Kc4, Black plays 3...Kc6;
B) If 3.Kd4, 3...Kd6;
C) And obviously, after 3.Ke4, he can continue 3...Ke6 (3...Kc6 also leads to a draw, e.g., 4.Kd4 Kd6).
The answer to Question 1 is that b5, c5 and d5 are the key squares.

1.Kc2!!
XIIIIIIIIY
9-+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+{}{+-+-0
9-+-+-+-+0
9+-zP-+-+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
1...Ke7
Black’s king too runs as quickly as it can in the direction of the crucial squares.
Question 3: What is White’s strongest continuation here?
2.Kb3!
This is the important finesse that brings White the win. His king, as it were, walks ‘out of the shadow’ of the
black king, and this enables him to gain full control of the essential squares.
2...Kd6
Of course, Black could also continue 2...Kd7, to which White replies in the same way as in the main line:
3.Kb4! Kd6 (Black is hoping for 4.Kc4, after which he would play 4...Kc6, but of course White doesn’t fall
for this; clearly, 3...Kc6 would be met by 4.Kc4) 4.Kb5!. Now, the white king occupies a key square,
securing the win. The continuation could be as follows: 4...Kc7 5.Kc5 Kb7 6.Kd6 Kc8 7.Kc6 Kd8 8.c4
(8.Kb7 Kd7 9.c4 Kd6 10.Kb6 Kd7 11.c5 Kc8 12.Kc6) 8...Kc8 9.c5 Kd8 10.Kb7 etc.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+{}{+-+-0
9-+-+-+-+0
9+KzP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Question 4: There is one more important decision White has to make here. What should he play?
3.Kb4!
Gaining the opposition, and with it the full point.
3...Kc6
If 3...Kc7, White obviously plays 4.Kc5.
4.Kc4 1-0
Mission accomplished.

§ 17.2 Shouldering the enemy king

An important endgame technique is keeping the enemy king at bay by obstructing it with your own king.
This occurs in various endgame types, and certainly also in pawn endings, which is why we discuss it here
as well. In several books, I have come across the term ‘shouldering’, or ‘shoulder budge’ or ‘bodycheck’, all
of which indeed provide insight into what this technique is about.

94
XIIIIIIIIY
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9zP-zpk+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Superficially, this position seems to be hopelessly lost for White. The black king is inside the square of the
white pawn, while the black pawn cannot be stopped in any way. But still, salvation is at hand!
1.Kd7!
By cutting off the black king, White can use his a-pawn anyway. Because the black c5-pawn is in the way at
the moment, Black cannot attack the white pawn.
After 1.a6?? Kc6!, the struggle would have been decided in Black’s favour.

1.Kd7! c4
The pawn has to run, but it’s too late already.
2.a6
The white pawn will also queen, and what is more, since the white king controls the essential c6-square, it
cannot be stopped. And thirdly, White will queen with check and even win!
2...c3
2...Kc5 is nonsense in view of 3.a7.
3.a7 c2 4.a8=Q+ Kd4
XIIIIIIIIY
9Q+-+-+-+0
9+-+K+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+p+-+-+0
9+-+-+-+-0
xiiiiiiiiy
What was it again with the bishop’s pawn? It was a draw, wasn’t it?
5.Qa1+!
No – not in this case. The queen reaches the promotion square of the pawn, and Black can forget about the
draw as well.
5...Kd3 6.Qc1 1-0
5.Qh1 Kd3 6.Qc1 also wins.

95
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-mk0
9+-+PmK-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
In this fragment, again, we have a race between two pawns, and Black’s pawn is even one square closer to
the promotion square than White’s. But here too, shouldering is the tune.
1.Kf4!
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+-mK-mk0
9+-+P+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
An excellent move, giving his opponent a firm shoulder budge. White not only prevents 1...Kg5 (which
would secure the draw), but also keeps Black’s king in an unfortunate position at the edge of the board.

Certainly not 1.d4? Kg5, with an immediate draw.

1.Kf4! Kh3 2.d4!

2.Kf3? makes no sense, since Black can force a draw with 2...h4 3.d4 Kh2 4.d5 h3 5.d6 Kg1 6.d7 h2 7.d8=Q
h1=Q+.
2...h4 3.d5
Also here, keeping the black king tied to the edge of the board is too slow: 3.Kf3? Kh2 4.d5 h3 5.Kf2 Kh1
6.d6 h2 7.d7 stalemate! (or 7.Kg3 Kg1 8.d7 h1=Q 9.d8=Q Qg2+, with a draw).
3...Kg2 4.d6 h3 5.d7 h2 6.d8=Q h1=Q
XIIIIIIIIY
9-+-wQ-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+q0
xiiiiiiiiy
The white pawn has queened first, and since Black’s king is very clumsily placed, the ensuing ending is lost
for him.
7.Qd2+ Kg1
7...Kh3 does not bring any relief in view of 8.Qd3+ Kg2 9.Qe2+ Kh3 10.Qg4+ Kh2 11.Qg3#; 7...Kf1
amounts to approximately the same thing: 8.Qd1+ Kg2 9.Qe2+ (9.Qf3+? Kg1 10.Qd1+ Kh2) 9...Kg1
(9...Kh3 10.Qg4+ Kh2 11.Qg3#) 10.Kg3, and White wins.
8.Kg3!
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-mK-0
9-+-wQ-+-+0
9+-+-+-mkq0
xiiiiiiiiy
Domination! White’s king, as it were, pushes its opposite number to the edge of the board, and Black cannot
prevent being checkmated.
We will take a look at a few other cases in exercise form.

96
Study by Josef Moravec
Ceskee Slovo 1940
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+p+-+k0
9-+-+-+-+0
9+K+-+-+-0
9-+-+P+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
1.Kb4!
Not 1.Kc3? Kg5 2.Kd4 Kf4, and things go wrong for White as after 3.Kxd5 Ke3, he loses his pawn.

1.Kb4! Kg5 2.Kc5 Kf4 3.Kd4!

Now, the same position arises as in the previous variation, but this time it’s Black’s move!
3...Kf5 4.Kxd5 1-0

97
Study by Josef Moravec
Shakhmaty v SSSR 1947
 XIIIIIIIIY
9-+-+-+-+0
9+p+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+K+-+-0
9P+-+-+-+0
9+-+-+-mk-0
xiiiiiiiiy
White to play and win
The solution is highly paradoxical!
1.Ke2!
An unexpected move! White’s king has to keep its opposite number at a distance first, before he can set off
to win the enemy pawn. In the meantime, White gains time to push his own pawn.
Not 1.a4? in view of 1...Kf2, and Black holds the draw, for instance: 2.Kd4 Ke2 3.Kc5 Kd3 4.Kb6 Kc4 5.a5
Kb4, and the black king is right on time to ‘cling to’ White’s pawn.

1.Kc4? Kf2 leads to the same.

1.Ke2! Kg2 2.a4 Kg3 3.Ke3!

Giving the opponent the ‘shoulder budge’.
3...Kg2
Black tries to reach the rear side of the pawn, but it won’t bring him any benefit this time.
Neither does 3...Kg4 help: 4.Ke4 Kg5 5.Ke5 Kg6 6.Kd6, and Black loses.
4.Kd4 Kf3 5.Kc5 Ke4 6.Kb6 Kd4 7.Kxb7 Kc5 8.a5
Black is too late: 1-0

98
Study by Samuel Boden
source unknown, 1851
XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+-0
9-+-+-+-+0
9+-+k+-+-0
9-+-+-zP-zP0
9+p+-+-+-0
9pmK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
White is a pawn down, and Black even has two connected passed pawns, but still he is going to draw the
short end of the stick. How?
1.h5!
The most distant pawn has to be set in motion first.
1...Ke6
The tricky try 1...Kc4 2.h6 Kd3 3.h7 a1=Q+ 4.Kxa1 Kc2 just barely fails; White is exactly on time with his
promotion, preventing the lethal push of Black’s b-pawn: 5.h8=Q.
2.h6 Kf6 3.f5
And the second pawn brings about zugzwang! 1-0

99
Study by Oscar Carlsson
Finales y Temas 2009
XIIIIIIIIY
9-+-+-+-+0
9+-zp-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-+-+P+0
9+-mk-+-+-0
xiiiiiiiiy
White to play and win
A difficult case. Should White start running with the pawn or should he do something with his king first?
1.Ke3!
A) Funnily enough, straightforwardly running with the pawn only leads to a draw: 1.g4? c5, and now:
A1) Using the king no longer helps: 2.Ke4 Kd2! (now, Black is ready for ...c7-c5-c4-c3) 3.Kd5 Kd3, and this
enables Black to play both ...c5-c4 and next, ...Kd3-e4: 4.Kxc5 Ke4 – a Réti-like manoeuvre. We will see
later what this entails;
A2) 2.g5 c4, and the position is a draw: 3.g6 c3 4.g7 c2 5.g8=Q Kb2, and here we refer you to the
explanation given in Chapter 10, section B.

B) The direct 1.Ke4 fails to 1...Kd2! 2.Kd4 c5+! 3.Kxc5 Ke3, eliminating the g-pawn.

1.Ke3! c5 2.Kd3!
Thus, White not only stops the pawn but also shoulders off the enemy king.
2...Kb2 3.Kc4! 1-0
3.g4? Kb3=.

§ 17.3 Zugzwang
The term ‘zugzwang’ indicates a situation in which any move a player makes will significantly worsen his
position. The obligation to move forces him to make one, often with a losing position as a result. In pawn
endings – but also in other endings – this theme occurs on a regular basis.
Let’s have a look at a quite common type of position, presented in the form of an exercise.
100
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-mk-zp-+-+0
9+-+P+K+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
We assume that this position is well-known.
1.Kf6!
After 1.Ke6??, White himself will fall victim to zugzwang: 1...Kc5, and he has to let go of the d5-pawn and
loses the resulting ending, for example: 2.Kf5 Kxd5 3.Kf4 Kc4 4.Ke3 d5 5.Kd2 Kd4, and according to the
‘key squares’ theory, Black obtains a winning position.

1.Kf6! Kb5
Moving back with the king makes no sense; after 1...Kb7 2.Ke6 Kc7 3.Ke7, Black also loses the pawn, and
again White forces the ‘standard winning position’: 3...Kc8 4.Kxd6 Kd8 5.Ke6 Ke8 6.d6 Kd8 7.d7 Kc7
8.Ke7+–.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-zp-mK-+0
9+k+P+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
2.Ke7!
White is exactly on time to attack Black’s pawn in the right way – otherwise, Black would do the same with
2...Kc4!.
Very stupid would be 2.Ke6?? in view of 2...Kc5–+.
2...Kc5 3.Ke6
And Black ends up in zugzwang. Due to the obligation to move, he has to let go of the d6-pawn and
therefore loses:
3...Kb6 4.Kxd6 Kb7 5.Ke7 Kc8 6.d6 1-0
Here is a study by Edward Lasker showing us another case.

101
Study by Edward Lasker
Schachstrategie 1911
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+k+-+0
9+-zp-+-+-0
9-zpP+-+-+0
9+P+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
1.Ke4!
Black is in zugzwang. The study ends here, but we will show you a little more of the conversion:
1...Kd6
After 1...Kf6, the white king enters with 2.Kd5 and conquers two pawns.
2.Kf5
White’s king starts an enveloping manoeuvre.
2...Kd7 3.Ke5 Kc6 4.Ke6 Kc7 5.Kd5 Kb6 6.Kd6
This is the second zugzwang. As soon as the c5-pawn is lost, the other pawn also goes: 1-0
A little more complicated is the next study. Zugzwang is accomplished by some complex tempo play on
what are called ‘corresponding squares’ in chess literature.

102
Study by Frantisek Dedrle
Ceskoslovensky Sach 1921
XIIIIIIIIY
9-+-+-+K+0
9+-+-+-+-0
9k+-+-+-+0
9+-+-+-+-0
9-+-zp-+-+0
9+-+-+-+-0
9-+P+P+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
This study comprises only one move for Black and one for White. But we need some more explanation.
1...Ka7
Black tries to outsmart his opponent, but he is going to be treated to a case of ‘distant opposition’. What
does this entail? Actually, this is a distant zugzwang, according to what is also called ‘the theory of the
corresponding squares’.
If 1...Kb7, 2.Kf7!.

Optically, the most logical move is 1...Kb5 to gobble up White’s pawns as quickly as possible. But White’s
king can run fast as well, and after 2.Kf7 Kc4 3.Ke6 Kc3 4.Kd5, it is right on time to snatch Black’s last
pawn: 4...Kxc2 5.Kxd4+–.
1...Ka7 2.Kg7!
The only move that wins! This situation is called the ‘distant opposition’. Please note that there is an odd
number of squares between both kings. We will take this a few moves further than the composer did. For
example, 2.Kf7 would not yield White more than half a point after 2...Kb7 3.Kf6 Kb6 4.Ke6 Kc6 5.Ke5
Kc5 6.Ke4 Kc4, and Black holds the draw.
2...Ka6
2...Kb7 3.Kf7 Ka6 4.Kg6. Here, 4.Ke6 is possible too! As long as the white king remains on the same colour
square as the black king.
3.Kg6 Ka5 4.Kg5 Kb4 5.Kf4 Kc5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mk-+-+-0
9-+-zp-mK-+0
9+-+-+-+-0
9-+P+P+-+0
9+-+-+-+-0
xiiiiiiiiy
6.Ke5! Kc4
And now the ‘distant opposition’ game has to be discontinued:
7.Kd6!
Conquering the d4-pawn by means of an enveloping manoeuvre with the king.
7...Kc3 8.Kd5 1-0
In the above example, we encountered the term ‘distant opposition’. We can define this as follows:

Rule of thumb: ‘distant opposition’

A player is said to have the distant opposition if there is an odd number of squares between
both kings, and the enemy king has to make a move. In principle, the kings are standing on the
same coloured squares. This ‘distant opposition’ can be either on a rank, a file or a diagonal.
The obligation to move forces one of the kings to concede ground.

This endgame in practice

Here is a fragment from my own practice.

103
Mehdi Aithmide Mohamed 2107
Herman Grooten 2365
Gibraltar 2006 (4)
 XIIIIIIIIY
9-mk-+-+-+0
9+-+-+-+-0
9-+-+-+-zp0
9zP-zP-+-+P0
9-+-+-zp-+0
9+-+-zpp+-0
9-+-+-+-+0
9+-+-mK-+-0
xiiiiiiiiy
Black to move
There is a tense fight between two pawn groups; Black has three pawns (albeit with one doubled pawn)
and White has two separate ones. Zugzwang is looming...
57...f2+
Now my opponent had two alternatives, and I had to react differently depending on his choice.
58.Ke2
After 58.Kf1, Black has to find a way to put his opponent in zugzwang again. He can do this, for example,
with 58...Kc7! (but 58...Kb7 59.Ke2 f3+ 60.Kf1 Kb8 61.a6 Ka7 also wins for Black), when after 59.a6 f3,
both white pawns will be lost again, e.g. 60.a7 Kb7 61.c6+ Kxa7 62.c7 e2+! (first remove the stalemate!)
63.Kxf2 Kb7, and White wins.
58...f3+ 59.Kf1
XIIIIIIIIY
9-mk-+-+-+0
9+-+-+-+-0
9-+-+-+-zp0
9zP-zP-+-+P0
9-+-+-+-+0
9+-+-zpp+-0
9-+-+-zp-+0
9+-+-+K+-0
xiiiiiiiiy
The white king is pinned down to the back rank, but now Black has to determine how to control the white
pawns. He can manage this with
59...Kb7!
putting White in zugzwang. White now has to push one of his pawns, leading to the loss of both. Black has
to avoid stalemate along the way.
60.c6+ Kxc6 61.a6 e2+!
And White surrendered. 61...Kb6 62.a7 e2+! also sufficed. After White takes on f2, the stalemate is gone
and the a-pawn is lost.
It is a marvellous thing when as an unknown player you can obtain immortality with one single move. This
is what happened to Fernando Saavedra, a mediocre player who saw his name forever connected to a
certain endgame motif when he made a fantastic discovery in 1895:

104
Georges Barbier & Fernando Saavedra
Glasgow Weekly Citizen 1895
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-mKP+-+-+0
9+-+r+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9mk-+-+-+-0
xiiiiiiiiy
Here, the original stipulation was: White to play and win.
1.c7 Rd6+ 2.Kb5
It is a draw after 2.Kb7? Rd7, while if 2.Kc5?, 2...Rd1 and Black can give up the rook for the pawn.
2...Rd5+ 3.Kb4 Rd4+ 4.Kb3
4.Kc3 Rd1 5.Kc2 leads to the same position.
4...Rd3+ 5.Kc2
This is the solution that Georges Barbier published in his newspaper The Glasgow Weekly Citizen in 1895.
XIIIIIIIIY
9-+-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+r+-+-0
9-+K+-+-+0
9mk-+-+-+-0
xiiiiiiiiy
But then someone came up with the idea that Black has a very neat defence here:
5...Rd4!
Initially, everybody thought that White had to settle for a draw now due to the stalemate trick 6.c8=Q?
Rc4+ 7.Qxc4. However, the Spanish priest Fernando Saavedra put his oar in with the following move:
6.c8=R!
A brilliant idea that has carried Saavedra’s name ever after. With this underpromotion, White avoids
stalemate and at the same time threatens to give mate on the a-file. Black’s reply is forced:
6...Ra4
And now the cat’s out of the bag:
7.Kb3!
With the double threat of 8.Kxa4 and 8.Rc1#.
7...Ra7 8.Rc1#
It’s time to present to you a special case, with zugzwang as the apotheosis in a simultaneous game. There is
also a nice story attached to it!
A stone’s throw from the city of Eindhoven, in a village called Best in the Dutch province of Brabant, lived a
chess journalist called Jules Welling (1949-2016) – no relative of the well-known International Master
Gerard Welling from Eindhoven, by the way. Jules was a club-level player whose GP had told him not to
play competition games because it was bad for his health. But if there was a grandmaster giving a
simultaneous exhibition in the area, Jules would always be present. Especially during the famous Interpolis
tournaments in Tilburg in the 1980s and 1990s (where he was a very frequent visitor), he would enter the
lists against many a grandmaster. Against one of them, Ludek Pachman, Jules invented a motif that in my
opinion puts him on a par with the famous Saavedra. Unfortunately for Jules, a mystification took place that
did him grievous injustice.

105
Ludek Pachman
Jules Welling
Tilburg sim 1973
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mKP0
9-+-trk+-+0
9+-+-+-+-0
9p+-+-+-+0
9zP-+-+-+-0
9-zP-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
Black to move
The reader will be surprised to see a rook on the board (we were talking about pawn endings, weren’t we?),
but it’s going to get sacrificed in an artful way!
1...Rd7+ 2.Kg6 Rd8 3.Kg7
XIIIIIIIIY
9-+-tr-+-+0
9+-+-+-mKP0
9-+-+k+-+0
9+-+-+-+-0
9p+-+-+-+0
9zP-+-+-+-0
9-zP-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
Black is a rook down, but a win seems impossible. However, Welling now sacrifices his rook to enter a
pawn ending in which he is no less than three pawns down!
3...Rh8!!
An incredibly beautiful idea based on zugzwang. Without this rook sacrifice, Black cannot win.
4.Kxh8
White could also wait with the acceptance of the rook sacrifice, but it wouldn’t have brought him anything.
After 4.h4, Black waits patiently: 4...Ke7 5.h5 Ke8 6.h6 Ke7, and now White has to take: 7.Kxh8 Kf7, with
the same continuation as in the game.
White can also refuse the rook sacrifice with 4.Kg6, but this is met by 4...Ke7 5.Kg7 Ke8, and now the king
can no longer move to g6 in view of 6...Kf8, so White is forced to make pawn moves. Black’s king keeps
commuting between e7 and e8 until White has run out of pawn moves... 6.h3 Ke7 7.h4 Ke8 8.h5 Ke7 9.h6
Ke8, and now he has to take!
4...Kf7 5.h3 Kf8 6.h4 Kf7 7.h5 Kf8 8.h6 Kf7
 XIIIIIIIIY
9-+-+-+-mK0
9+-+-+k+P0
9-+-+-+-zP0
9+-+-+-+-0
9p+-+-+-+0
9zP-+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Zugzwang – a cruel fate for the White player...
9.b4 axb3 10.a4 b2 11.a5 b1=Q 12.a6 Qa1#
Certainly an idea to be proud of! And Jules was not above such human feelings. He wrote about it
extensively in De Raadsheer, the club bulletin of the Eindhoven Chess Club. But his fate was cruel. The
fragment was included as a special endgame in Chess Informant, a prestigious chess periodical in book form
in which every six months (later this became four times a year) the best games of the period appeared,
complete with analyses. However, in that special endgame chapter, if I remember well, the colours of the
players were reversed. Apparently the authors had assumed that it had to be the grandmaster who had
won this game so brilliantly! But Jules had his pride, and I think he put in quite a bit of effort to get the
mistake rectified. I cannot remember whether he succeeded in this, but I think that he should be
immortalized in the chess world with the term ‘the Jules motif’.

This endgame in practice

Now I would like to present to you a fragment from a game that took place between two club players I have
(had) under my wings as pupils in different ways. By coincidence, they faced each other in a match, playing
for their respective club teams.
After a hectic game with chances going to and fro, the following position arose.

106
Armin Kohlrausch 1715
Jan Otten 1678
Tilburg tt 2021
XIIIIIIIIY
9-+-mk-+-+0
9zpp+-sn-+-0
9-+p+-+p+0
9+-+-+r+-0
9-+-zP-vLK+0
9+-+-+-zP-0
9PzP-+R+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Black had been a little better, but at this moment his pieces are clumsily placed. The king has moved from
d7 to d8, and the knight would prefer to be on d5. White could exploit this here by introducing a deadly pin:
43.Bg5
The game saw 43.Rh2? and later White even lost. Of course, the White player had seen the pinning move,
but he didn’t realize it would have given him a decisive advantage.
The reply
43...Rf7
would have been forced, after which White could liquidate into a pawn ending:
44.Rxe7! Rxe7
XIIIIIIIIY
9-+-mk-+-+0
9zpp+-tr-+-0
9-+p+-+p+0
9+-+-+-vL-0
9-+-zP-+K+0
9+-+-+-zP-0
9PzP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White had envisaged this position, but he hadn’t seen what use it would have been to him. However, here
he could have gained a crucial tempo with
45.Kf4!
since Black has to ‘lose’ a move to get out of the pin. Indeed, 45.Bxe7+? brings White nothing, as he had
rightly concluded. After 45...Kxe7 46.Kg5 Kf7, the position is a draw.
45...Kd7 46.Bxe7
Now White can liquidate.
46...Kxe7
The White player said afterwards: ‘I had seen this trade, but I only saw 46...Kxe7 47.Kg5 here, when Black
can just play 47...Kf7, and White doesn’t achieve anything. Therefore I played differently, but it didn’t win.’
47.Ke5!
XIIIIIIIIY
9-+-+-+-+0
9zpp+-mk-+-0
9-+p+-+p+0
9+-+-mK-+-0
9-+-zP-+-+0
9+-+-+-zP-0
9PzP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White’s conclusion is correct, but with two kings standing opposite each other, the question is which one of
them has to give way. So, it all comes down to the available pawn moves.
White has possession of the mighty central square for his king. Now, Black will slowly but surely lose on
tempo. White has many pawn moves left to put his opponent in zugzwang.
Let’s select a possible continuation here:
47...a6 48.g4 a5 49.b3 b6 50.g5 b5 51.a3 b4 52.a4
Now, Black’s king has to give up the opposition, allowing its opposite number to walk in triumphantly;
White wins.

107
 XIIIIIIIIY
9-+-+-+-+0
9mK-mk-+-+-0
9P+P+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Here, the zugzwang is mutual: whoever is to move, has bad luck. White to move can only commute between
the a8- and a7-squares, and Black’s king cheerfully trails along. If it is Black’s move, there follows:
1...Kc8
Or 1...Kxc6 2.Kb8+–.

1...Kc8 2.Kb6 Kb8 3.a7+

Or 3.c7+ Kc8 4.a7 Kd7 5.a8=Q+–.
3...Ka8 4.Kc7 1-0
XIIIIIIIIY
9k+-+-+-+0
9zP-mK-+-+-0
9-+P+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Rule of thumb: ‘reserve tempi’
If two kings are standing opposite each other, all that matters is which of the two has to give
way. Therefore, everything depends on the remaining pawn moves on both sides. It is
important to have a number of tempi left at the end (so-called ‘reserve tempi’), and therefore
both sides have to be very careful with pushing their pawns.

We conclude with two exercises:

108
Study by Domenico Ponziani
Giuoco Incomparibile 1769
 XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-zP-+0
9+-+-+-+P0
9pzpp+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9mK-+-+-+-0
xiiiiiiiiy
White to play and win
1.h6 Kg8 2.Kb1! b3
If 2...a3 3.Ka2 c3 4.Kb3 or 2...c3 3.Kc2 a3 4.Kb3, Black is in zugzwang in both cases.
3.Kb2
Zugzwang! Black loses all three of his pawns.
1-0

109
Study by Walter Bähr
Opposition und kritische Felder 1936
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-zp0
9+-+pzP-+P0
9-+-+-+-+0
9+K+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
1.Kb4!
White has to be able to attack the d5-pawn in the right way.
1.Kc3? is not correct here: 1...Kf7 2.Kd3 (2.Kd4 Ke6) 2...Ke7, and Black is ready to meet 3.Kd4 with
3...Ke6, after which he would even win!

1.Kb4! Kf7 2.Kc5

This forces Black to play the following move:
2...Ke6 3.Kd4
Thus, Black lands in zugzwang: 1-0

§ 17.4 Triangulation

110
Study by Hans Fahrni
Das Endspiel im Schach 1917
 XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-+-0
9p}P}-+-+0
9zP-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White would like to conquer either of the squares b6 or d6 with his king, but the black king cuts him off.
How can White win here?
Essential for the understanding of this problem is that White wants to reach this position with Black to
move. For this purpose, he has to ‘triangulate’ his king. In other words, he wants to lose a tempo on purpose
to realize his goal. To completely understand how this works, we will here present the theory of
‘corresponding squares’ in a nutshell, using a ‘piece of scrap paper’ on which we fill in a few numbers. It’s
just like mathematics!

We can simply read the problem from this diagram. In the first diagram, White’s king is on c5, and then
Black’s king must be on c7 to keep its opposite number at bay. So if the white king moves to that ‘1’ square
at any moment, Black’s king has to move to the corresponding square ‘1’ to keep the white king away from
the squares numbered ‘2’. There are also squares numbered ‘3’ to which the kings can move (within their
own lines). Now, the trick is that as soon as the white king stands on one of the numbered squares, the
black king has to stand on a square with the same number to keep the balance. The numbers in the diagram
border one another, and now the trick for White is to step on a different square to be able, after that, to
choose the type of square it moves to. These two squares are indicated with the number ‘4’. Black’s king
does not have such ‘No. 4’ squares at its disposal, and so it has to move to a different type square. And this
is exactly what White wants! After this has happened, his king moves to the corresponding square after
every Black move, until he wins the game on tempo.
Expressed in moves, the win can be forced as follows:
1.Kd5
The white king goes to a ‘No. 3’ square, so Black does the same:
1...Kc8
This position is essential for the understanding of this problem. White wants to reach this position with
Black to move. For this purpose, he has to ‘triangulate’ with his king. In other words, he wants to lose a
tempo on purpose to achieve his aim.
2.Kd4!
So White says goodbye to the numbered squares and steps on a square marked ‘4’. Now Black, with great
regret, has to move to a numbered square, and then White will know exactly how to manoeuvre his
opponent into trouble.
2.Kc4 leads to the same goal: 2...Kd8 3.Kd4 Kc8 4.Kd5. What matters is that this position is reached with
Black to move.
2...Kd8
The black king is on ‘2’ now, but it is precisely this square type (d6 or b6) that White cannot reach at the
moment. So his king ‘triangulates’, giving up a tempo on purpose to force Black to move to a square of the ‘3’
or ‘1’ type.
3.Kc4 Kc8
A ‘3’ type square it is then!
4.Kd5!
XIIIIIIIIY
9-+k+-+-+0
9+-+-+-+-0
9p+P+-+-+0
9zP-+K+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This is also a ‘3’ square. Now, Black has a choice between three different moves, all three of which are
losing. After all, from this ‘3’ type square White has a choice between moving to a ‘1’ or a ‘2’ square.
4...Kc7
Black opts for ‘1’, the square on which the king started in this study.
He could have opted for a ‘2’ type square in two ways:
A) 4...Kd8 (‘2’) 5.Kd6 (also ‘2’) 5...Kc8 6.c7, which means that White drives the enemy king away from the
promotion square. This is winning for White, but he has to evade stalemate as follows: 6...Kb7 7.Kd7 Ka7
8.Kc6!+–. White should certainly not promote to a queen or rook, as it would then be stalemate!;
B) 4...Kb8 also doesn’t make it hard for White to win along the following route: 5.Kd6 Kc8 6.c7 Kb7
7.Kd7+–.
5.Kc5
 XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-+-0
9p+P+-+-+0
9zP-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
To his horror, now Black has the move. This time he really has to concede ground: White invades on b6 and
thus conquers the second pawn.
5...Kc8 6.Kb6
White has done it! His king penetrates and conquers the a6-pawn, after which it is definitively over.
6...Kb8 7.Kxa6 Kc7 8.Kb5
And White is winning.
Certainly not 8.Ka7?? since it is a draw after 8...Kxc6 9.a6 Kc7.
So, this is ‘just’ endgame theory. We can refer to the concept of ‘distant opposition’ that we already touched
upon earlier in this chapter. Ergo, triangulation only works if the defending side is not able to ‘move along’
with the opponent owing to certain squares not being accessible. In the above example, the b7- and d7-
squares are taboo for the black king.

This endgame in practice

Positions like the above can arise in real games, even by great players. See the following fragment, with –
rather humorous – commentary by Garry Kasparov.

111
Lev Alburt
Garry Kasparov
Daugavpils 1978
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+k+-zp0
9+-+-+p+P0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
Black to move
Kasparov: ‘What now? The endgame is not simple. I need to gain the opposition.’
55...Kf5
‘This is the key. I am going to lose some time so that I can switch the player who has to move. Now I can set
up a situation similar to that of the previous note, but with an important difference.
Look at the variation that will take place if I go to f4: 55...Kf4? 56.Kf2 Ke4 57.Kf1 Ke3 58.Ke1 f2+ 59.Kf1
Kf3 – oops! It is stalemate!’
55...Kf5 56.Kg1 Ke5 57.Kf1 Ke4
‘We have already seen this position, but now it is White to move. My opponent, a future champion of the
United States, was a good enough endgame player to realize that resistance was useless, and he resigned,
not waiting for 58.Ke1 (58.Kf2 Kf4 59.Kf1 Kg3 60.Kg1 Kxh3 61.Kf2 Kg4) 58...Ke3 59.Kf1 f2 60.Kg2 Ke2
61.Kh2 Kf3 62.Kh1 f1=Q+ 63.Kh2 Qg2#.’
Here is another position in which White has to apply the theory of corresponding squares and triangulate
with his king.

112
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+p+-+-+0
9mk-zp-mK-+-0
9p+P+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
To win, White has to conquer the c6-pawn. This can only be done with:
1.Kf5!

Actually, this is a strange move at first sight if you don’t know why it is played. Black loses the distant
opposition because the b5-square is not accessible to him, and Black’s king cannot move along with White’s
king.

The direct 1.Kf6? doesn’t work due to 1...Kb6 2.Ke6 Ka6!. This is a case of ‘distant opposition’, this time
favouring the defender: 3.Kd7 Kb7 4.Kd8 Kb8 5.Ke7 Ka7 6.Kd6 Kb6, with a draw since White cannot
make any progress.

1.Kf5! Kb6
1...Ka6 would abandon the opposition without a fight (the number of squares between the kings is even):
2.Ke6 Ka7 3.Ke7! Ka8 (3...Ka6 4.Ke6 Kb7 5.Kd7 Kb6 6.Kd6+–) 4.Kd6 Kb7
 XIIIIIIIIY
9-+-+-+-+0
9+k+-+-+-0
9-+pmK-+-+0
9+-zp-+-+-0
9p+P+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
5.Kd7. The simplest solution, analogous to the main line. Here, 5.Kxc5 also wins since the a4-pawn will be
lost, for example: 5...Kc7 6.Kb4 Kb6 7.c5+ Ka6 8.Kxa4 Kb7 9.Kb3 Ka6 10.Kb4 Kb7 11.Kc4 Ka6
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9k+p+-+-+0
9+-zP-+-+-0
9-+K+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
12.Kd3! (another continuation that is extremely hard to find – the king moves towards d6 but it doesn’t
take the shortest route; 12.Kd4?? Kb5!) 12...Ka5 13.Ke4 Ka4 14.Ke5 Kxa3 15.Kd6, and Black is too late.
2.Kf6
Now, White seizes the distant opposition.
2...Kb7 3.Kf7 Kb6
Or 3...Kb8 4.Ke6 Kc7 5.Ke7 Kb7 6.Kd7 Kb6 7.Kc8 and again, the ‘enveloping manoeuvre’ works.
4.Ke8!
XIIIIIIIIY
9-+-+K+-+0
9+-+-+-+-0
9-mkp+-+-+0
9+-zp-+-+-0
9p+P+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Also here, the white king embarks on an enveloping manoeuvre.
4...Ka7 5.Ke7!
The distant opposition.
5...Ka8 6.Kd6 Kb7 7.Kd7 Kb6 8.Kc8
And White will gobble up the c6- and c5-pawns, for example:
8...Ka6 9.Kc7 Ka7 10.Kxc6 Ka6 11.Kxc5 Ka5 12.Kd6
And the c-pawn decides the issue.
1-0

113
Study by Ludwig Rödl
Endspiele 1936
XIIIIIIIIY
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-+-+0
9zp-+-+-zp-0
9P+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
xiiiiiiiiy
White to play and win
1.Kd2
Now that we are familiar with the motif of ‘distant opposition’, we know how to handle such positions. We
put our king on the same colour square as the enemy king, keeping an odd number of squares between
both kings (in this case, five).
1...Ke7
And so we meet 1...Kd7 with 2.Kd3.
2.Ke3 Kd7 3.Kd3 Kd6
If Black chooses another square, we have to be accurate with our reply: 3...Kc6 4.Kc4! Kd6 5.Kb5 Ke5
6.Kxa5 Kf4 7.Kb5 Kxg4 8.a5, and White gets there first and therefore wins; or 3...Ke6 4.Ke4! Kf6 (4...Kd6
5.Kf5 Kc5 6.Kxg5 Kb4 7.Kf5 Kxa4 8.g5 and again, White is easily on time) 5.Kd5 Ke7 6.Kc6 Ke6 7.Kb5
Ke5 8.Kxa5 Kf4 9.Kb5, and White wins.
4.Kd4 Ke6
Now we have to watch out.
5.Ke4
Maintaining the opposition is best here.
Not 5.Kc5? because then White will need just too much time to win, viz. 5...Ke5 6.Kb5 Kf4 7.Kxa5 Kxg4
8.Kb5 Kh3 9.a5 g4 10.a6 g3 11.a7 g2 12.a8=Q g1=Q=.
5...Kd6
Black has to concede decisive ground, and so the struggle is decided.
Also after 5...Kf6 6.Kd5 Ke7 7.Ke5 Kf7 8.Kf5 Kg8 (8...Kg7 9.Kxg5+-) 9.Kxg5 Kg7 10.Kf5 Kf7
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9zp-+-+K+-0
9P+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
White is exactly on time, e.g. 11.Ke5 Kg6 12.Kd5 Kg5 13.Kc5 Kxg4 14.Kb5 Kf5 15.Kxa5 Ke6 16.Kb6
Kd7 17.Kb7. We will encounter this type of position further on when we discuss Bähr’s position.
6.Kf5 Kc5 7.Kxg5 Kb4 8.Kf5
Here, it doesn’t make any difference, but it is never smart to put the king on the diagonal on which the
enemy pawn might promote.
8...Kxa4 9.g5 Kb4 10.g6 a4 11.g7 a3 12.g8=Q
And White wins.

§ 17.5 Various aspects of the passed pawn

• The square
Every beginning chess player will, in one way or another, be confronted with the idea of the square of the
passed pawn in one of his games. This term has been mentioned a few times already in this book, since it
should be part of the elementary knowledge of every club player. Nevertheless, it is useful to dive a little
deeper into this matter. Sometimes, it is of crucial importance to be able to see quickly whether a king can
stop an enemy pawn’s promotion or not. We will repeat the basic knowledge of the square one more time.

114
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
9P+-+-+-zp0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Ke3!
If White pushes his pawn with 1.a5?, the square is moved up, and the new corner squares are a5/a8/d8/d5.
Black’s king could make it to that square if he wants, but in this position he has a much stronger reaction:
1...h3.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+k+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+{+p0
9-+-mK-+-+0
9+-+-+{+{0
xiiiiiiiiy
analysis diagram
Now, the corner squares of the black pawn’s square are h3/h1/f1/f3, which means that the white king can
no longer stop the pawn: 2.a6 h2 3.a7, and it is clear that Black is just a little quicker: 3...h1=Q, and wins.

1.Ke3!
XIIIIIIIIY
9{+-+{+-+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
9P+-+{+-zp0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-}-+{0
xiiiiiiiiy
By moving inside the square of the h4-pawn, White avoids the threatening loss in any case.
1...h3 2.Kf3
With a draw.
Especially when you are in time-trouble, it is important to be able to determine in a flash whether an enemy
passed pawn can be stopped by your king or not. It doesn’t always work well, for example, if one of the
sides can ‘interrupt’ the enemy king’s route, as in the following position:

115
Interference
Constructed position
 XIIIIIIIIY
9-+-+-+-+0
9+-+p+-+-0
9-+-+-+-+0
9+-zP-zp-+-0
9P+-+k+-+0
9+-+-+-+-0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
White to move
In this position, the black king is inside the square of White’s a-pawn (the corner points being
a4/a8/e8/e4). But White has a way to cause havoc inside this square.
1.c6!
XIIIIIIIIY
9{+-+{+-+0
9+-+p+-+-0
9-+P+-+-+0
9+-+-+-+-0
9P+-+k+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
After 1.a5 Kd5 2.a6 Kc6 3.Kg2, it seems that White can win after all, but Black is in the nick of time to
salvage the half point. First, he has to collect the a6-pawn, and then he has to return in time to ‘cling on’ to
the c5-pawn. He succeeds with 3...Kc7 4.Kf3 Kb8 5.Ke4 Ka7 6.Kxe5 Kxa6 7.Kd6 Kb5, with a draw.

1.c6! dxc6
Blocking the black king’s route to b7.
2.a5 Kd5 3.a6
XIIIIIIIIY
9{+{+-+-+0
9+-+-+-+-0
9P+p+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
And now Black can’t enter the c6-square since he has his own pawn there: White wins.
As a 15- or 16-year-old schoolboy, I became a member of the local chess club and started playing in the
seniors section. On a certain evening, I witnessed a curious scene. Two very old gentlemen were finishing
their game. I stood there watching with increasing amazement:

116
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-+-0
9-mk-+-+-zp0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White started with
1.Kh7
which was of course met by
1...h5.
I was expecting White to resign here, but no – he continued:
2.Kh6 h4 3.Kh5 h3
With increasing annoyance, White hurled his moves on the board:
4.Kh4 h2 5.Kh3
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-mk-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
9-+-+-+-zp0
9+-+-+-+-0
xiiiiiiiiy
At this moment, Black started looking in the box for a queen. When he had found it and put it on the board
with
5...h1=Q+
White looked at him in astonishment, uttering the following words that have become immortal within this
club’s circles: ‘That’s odd, I nearly had you there, I was just one move short!’ His opponent looked at him,
wondering if he was serious, and I remember that I quickly ran outside because I couldn’t hold back my
laughter...
Obviously, this gentleman wasn’t familiar with the ‘square’ idea, but still it gave me an association with one
of the most famous, and most baffling, endgame studies ever made. As young as I was, I already knew Réti’s
study from 1921. I had mathematics as a subject in secondary school, and my teacher had explained that in
mathematics a straight line was the shortest distance between two points. But I knew better than that. In
chess, the diagonal march of the king that Réti introduced in his study is the shortest distance between two
points! This motif has found its way into the collective memory as the ‘Réti manoeuvre’, doing justice to the
imaginary power of the former top player from Hungary. Clearly, the old man from the previous example
had never gotten beyond conventional mathematics!
Even though Réti’s study is very well known, we cannot omit showing it here.

117
Study by Richard Réti
Kagan’s Neueste Schachnachrichten 1921
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+-+-0
9k+P+-+-+0
9+-+-+-+p0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
White’s king is in the corner, and the black pawn on h5 seems to be beyond reach. With the example of the
two old gentleman in mind, it seems wise for White to resign here, to avoid running the risk that your club
mates will be chuckling over your moves for years to come...
1.Kg7!
White continues the struggle – against better judgement?
Of course, 1.Kh7? is useless since after 1...h4 2.Kh6 h3, White would be not one tempo short but two!

1.Kg7! h4
The most logical move, since with an extra queen the rest is simple – or so you would think...
Since the black pawn is still out of the white king’s reach, Black could also reason that it is clever to take the
white pawn off the board first with 1...Kb6. However, after 2.Kf6!, he would wake up to a curious fact: now,
White is threatening 3.Kg5, eliminating Black’s pawn. 2...h4 3.Ke5!. White continues his king’s march along
the h8-a1 diagonal. Now, he threatens to play 4.Kf4, stepping inside the square and controlling the h4-
pawn. 3...h3 4.Kd6!. And here we see the whole idea of Réti’s manoeuvre. Suddenly, the white king is in the
vicinity of its own pawn, supporting its promotion to a queen: 4...h2 5.c7 Kb7 6.Kd7, with a draw!
2.Kf6!
The king has to keep moving along the long diagonal; not 2.Kf7? Kb6 3.Ke7 Kxc6, and the battle is lost.
2...Kb6
If 2...h3, 3.Ke6 and White’s king reaches the vicinity of its own pawn: 3...h2 4.c7, with a draw!
With the text move, for a moment, Black thinks he can consume the White pawn first, but...
3.Ke5!
Again, a move on the long diagonal.
3...h3
Necessary, as after 3...Kxc6, the white king is right on time to stop the h4-pawn: 4.Kf4 h3 5.Kg3.
4.Kd6!
The key point! The king supports its own pawn, thus forcing the draw.
4...h2 5.c7 Kb7 6.Kd7 ½-½
This marvellous escape seems to defy all the physical and mathematical laws and is justifiably known as the
Réti manoeuvre.
Less well-known is the following study by Réti. It is derived from his own discovery given above, but it is at
least as baffling as the original composition.
118
Study by Richard Réti
Deutsch-Österreichische Tageszeitung 1921
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zp-0
9k+P+-zp-zp0
9+-+-+-+K0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
In this seemingly hopeless position, White comes up with a final twitch:
1.Kg6! Kb6
Black decides to eat that white pawn first of all...
Of course, he could also try to lure White’s king back with 1...h5, but White lets this pawn live! He continues
2.Kxg7! h4 3.Kxf6 Kb6 (running with the h-pawn doesn’t win either: 3...h3 4.Ke7 h2 5.c7 Kb7 6.Kd7, and
the king is exactly on time to support its own pawn) 4.Ke5! (threatening to step inside the square of the
black pawn) 4...Kxc6 5.Kf4, with a draw.
The march with the other pawn cannot confuse White either: 1...f5 2.Kxg7! (again, this is the pawn White
has to take! Certainly not 2.Kxf5?? Kb6!, and all of White’s counterplay disappears like snow in summer)
2...f4 3.Kf6, but here we get the same scenario as in the other variation: 3...f3 (3...Kb6 4.Ke5!, threatening
to eliminate the f4-pawn and at the same time controlling the h6-pawn: 4...f3 5.Kd6, and once more the
white king is in time to support its own pawn: 5...f2 6.c7 f1=Q 7.c8=Q, with a draw) 4.Ke7 f2 5.c7 f1=Q
6.c8=Q+.
2.Kxg7!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mK-0
9-mkP+-zp-zp0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White’s king starts on its ‘mission impossible’. It takes the g7-pawn from the board, taking a voluntarily
step out of the square of both the f6- and the h6-pawns.
2...h5
Running with the other pawn doesn’t make any difference: 2...f5 3.Kf6 f4 4.Ke5 f3 5.Kd6, and we have seen
this theme above. By the way, here Black can reach a queen ending with an h-pawn that is a draw: 5...f2 6.c7
f1=Q 7.c8=Q.
3.Kxf6
Threatening 4.Kg5, eliminating the h5-pawn.
3...h4
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-mkP+-mK-+0
9+-+-+-+-0
9-+-+-+-zp0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
4.Ke5!
The king crosses the board diagonally (from g7), thereby performing two tasks: on every move it is
threatening to enter the square of Black’s h-pawn to eliminate it, and at the same it threatens to arrive
exactly in time to support its own passed pawn, which throws in enough weight to secure the half point.
4...h3
4...Kxc6 is met by 5.Kf4.
5.Kd6 h2 6.c7 Kb7 7.Kd7 ½-½
Both sides promote to a queen. A great piece of art in which Richard Réti’s imaginative power defies some
of the ‘chess laws’.
• Protected passed pawn
It can be an enormous advantage in a game to possess a protected passed pawn. The reason is simple: the
pawn is invulnerable, and the enemy king has to stay inside the square, since otherwise the pawn gets to
the back rank unhindered. In most cases, converting such positions to a win is child’s play.

119
XIIIIIIIIY
9-+-+-+-+0
9+-+-mk-+-0
9-+-+-+-+0
9+p+-+Pzp-0
9-+-+-+P+0
9+K+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
In this position, we discover that there is an important difference between a protected passed pawn and a
‘normal’ passed pawn. White’s pawn on f5 is defended by its colleague pawn and can never be conquered
by the opponent on this square. Since Black’s king has to stay inside the square of this pawn, it cannot
support its own pawn. So, the winning procedure may run as follows:
1.Kb4 Kd6 2.Kxb5 Kd5 3.Kb4 Kd6 4.Kc4 Kc6 5.Kd4 Kd6 6.Ke4 Ke7 7.Ke5 Kf7 8.f6 Kf8
8...Kg6 9.Ke6.
9.Kf5 Kf7 10.Kxg5
And White wins.

120
Study by Mikhail Zinar
Shakhmaty v SSSR 1989
 XIIIIIIIIY
9-mk-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9zPp+-zp-+-0
9-zP-+-+-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
White has a protected passed pawn on a5. This means that Black has to stay exclusively inside the square of
that pawn (a5/a8/d8/d5), since otherwise the pawn moves to the back rank unhindered. To win, White
needs to find a trick along the way:
1.Kg4!
White’s king starts an enveloping manoeuvre. We will soon see why this is necessary.
1.Kf3? looks attractive, but it doesn’t win: 1...Kc7 2.Ke3 Kc6 3.Ke4 Kd6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9zPp+-zp-+-0
9-zP-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This is actually a kind of mutual zugzwang. If it were Black’s move, he would lose since his king has to
abandon the protection of the e5-pawn, or else it would leave the square of White’s protected passed pawn.
But it is White’s move, and the e5-pawn becomes an important factor after 4.Kf5 Kd5!. All of a sudden,
Black threatens 5...e4, and after 5.a6 the resulting ending is a draw: 5...Kc6 6.Kxe5 Kb6 7.Kd6 Kxa6 8.Kc6
Ka7 9.Kxb5 Kb7=.

1.Kg4!
 XIIIIIIIIY
9{mk-}-+-+0
9+-+-+-+-0
9-+-+-+-+0
9zPp+{zp-+-0
9-zP-+-+K+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
1...Kc7 2.Kf5 Kd6 3.Ke4
We have seen this position above, but now it is Black’s move. In fact, White has made a triangulation to
force the opponent to make a move! Of course, 3...Ke6 is no good in view of 4.a6, and so White gains the full
point.

• Floating squares
We speak of ‘floating squares’ if a player has two distant passed pawns. Sometimes, the enemy king is
inside both squares, but of course it is impossible for it to stop both pawns if they both move forward.

121
Study by Robert Fontana
National-Zeitung 1948
XIIIIIIIIY
9-+-+-+-+0
9+-+p+-+-0
9-+-+-+-+0
9+-zp-zP-+-0
9-+-mk-+P+0
9zP-+P+-+p0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
White to play and win
At first sight, this position looks chaotic. But if we do a little orientation, we will see that White has two
passed pawns, on a3 and g4, while Black has a far-advanced h-pawn. He would like to create a second
passed pawn, which can be done in two ways: with ...Kxd3 followed by ...c5-c4, or with ...Kxe5 followed
by ...d7-d5 and ...c5-c4. The advantage of the latter option is that Black’s king can still keep an eye on the
white g-pawn, hoping that his two passed pawns will be just as fast as White’s. But White finds a way to
thwart that second plan:
1.e6!
This pawn was about to fall anyway, so White gives it up in a clever way.
The immediately 1.g5? is met with 1...Kxe5! (not 1...Kxd3? in view of 2.g6 c4 3.g7 c3 4.g8=Q c2 5.Qg5,
winning easily). Now, White has to send his second pawn to the other side, but it turns out that Black gets
there right on time: 2.a4 d5 3.a5 c4 4.dxc4 dxc4 5.a6 h2 6.Kg2 c3 7.a7 h1=Q+ 8.Kxh1 c2 9.a8=Q c1=Q+
10.Kh2 Qxg5, with a draw.

1.e6! dxe6
Thus, White has managed to separate Black’s pawn pair d7/c5, depriving Black of any counterplay.
2.g5 Ke5
Also here, 2...Kxd3 is too slow after 3.g6 c4 4.g7 c3 5.g8=Q c2 6.Qg5.
3.a4 1-0
The two distant passed pawns win.
Sometimes you have to defend against a ‘floating square’.

122
End of a study by Wouter Mees
De Schaakwereld 1940
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-zp-+-+0
9+-mk-+-+P0
9-+-+-+-+0
9+-+-+-zp-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White has a distant passed pawn that can no longer be stopped. But since Black has two passed pawns, the
game isn’t over yet... The question is: with which move should White start? If he pushes his own pawn,
Black’s g-pawn will be too fast.
1.Ke1!
The right square for the king, after which Black is too late.
1.h6? g2 2.h7 g1=Q 3.h8=Q Qd4+!, and Black wins the ensuing pawn ending!

Also wrong is 1.Ke2? in view of 1...d5 2.h6 d4 3.h7

XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+-+0
9+-mk-+-+-0
9-+-zp-+-+0
9+-+-+-zp-0
9-+-+K+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
3...d3+!, and since this pawn advances with check, Black wins a crucial tempo: 4.Ke3 g2 5.Kf2 d2 6.h8=Q
g1=Q+ 7.Kxg1 d1=Q+, with a draw.

Illogical, and also plain bad, is 1.Ke3? in view of 1...d5 2.h6 d4+, and again Black gains an important tempo:
3.Kf3 d3 4.h7 d2 5.Ke2 g2 6.h8=Q d1=Q+ 7.Kxd1 g1=Q+, with a draw.

1.Ke1! d5 2.h6 g2 3.Kf2 d4 4.h7 d3 5.h8=Q d2

Now the queen can stop the pawn:
6.Qd8 1-0

• Creating a passed pawn

We have often mentioned that a passed pawn is one of the most important weapons in the endgame.
Therefore, you should try to create a passed pawn whenever possible. But this has to be done with care, as
we will see.

123
Study by Artur Mandler
Sach 1942
XIIIIIIIIY
9-+-+-+-+0
9mk-+-+-+-0
9-+-zp-+-+0
9+-+-+-mK-0
9-+P+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
To win this position, we have to use the theory of the ‘distant opposition’:
1.Kg6!
Also attractive is 1.Kf6?, but this doesn’t win with good defence by Black: 1...Kb6 2.Kf7 (White’s idea is to
lure the black king to the back rank, so that he can promote with check) 2...Kb7 3.Kf8 Kb8 4.c5. This has to
be the winning idea: sacrificing a pawn to gain free passage for the other one. But Black replies 4...Kc7!
5.Kf7 (5.cxd6+ Kxd6) 5...dxc5 6.e5 c4 7.e6 c3 8.e7 c2 9.e8=Q c1=Q, and promotes just in time.

1.Kg6! Ka6
The ‘normal’ way to defend against distant opposition. But again, White now lures the enemy king to the
back rank.
1...Kb6 loses to 2.Kf6 Kb7 3.Kf7 Kb6 4.Ke8 (again, an enveloping manoeuvre makes the difference)
4...Kc7 5.Ke7 Kc6 6.Kd8 Kc5 7.Kd7, and White wins.
2.Kg7! Ka7
XIIIIIIIIY
9-+-+-+-+0
9mk-+-+-mK-0
9-+-zp-+-+0
9+-+-+-+-0
9-+P+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black’s king has to keep ‘following’ his white colleague.
3.Kg8! Ka8
Now, the black king is on the back rank, and the breakthrough wins for White!
Again, Black could try 3...Ka6, but then White strikes home with the well-known ‘surrounding strategy’:
4.Kf8! Kb6 (4...Kb7 5.Kf7) 5.Ke8! Kc7 6.Ke7 Kc6 7.Kd8 etc.
4.c5! 1-0
4...dxc5 5.e5 c4 6.e6 c3 7.e7 c2 8.e8=Q+, and White queens with check.

124
Study by Ramon Rey Ardid
Finales de Ajedrez: Reyes y Peones 1944
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zp-mk-+-+0
9+-+-+-+-0
9-zPPmK-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
The correct move to win is:
1.b5! Ke6
After 1...Kd7, the finish is also simple: 2.Kd5 Kc7 3.Ke6+–.
2.c5
Now that Black’s king is just too far away, White can exchange the pawns; he is conquering a key square!
2...bxc5+ 3.Kxc5 Kd7 4.Kb6 Kc8 5.Ka7
And White wins.
• Breakthrough
The first time I saw this position, my eyes popped out of my head!

125
Study by Jacob Sarratt
A Treatise on the Game of Chess 1808
XIIIIIIIIY
9-+-+-+-+0
9zppzp-+-+-0
9-+-+-+-+0
9zPPzP-+-+-0
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
White to play and win
There is only one way to win here:
1.b6!
White has to start with the middle pawn.
1...axb6
Black has to take, otherwise one of White’s pawns will reach the other side anyway. In this case, it doesn’t
matter how he takes, since 1...cxb6 is met by the same type of breakthrough: 2.a6! bxa6 3.c6. But if the black
king were able to step inside the square of this pawn now, all would have been for nothing...
2.c6!
This is the idea. White threatens 3.cxb7 and at the same time lures away the final defending pawn.
2...bxc6 3.a6
And the pawn promotes.
How can it be possible that three soldiers, standing right opposite three equivalent enemy soldiers, can still
break through the defensive line with so little effort? This was what I thought when I saw the solution for
the first time. As a side note, the position of the defending king is of course important! If it is inside the
square of the passed pawn, White can forget about the whole idea.
If it can be done with three pawns on both sides, then is it also possible with four pawns standing in line,
facing four enemy pawns? The following remarkable study shows that it is possible. But it requires a very
delicate touch, because White has to shepherd the right pawn to the other side. And which one is the right
pawn?

126
Study Josef Kling & Bernard Horwitz
The New Chess Player 1851
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+pzppzp0
9+-+-+-+-0
9-+-+PzPPzP0
9mk-+-+-+-0
9-+-+-+-+0
9mK-+-+-+-0
xiiiiiiiiy
White to play and win
A chess player’s head will be filled with chaos if he has never seen this before... where do you start? Which
pawn should be pushed first? One thing is certain: the promotion square has to be f8. Why can we tell this
with so much certainty? If one of Black’s pawns reaches the promotion rank, White will be mated
immediately. This means that the only good square for White to promote on is f8, since that is with check!
1.h5!
Perhaps quite surprisingly, this is the only move.
On 1.g5?, there follows 1...fxg5 2.h5 gxf4!, and White doesn’t get a passed f-pawn, so after 3.hxg6 f3 4.g7 f2
5.g8=Q f1=Q# it is Black who wins.

Also not good is 1.e5? fxe5 2.h5 exf4! and again, no passed f-pawn for White.

The same goes for 1.f5? exf5 2.h5 (2.g5 makes no sense due to 2...fxg5 3.hxg5 hxg5 4.e5 g4–+; 2.exf5)
2...fxe4 3.hxg6 e3 4.g7 e2 5.g8=Q e1=Q#.

1.h5! gxh5
After 1...g5 2.e5! fxe5, White gets the desired f-pawn with 3.f5! (as is also the case after 2...gxf4 3.exf6):
3...exf5 4.gxf5, and White’s pawn queens with check.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+pzp-zp0
9+-+-+-+p0
9-+-+PzPP+0
9mk-+-+-+-0
9-+-+-+-+0
9mK-+-+-+-0
xiiiiiiiiy
Another essential move:
2.e5! fxe5
2...hxg4 3.exf6 doesn’t help either.
3.f5!
Again, an f-pawn breaking through!
3...exf5 4.gxf5 h4 5.f6 h3 6.f7 h2 7.f8=Q+
And the win is easy with an extra queen: 1-0
Here are three playful exercises on this theme:

127
Study by Hans Fahrni
Das Endspiel im Schach 1917
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpp0
9-+-+-+-+0
9+-+-mkPzPP0
9-+-+-+-+0
9+-+-+-+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
1.f6!
Since the black king is close to White’s pawns, something has to be done quickly.
1...gxf6
After 1...Ke6 2.fxg7 Kf7 3.h6, White has a protected passed pawn, but the win is not entirely trivial yet.
White hauls in the loot as follows: 3...Kg8 4.Kd3 Kf7 5.Ke4 Kg8 6.Ke5 Kf7 7.Kf5 Kg8 8.g6 hxg6+ 9.Kf6!
(taking the pawn means stalemate). 9...g5 10.Kg6 (and now this pawn is harmful material! Another
winning idea was 10.h7+ Kxh7 11.Kf7) 10...g4 11.h7#.
2.g6 hxg6 3.h6
And White wins. The h-pawn cannot be stopped any more since the square has been interrupted by the
pawn on f6.

128
Study by Ladislav Prokes
Obrana Lidu 1949
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+-0
9-+-+-+-+0
9+-mk-zP-+-0
9-+-zp-+P+0
9+-zP-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
There are three moves that need to be analysed here:
First, sacrificing the e-pawn with 1.e6? fails to the zwischenzug 1...dxc3+! 2.Kxc3 fxe6 3.g5 Kd6 4.g6 Ke7,
and Black’s king is in time to stop the pawn.

After 1.cxd4+ Kxd4 2.e6 fxe6 3.g5, Black’s king enters the square of White’s g-pawn with 3...Ke5.

1.c4!
The only move that wins!
1...Kxc4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+-0
9-+-+-+-+0
9+-+-zP-+-0
9-+kzp-+P+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
No choice. But now the king is side-tracked, and can’t get to the g-pawn:
2.e6!
This second pawn sacrifice is necessary to accomplish the breakthrough; if 2.g5, 2...Kd5 and there is no
breakthrough.
2...fxe6 3.g5 Kd5 4.g6
And again, we see that the square has been ‘interrupted’ – the g-pawn cannot be stopped: 1-0

129
Study by Ernest Pogosyants
Shakhmaty v SSSR 1976
 XIIIIIIIIY
9-+-+-+-+0
9+-+-zp-+-0
9-+-+p+-+0
9+-+P+p+-0
9-+-zPP+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
White to play and win
1.d6!
The only correct continuation; 1.dxe6? Kxe4! would even lose!

1.d6! exd6
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-zpp+-+0
9+-+-+p+-0
9-+-zPP+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
2.d5!
A typical breakthrough that is regularly seen.
2...fxe4
Obviously, 2...exd5 doesn’t make any sense either in view of 3.exf5.
3.dxe6
And White’s pawn has broken through the lines: 1-0

• Most distant passed pawn

The concept of the most distant passed pawn often plays a role already in the middlegame. If a player
knows that the resulting pawn ending is winning, he will obviously make use of this knowledge. Let’s see
how this works.

130
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+k+-+p+0
9+-zp-+pzPp0
9-+K+-zP-zP0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The idea of the ‘most distant passed pawn’ can be aptly illustrated with this standard scheme: on the king’s
wing, a number of pawns has been fixed, and this is the wing to which both kings would like to walk to raid
the enemy pawns. On the other side of the board, both sides have a passed pawn, but White’s pawn is
farther removed from the battle scene. This means that Black’s king will need more time to eliminate it, and
therefore White’s king will get to the other wing first.
Here is this scheme expressed in moves:
1.a4 Kb6 2.a5+! Kxa5 3.Kxc5
Here, we see very clearly that the exchange of the passed pawns works strongly in White’s favour.
3...Ka4 4.Kd5 Kb4 5.Ke5 Kc4 6.Kf6 Kd4 7.Kxg6 Ke4 8.Kxh5 Kxf4 9.g6 and White queens first.
This brings us to the following rule of thumb:

Rule of thumb
If there are pawns on two wings, and one side possesses the ‘most distant passed pawn’, he can
use it to lure the enemy king away, allowing his own king to get to the pawns on the other wing
first.
Another example, in which White has to play very accurately – especially in the beginning:

131
Study by Emil Richter
Sachove Koncici Hry 1958
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+-zP-0
9P+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
White to play and win
It’s clear that White is going to use his a-pawn to lure the black king to the queenside. However, since Black
has one extra pawn on the kingside, White has to prevent the trade of Black’s final pawn as well as the
creation of a passed pawn. So:
1.g4+!
This move is essential for the win. Too early is 1.a4? because now Black’s king is in time with 1...Ke4. But
what is more important is that Black will manage to trade his kingside pawns for White’s g-pawn: 2.Kg2
(2.g4 is met by 2...h5!, when White has to trade, since otherwise he would be saddled with a protected black
passed pawn, which would even be losing for him) 2...h5 (Black has to work quickly to trade his pawns)
3.Kh3 g5, and ...h5-h4 cannot be prevented.

1.g4+! Ke5
1...Kxg4 is met by 2.a4...
XIIIIIIIIY
9{+-+{+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+-+-+-0
9P+-+{+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
analysis diagram
and Black’s king is no longer in the square of White’s a-pawn.
2.g5!
Fixing Black’s pawns and deciding the game in White’s favour.
2...Kd5 3.Kg2
The composer stopped here (see diagram on the right). A possible winning line is 3...Kc5 4.Kf3 Kb4 5.Ke4
Ka3 6.Ke5 Kxa2 (Black has eliminated White’s passed pawn with his king, but in every other respect he
has no chance...) 7.Kf6 Kb3 8.Kg7 Kc4 9.Kxh7 Kd5 10.Kxg6 Ke6 11.Kh7, and White’s pawn decides the
battle.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+k+-zP-0
9-+-+-+-+0
9+-+-+-+-0
9P+-+-+K+0
9+-+-+-+-0
xiiiiiiiiy
This endgame in practice
You wouldn’t expect it, but this theme occurs in practice rather frequently. First, we will look at an
‘elementary case’ from a German team competition, and then we will look at some heavier stuff – a game
with a faster time control between Nakamura and Carlsen.

132
Ralf Müller 2300
Axel Schmitt 2340
Germany tt 1996/97
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zp-+-zpp+0
9zpP+k+-+-0
9P+-+-+PzP0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This game was played in the Oberliga Württemberg 1996/97 during the team match between Ditzingen
and Ulm.
58.h5
Creating the most distant passed pawn. 58.Kf4 would have won too.
58...gxh5 59.gxh5 Ke6 60.Kf4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zp-+kzp-+0
9zpP+-+-+P0
9P+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black can stop the h-pawn, but this will cost him his passed pawn on f6. What it comes down to is that
White trades his h-pawn for Black’s f-pawn, after which his king reaches the queenside much more quickly
than its opposite number. Black gave up here in view of 60...Kf7 61.Kf5 Kg7 62.h6+ Kxh6 63.Kxf6,
followed by a king’s march to b6.

133
Hikaru Nakamura 2736
Magnus Carlsen 2847
Crypto Cup 2021 (1.13)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+p+-mk-zp0
9+p+rtR-+P0
9-zP-zP-+P+0
9+-+K+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Nakamura has just played 46.Re5!, which is an invitation to a pawn ending. Carlsen didn’t accept the
invitation, and later lost the rook ending anyway. But what if he had exchanged the rooks?
46...Rxe5
The game continued 46...Rd8 47.Rc5 Rd6 48.Ke4 Re6+ 49.Re5 Rd6 50.d5 cxd5+ 51.Rxd5 Rb6 52.Kd4 Rb8
53.Rd6+ Kg5 54.Rg6+ Kh4 55.Kc5 Kg3 56.Rxh6 Kxg4 57.Rb6 Rc8+ 58.Kxb5 1-0.

46...Rxe5 47.dxe5+ Kxe5 48.Ke3 Kf6 49.Kf4

XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+p+-mk-zp0
9+p+-+-+P0
9-zP-+-mKP+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Despite the equal number of pawns, White is winning here. The most important factor is that the h5-pawn
is further advanced than Black’s b5-pawn.
49...Ke6
49...Kf7 50.Kf5; please note that Black can almost never play 49...c5 due to the ‘floating square’: 50.bxc5 b4
51.Ke4 (White’s king is inside the square of the b4-pawn) 51...b3 52.Kd3 Ke5 53.g5 hxg5 54.h6, and one of
the two white pawns will queen, before Black can achieve this.
50.g5 hxg5+ 51.Kxg5 Kf7
51...c5 52.h6! (certainly not 52.bxc5?? b4 53.h6 b3 (53...Kf7 54.h7 Kg7 55.c6 b3 56.c7 b2 57.c8=Q b1=Q
58.Qg8#) 54.h7 b2 55.h8=Q b1=Q, and Black seems to be able to draw the ensuing queen ending) 52...cxb4
(52...Kf7 53.bxc5 b4 54.h7 Kg7 55.c6 etc.) 53.h7 b3 54.h8=Q, and White is on time.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9-+p+-+-+0
9+p+-+-mKP0
9-zP-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The text move indicates that Black is worried about the h5-pawn. Now, there are two ways for White to win
this ending:
A) By zugzwang: White can stalemate the enemy king;
B) Using the principle of the most distant passed pawn.
We will discuss the latter in the main line:
52.Kf5
The king walks to the queenside and will pick Black’s pawns off the board.
52.h6 was possible too: 52...Kg8 53.Kg6 Kh8 54.h7 c5 55.bxc5 b4 56.c6 b3 57.c7 b2, and White is right on
time with 58.c8=Q#.
52...Kg7 53.Ke4 Kh6 54.Kd4 Kxh5 55.Kc5 Kg6 56.Kxc6 Kf7 57.Kxb5 Ke7 58.Kc6 Kd8 59.Kb7
And White wins.
• Promotion with check
We’ve seen several times that it is very important to determine on which square a pawn will promote. If the
enemy king is on the same line or diagonal (or if it can be lured there!), this may mean a world of difference.
Let’s have a look at an example in which White has to play very ingeniously to enable his pawn to promote
with check.

134
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+p+-+-+0
9+k+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The pawns on both sides have advanced equally far. White has to play very ingeniously to win here:
1.h6!
1.Kd2 is met by 1...Kb2!. In this way, Black can push his pawn with check and gain a crucial tempo: 2.h6
c3+ 3.Ke2 c2 4.h7 c1=Q 5.h8=Q+, with a draw.

1.Ke2 is not a good winning attempt either: 1...c3! 2.h6 c2 amounts to the same thing as the previous
variation.

1.h6! c3 2.h7 c2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+k+-mK-+-0
9-+p+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, Black threatens to queen (with check, even!), but this is precisely the right moment for White to stop
the black pawn:
3.Kd2!
Now, Black has no better than to support his own pawn, but this means that his king has to move to a
square where it will be in check:
3...Kb2 4.h8=Q+
After the promotion with check, mate follows quickly:
4...Kb1 5.Qb8+ Ka2 6.Kxc2 Ka3 7.Qb3#
So now you can test if you can see exactly what is going on in the following position:
135
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+p+k+K+-0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White plays 1.g5. How will this end?
1.g5
Black has to make a momentous decision here: push his own pawn, or do something with his king?
1...Kd6!
XIIIIIIIIY
9-+-}-+{+0
9+-+-+-+-0
9-+-mk-+-+0
9+p+{+KzP-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+{+-+{+-0
xiiiiiiiiy
The correct defence! The king is in the square of the white pawn and can try to stop it in this way.
Running straight to the other side with the pawn would cost Black dearly. Black’s king is on the same
diagonal as the promotion square g8, and a player always has to be aware of this: 1...b4?? 2.g6 b3 3.g7 b2
4.g8=Q+, and White queens with check and wins.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+p+-+KzP-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
2.Kf6
Of course, White cuts off the black king. But the situation has changed: White will not queen with check
now.
2.g6 is met by 2...Ke7, and the king will reach the pawn.
2...b4 3.g6 b3 4.g7 b2 5.g8=Q b1=Q ½-½
The following study is worth looking at as well. We will present it in exercise form.
136
Study by Aristide Dall’Ava
L’Italia Scacchistica 1947
XIIIIIIIIY
9-+-+-+-+0
9+-+p+p+-0
9-+-+P+-+0
9+-+-zP-+-0
9-+-+k+-+0
9+-+-zp-+-0
9-+-zP-+-+0
9+-+-+-+K0
xiiiiiiiiy
White to play and win
1.d3+!
The only winning move. The most important reason is that Black was ready to promote with check by
1...exd2 followed by 2...d1=Q+. The other point will become clear in the following variations.
In fact, Black has three options for the king. It can either move to the d-file (for example, by taking on d3),
or to the f-file, or it can stay on the e-file. However, in all three cases, White can arrange it so that he can
promote on the file where Black’s king is standing. Here are the variations:
1...Kxd3
1...Kf3 2.exf7 e2 3.f8=Q+ is check! 3...Ke3 (now, White has to watch out for a bit since 4...e1=Q+ is
threatened) 4.Qb4 Kf2 5.Qd2 (the quickest path to the win; of course, the prosaic 5.d4 also wins) 5...Kf1
6.Kh2 e1=Q 7.Qg2#.
1...Kxe5 2.e7, and White queens on e8 with check: 2...Kd4 3.e8=Q Kxd3 4.Qxd7+, and wins.
2.exd7 e2 3.d8=Q+ 1-0
Also here the finish is simple: 3...Ke3 4.Qh4 Kd2 5.Qd4+ Kc2 6.Qe3 Kd1 7.Qd3+ Ke1 8.Kg2+–.

This endgame in practice

The following position has been published in many books. It is a curious case from a game between two
strong grandmasters.

137
Ljubomir Ljubojevic 2510
Walter Browne 2500
Amsterdam 1972 (9)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+-0
9-+k+-+-+0
9mK-+-+-+-0
9-+-+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Here, Browne didn’t manage to find the win. He played:
39...f5?
With 39...Kd5!!, Black could have secured the win. However, this move looks paradoxical since it loses a
tempo. What exactly is going on here?
A) If White tries to use his own pawn immediately with 40.b4, Black has the well-known idea: 40...f5 41.b5
f4 42.b6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-+-+0
9mK-+k+-+-0
9-+-+-zp-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This is all about timing! Black now forces White’s king to the fatal a6-square so as to queen with check a
little later: 42...Kc6! 43.Ka6 f3 44.b7 f2 45.b8=Q (White even gets there first, but he cannot prevent being
mated or losing the queen) 45...f1=Q+ 46.Ka5 (46.Ka7 Qa1#) 46...Qa1+ 47.Kb4 Qb2+ – here is the X-ray
again – 48.Kc4 Qxb8;
B) 40.Kb4. Since the direct push of White’s pawn fails, he tries to reach the square of the black pawn with
his king, and now we see why Black had to start with the king move: 40...Kd4!. Outflanking White’s king.
With this ‘shouldering’ technique, Black manages to bring his pawn to the other side in time: 41.Ka5 f5
42.b4 f4 43.b5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9mKP+-+-+-0
9-+-mk-zp-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
43...Kc5! (here also, Black’s king has to force its opponent to the unfavourable a6-square) 44.b6 Kc6!
45.Ka6 f3, and we have already seen the rest in the other variation: 46.b7 f2 47.b8=Q f1=Q+ 48.Ka5 Qa1+
49.Kb4 Qb2+–+.

39...f5? 40.Kb4!
The correct continuation. White’s king reaches the square of the black pawn, thus pulling a half point out of
the fire.
40...f4
At this moment, 40...Kd5 makes no sense in view of 41.Kc3. White’s king takes care of Black’s f-pawn, and
he will be able to use his own pawn: 41...Ke4 42.b4 (42.Kd2 is possible too, e.g. 42...Kf3 43.b4 f4 44.b5 Kg2
45.b6 f3 46.b7 f2 47.b8=Q f1=Q, with a draw) 42...f4 43.b5 f3, and now White has to watch out for a bit:
44.Kd2! (certainly not 44.b6?? f2 45.b7 f1=Q 46.b8=Q Qc1+ 47.Kb4 Qb2+, and Black wins) 44...Kd5, and
both pawns disappear.
41.Kc4
With a draw.
• X-ray check
In many endgames, the so-called ‘X-ray check’ plays a role. In pawn endings, both players have to look on
which square a pawn is going to promote. If the king is on a fatal diagonal, file or rank, the new queen may
be lost at once.

138
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9K+-+-+-+0
9zP-+p+-+-0
9-+-+-+-+0
9+k+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Kb5!
White plays the king to this square so that it already controls the a4-square for the protection of the future
queen.
1...d4 2.a6 d3 3.a7 d2 4.a8=Q d1=Q
Relatively better is 4...Kc2, but we know that this queen vs pawn ending is winning for White, for example:
5.Qe4+ Kc1 6.Qc4+ Kb2 7.Qd3 Kc1 8.Qc3+ Kd1 9.Kc4 Ke2 10.Qd3+ Ke1 11.Qe3+ Kd1 12.Kc3 Kc1
13.Qxd2+ Kb1 14.Qb2#.
XIIIIIIIIY
9Q+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+K+-+-+-0
9-+-+-+-+0
9+k+-+-+-0
9-+-+-+-+0
9+-+q+-+-0
xiiiiiiiiy
Black’s queen and king are both on a fatal diagonal, enabling White to give an X-ray check:
5.Qa4+ Kc3 6.Qxd1 1-0
Naturally, this theme is thankfully used by study composers. Personally, I think the following two studies
are very beautiful.

139
Study by A.Melnikov
Sachove Umenie 1976
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+K0
9-zp-+p+-+0
9+-zp-+-mk-0
9PzP-+-+-+0
9+-+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White to play and win
Both sides can promote a pawn:
1.a5!
The correct move. The liquidation with 1.bxc5 bxc5 2.a5 c4 3.a6 c3 4.a7 c2 5.a8=Q c1=Q brings White
nothing.

1.a5! bxa5
If 1...c4, White takes on b6 because he prefers to promote on the b8-square. Obviously, 1...cxb4 ends badly
for Black since after 2.axb6, he can promote on the same file, but one move too late.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+K0
9-+-+p+-+0
9zp-zp-+-mk-0
9-zP-+-+-+0
9+-+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
2.b5!
This push is essential for the win. White shows great insight by pushing the pawn to the promotion square
b8. We will see why he does this as soon as both players have a queen.
2...a4
2...c4 has to be investigated as well: 3.b6 c3 4.b7 c2 5.b8=Q c1=Q 6.Qe5+ (this check is what it’s all about...)
6...Kh4 7.Qh2+!, and after 7...Kg5, the nasty X-ray check 8.Qh6+ wins the queen.
3.b6 a3 4.b7 a2 5.b8=Q a1=Q
 XIIIIIIIIY
9-wQ-+-+-+0
9+-+-+-+K0
9-+-+p+-+0
9+-zp-+-mk-0
9-+-+-+-+0
9+-+-+P+-0
9-+-+-+-+0
9wq-+-+-+-0
xiiiiiiiiy
Both sides have queened; now, how can White force the black king to move to the fatal long diagonal?
6.Qg3+!
Only thus:
6...Kf5 7.Qg4+
Please note the importance of the f3-pawn.
7...Ke5
7...Kf6 8.Qg7+ would make no difference.
8.Qg7+
And White wins the queen.
1-0

140
Study by Emil Melnichenko
Sinfonie Schacchistice 1981
XIIIIIIIIY
9-+-+-+k+0
9+p+-+-+-0
9-zP-+P+p+0
9zpK+-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
1.Kc5!
The king hurries to c7 via d6. Taking on a5 (1.Kxa5) only leads to a draw after 1...Kf8 2.Kb5 Ke7 3.Kc5
Kxe6.

1.Kc5! Kf8
Black has to take control of the e6-pawn first. Going for promotion with 1...a4 loses as well: 2.Kd6 a3 3.e7
Kf7 4.Kd7, and White queens with check.
2.Kd6 Ke8
Forced; otherwise 3.Kd7 would have followed.
3.Kc7 a4 4.Kxb7 a3 5.Kc6
We will soon see why the king has to go to this square.
5...a2 6.b7 a1=Q
Black queens first, but White promotes with check and then has the initiative.
7.b8=Q+ Ke7
Now, White has to determine how to drive the enemy king into a corner:
8.Qd6+ Kf6
Forced; here, 8...Ke8 would lead to a simple mate with 9.Qd7+ Kf8 10.Qf7#.
9.Qf8+
This is the right check; not 9.e7+? Kf7 10.Qxg6+ Kxg6 11.e8=Q+ Kf5, and the f4-pawn will be lost.
9...Kxe6
XIIIIIIIIY
9-+-+-wQ-+0
9+-+-+-+-0
9-+K+k+p+0
9+-+-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9wq-+-+-+-0
xiiiiiiiiy
Again, forced.
10.f5+!
Only by depriving Black’s king of the escape square f5 can White force the intended X-ray check.
10...gxf5 11.Qe8+ Kf6 12.Qh8+
Mission accomplished! 1-0
Time for an exercise.

141
Study by Alexey Troitzky
Shakhmaty Zjurnal 1896
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+KzP-+0
9+-+-+-+-0
9-+-+k+-+0
9+-zp-+-zp-0
9-zP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
Both sides have a pawn that is about to promote, but White will come first.
The nice thing about this study is that wherever Black queens (on b1, c1 or g1), he will fall victim to an X-
ray check. See the following variations:
1.f7 cxb2
A) 1...c2 2.f8=Q c1=Q 3.Qf5+ Kd4 (moving to the c1-h6 diagonal loses immediately: 3...Ke3 4.Qg5+)
4.Qd5+ (here, we see the importance of the white pawn on b2: the king has to move to the fatal diagonal)
4...Ke3 5.Qg5+;
B) 1...g2 2.f8=Q g1=Q 3.Qf5+ (again, this nasty check) 3...Kd4 (3...Ke3 clearly fails to 4.Qc5+) 4.Qd5+ and
again, the king has to move to the same diagonal as its queen: 4...Ke3 5.Qc5++–.
2.f8=Q b1=Q
 XIIIIIIIIY
9-+-+-wQ-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-zp-0
9-+-+-+-+0
9+q+-+-+-0
xiiiiiiiiy
3.Qf5+
And White wins.
• Mating lines
As we saw in the previous positions, a race in a pawn ending can lead to a queen ending. If the pawns
promote on the same move, what matters is not just who gets there first, but also if a pawn queens with
check! With the new queen on the board, there may be something in it for you if you can act quickly.
Let’s have a look at the following examples, in which White promotes one move later than his opponent, but
with check – and that makes all the difference! First, we look at an elementary case.

142
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9Pzp-+-+-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+k0
xiiiiiiiiy
White to move
To quickly see whose pawn will arrive on the other side first, the horizontal middle line can serve as a
distinguishing point. The black pawn has already passed it, and White will pass it on the next move. Black
will arrive first, but the white pawn queens with check:
1.a5 b3 2.a6 b2 3.a7 b1=Q 4.a8=Q+ Kg1
Which means that the black king can be checkmated:
5.Qg2#
In the next study, things are a lot more complicated. White has to win a queen vs queen ending, and he
needs to use domination to force checkmate.

143
Study by Jindrich Fritz
Svobodne Slovo 1964
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+K+-+-+0
9+-+-+-+-0
9-+P+-+-+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
In this endgame study, we have to use the knowledge we have just acquired in this chapter to force the win:
1.Kd5! h5
Black tries to push the pawn to the other side as quickly as possible: it would be with check.
Another idea is to use the king to help the h-pawn promote. But a move with the king costs a tempo, and
this is enough for White to win: 1...Kf3 2.c5 h5 3.c6 h4 4.c7 h3 5.c8=Q h2 (if 5...Kg3, White has a technical
win with 6.Qc3+ Kg2 7.Qd2+ Kg3 8.Qe3+ Kg2 9.Qg5+ Kf3 10.Qh4 Kg2 11.Qg4+ Kh2 12.Ke4, and the h3-
pawn will be lost) 6.Qh3+.
1...Kf2 2.Ke4 transposes to the main line.
2.Ke4!
The only good move, after which Black’s pawn is within reach of the white king.
On 2.Ke5?, Black could force a draw with 2...Ke3!, for example: 3.c5 h4, and both pawns promote in turn,
after which there is no win for White.
2.c5? would spoil the win after 2...h4 3.c6 h3 4.c7 h2 5.c8=Q h1=Q+, since this is a draw.
2...Kf2
2...h4 would be too early in view of 3.Kf4, and the white king will gobble up Black’s pawn.
3.Kf4
Now, White threatens to conquer the black pawn with 4.Kg5, while the black king is outside the square of
the white pawn. However, Black manages to complicate things with a ‘Réti manoeuvre’:
3...Kg2!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+p0
9-+P+-mK-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
xiiiiiiiiy
4.c5!
The only move that wins; both pawns are approaching the back rank.
Superficially, it looks like 4.Kg5? could be a simple win. But things turn out badly for White after 4...Kg3!
(threatening to support the pawn again after 5...h4. After 5.Kxh5 (5.c5 h4), Black’s king steps inside the
square of the white pawn again: 5...Kf4 with a draw.
4...h4 5.c6 h3 6.c7 h2 7.c8=Q h1=Q
 XIIIIIIIIY
9-+Q+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+q0
xiiiiiiiiy
Thanks to the king manoeuvres, White’s king has ended up in a favourable position, and with a couple of
checks White eventually creates an unstoppable mate threat:
8.Qc2+ Kh3
The same mating line would come on the board after 8...Kf1 9.Qd1+ Kg2 10.Qe2+ Kg1 11.Kg3! –
domination.
9.Qd3+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+Q+-+k0
9-+-+-+-+0
9+-+-+-+q0
xiiiiiiiiy
The right square to give check.
9...Kg2
Black is mated at once after 9...Kh2 10.Qg3# or 9...Kh4 10.Qh7#.
10.Qe2+ Kg1
Similar is 10...Kh3 11.Qg4+ Kh2 12.Qg3#.
11.Kg3!
Domination – Black’s king and queen have been pushed to the edge of the board, and mate cannot be
prevented:
1-0

• Bähr’s position
Sometimes, there are endgame experts who only study one small aspect of endgame theory that they think
has been underexposed. Sometimes, optical illusion plays a role. In an earlier example, we discussed the
diagonal king’s march (the Réti manoeuvre); in this paragraph we will look at a quite special type of
position investigated by Walter Bähr (1905-1986).

144
Bähr’s position – 1
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9p+-+-+-+0
9+-+-+PmK-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White wins if it’s his move. The win is obtained as follows:
1.a5! Kg7 2.Kf4 Kf6 3.Ke4 Kf7 4.Kd5 Kf6 5.Kc6 Kxf5 6.Kb6 Ke6 7.Kxa6 Kd7
And White’s king cuts off its opposite number just in time:
8.Kb7
And White wins.
Okay, we can understand this. But what about the next position?

145
Bähr’s position – 2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9p+-+-+-+0
9+-+-+PmK-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
With Black to move, the story is completely different:
1...a5!
What matters is that Black has just enough time in the race to reach c8, holding the draw.
2.Kf4 Kf6 3.Ke4 Kf7 4.Kd5 Kf6 5.Kc5 Kxf5 6.Kb5 Ke6 7.Kxa5 Kd7 8.Kb6
And Black is exactly in time:
8...Kc8 with a draw.
If we move both kings and the f-pawn back one rank, what is the situation then?

146
Bähr’s position – 3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mk-+0
9zp-+-+-+-0
9P+-+-zPK+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Kf3 Kf5 2.Ke3
Black has to lose time.
2...Ke6 3.Kd4 Kf5 4.Kc5 Kxf4 5.Kb5 Ke5 6.Kxa5 Kd6 7.Kb6
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-mK-mk-+-+0
9+-+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
With this ‘shoulder budge’, Black’s king is kept out of the corner: 1-0
It was the merit of Bähr that he formalized situations like these into rules, already in 1936. Bähr
investigated situations in which both rook’s pawns were fixed against each other at different ‘levels’. Also,
he investigated what happened when he put White’s pawn on different files and ranks. Each and every new
position leads to differences, and of course it is not easy for a practical player to familiarize himself with all
situations. In practice, such endgames always appear on the board at the end of the game, so fatigue will
play a role. Therefore, knowledge of the schemes should be stored in the memory in a clear and systematic
way. That is why Mark Dvoretsky in his famous book Endgame Manual speaks of a few confusing rules, and
made something else of it.
Let’s have a look at the scheme first.

147
Bähr 4 – schematic diagram
There are three conditions for Bähr’s rule:
Condition 1: The rook’s pawn of the attacking side has not yet passed the middle of the board (i.e. it has not
yet reached the fifth rank).
Condition 2: The attacking king should stand right next to the passed pawn.
Condition 3: The defending king has to stand right opposite the passed pawn or to the attacking king.
With the help of this schematic diagram, we can formulate a kind of general rule that also offers visual
insight (subject to the three above-mentioned conditions):

Rule of thumb: Bähr’s rule

After having drawn (in our minds) the borderlines as they were discovered by Bähr (see the schematic
diagram), we can establish the following:
• If the passed pawn is on or below the borderline, the stronger side wins.
• If the passed pawn is above the borderline, the position is a draw.
In essence, we can claim that the problem for the defender is as follows: if the pawn is below the line, the
king’s route back to c8 is longer, and White’s king can cut it off from the corner.
Let’s check this right away by putting the pawn on g4. And yes, the rule works here!

148
Bähr’s position – 5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+k+0
9zp-+-+-+-0
9P+-+-+PmK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Kg3 Kg5 2.Kf3 Kg6 3.Ke4 Kg5 4.Kd5 Kg4 5.Kc5 Kf5 6.Kb5 Ke6 7.Ka5 Kd7 8.Kb6 Kc8 ½-½
The following position was included in the thesis of my partner, FM Petra Schuurman, (titled
Approximating Schedules at the TU/Eindhoven on 9 January 2001) on the occasion of her doctoral research
that was honoured in 2001 with a ‘doctor of mathematics’ degree. This is theorem No. 6 in her thesis.
After conducting a small investigation into Bähr’s position as an assignment for the highest-level trainers’
course in the Netherlands, Petra composed the following endgame study with which she managed to
wrong-foot quite a few strong players:
149
Petra Schuurman (after Bähr)
2001
XIIIIIIIIY
9-+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9zp-+-+-+-0
9P+-+p+p+0
9+-+-+P+-0
9-+-+-mK-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
1.fxe4!
This is in defiance of all the endgame rules – but it’s the only move that wins! Superficially, it looks like
White should create a distant passed pawn, but not here! What actually matters here is that White has to
take one of the two black pawns, and win the other one after that. But then, of course, the question is which
pawn he should end up with. Bähr provided us with the insight that White should not be left with a pawn
on g4 due to the diagonal line (g4-c8) he had drawn; with the pawn on that line, Black’s king will be able to
get back to c8 in time. It is an optical illusion that White should end up with the pawn on e4, not g4. If Black
starts the king’s race on the e4-square, it cannot reach the c8-square in time. A curious case!
Every strong chess player will, in principle, follow the rule of the ‘most distant passed pawn’ and will
continue with 1.fxg4? here. But thanks to Bähr’s principle, we know that precisely this continuation is
wrong. Black holds the draw with 1...Kf7 2.Ke3 Kf6 3.Kxe4 Kg5 4.Kd5 Kxg4 5.Kc5 Kf5 6.Kb5 Ke6
7.Kxa5 Kd7 8.Kb6 Kc8, and the black king reaches c8 exactly in time to prevent the cut-off by its opposite
number – draw!

1.fxe4! Kf7 2.Kg3 Kf6 3.Kxg4 Ke5 4.Kf3 Ke6

XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+k+-+0
9zp-+-+-+-0
9P+-+P+-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, the white king walks in the direction of the a-pawn by manoeuvring behind its own pawn:
5.Ke3 Ke5
Other king moves are useless too: 5...Kd6 is met by 6.Kd4 Ke6 7.Kc5 Ke5 8.Kb5 Kxe4 9.Kxa5 Kd5
10.Kb6, and clearly Black is too late also here.
6.Kd3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9zp-+-mk-+-0
9P+-+P+-+0
9+-+K+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
6...Ke6
This is the problem for Black: he has to lose a tempo before he can conquer the e4-pawn.
7.Kc4 Ke5 8.Kb5 Kxe4 9.Kxa5 Kd5 10.Kb6
And White cuts off the enemy king with the ‘shoulder budge’: 1-0
There is more to be said about Bähr’s rule, but we will keep it at this.
Chapter 18
Knight endings
§ 18.1 Anecdote
Some people say that knight endings can be assessed as pawn endings. There is a grain of truth in this, but
it isn’t a one-on-one case. In many cases where a pawn ending is a simple win, things are definitely not
trivial in a knight ending with the same configuration. Perhaps one side’s knight can be sacrificed for the
enemy’s last passed pawn. In other cases, a knight may be able to jump around the board with reckless
abandon, causing a lot of headaches for a player who seems to be winning. The knight jump in chess is so
special, sometimes creating such strange curves, that a player may lose the thread entirely. This reminds
me of the following anecdote.
Imagine your club has a first team that plays in one of the higher leagues in your country. For this club, it is
important that the other teams have a close connection with the first team. The second team is playing two
leagues lower but is in danger of relegation. In the final match of the season, a meagre tie will allow them to
save their skin – but in case of a loss, the curtain will fall. The match doesn’t proceed very successfully, but
fortunately for the team a few points are scored, and with a 3½-3½ score only the team captain is still
glued to the board. Beforehand, he has warned his teammates to avoid unnecessary risks, and now he is
required to notch up the half point that guarantees the team’s prolonged stay in the league. The future
might be bright, since there are a few new strong players who may want to join the team, and perhaps in
the next season a renewed outfit might be a contender for the top of the league instead of the bottom, as is
the case now.
After a heavy struggle in which the team captain has managed to avoid any kind of risk throughout, he has
reached a favourable knight ending with two extra pawns.
We will cover the names of the players with the cloak of charity.

150
Team captain
Opponent
Eindhoven match
XIIIIIIIIY
9-+-+-+-+0
9+-+-+nsN-0
9-+-+-zPp+0
9+-+-+-zP-0
9-+-+k+P+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
True, White is two pawns up, but still this position is a dead draw. White cannot really play for a win, and
he doesn’t have to. His teammates nudge him: accept the draw, then we can open the champagne bottles.
But the team captain laughs off these comments: ‘I can never lose this position! My opponent has only one
pawn, and I’m blocking it well with my doubled pawns. And since I’ve scored so badly this season, I want to
make up for it by winning this game at least.’ He returns to the table and hurls a knight sacrifice on the
board:
1.Nf5?!
In light of the following, this may be the only practical chance to get something out of this position, but
objectively the move amounts to nothing. Black can draw simply with 1...Nxg5 after which there is nothing
special. Black does have to find this move; there are many mistakes he can make here.
The only move to play for a win was 1.Kh4, but it doesn’t bring any benefits if Black does ‘nothing’: 1...Ke5
2.Ne8 Kf4 3.Nc7 Ke5, and White doesn’t get any further at all.

1.Nf5?! gxf5??
Black lends a helping hand – quite a big one! Accepting the knight sacrifice is not only unnecessary, but
even loses immediately! As said, 1...Nxg5! would have drawn at once.
2.g6!
We all know that two connected passed pawns on the sixth rank are generally stronger than a rook, let
alone a knight. Here this is also the case, but...
2...f4+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+n+-0
9-+-+-zPP+0
9+-+-+-+-0
9-+-+kzpP+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
A desperate deed – a spite check. Probably Black just found out what horrible crime he had committed, and
in his distress came up with this zwischenschach. And now give yourself three minutes at most to
determine where the white king should go:
A) 3.Kf2
B) 3.Kg2
C) 3.Kh2
D) 3.Kh3
E) 3.Kh4.
3.Kf2??
How cruel our game can be! White chooses the wrong square!
3.Kg2! was the winning move that was pointed out by the team captain’s club mates immediately after the
game. He protested, ‘Yes, I considered that, but of course I thought I shouldn’t move my king into the pawn
check.’ Precisely this is not a problem, however, since after 3...f3+ 4.Kf1, Black runs out of options. His
pawn is under control, the knight has no checks, and the promotion of one of White’s passed pawns cannot
be prevented.
3.Kh4 was winning too. It is important that after 3...Ne5 4.g7, White can promote a pawn and Black can do
nothing about it.
The illogical 3.Kh3 is a move almost nobody would play, but it does win! At it turns out, it’s not a problem
to move the king ‘into a direct check’: 3...Ng5+ 4.Kg2 (even 4.Kh2 is possible now) 4...Ke5 5.g7 and here
too, White’s pawn cannot be stopped.
3.Kh2?? loses in the same way as in the game.
3...Ne5!
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-zPP+0
9+-+-sn-+-0
9-+-+kzpP+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+-+-0
xiiiiiiiiy
After this miraculous knight move, suddenly the full point goes to the opponent! The knight not only
attacks the g6-pawn, but it can also move back in time via g4 to stop the other passed pawn.
4.g7
4.f7 is simply met by 4...Nxg6, when the knight not only keeps the f7-pawn from promoting, but also blocks
White’s other passed pawn on g4. It’s quite pathetic: White has to sit and watch while Black promotes his
only passed pawn, while his own passed pawns are powerless. A possible winning line is 5.Ke2 f3+ 6.Kf2
Kf4 7.g5 Nf8 8.Kf1 Ke3 9.Ke1 f2+ 10.Kf1
XIIIIIIIIY
9-+-+-sn-+0
9+-+-+P+-0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-mk-+-0
9-+-+-zp-+0
9+-+-+K+-0
xiiiiiiiiy
analysis diagram
10...Ng6 11.f8=Q (11.Kg2 Ke2) 11...Nxf8 12.g6 Nxg6 13.Kg2 Ke2, and it’s over.
Now follows the essential in-between move:
4...Nxg4+
With check!
5.Kf1 Nxf6
A gruesome sensation for the White player. He threw in the towel and remained seated at the board in
despair.
The agony was visible not only with the team captain, who was immediately nicknamed ‘Team Kaput’. The
entire team was standing there, hands before their eyes, and none of them could believe the drama that had
just unfolded. A few members of the first team even came up to rub salt into the wound by casually
remarking that he should have put the king on g2 to be checked by the pawn! That evening, the team
captain went through some tenebrous hours – and also, especially, the next day.
For it was at the team captain’s home that on Sundays after every league match, a couple of fanatics
always came to make the club bulletin. Those were always good-natured and entertaining sessions that
were concluded with a copious meal, and not rarely going on into the small hours. But now everything was
very different. The team captain opened the door to allow the people inside his home, but didn’t show up
for the rest of the day. When one of the editors resolved to ask him whether he was still capable of handing
in a report on the match, as he had always done with great gusto, the door remained hermetically shut. At a
certain moment, a sheet of paper was shoved under the door with the following text:
THE CAPTAIN OF THE SECOND TEAM HAS DECIDED TO JOURNEY TO A TIBET MONASTERY IN WHICH HE
WILL CONTEMPLATE THE MEANING OF LIFE.
These ‘philosophical’ words were all the second team’s report was made up of. A malicious editor printed
the diagram with the accompanying moves next to it. And that was a picture that kept haunting the team
captain for years... it’s not very strange that he made his position available at the end of the season.
(The club bulletin also featured the following text: ‘any similarity with true occurrences and/or persons is pure
coincidence.’ The editorial staff apparently honoured the principle that the truth should never stand in the way
of a good story!)

§ 18.2 Knight fork

Every beginning chess player will soon come across one of the first combinational themes in chess: the
knight fork. This motif always causes great merriment, if it means that a rook or even a queen is lost in the
opening phase. As for me, I’ve always been enraptured by the bizarre performances of the knight. It has
brought me a lot of fun, but sometimes atrocious blunders too...

151
Herman Grooten 2376
Kai Rooda 2141
Hoogeveen 2008 (6)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-sn-+K0
9-+-sN-mkp+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
I had been defending a pawn-down endgame for a long time. In this position, the draw is almost in the
pocket. Just a few little pitfalls...
87.Kh4??
A ghastly move that nullifies all White has achieved. This is the last move a player should even think of.
White could have saved his skin with, for example, 87.Ne6+, after which Black can make no progress:
87...Ke3 (nor does 87...Kf3 help: 88.Nd4+ Kf2 89.Nf5, with a draw) 88.Kh4, and the draw is secured.

White even draws with 87.Ne2+ after 87...Kf3 88.Nd4+ Ke4 89.Ne2.

Black replied with the sobering

87.Kh4?? Nf3+!
and I resigned immediately. The liquidation into a pawn ending ends it all.
Actually, the disgrace of this blunder pales in comparison with the beautiful giant leaps my knight made
against a strong German grandmaster. The remarkable thing about this game is that at various points, a
pawn race arose, which, after both sides’ pawns had promoted to queen, culminated in a new knight ending
with another pawn race. These adventures, which are so typical of knight endings, are an attractive sight:
152
Herman Grooten 2386
Lothar Vogt 2469
Germany Bundesliga B 2000/01 (6)
XIIIIIIIIY
9-+-+-+-+0
9+-+P+-+-0
9-zp-+-mk-+0
9+P+-zp-sNp0
9-zpPtr-+-zP0
9+n+-+-+-0
9-+-+-+-+0
9+K+-+-tR-0
xiiiiiiiiy
White to move
A blunder just slipped from Black’s fingers with 57...Rf4-d4. Placing the rook behind an enemy passed pawn
is the standard way to defend, but here there is a nasty trick in the position that was missed by the
grandmaster from East Germany:
58.Rf1+!
This in-between check is essential. Black’s king doesn’t have a single safe square.
58...Kg6
Black’s problem is that he cannot move his king toward the pawn with 58...Ke7 in view of 59.Rf7+ Kd8
60.Ne6#, while 58...Kg7 runs into the knight fork 59.Ne6+.
XIIIIIIIIY
9-+-+-+-+0
9+-+P+-+-0
9-zp-+-+k+0
9+P+-zp-sNp0
9-zpPtr-+-zP0
9+n+-+-+-0
9-+-+-+-+0
9+K+-+R+-0
xiiiiiiiiy
Also here, a knight fork decides the battle in White’s favour:
59.Ne6!
Forcing a decisive liquidation.
59...Rxd7
After 59...Rd6, White has to watch out for a bit, but he can win with 60.Ka2!, e.g. 60...Na5 61.c5 (again, this
breakthrough guarantees a simple win as the d7-pawn is still untouchable; winning the rook with 61.d8=Q
Rxd8 62.Nxd8 wins as well for White, but this would still require some effort) 61...bxc5 62.d8=Q Rxd8
63.Nxd8+–.
60.Nf8+ Kg7 61.Nxd7 Nd2+
With this knight fork, Black regains the rook, but his knight will be quite off-side, and in the meantime
White has a breakthrough:
62.Kb2 Nxf1 63.c5!
 XIIIIIIIIY
9-+-+-+-+0
9+-+N+-mk-0
9-zp-+-+-+0
9+PzP-zp-+p0
9-zp-+-+-zP0
9+-+-+-+-0
9-mK-+-+-+0
9+-+-+n+-0
xiiiiiiiiy
63...e4
It is essential that after 63...Ne3 64.cxb6 Nc4+ 65.Kb3 Nd6 66.Nxe5 Kf6 67.Nc6 Nb7 68.Kxb4 followed by
Na5, White can nurse a pawn to the other side.
64.cxb6 e3
Now, things become tense after all; Black will have a new queen as well!
65.b7 e2 66.b8=Q e1=Q
XIIIIIIIIY
9-wQ-+-+-+0
9+-+N+-mk-0
9-+-+-+-+0
9+P+-+-+p0
9-zp-+-+-zP0
9+-+-+-+-0
9-mK-+-+-+0
9+-+-wqn+-0
xiiiiiiiiy
And now, White liquidates into a new knight ending!
67.Qe5+! Qxe5+ 68.Nxe5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+-+0
9+P+-sN-+p0
9-zp-+-+-zP0
9+-+-+-+-0
9-mK-+-+-+0
9+-+-+n+-0
xiiiiiiiiy
Black’s knight is too far removed.
68...Kf6
After 68...Ne3 69.b6, Black does not have the c4-square available, and the pawn cannot be stopped; also,
after 68...Nd2 69.b6, ...Nc4+ is impossible, and if 69...Ne4, 70.b7.
69.b6 Kxe5 70.b7 Kf5 71.b8=Q Ne3
With queen vs knight, the win is certain. Seeing he was going to lose the h5-pawn, Black immediately
resigned here, since after 72.Qb5+ Kg4 73.Qg5+ Kf3 74.Qxh5+, there are no technical problems anymore.
§ 18.3 Deflecting the enemy knight or king
We will come straight to the point with two rules of thumb:

Rules of thumb
• Kasparov: ‘The power of a passed pawn increases as it comes closer to the promotion rank.’
• Nimzowitsch: ‘A passed pawn is a criminal, who should be kept under lock and key.’
Kasparov’s remark was intended for the middlegame too (as he demonstrated in several games, see e.g.
Kasparov-Pribyl, Skara 1980), but it will be clear that his advice also applies to the endgame. Nimzowitsch’s
famous statement makes clear once more that passed pawns can be of great influence in a game.
In each of the following examples, we will see a knight, or sometimes the king, trying to keep under control
one or several enemy passed pawns. The way that White succeeds in wrong-footing (i.e., deflecting) these
defenders in each example is highly instructive:

153
End of a study by Ernest Pogosyants
Shakhmaty v SSSR 1976
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-mk0
9+-+-+-zp-0
9-+-+-+KzP0
9+-+-sn-sN-0
xiiiiiiiiy
White to play and draw
White’s king is in check and finds itself in a delicate situation, since it has to keep the black pawn under
control. What should White do?
1.Kf1
Obviously, 1.Kh1? was not possible in view of 1...g2#.

1.Kf1 gxh2
White seems to be done for. Black is threatening 2...h1=Q, and the g2-square is not accessible for White’s
king. So – should he resign?
2.Nf3+!
No! White sacrifices his knight (with check!), and it has to be taken as otherwise White can play 3.Nxh2.
2...Nxf3 3.Kg2
And this is a theoretical draw that also appears elsewhere in this book (only with a knight on g4, which
doesn’t make a difference). White’s king can’t be driven out of the corner without being stalemated or
without giving up the pawn. Good to remember!

154
Study by Alexey Troitzky
The Chess Amateur 1916
 XIIIIIIIIY
9-+n+-+-+0
9+-+-+-+-0
9Pzp-+-+-zp0
9+-+-+N+-0
9-+-+-+-+0
9+-+-+k+-0
9-+-zP-+-+0
9+-+-+-+K0
xiiiiiiiiy
White to play and win
White’s a-pawn is being held off by the black knight, but this means that this knight’s action range is limited.
White sets his other pawn in motion:
1.d4
1.Ne7? was also tempting, but it doesn’t lead to the desired result: 1...Na7 2.d4 Ke4 3.d5 Ke5 4.Nc6+.
XIIIIIIIIY
9-+-+-+-+0
9sn-+-+-+-0
9PzpN+-+-zp0
9+-+Pmk-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
analysis diagram
You’d swear that White is winning here; however... there follows 4...Kxd5 5.Nxa7 b5!, and Black reaches a
theoretically drawn position! The line runs as follows: 6.Nxb5 Kc6 7.Kg2 Kb6 8.a7 Kb7, and White cannot
win this, as we already established in the previous example!

1.d4 Kf4
Black might consider 1...b5 to vacate the b6-square for his knight, but it doesn’t help him: 2.d5 b4 3.d6 b3,
and here White is saved by a knight fork: 4.Nd4+.
Obviously, 1...Ke4 fails to 2.Nd6+ Nxd6 3.a7, and the pawn promotes.
2.Ne7 Na7
The a-pawn would get through after 2...Nxe7 3.a7.
3.Nc6 Nb5
Also 3...Nc8 doesn’t work. After 4.d5 Kf5, White again has a knight fork to deflect Black’s knight from one of
his passed pawns: 5.Ne7+.
It is important that White wins an essential tempo if Black takes the knight: 3...Nxc6 4.d5 Na7 5.d6, and the
pawn is out of the black king’s reach.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9PzpN+-+-zp0
9+n+-+-+-0
9-+-zP-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
4.d5 Kf5
Black’s king has to move to this unfortunate square because he wants to meet d5-d6 with ...Kf5-e6. But it’s
clear that a calamity is in store for him...
The alternative doesn’t help Black: 4...Kg5 5.d6! Kf6 6.d7, and his king can’t stop the pawn.
5.Nd4+
The final blow; White wins.

155
Study by Jan Foltys & Jaroslav Genttner
Ceskoslovensky Sach 1935
XIIIIIIIIY
9-+-+-+-+0
9+-+-zp-+-0
9-sN-+k+P+0
9+-sn-zP-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
xiiiiiiiiy
White to play and win
White has a passed pawn on g6 that cannot be exploited for the moment. It is kept under control by Black’s
king, since the pawn is lost in case of 1.g7? Kf7. But the e5-pawn is an important trump card for White. At
this point, it controls f6, but there is a more important role for this pawn.
The question is, however: how can White win? Before we establish this, we may ask what Black would play
if it were his move. After 1...Ne4!, Black is ready to continue with 2...Nf6!, rendering the g6-pawn harmless.
In short: White has to be quick, and he has to start with a move with a forcing character. We want to try to
lure both Black’s king and knight away from the g-pawn, and to that end we evaluate the following move:

1.Nd7!
A marvellous idea! White puts his knight on a square that is defended twice.
1...Nxd7
1...Nd3+ 2.Kf1 Nxe5 also loses for Black after 3.Nxe5 Kxe5 4.g7.
It is important to examine what follows after 1...Ne4 here. White’s strong reply is 2.Nf8+ Kxe5 3.g7 Nf6;
now he has 4.Nd7+, luring the defender away from the g8-square.
Finally, 1...Kxd7 clearly fails to 2.g7.
2.g7
Here, we see the importance of the pawn on e5. It covers the f6-square, preventing Black’s knight from
helping out.
2...Kf7
If 2...Nf6 3.exf6 Kf7, 4.fxe7 wins.
XIIIIIIIIY
9-+-+-+-+0
9+-+nzpkzP-0
9-+-+-+-+0
9+-+-zP-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
xiiiiiiiiy
3.e6+!
The point of the entire combination. Thanks to this pawn fork, one of the pawns will reach the back rank.
3...Kxe6
Or 3...Kxg7 4.exd7.
4.g8=Q+ 1-0

156
Study by Alexey Troitzky
Novoye Vremja 1896
XIIIIIIIIY
9-+-+-+-+0
9+-+-zp-+-0
9-+-+-+-+0
9+n+P+-+-0
9-+k+-+-zP0
9+-+-+-+-0
9-+-+-+-+0
9+N+-+-mK-0
xiiiiiiiiy
White to move
White to play and win
White is focusing all his efforts on the rook’s pawn – the arch-enemy of the knight, as we know.
1.Na3+!
By luring Black’s knight further away from the rook’s pawn, we increase the chances of the latter pawn
doing the job.
1...Nxa3
This is a crucial moment.
2.d6!
All of a sudden, the other pawn does the dirty work, haring off while both Black’s knight and king are forced
to sit and wait. If White started with 2.h5, an adequate reply would be 2...Nb5 3.h6 Nd6 4.h7 Nf7, and the
knight is right on time.
2...exd6
2...Nb5 is, of course, met by 3.d7, and the pawn promotes.
3.h5
But now the knight can no longer follow the mapped-out route via d6. Neither is Black’s king able to catch
up with the pawn: 1-0

§ 18.4 Zugzwang
We have encountered the notion of ‘zugzwang’ several times in this book. This is a theme that can play a
role in almost any type of ending. With knights on the board that have to perform certain tasks, the
zugzwang theme is likely to appear. Let’s study the following attractive examples.

157
Study by Hans Fahrni & Hans Keemink
Het Eindspel 1928
XIIIIIIIIY
9k+-sn-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+p+-sN-+-0
9-zP-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
Both sides have only one knight and one pawn left, but Black’s men are so unfortunately placed that the
defeat is inevitable:
1.Kd7
The knight has only one escape square:
1...Nb7 2.Nc6!
Putting Black’s pieces in a jam. The king cannot move; the knight can, but it will be immediately lost –
zugzwang!
1-0

158
End of a study by Frantisek Richter
Ceskoslovensky Sach 1954
XIIIIIIIIY
9k+-+-+-sN0
9zPp+-zp-+P0
9-zP-+p+-+0
9+-+-sn-+p0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
xiiiiiiiiy
White to move
White to play and win
Black’s king is confined to the corner, and the black and white knights are keeping each other in check. If
Black’s knight leaves, White can move his knight out of the corner to queen his pawn next. Since Black’s
knight is standing in front of the pawns on e6 and e7, they can’t move either. The only mobile pieces are the
pawn on h5 and White’s king. White is more or less forced to move his king into the square of the h5-pawn,
and for this, two squares are available to him: d1 and d2. Of course, only one of them is the correct square!
1.Kd1! h4
Black sets his pawn in motion, thereby forcing a second king move – again, only one of the two is good!
2.Ke2!
Certainly not 2.Ke1?, for the same reason: 2...Nf3+ 3.Kf2 Ng5, and the h7-pawn leaves the board.
2...h3
XIIIIIIIIY
9k+-+-+-sN0
9zPp+-zp-+P0
9-zP-+p+-+0
9+-+-sn-+-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+K+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, White has to find the last ‘problem move’:
3.Kf1!
And the battle is decided: Black is in zugzwang. White could still go wrong with 3.Kf2?, which again leads to
a drawn endgame after 3...Ng4+ 4.Kg3 h2 5.Kg2 Nf6=.

§ 18.5 The rook’s pawn is the enemy of the knight

In an earlier paragraph, we have drawn attention to the adage, ‘The rook’s pawn is the enemy of the knight’.
There is a simple explanation for this phenomenon. In the centre of the board, in principle, a knight can
jump to eight different squares – it is maximally active there. As the knight gets closer to the edge of the
board, it has less and less squares at its disposal, and the absolute worst is the corner of the board. In the
following table, we show the number of squares a knight can go to from a certain square. The following
diagrams serve to chart the activity of the knight schematically, and they clearly show that the knight’s
activity decreases as it comes closer to the edge of the board.

Or as we can see in the following diagrams:

XIIIIIIIIY
9-+-+-+-+0
9+{+{+-+-0
9{+-+{+-+0
9+-sN-+-+-0
9{+-+{+-+0
9+{+{+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
8 squares
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-}-}-+-+0
9}-+-}-+-0
9-+N+-+-+0
9}-+-}-+-0
xiiiiiiiiy
6 squares
XIIIIIIIIY
9-+-+{+-+0
9+-+-+-sN-0
9-+-+{+-+0
9+-+-+{+{0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
4 squares
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+{+{0
9-+-+{+-+0
9+-+-+-sN-0
xiiiiiiiiy
3 squares
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-}-0
9-+-+-}-+0
9+-+-+-+N0
xiiiiiiiiy
2 squares

The stronger side can exploit this to the full, as the next two studies show:

159
Study by Hans Kohser
Dresdener Anzeiger 1923
XIIIIIIIIY
9-+-mk-+-+0
9+-+-+-+-0
9-+-sN-+-+0
9zP-sn-+-+-0
9-+p+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
If one side has a rook’s pawn, the opponent always has to be alert in a knight ending. The following
‘standard trick’ is an important one to have in your arsenal of tactical endgame weapons:
1.Nb7+! Nxb7 2.a6
Black’s knight is extremely unfortunately placed. If it moves, White’s pawn marches on unhindered.
2...c3+
A final attempt to get the knight to a good square; 2...Kc8 3.a7.
3.Kc2!
3.Kxc3? would be mistaken, presenting Black with half a point after 3...Nd6 4.a7 (4.Kb4 Kc7, and the pawn
is stopped) 4...Nb5+.
3...Kc8
3...Kc7 doesn’t make any difference.
4.a7 1-0
Thus, we see how unfortunate the knight is on b7...

160
Study by Josef Louma
Ceskoslovensky Sach 1937
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+N+-+-+0
9+-+k+-+P0
9p+-+-zP-sn0
9+p+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
The h-pawn is quickly set in motion:
1.h6 Ng6
Thus, it seems that the knight can control both pawns, but White has a few more strings to his bow.
It is important that the white knight is taboo: 1...Kxc6 fails to 2.f5!, blocking the retreat of the black knight.
Another important point appears after 1...Ke6 2.h7 Ng6:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+P0
9-+N+k+n+0
9+-+-+-+-0
9p+-+-zP-+0
9+p+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
3.f5+! lures the black king to a fatal square: 3...Kxf5 4.Ne7+.
Also after 1...Ke4, White proceeds energetically:
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+N+-+-zP0
9+-+-+-+-0
9p+-+kzP-sn0
9+p+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
2.f5!, and neither of Black’s pieces can take on f5 due to obvious drawbacks: 2...Nxf5 3.h7 or 2...Kxf5 3.h7
Ng6 4.Ne7+.
2.Ne7+! Nxe7 3.f5!
Again, depriving Black’s knight of the g6-square, after which the h-pawn is unstoppable. 1-0

§ 18.6 Passed pawns

In the many chess videos that can be found on the Internet nowadays, increasingly often we can hear
someone cry, ‘Push the pawn!’. This is even done in middlegame situations (doubtlessly under the influence
of the strong computer engine AlphaZero), but obviously it applies even more strongly in the endgame. In
fact, the potential of a passed pawn can only be exploited if it is pushed forward. Such a pawn can be a real
weapon, as is demonstrated in the following two beautiful endgame studies.

161
Study by Ladislav Prokes
Svobodne Slovo 1949
XIIIIIIIIY
9-+-+-+-+0
9+-+N+-+-0
9-+-zp-+n+0
9+P+p+-+-0
9-+-zP-+-+0
9+-+k+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
A passed pawn must be pushed as soon as possible. Therefore:
1.b6 Ne7
The only move to try to stop the pawn. At this point, White has to come up with something special,
otherwise the draw is secured.
2.Ne5+!
Fantastic! Here, 2.b7 brings White nothing in view of 2...Nc6.
2...dxe5
Of course, the crucial c6-square is protected after 2...Kxd4 3.b7, and Black’s knight cannot stop the pawn.
3.dxe5 Nc6
3...Kd4 doesn’t help either: 4.b7 Nc6 5.e6.
4.e6 1-0
In this case, two pawns on the sixth rank beat the knight, mainly because the black king was unable to lend
a hand.

162
Study by Olavi Riihimaa
Uusi Suomi 1941
XIIIIIIIIY
9-+-+-sn-+0
9+-+p+-+-0
9NzPk+-+-+0
9+-+-zPK+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
At the first sight, the draw seems to be within reach for Black, but White has a fantastic concept to force a
win:
1.Nc5! Kxb6
The most important continuation. There are two alternatives:
White meets 1...d5 with 2.b7! Kc7 3.b8=Q+! Kxb8 4.Nd7+ Nxd7 5.e6, forcing the promotion of the e-pawn.
We should add that after 5...Nb6 6.e7 Nc8 7.e8=Q, Black has no knight fork on d6 since his knight is pinned!
Otherwise, the defending side will want to exchange pawns, but 1...d6 also fails to 2.b7 Kc7 3.exd6+
(importantly, 3.b8=Q+ Kxb8 4.Nd7+ doesn’t work in view of 4...Kc7!, with a draw since White’s last pawn
is traded off) 3...Kb8, with a technical win that may run as follows: 4.Kf6 Ka7 5.Kf7 Nh7 6.d7 Ng5+ 7.Ke7.
XIIIIIIIIY
9-+-+-sn-+0
9+-+p+-+-0
9-mk-+-+-+0
9+-sN-zPK+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
2.Nxd7+!
This beautiful move decides the issue.
2...Nxd7 3.e6 Nc5
An ultimate attempt to try to stop the pawn.
After 3...Kc7, we again see a phenomenon we encountered earlier with the rook’s pawn: 4.e7, and the black
knight is very unfortunately placed. 4...Nf6 no longer helps due to 5.Kxf6 Kd7 6.Kf7.
The knight cannot retreat: 3...Nf8 4.e7.
4.e7 Nb7 5.Ke5!
Thus, White keeps off the knight, and he will promote to a queen next.
5.Ke6 is winning too: 5...Nc5+ 6.Ke5 Nd7+ 7.Kf5; but certainly not 5.e8=Q?? Nd6+, with a draw.

§ 18.7 Poor placement of pieces

Most of the pieces have less squares available to them as they move closer to the edge of the board. Not
only that, but also they become more vulnerable at the edge or in a corner. Only to the rook it doesn’t
matter, in principle, whether it is posted in the centre or in the corner – provided, of course, that it is not
hindered by its own pieces or pawns, because then a rook can be inactive too.
The knight is the piece that suffers most from a lack of squares at the edge or in the corner. The following
two studies are striking examples.

163
Study by Ernest Pogosyants
Shakhmaty v SSSR 1988
XIIIIIIIIY
9n+-mK-+-+0
9+-+-+-+-0
9-zp-+-+-+0
9+-+-+-+-0
9-sN-+-+-+0
9+P+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
As we know, a knight in the corner has only two squares to move to. A worse position on the board is
inconceivable, and if one of those two squares is occupied by its own pawn, disaster seems to be imminent.
But this case is not so simple at all. White too has only one pawn, and if he loses it, it will immediately be a
draw. Therefore, White’s task is twofold: he has to save his pawn and at the same time try to capture the
enemy knight.
1.Kc8 Kd2
The attempt with 1...Ke3 brings Black nothing: 2.Kb7 Kd4 3.Na2 (3.Na6 also wins, viz. 3...Kc3 4.b4 Kc4
5.Kxa8) 3...Kd3 4.Kxa8 Kc2 5.b4 Kb3 6.Kb7 Kxa2 7.b5, and White wins.
2.Nc6!
Logical, and yet, not obvious.
Not 2.Na6? in view of 2...Kc3 3.b4 Kc4 4.Kb8 Kb5 5.Kb7, and here too Black seems to be in zugzwang. But
he saves himself from this predicament with a miraculous move: 5...Nc7! 6.Kxc7 Kxa6 7.Kc6 b5! 8.Kc5
Ka7! 9.Kxb5 Kb7, with a draw!
2...Kc3 3.b4 Kc4
XIIIIIIIIY
9n+K+-+-+0
9+-+-+-+-0
9-zpN+-+-+0
9+-+-+-+-0
9-zPk+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Again, the issue is decided by tempo play:
4.Kb8!
After 4.Kb7? Kb5, we have a reciprocal zugzwang situation; White can make no progress here.
4...Kb5
There is no good alternative; for instance, 4...Kd5 loses to 5.b5 Kc5 6.Na7.
5.Kb7!
Now, it’s Black’s move and he is going under! 1-0

164
Study by Ernest Pogosyants
Shakhmaty v SSSR 1978
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9p+-+-+-+0
9mk-+-+-+-0
9-+-+-+-+0
9+-+KsN-+-0
9P+-+-+-+0
9+n+-+-+-0
xiiiiiiiiy
White to play and win
Each side has only one knight and one pawn, but again, the poor placement of Black’s pieces decides the
battle:
1.Kc2!
The knight on the edge is driven further into a tight spot.
1...Na3+ 2.Kb3
White causes trouble for his opponent with this clever move.
2...Nb1
Forced, in order not to have to resign right away.
If the knight wants to go into ‘the wide world’ with 2...Nb5, Black will be rudely awakened – the reply
3.Nc4# would be quite a shock!
3.Nf1!
But with this ‘modest’ retreat of the knight, the curtain falls after all.
3...Kb5 4.Kb2
And so, Black’s knight turned into a sitting duck and was hunted down... 1-0
Chapter 19
Bishop endings
§ 19.1 Tactics in bishop endings
The bishop ending is a special type of endgame. As you will understand, diagonals play the main role in
these endings, which makes the placement of both sides’ pawns an important factor.
We distinguish between two types of bishop ending, each of which demands a totally different approach.
As we know, the activity of a bishop is seriously limited if its own pawns are fixed on the same colour
squares.

Rule of thumb
A bishop is called a ‘bad bishop’ if its own pawns are fixed on the colour of the bishop.
Now, it’s time to focus on a number of stereotypical techniques. But before we do this, first we will show a
compelling example to get going.

• Plugging a diagonal

165
Study by T.Flores
L’Italia Scacchistica 1959
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+L0
9P+-+-+-+0
9+-+p+-+-0
9k+-+-+-+0
9+-+-+-+-0
9-+-mK-+l+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
White has a passed pawn that cannot be caught by the enemy king. The bishop on g2 will have to do it, but
White has a fantastic tactical shot to prevent it from doing the job:
1.Be4!! Bxe4
1...dxe4 is met by 2.Ke3. White blocks the d-pawn so that Black cannot open the long diagonal, allowing
White’s pawn to march on.
2.Ke3
This move fulfils two functions.
2...Bg2
If 2...d4+, White takes the bishop – 3.Kxe4. After this bishop move, White again finds a way to ‘plug’ the
diagonal:
3.Kd4! 1-0
It won’t surprise you that endgames too contain many tactical tricks, as we have seen much proof of this
already. The following series of endgame studies highlights a good arsenal of tactical finesses that every
club player can use to his advantage. Let’s make an inventory of themes.

• Stalemate
166
Study by Lazar Zalkind
La Stratégie 1914
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpp0
9-+-+lzp-+0
9+-+-+-zPP0
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-mkL+0
9+-+-+-+-0
xiiiiiiiiy
White to play and draw
1.Be4!
A move like 1.Bc6? is met by 1...Ke3 2.gxf6 (White is immediately lost after 2.h6 g6 3.gxf6 Kf4!, and there is
no defence to mate with ...g7-g5+) 2...gxf6 3.h6 f5 4.Kg5 f4 5.Kf6 Bc4 6.Kg7 Bd3, and Black wins.

1.Be4! f5
Another winning attempt is 1...h6 2.gxf6 gxf6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+lzp-zp0
9+-+-+-+P0
9-+-+L+-mK0
9+-+-+-+-0
9-+-+-mk-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
But here, White has an artistic motif to save the draw: 3.Bf5! Bd5 (3...Bxf5 is stalemate!) 4.Kg4 (the king
escapes from its predicament, but it does cost another pawn) 4...Bf3+ 5.Kf4 Bxh5. Two pawns less, and yet
it is a simple draw after 6.Bd3!, when Black cannot prevent Kf4-f5: 6...Kg2 (6...Be2 7.Bg6) 7.Kf5 – the f6-
pawn will be lost, and then White can always sacrifice his bishop for Black’s h-pawn.
No use is 1...fxg5+ 2.Kxg5 h6+ (2...Bg8 3.h6) 3.Kg6, with a draw in all lines.
1...Bg8 2.gxf6 gxf6 3.Kg4 Ke3 4.Kf5, and Black cannot avoid his pawns disappearing from the board, e.g.
4...Be6+ 5.Kxf6 Kxe4 6.Kxe6.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpp0
9-+-+l+-+0
9+-+-+pzPP0
9-+-+L+-mK0
9+-+-+-+-0
9-+-+-mk-+0
9+-+-+-+-0
xiiiiiiiiy
2.Bxf5!
Also here, this stalemate motif.
2...Bxf5 3.g6! hxg6
A waiting move like 3...Bd3 also yields Black nothing after 4.gxh7 Bxh7, and now White has 5.h6! g6 6.Kg5,
and the king walks to g7.
The bishop sacrifice 3...Bxg6 4.hxg6 h6 5.Kg4 Ke3 6.Kf5 actually only risks losing, although Black can
easily force a draw with 6...h5 7.Kg5 (7.Ke6?? loses to 7...h4 8.Kf7 h3 9.Kxg7 h2 10.Kf7 h1=Q 11.g7 Qd5+)
7...h4 8.Kxh4 Kf4 or 7...Ke4.
Also after 3...Kf3 4.gxh7 Bxh7 5.Kg5 Ke4, White forces a split of the point with 6.h6 g6 7.Kf6.
4.h6! gxh6
And again, stalemate! ½-½

• Avoiding stalemate

167
Study by Anton Tebben
Tijdschrift v.d. KNSB 1938
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+k0
9-+-zP-+-+0
9+-+-+-+-0
9-zpr+P+K+0
9+-+-+-+-0
9-vL-vl-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
1.d7 Rxe4+ 2.Kh5 Rh4+! 3.Kxh4 b3!
 XIIIIIIIIY
9-+-+-+-+0
9+-+P+-+k0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-mK0
9+p+-+-+-0
9-vL-vl-+-+0
9+-+-+-+-0
xiiiiiiiiy
Here, there is a snake in the grass:
4.Bc3!
Winning. If White unsuspectingly queens, he will be unpleasantly surprised: 4.d8=Q? Bg5+!, and both
5.Kxg5 and 5.Qxg5 are stalemate!

• Fighting against the defender – X-ray check

168
Alexey Troitzky
(500 Endspielstudien 1924)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9vLpvl-+-+P0
9-zp-zp-+p+0
9+-+P+-mk-0
9-+-+-+P+0
9+-+-+-mK-0
xiiiiiiiiy
White to play and win
1.h6
Obviously, White puts all his efforts into his only passed pawn. Black can now continue in various ways:
1...Bf8
An attempt to eliminate the pawn, but White has more strings to his bow.
Another idea for Black is to try to reach the a1-h8 diagonal with 1...Be7 2.h7 (2.Bxb4? Bf6 3.Bf8 Kf4 4.Bg7
Kg5 5.Kf2 Kg6) 2...Bf6, but then again the poor placement of Black’s king will tell: the bishop is deflected
from the promotion square with 3.Bc7+ Kh4 4.Bd8!.
Black can also choose to play his own trump card with 1...b3, since this pawn cannot be stopped by White,
and moreover Black is going to queen with check. 2.h7 b2. Hopeless? No! Again, White wins by a
‘coincidental shot’: 3.Bc7+!. With this check, Black’s king is driven to the fatal h-file, and so White will queen
with check! Chess can be so cruel... 3...Kh4 4.h8=Q+. Now, we should check if White can actually win the
game with the extra queen, since the b2-pawn is still extremely dangerous: 4...Kg5, and now Black’s king is
driven into a mating net in an artistic way: 5.Qg7+ Kf5 (5...Kh5 6.Qf7+ Kg5 7.Bf4+ Kh4 8.Qh7#) 6.Qf7+
Kg5 7.Bf4+ Kh4 8.Qh7#.
The bishop move 1...Bd6 is refuted in a similar way as in the line with 1...Be7: 2.h7 Be5 3.Bc7!, and again
the bishop is deflected.
2.Bxb4! Bxh6
Forced, since Black can’t take the white bishop. But this is a tactically unfavourable square for the bishop:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-vl0
9+p+-+-+-0
9-vL-zp-+p+0
9+-+P+-mk-0
9-+-+-+P+0
9+-+-+-mK-0
xiiiiiiiiy
3.Be1+! Kf4 4.Bd2+
With this X-ray check, White wins the bishop and the game.

169
Study by Ladislav Prokes
Rude Pravo 1950
XIIIIIIIIY
9-+-+-+-+0
9+-zPl+-+-0
9-+pzp-+-+0
9+-+kzp-+-0
9-+-+-zPp+0
9+-+-+-+-0
9-+-+P+K+0
9+-+L+-+-0
xiiiiiiiiy
White to play and win
White’s pawn on c7 has broken through, but it cannot do anything on its own. Therefore, White tries to
bring on the reserves:
1.f5
The pawn threatens to march on with f5-f6, preceded by a check.
1...Bxf5
Now, White has a quite special move that forces the win:
2.e4+!
Some of us play this on move 1!
Not 2.Kg3? in view of 2...Kc5 3.e4 Bc8 4.Bxg4 Kb6 5.Bxc8 Kxc7, and on the next move, Black trades off
White’s final pawn: 6.Ba6 d5.
2...Bxe4+
2...Kxe4 is met with the X-ray check 3.Bc2+ followed by an exchange and promotion.
3.Kg3 Bf5 4.Bxg4
Trading off the defender and forcing promotion.

• Fighting against the defender – blocking

170
Study by Ladislav Prokes
Vecherni Praha 1957
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9p+l+-+P+0
9mk-mK-+p+-0
9p+-+-+-+0
9+-+-+-+p0
9-zP-+-+-+0
9+L+-+-+-0
xiiiiiiiiy
White to play and win
1.g7
This move seems to do the job already since the pawn cannot be stopped – or can it?
1...Bd5!
A clever trick. 1...h2 2.g8=Q h1=Q 3.Qd8# leads to mate.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zP-0
9p+-+-+-+0
9mk-mKl+p+-0
9p+-+-+-+0
9+-+-+-+p0
9-zP-+-+-+0
9+L+-+-+-0
xiiiiiiiiy
2.Ba2!!
A formidable move with a brilliant idea.
The composer assumed that taking the bishop would lead to a promotion of Black’s h-pawn with check. But
he is wrong. My pupil Armin Kohlrausch made a wonderful discovery here: 2.Kxd5 h2 3.Be4!. By first
giving up the bishop, White prevents Black from queening with check (after 3.g8=Q h1=Q+ 4.Kc5 Qc1+
5.Qc4 Qxb2, the endgame of queen+bishop vs queen – with a few pawns added in this case – is a theoretical
draw; White cannot create sufficient dangerous threats with, for instance, 6.Bd3, since after 6...Qb6+ 7.Kd5
Qb5+, the draw is easy), and then we see the following: 3...fxe4 4.g8=Q h1=Q. As White can now start
giving some checks, Black gets caught in a mating net: 5.Qd8+ Kb5 (5...Kb4 6.Qb6#) 6.Qb8+ Ka5 7.Qc7+
Kb5 (7...Kb4 8.Qb6#) 8.Qc5#. Beautiful!
Here, 2.Be4? immediately doesn’t work in view of 2...Bg8.
2...Bxa2 3.b4+
Black can parry the mate by taking this pawn en passant. However, thereby he causes his own pawn to
block his bishop’s diagonal and this kills him as well!
3...axb3 4.g8=Q
A pretty motif. White wins as Black will be quickly mated after 4...h2 5.Qc4 h1=Q 6.Qb4#.

171
Study by Theo Kok
Eindspelen en Problemen 1938
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zp-0
9-zP-+-zp-+0
9+-+-+-+-0
9-vl-+P+-+0
9+-+-+-+-0
9-vL-+-+-+0
9+-+-mk-+K0
xiiiiiiiiy
White to play and win
With his next move, White tries to create a second passed pawn:
1.e5! fxe5
But it is immediately eliminated by Black, so everything in the garden is lovely, isn’t it...?
2.Bc3+!
Anyway! The e5-pawn is quite in the way of Black’s bishop, since after
2...Bxc3 3.b7
White’s pawn cannot be stopped anymore: 1-0
• Zugzwang, X-ray check, discovered check, blockade

172
Study by Leonid Kubbel
Shakhmatny Listok 1928
XIIIIIIIIY
9-+-+-+-+0
9+-+p+-+-0
9-+-zP-+-+0
9+-+p+-+-0
9-vL-mk-mK-+0
9+-+-+-+-0
9-+PzP-+-+0
9+-+-vl-+-0
xiiiiiiiiy
White to play and win
It sometimes occurs that a bishop is threatened to be lost on all sides of the board due to a ‘concurrence of
circumstances’. White plays a waiting move, after which Black, curiously, lands in a kind of zugzwang:
1.Ba5! Bh4!
After other bishop moves, White has either a discovered check with his d-pawn or an X-ray check, winning
the black bishop:
1...Bf2 2.Bb6+;
1...Kc4 2.d3+;
1...Kc5 2.d4+.
2.Kg4 Be1!
2...Bf6 3.Bc3+.
3.Kh3!
White plays another waiting move, guarding the h4-square at the same time.
Suddenly, Black has no moves!
3...Ke4
3...Ke5 4.d4+;
3...Kc4 4.d3+;
3...Kc5 4.d4+;
3...Bf2 4.Bb6+.
4.d3+ 1-0
We get to see a beautiful motif in the diagram below.

173
Study by Max Karstedt
Deutsches Wochenschach 1915
XIIIIIIIIY
9-+-+k+-+0
9+-+-zpp+-0
9P+-zp-+-+0
9+-+-+-+-0
9-+lzP-+-+0
9zp-+-+-+-0
9-+P+-+-+0
9mK-+L+-+-0
xiiiiiiiiy
White to play and win
1.a7 Bd5
The bishop stops the pawn from a distance, but now look what happens:
2.c4 Bb7
The bishop still guards the a8-square, but...
3.Bf3!
With this motif, the black bishop is ‘pulled over’ the critical square d5. Black has no choice:
3...Bxf3 4.d5
Cutting off the bishop. With this blocking motif, the defender is rendered powerless: 1-0
• Passed pawn

174
Study by Ladislav Prokes
Helsingin Samomat 1941
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9LzP-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-zP-0
9-+-+-+l+0
9+K+k+-+-0
xiiiiiiiiy
White to play and win
1.Bc8 Ke1
Black tries to remove the g3-pawn from the board as soon as possible, which would secure the draw.
Black cannot save the g4-pawn with 1...Bf3 in view of 2.Bxg4 Bxg4 3.b7, and the b-pawn gets through.
2.Bxg4 Kf2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
9-+-+-+L+0
9+-+-+-zP-0
9-+-+-mkl+0
9+K+-+-+-0
xiiiiiiiiy
Mission accomplished – or...?
3.Bf3!
With this weird move, placing the bishop on a square that is covered twice, White hauls in the loot after all.
3.Bh3? Be4+! followed by 4...Kxg3 would lead to a draw.
3...Bxf3 4.g4
And White wins.
A possible variation is 4...Ke3 5.g5 Kd4 6.g6 Kc5 7.g7 Bd5 8.b7, and indeed one of the white pawns
promotes.

• Pin

175
Study by Franz Sackmann
Deutsche Schachblätter 1910
(corrected by W. Speckmann, Eur Rochade 1995)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9L+-+P+-+0
9+-+k+-+-0
9-+-+pmK-+0
9+-zP-+-+l0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
The e6-pawn is attacked twice.
1.Bc8!
Now White is ready for 2.e7!.
1.c4+? doesn’t yield the desired result after 1...Kd6! (not 1...Kxe6?? 2.Bc8+ Kd6 3.Bxh3 Kc5 4.Bf1) 2.Bc8
Bf1 3.c5+ Kxc5 4.e7 Bb5 5.Ke5 (threatening 6.Ke6 and 7.Bd7) 5...Be8!, with a draw.

After 1.e7 Bd7, White has no options left to win; for example, 2.c4+ doesn’t work after 2...Kd6 3.Bb5 in
view of 3...Kxe7.

1.Bc8! Bf1!
1...Bxe6 2.Bxe6+ Kxe6 3.Kxe4; 1...Kd6 2.e7 Bd7 3.Bxd7 Kxd7 4.Kxe4 Kxe7 5.Kd5.
2.e7 Bb5
XIIIIIIIIY
9-+L+-+-+0
9+-+-zP-+-0
9-+-+-+-+0
9+l+k+-+-0
9-+-+pmK-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, White has a beautiful shot on which this entire study is based:
3.c4+! Kxc4 4.Ba6!
And White wins. Due to the pin, Black can no longer keep the e7-pawn from promoting.

§ 19.2 Same-coloured bishops

Bishop endings look simpler than they really are. Still, there are certain positions that can be easily won or
held drawn with straightforward and methodical play. So, it will be useful to formulate a few principles and
techniques that may come in handy:
• Length of the diagonal
In the examples given below, the lengths of two diagonals determine the difference between a win and a
draw. Also the position of the weaker side’s king is essential for the question whether the position can be
kept within drawing margins.

176
Bishop + pawn vs bishop – 1
XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-zP-+-+0
9+lmk-+-+-0
9-+-+-+L+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This is an important type of position in the endgame with same-coloured bishops. On c5, Black’s king has
managed to occupy an ideal position to just hold this position to a draw. White cannot win:
1.Bd7
The only attempt to drive the black bishop from the a4-e8 diagonal.
1...Be2 2.Bc6 Bg4
White cannot make progress. The black bishop has two diagonals at its disposal (a4-e8 and c8-h3), giving
him sufficient squares to avoid ending up in zugzwang. It is different with another pawn, the b-pawn, for
example, as we shall see further on. Also,
3.Bb7
with the intention
3...Bf5 4.Bc8
doesn’t help since after
4...Bxc8
White has to let go of the pawn. Here, we see how important it is for the black king to ‘hang onto the tail’ of
the enemy pawn.
½-½

177
Bishop + pawn vs bishop – 2
XIIIIIIIIY
9-+-+-+-+0
9mK-+-+-+-0
9lzP-+-+-+0
9mk-+-+L+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
In this position, White has a b-pawn – otherwise the situation is approximately the same as in the previous
diagram. But there is a subtle difference between the two positions. White puts his opponent into zugzwang
with a simple waiting move (e.g. 1.Bg4), compelling Black to take in the sail. The a6-c8 diagonal is too short,
and Black’s king doesn’t have any moves either (1...Kb5 2.Be2+).

1.Bg4 Kb5 2.Be2+

And White wins.

• Building a bridge
Just like in rook endings (as will be discussed in the next chapter), we have the bridge-building technique
also in bishop endings:

Technique: ‘building a bridge’

With ‘building a bridge’, we mean placing our bishop on a diagonal to drive the enemy bishop
to another diagonal where it no longer prevents the promotion of our pawn.

178
Bishop + pawn vs bishop – 3
End of a study by Yury Averbakh
1972
 XIIIIIIIIY
9-+-+-mk-+0
9+-+l+-+-0
9-+-zP-+-+0
9+-+K+-+L0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
This position arose on the third move of a study by the famous endgame theoretician Yury Averbakh in
1972. Averbakh indicated that White can win this position, but that turns out not to be the case. After I had
published this study on the Dutch chess website Schaaksite, I received a reaction from the Dutch FIDE
master and chess trainer Frank Erwich. He indicated that Black can force a draw in a miraculous way, with
an impressive king manoeuvre that slightly resembles the famous Réti manoeuvre:
1...Kg7!
The only correct move, even forcing the draw! So, I have turned this into the main line. Black’s king is
heading for c5, to hang onto the tail of White’s pawn.
Earlier on in Schaaksite, I had indicated 1...Be8? as one of the alternatives. I also showed that this option is
obviously a mistake since the pawn ending loses for Black: 2.Bxe8 Kxe8 3.Ke6! Kd8 4.d7 Kc7 5.Ke7.

But from Averbakh, I had indiscriminately copied the alternative 1...Ba4? and presented it as the main line,
without consulting the Tablebases. Frank Erwich correctly indicated that 2.Kc5! is winning here. The white
king heads for c7 to later build a bridge on c6, more or less as follows: 2...Bd7 3.Kb6 Ba4 (3...Bf5 would be
met with 4.Kc7, e.g. 4...Bh3 5.Be2, and the bishop heads for c8 where the win would be secured; for
example, 5...Kf7 6.Ba6 Bf5 7.Bc8 Bxc8 8.Kxc8, and the pawn walks through) 4.Kc7 Bb5 5.Bf3 Ba4 (5...Ke8?
6.Bc6+ Bxc6 7.Kxc6 Kd8 8.d7, and wins, e.g. 8...Ke7 9.Kc7).
XIIIIIIIIY
9-+-+-mk-+0
9+-mK-+-+-0
9-+-zP-+-+0
9+-+-+-+-0
9l+-+-+-+0
9+-+-+L+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
In this variation, we see how the ‘bridge-building’ works: 6.Bc6. Here is the bridge, and now the win is
secured: 6...Bxc6 7.Kxc6 Ke8 8.Kc7.

1...Kg7! 2.Kc5
White follows the same plan: king to c7. That is the only way to drive away the black bishop.
2...Kf6 3.Kb6 Ke5
Black’s king hurries closer, but it still seems to be relatively far removed from its ideal square – all the more
so because White wins a tempo with his next move:
4.Kc7
XIIIIIIIIY
9-+-+-+-+0
9+-mKl+-+-0
9-+-zP-+-+0
9+-+-mk-+L0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
4...Be6!
The bishop has to stay on the c8-h3 diagonal – this is the instructive thing about this position. Diverting to
the other diagonal (a4-e8) loses for Black: 4...Ba4? 5.Bf3.
XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-zP-+-+0
9+-+-mk-+-0
9l+-+-+-+0
9+-+-+L+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
The bishop builds a bridge again! 5...Be8 6.Bc6, and White wins.
Neither should the black bishop move to c8 – curiously, this loses here: 4...Bc8? 5.Be8 and now:
A) If 5...Kd5 6.Bd7 Bb7 (or 6...Ba6 7.Bc6+ Ke6 8.d7, and the pawn marches on) 7.Bh3 (or 7.Be6+ Kxe6
8.d7) 7...Bc6 8.Bg2+, and the black bishop is lost;
B) 5...Ke6 is the most tenacious defence, but White can force the win anyway by slowly putting Black in
zugzwang: 6.Bc6 Ke5
XIIIIIIIIY
9-+l+-+-+0
9+-mK-+-+-0
9-+LzP-+-+0
9+-+-mk-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
7.Bd7. Now that Black’s king no longer covers e6, his bishop is dispelled: 7...Ba6 (7...Bb7 is also possible,
but this fails more quickly; for example, 8.Bh3 and d6-d7 can no longer be prevented), and now White wins
in the well-known thematic way: 8.Bh3 Bb5 (now that Black’s bishop has been driven to the a4-e8 diagonal
again, White builds a bridge on c6 once more) 9.Bg2, and the job is done again!
5.Be8
Neither does the alternative 5.Bf3 Kd4 6.Bc6 Kc5 7.Bd7 Bc4 yield White anything. As soon as his bishop
leaves one diagonal, the black bishop takes possession of that same diagonal to stop the promotion of the
pawn. It is essential that a bridge on the c6-square, which we saw in the diagrams given above, is no longer
possible since Black’s king also covers this square here.
XIIIIIIIIY
9-+-+L+-+0
9+-mK-+-+-0
9-+-zPl+-+0
9+-+-mk-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
I had also given this line on Averbakh’s authority, adding: ‘After this move, White continues with Bd7 and
the black bishop has to leave the diagonal.’ The main line was meant to show how White wins after that.
However, as we have seen, Erwich correctly claimed that the position is a draw anyway!
5...Kd4!
Black’s king pursues a heroic journey to its destination.
Incidentally, not 5...Kd5?? as White has a little tactical joke here: 6.Bf7!? (6.Bd7 Bf7 7.Bc6+ is even simpler)
6...Bxf7 7.d7, and the pawn promotes.
6.Bd7
Forcing Black’s bishop to change diagonals.
6...Bb3 7.Bg4 Ba4
White has driven the enemy bishop to the other diagonal, but he won’t get around to building a bridge...
8.Bf3
XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-zP-+-+0
9+-+-+-+-0
9l+-mk-+-+0
9+-+-+L+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, White threatens to build the ultimate bridge on the c6-square, but Black’s king is just in time to stop
this:
8...Kc5! ½-½
Indeed, there is no way for White to make progress. Black’s king is fulfilling a double task: it prevents the
bridge on c6 and keeps hanging on to the d6-pawn.

179
Bishop + pawn vs bishop – 4
Luigi Centurini
La Régence 1856
XIIIIIIIIY
9-+KvL-+-+0
9+P+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-vl0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
White has only one way to try to make progress here: he has to drive Black’s bishop from the b8-h2
diagonal. To that end, he will try to put his own bishop on b8, forcing the enemy bishop to move away.
1.Bh4 Kb6
Black puts a stop to White’s winning plan. If he makes a tempo move with his bishop, White wins as follows:
1...Be5 2.Bf2 Bh2 3.Ba7 Bg3 4.Bb8 Bf2 5.Bf4 Ba7, and now the little trick 6.Be3!, when the black bishop
has to abandon the protection of the b8-square.
2.Bf2+ Ka6
XIIIIIIIIY
9-+K+-+-+0
9+P+-+-+-0
9k+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-vL-vl0
9+-+-+-+-0
xiiiiiiiiy
The king is just in time to defend the a7-square, and so it seems at first sight that White’s plan has failed.
Now, a new winning plan has to be found that consists, again, in driving Black’s bishop from the b8-h2
diagonal:
3.Bc5!!
The only move that wins. Below, we will show why this square is crucial by looking at other bishop moves.
3...Bg3
The bishop has to leave its hiding-place. It doesn’t matter here whether it moves to e5, f4, or g3.
4.Be7
Striving to accomplish the second part of the winning plan: the bishop is on its way to d8 to build a bridge.
4...Kb6 5.Bd8+ Kc6
 XIIIIIIIIY
9-+KvL-+-+0
9+P+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-vl-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
So far, Black seems to have defended adequately. The only difference with the starting position is that his
bishop is not on h2 but on g3 (or e5, or f4), and this subtlety determines the difference between a win and a
draw!
6.Bh4!
Gaining the essential tempo to carry through his original winning plan.
6...Bh2 7.Bf2 Kb5 8.Ba7 Ka6 9.Bb8 Bg1 10.Bf4 Ba7 11.Be3!
XIIIIIIIIY
9-+K+-+-+0
9vlP+-+-+-0
9k+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-vL-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
And it’s game over.
Now that we have seen the winning line, we can see why c5 was the only good square for White’s bishop. If
White moves his bishop to b8 to build a bridge on c7, Black, again, is just in time to keep the bishop from c7:
3.Bh4 Kb6 4.Bd8+ Kc6, and again progress is impossible. So, White will have to gain a tempo somewhere
to prevent the black king from reaching the c6- or the a6-square in time.
If White plays a ‘waiting move’ (= tempo move), for instance, 3.Be3?:
XIIIIIIIIY
9-+K+-+-+0
9+P+-+-+-0
9k+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-vL-+-0
9-+-+-+-vl0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
The only move for Black to defend is 3...Bd6! (not 3...Bg3? 4.Bg5 Kb6 5.Bd8+ Kc6 6.Bh4, and now White
reaches the other diagonal with tempo: 6...Bh2 7.Bf2, and Black can no longer keep the white bishop away
from a7: 7...Bf4 8.Ba7 Bh2 9.Bb8 Bg1 10.Bg3 Ba7 11.Bf2, and White wins; this is the trick from the main
line) 4.Bg5 (picking up the second plan) 4...Kb6 5.Bd8+ Kc6. Again, Black has thwarted White’s plan.
However, since the bishop has now been forced to leave the h2-square (which was a great hiding-place for
it), White can win a tempo to try to get to a7: 6.Be7 (this has to be the idea) 6...Bh2. The bishop returns to
its shelter. Here, we see that e3 was the right square for the white bishop. White cannot play the bishop to
a7 via c5 since the black king is controlling that square.

180
Bishop + pawn vs bishop – 5
Constructed position
XIIIIIIIIY
9-+-mKL+-+0
9+-zP-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+l0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
We have moved the pieces from the starting position of the previous fragment one file to the right. The c8-
h3 diagonal has been shortened by one square, but on the other hand the a6-c8 diagonal has been extended
by one square, and again, this makes the difference between a win and a draw. We try to force the win in a
similar way to the first example:
1.Bh5 Kc6 2.Bf3+ Kb6 3.Bd5
The same tempo moves as in the previous position.
3...Bg4 4.Bf7 Kc6 5.Be8+ Kd6 6.Bh5
Again, we win a tempo now.
6...Bh3 7.Be2 Bf5 8.Ba6 Bg4 9.Bc8
The bishop is driven from the c8-h3 diagonal.
9...Bf3 10.Bh3 Bb7 11.Bg2
In the previous example, this trick was sufficient to force the promotion of the pawn, but now it turns out
that the a6-c8 diagonal is long enough!
A clever move like 11.Bf1 doesn’t help either as long as Black reacts in the right way: 11...Ke6, with a draw
since the a6-c8 diagonal is just long enough: 12.Bh3+ Kd6 13.Bc8 Bg2 14.Ba6 Bh3, and it is clear that
White can make no progress.
XIIIIIIIIY
9-+-mK-+-+0
9+lzP-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+L+0
9+-+-+-+-0
xiiiiiiiiy
11...Ba6 with a draw!
§ 19.2 Opposite-coloured bishops
Before discussing this endgame type, we have to establish that many players think that these endings are
almost always drawn. This is absolutely incorrect, even though there are tendencies that increase the
probability of a draw.
I must admit that I was sometimes ‘groping in the dark’ whenever I had to play, or analyse, an opposite-
coloured bishops ending. In some way, these endings initially contained a touch of mystery for me. Material
seems to be of secondary importance in some of the cases; there are positions in which the weaker side
doesn’t have to lose even with two or three pawns less. There are known cases of players who became
desperate, having conquered material and thinking they had a simple win. Since I have experienced this too,
I can sympathize with club players when they have to play this ending type. Any general orientation is
lacking, and each move seems like a shot in the dark since you don’t know what to play for. We can notice
that very strong players, in one way or another, seem to have the knowledge that tells them which strategy
they should use. Fortunately, there are a few known rules, and we will categorize them.

• Blockade on one colour squares

The big difference between same-coloured bishops endings and opposite-coloured bishops endings is that
in the second case, one side may be superior on one colour and the opponent cannot fight him on that
colour. Therefore, the weaker side may be able to draw even with many pawns less by putting up a
blockade on the colour on which he is stronger – the colour of his bishop!

181
Constructed position
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mkp+-+-0
9-+p+l+-+0
9+-vL-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Even with two extra pawns, Black cannot win here. This blockade on the dark squares leads to a draw –
under the condition, of course, that White maintains the blockade. So, after
1.Bd4+ Kb4
White has to prevent 2...c3+ with
2.Kd2
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+p+-+-0
9-mkpvLl+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
and Black is not able to make an inch of progress and has to settle for the draw.
This example also shows that it is important for the stronger side to have two pawns next to each other on
the same rank, so that the enemy bishop cannot block them. All the same, the defence is still not obvious.
Have a look at the following diagram.

182
End of a study by Olavi Riihimaa
Suomen Sosialidemokraati 1942
XIIIIIIIIY
9-+-vl-+-+0
9+-+-+-+-0
9-+-+k+-+0
9+LzP-+p+-0
9-+-+p+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
What is the problem in this position? For the moment, White’s pawn on c5 is not important, but it is clear
that Black will soon have two pawns on the same rank by playing ...f5-f4. Then, with support of his king, he
will try to push these pawns forward in an effective way. It is preferable to push the pawns to the colour of
the enemy bishop – in this case, the light squares. White will always be able to get his king in front of the
pawns, but the big question is: where should he place his bishop? There are two possible strategies:
A) In front of the black pawns (the g2-square!), taking aim at the pawn on e4;
B) Clinging on to the rear of the black pawns.
These two strategies involve two respective manoeuvres:
A) The manoeuvre Bb5-f1-g2 to take aim at the e4-pawn and post the king on e2, so as to meet ...f4-f3+ with
Bxf3;
B1) Bb5-c6, hanging on to the e4-pawn with the bishop;
B2) Bb5-e8-h5 (in combination with Ke2) so as to then meet ...f4-f3+ with Bxf3.
The choice White makes is very important here!
1.Bf1!!
Method A) is the only right one. Method B) fails miserably.
B1) After 1.Bc6, Black can win as follows: 1...Be7 2.Ke2 Bxc5 3.Bb7 Ke5 4.Bc6 f4 5.Bb7 f3+, and the pawns
will inevitably reach the third rank, forcing promotion after 6.Kd1 Kf4 7.Bc6 e3 8.Bb5 Bb4 9.Bc4 Kg3
10.Bb5 Kf2, and ...e3-e2+ can no longer be prevented;
B2) 1.Be8 Ke5 2.Ke2 f4 3.Bh5.
XIIIIIIIIY
9-+-vl-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zP-mk-+L0
9-+-+pzp-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
It seems like White can stop the pawns for now. 3...Kd4 4.c6 Ke5 5.Bg4 Bb6 6.Bh5 Kf5 7.c7 Bxc7 8.Kf2
Bb6+ 9.Kg2. Now, the white bishop is ‘doomed’ to stay on the d1-h5 diagonal since otherwise ...f4-f3+ can
be played with check. This means that the black king can now go off on a new walk: 9...Ke5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-vl-+-+-+0
9+-+-mk-+L0
9-+-+pzp-+0
9+-+-+-+-0
9-+-+-+K+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
10.Bd1 Kd4 11.Bh5 Ke3, and in this way he can push through ...f4-f3+ after all, winning the game: 12.Bd1
f3+ 13.Kf1 Kf4 followed by ...e4-e3, after which the d1-e2 diagonal is too short, and White will end up in
zugzwang.

1.Bf1!! f4 2.Bg2!
This is the right square for the bishop; 2.Ke2? f3+.
2...Ke5
Now, 2...f3+ fails to 3.Bxf3 exf3 4.Ke3, and Black’s last pawn perishes.
3.Ke2
 XIIIIIIIIY
9-+-vl-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zP-mk-+-0
9-+-+pzp-+0
9+-+-+-+-0
9-+-+K+L+0
9+-+-+-+-0
xiiiiiiiiy
This is the right set-up, guaranteeing the draw for White. His bishop seems to be very passive, but it is
essential for it to take aim at the e4-pawn from the front. The reason is that now Black’s king cannot walk
around to g4 since it is tied to the defence of the e4-pawn. Also, White has just enough space to not end up
in zugzwang (Bg2-h1-g2). White’s bishop fulfils a double function, preventing ...f4-f3+ and at the same time
taking aim at the e4-pawn. Black cannot strengthen his position and has to settle for a draw (which would
also be the case without White’s c5-pawn).
It is important for any practical player to internalize this image so as to be able to make the right choice
during a game.
In the previous example, we saw how White can secure the half point by setting up a kind of fortress with
the king and bishop. Here, it was essential that the enemy pawns had not advanced too far. White always
had a tempo move with his bishop, but that wouldn’t have been the case if the enemy pawns had advanced
one rank further.
With this in mind, Magnus Carlsen knew how to liquidate successfully in the following position; his
opponent wasn’t able to achieve the given set-up with his king and bishop.

183
Magnus Carlsen 2862
Alireza Firouzja 2749
Skilling Prelim 2020 (5)
XIIIIIIIIY
9-+-+-+-+0
9+-zPk+-+-0
9l+-+-zp-+0
9+-+-+P+-0
9-+-vL-mKP+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Carlsen liquidates flawlessly since he has seen that Black is not able to organize his defence in time:
57.Bxf6! Kxc7 58.g5 Kd7 59.g6
With two connected passed pawns on the sixth rank, supported by the king, the win is assured.
59...Ke8 60.Kg5 Kf8 61.Bc3 Bc4 62.f6
 XIIIIIIIIY
9-+-+-mk-+0
9+-+-+-+-0
9-+-+-zPP+0
9+-+-+-mK-0
9-+l+-+-+0
9+-vL-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Even if Firouzja had been able to achieve the set-up with Kg8/Be8, he wouldn’t have been able to save the
game. The bishop’s diagonal is too short, and he would have ended up in zugzwang.
62...Kg8 63.Bb4
In this way, Carlsen controls the promotion square, thereby forcing Black’s bishop to stay on the a2-g8
diagonal. Then, the march of the white king to e7 suffices for the win.
63...Bb3 64.Kf4
And the former Iranian had seen enough. After, for example, 64.Kf4 64...Bc4, the bishop is tied to the a2-g8
diagonal since otherwise f6-f7+ wins immediately: 65.Ke5 Bb3 66.Kd6 Bc4 (66...Bc2 67.f7+ Kg7 68.Ke5
Kxg6 69.f8=Q) 67.Ke7 Bb3 68.f7+, and all is clear.
Now that we have seen the double function of the defending bishop, we can also apply it in a different
situation.

184
Constructed by Stefan Kuipers
2014
XIIIIIIIIY
9-+-+-+-+0
9+-vlL+-+-0
9-+P+-+-+0
9+-+-+-mk-0
9-+-+K+-+0
9+-+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This position was created by the Dutch master Stefan Kuipers for an assignment he was doing for a
trainer’s course that he rounded off successfully under my tuition. In this endgame, we see again the double
function of the defending bishop: it fulfils the task of keeping the c6-pawn under control as well as prevents
the march of the f3-pawn (together with the king). This is sufficient for a draw:
1.Kd5 Kf6
And White has no way to make progress: ½-½
This motif doesn’t work in the next position:

185
Constructed by Herman Grooten
2021
 XIIIIIIIIY
9-+-+-+-+0
9+-vlL+-+-0
9-+P+-+-+0
9+-+-+-mk-0
9-+-+K+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
If we ‘lift the pawn over the b8-h2 diagonal’, the situation changes drastically. The task of Black’s bishop is
reduced to controlling the c6-pawn, and his king now has to take care of the g4-pawn. This turns out to be
more than Black can handle: the white king will leisurely stroll towards b7, and if the black king tags along
to d8 to keep the c7-square covered, White’s g-pawn will decide the issue.
A sample line:
1.Kd4 Kf6 2.Kc5 Ke7 3.Kb5 Bf4
3...Kf6 is no solution either: 4.Ka6 Ke7 (4...Bf4 5.Kb7) 5.Kb7 Kd8 6.g5 Bd6 7.g6 Be5 8.Bf5, and one of
White’s pawns will march on the next move (of course, 8.g7 also wins, as well as 8.c7+).
4.Ka6 Kd8 5.Kb7
Putting Black in zugzwang. 1-0

• Route of the king

What I have learned through the years in opposite-coloured bishops endings is that sometimes material is
of secondary importance only, and control of squares is the main issue. The king plays an important role. If
His Majesty can penetrate the enemy regions, it will be able to fulfil an important task by supporting its
(future) passed pawns. Sometimes, sacrificing a pawn is needed to activate the king.

This endgame in practice

Let’s see how Dutch grandmaster Benjamin Bok outwitted his opponent in a tricky ending, in a blitz game.

186
Benjamin Bok 2884
Homayoon Tofighi 2767
Titled Tuesday blitz 2019
XIIIIIIIIY
9-+-vl-+-+0
9+-+-+-+-0
9-+-mk-+-zp0
9+-+-+P+-0
9-zPK+-+P+0
9+-+-+-+-0
9-+L+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White has two extra pawns, but the win looks anything but simple. Black’s king is quite dominantly placed
on d6, preventing White’s king from walking to the kingside. Therefore, it seeks employment on the
queenside:
63.Kb5! Kc7 64.Ka6
From this square, White’s king supports the advance of his b-pawn.
64...Bf6 65.b5 Bd4
For a moment, Black seems to have everything in order. The b5- and f5-pawns are under control – but not
for long!
66.Be4
XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-+-0
9K+-+-+-zp0
9+P+-+P+-0
9-+-vlL+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
With this tempo move, White puts his opponent in zugzwang! Both Black’s king and his bishop have to stay
on the spot to prevent the march of one of White’s pawns, and that is not allowed.
66...Kb8 67.b6
The pawn is quickly shepherded over the dark square b6.
67...Bf6 68.Kb5 Bd8 69.Kc6 Kc8 70.b7+ Kb8 71.Kd7 1-0
With Black’s king having been driven off its dominant post in the centre, White’s king makes optimal use of
this. With the help of the king, the f-pawn now walks through to the other side. It is striking how important
the role of the white king is in this position.
The power of a king behind enemy lines is also aptly illustrated in the next study:

187
Study by Alexander Herbstman
Shakhmatny Listok 1930
XIIIIIIIIY
9-+-+-+-+0
9+-mK-mk-+-0
9-+-zpP+-vl0
9+-+L+P+-0
9-+-+-+-zp0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
To make progress, White has to find a way to use his foremost passed pawn:
1.f6+! Kxf6 2.Kd7! Bf8 3.Ke8
This forces 3...Be7. Not 3.e7? Bxe7 4.Ke8 – this wins the bishop, but White is left with the bishop of the
wrong colour, and so, even if he wins all the black pawns, a draw is the maximum result.
3...Be7
If 3...Bg7, 4.e7 wins; 3...Kg7 4.h3! (zugzwang!) 4...Kg8 5.e7+.
XIIIIIIIIY
9-+-+K+-+0
9+-+-vl-+-0
9-+-zpPmk-+0
9+-+L+-+-0
9-+-+-+-zp0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
4.h3!
Zugzwang! 1-0
This latest study actually serves quite well as a bridge to the next paragraph.

• Zugzwang
Zugzwang is just about the most horrible scenario for a chess player. It means nothing less than that every
(legal) move you can play fails for some reason. It is hard to describe how helpless you feel at such
moments. For the opposite side, it is a powerful weapon to win from certain positions, simply by forcing the
opponent to move and play something that loses immediately.
We see a quite dramatic example in the following blitz game.

188
Narayanan Sunilduth Lyna 2914
Joshua Colas 2566
Titled Tuesday blitz 2020
XIIIIIIIIY
9-+-+-+-mk0
9+-zp-+-zp-0
9-+P+K+-+0
9zp-+-+-zPP0
9P+Lvl-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
57.Kd7
Now, Black is forced to protect the c-pawn. He does this in the wrong way, and as a result ends up in
zugzwang:
57...Bb6??
With 57...Be5!, Black could have achieved a relatively simple draw.
58.g6!
 XIIIIIIIIY
9-+-+-+-mk0
9+-zpK+-zp-0
9-vlP+-+P+0
9zp-+-+-+P0
9P+L+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black’s king is trapped, and his bishop has to let go of c7, so Black is definitively lost.
How cruel our game can be...
The next study is another instructive case:

189
Study by Ladislav Prokes
Schweizerische Arbeiter Schachzeitung 1949
XIIIIIIIIY
9-+K+-+-+0
9+-+-+-+-0
9-+P+kvL-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+l+-+-0
xiiiiiiiiy
White to play and win
Both sides have only one pawn left, and a draw seems to be within Black’s reach. Nevertheless, there are a
few factors that bring about his downfall after all:
1.c7
Obviously, White pushes his passed pawn as quickly as possible, and now Black’s bishop has to stop it in
one way or another.
1...Ba4
The alternatives lose as well: after 1...Be2, Black threatens 2...Ba6+, with a draw, but White replies strongly
with 2.Kb7! Bf3+ 3.Kb8, and 4.c8=Q(+) cannot be prevented. Also 1...Bb3 fails immediately, this time
because of 2.Kb8.
2.Kd8
This is the right square for the king.
2...Bd7
But this is an extremely unfortunate square for the bishop. White’s reply is quite adequate:
3.Be5!
 XIIIIIIIIY
9-+-mK-+-+0
9+-zPl+-+-0
9-+-+k+-+0
9+-+-vL-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Underlining the poor positioning of Black’s pieces.
3...g3 4.Bxg3
With zugzwang. Neither Black’s king nor his bishop can make a move without losing immediately.

• Creating a (second) passed pawn

An important strategy for the stronger side is to create a second passed pawn (often he already has one) to
force the decision. As a rule, it is best if the passed pawns are as far removed from each other as possible.
First, we will look at a nice little example from practice, and then at a beautiful study.

190
E.Enkhnar 2309
Agastya Kalra 1956
Al-Ain WYCC U16 2013 (1)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+-0
9k+p+-+p+0
9zP-vL-+lmK-0
9-zP-+-zP-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
White to move
White realizes that he won’t be able to win on the queenside (due to the blockade on the light squares), so
he will have to strike on the kingside. With his following move, he threatens 43.h5:
42.h4 Be6
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+-0
9k+p+l+p+0
9zP-vL-+-mK-0
9-zP-+-zP-zP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The bishop leaves the f5-square, to avoid ‘hanging’ after h4-h5. But now, Black is in for a nasty surprise!
43.f5! Bxf5
After 43...gxf5, the h-pawn walks through, e.g. 44.h5 f6+ (Black is too late if he plays the bishop to the b1-h7
diagonal, since he also has to spend a tempo on ...f5-f4, and he doesn’t have the time for that) 45.Kxf6 Bg8
46.h6 f4 47.Kg7, and the g8-h7 diagonal is too short for the Bg8.
44.h5
The second passed pawn has been created, and stopping it will cost Black a piece. Therefore, he resigned. A
possible continuation is 44...gxh5 (44...Bd3 45.h6 f6+ 46.Kxf6 g5 47.Kxg5 followed by Kg5-g6-g7, and h6-
h7 will eventually cost Black his bishop too) 45.Kxf5, and the remaining endgame is a simple win because
White can easily eliminate Black’s two kingside pawns.

191
Study by Gerd Hörning
Problem Forum 2007
XIIIIIIIIY
9-+-+-+-+0
9+K+-+-+-0
9-+-zp-mkP+0
9zp-zp-+-+-0
9P+-+Pvl-+0
9+-+L+-+-0
9-zP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
White is a pawn up, but that doesn’t seem very significant. It looks as if Black will be able to protect his
pawns with his bishop, and moreover the g6-pawn threatens to be lost as well. However, an unexpected
catastrophe awaits him...
1.b4!
A brilliant concept!
The move order is very intricate here. First 1.e5+? isn’t correct:
A) Black mustn’t take with the bishop since he will then lose the a-pawn, and that is what White has been
aiming for all along: 1...Bxe5? 2.Kb6! c4 3.Bb1! Bxb2 (3...d5 4.Kxa5 d4 5.Kb5 d3 6.b4, and White wins)
4.Kxa5 d5 (Black’s pawns look impressive, but they will be easily controlled by White’s pieces, after which
the two distant passed pawns decide the game in White’s favour) 5.Kb5 Bc3 6.Kc5 Ke6 7.Bc2 Ba5
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+k+P+0
9vl-mKp+-+-0
9P+p+-+-+0
9+-+-+-+-0
9-+L+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
8.Bf5+! Kf6 9.Kxd5;
B) 1...dxe5!, and here, 2.b4 is nonsense (though 2.Kb6 Bd2 brings White nothing either). After 2...axb4! 3.a5
XIIIIIIIIY
9-+-+-+-+0
9+K+-+-+-0
9-+-+-mkP+0
9zP-zp-zp-+-0
9-zp-+-vl-+0
9+-+L+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
3...c4! 4.Bxc4 (4.Be4?? b3 5.a6 Be3 6.a7 Bxa7 7.Kxa7 c3, and Black even wins) 4...Kxg6, the crucial g6-
pawn leaves the board.

Also 1.Kc6? Bd2! 2.Kxd6 (2.Bc4 Kxg6 3.Kxd6 Kg7 4.e5 Kf8 5.e6 Bg5) 2...Kxg6 3.e5+ Kg7 leads to a draw,
e.g. 4.e6 c4! 5.Bxc4 Bb4+ 6.Kd7 Kf6.

If 1.Kb6?, 1...Bd2!;

1.Kc7? Bd2!.

1.b4! cxb4
White’s win after 1...axb4 is almost a study in itself: 2.a5 c4 3.Bxc4 Kxg6 4.a6 Be3 5.Kc6! Bc5
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9P+Kzp-+k+0
9+-vl-+-+-0
9-zpL+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
6.e5!. Thus, White obtains a second passed pawn, breaking all further resistance.
2.e5+!
This zwischenzug is essential, since White is banking on his two distant passed pawns.
2...dxe5
XIIIIIIIIY
9-+-+-+-+0
9+K+-+-+-0
9-+-+-mkP+0
9zp-+-zp-+-0
9Pzp-+-vl-+0
9+-+L+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
3.Kb6
Now, Black cannot save the a5-pawn, and thus White creates the necessary second passed pawn.
3...b3 4.Kxa5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mkP+0
9mK-+-zp-+-0
9P+-+-vl-+0
9+p+L+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Here, there is no longer any salvation for Black. It is essential that White’s bishop fulfils several important
tasks on the b1-h7 diagonal. Besides keeping the g6-pawn covered, it also controls both of Black’s passed
pawns, so that they won’t be able to stop White.
• The fortress
A special phenomenon in the endgame is the fortress. We may encounter this motif in many different
ending types, but it is especially frequent in opposite-coloured bishops endings.
We will limit ourselves to two examples.
192
Joel Lautier 2666
Sergey Rublevsky 2670
Poikovsky 2003 (7)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+pmK-0
9-+-+k+-+0
9+-+-+-vL-0
9-+-+-zPlzP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
This position is presented by Karsten Müller and Wolfgang Pajeken in their book How to Play Chess Endings
(2008) as an example of a fortress.
Black shows here what can sometimes be possible in this type of ending. He ostentatiously gives up the f7-
pawn to erect a blockade on the light squares with no less than three minus pawns:
47...Kf5!
Not necessary, but nice!
48.Kxf7 Bh5+ 49.Kg7 Bd1 50.Be7
And draw agreed.
If White plays 50.Kh6 and after 50...Be2, 51.h5, Black simply keeps his bishop glued to the h5-pawn along
the d1-h5 diagonal to prevent White’s king from moving.
Former World Champion Vasily Smyslov, who was known for his love of endgame studies, composed the
following gorgeous one.

193
Study by Vasily Smyslov
New In Chess #7, 2002
XIIIIIIIIY
9-+-+-+-+0
9+-+pmk-+-0
9-+-+p+lvL0
9+-+-zPp+-0
9pzP-+-zPp+0
9+-+-+-zP-0
9-+-zP-+-zP0
9+-+-+-mK-0
xiiiiiiiiy
White to play and draw
1.Bg5+!
White’s bishop is not able to stop Black’s dangerous a-pawn. But White has to start with this move, to
prevent Black’s king from entering the game via d8.
1.b5? would be nonsense in view of 1...a3, and Black queens with check;
and 1.d4? is met by 1...Kd8!.

1.Bg5+! Ke8
Now the big question is; what can White undertake here? There is no remedy against the march of the black
a-pawn.
2.d4
This looks mysterious... what is White planning to do?
2...a3
Importantly, if Black tries to free himself with 2...d5, White replies with 3.exd6 a3 4.Bf6 a2 5.d5, when he is
just in time to open the long diagonal, keeping the a-pawn under control.
3.d5
He continues unruffled.
3...a2
If 3...exd5, again the long diagonal is opened just in time: 4.e6 dxe6 5.Bf6, and White has a blockade on the
dark squares.
XIIIIIIIIY
9-+-+k+-+0
9+-+p+-+-0
9-+-+p+l+0
9+-+PzPpvL-0
9-zP-+-zPp+0
9+-+-+-zP-0
9p+-+-+-zP0
9+-+-+-mK-0
xiiiiiiiiy
4.d6!
White closes the pawn chain and simply allows Black to promote!
4...a1=Q+ 5.Kg2!
Black has an extra queen, but there is no way he can win this position. White cannot be put in zugzwang
since he can always make a move with his bishop. A curious fortress!
5.Kf2? Qh1! 6.Bh4 Qxh2+ 7.Ke3 Qxh4.

This endgame in practice

It is always interesting to watch endgames played by (super-)grandmasters. In the databases, I found a
superb but also highly complicated endgame between Magnus Carlsen and Alexander Morozevich, who was
a former top-10 player. The endgame was extensively and instructively analysed by Mihail Marin. I am
always charmed by the analyses and explanations of this diligent Romanian grandmaster, who knows how
to serve a large audience with his commentaries. He can be forgiven for missing something with the
engines of the day. With the help of the latest technology, I found something very beautiful!
White has an extra pawn, but Black is active, and due to the presence of opposite-coloured bishops, the win
looks far away, if not impossible. But thanks to a ‘coincidence’ in the position, things suddenly look a lot
rosier for White after the next move!

194
Magnus Carlsen 2690
Alexander Morozevich 2741
Morelia/Linares 2007 (1)
 XIIIIIIIIY
9-+-+-+-+0
9+-+-mk-+p0
9-+-+-+p+0
9zp-+-+-+-0
9-+LtrpvlP+0
9+P+-+-+P0
9P+-+RzPK+0
9+-+-+-+-0
xiiiiiiiiy
White to move
43.Bd3!!
A brilliant discovery, winning a second pawn and thereby increasing his winning chances enormously.
43...Kf6
It is clear that Black would have ended up in a totally lost rook ending after 43...Rxd3 44.Rxe4+.
44.Bxe4 Rd2
White cannot prevent the rooks being exchanged, but with two potential passed pawns at a respectful
distance from each other, his chances of winning are considerable.
45.Rxd2 Bxd2 46.Kg3 Be1 47.Kf3 Bb4 48.h4 h6 49.Ke2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mkpzp0
9zp-+-+-+-0
9-vl-+L+PzP0
9+P+-+-+-0
9P+-+KzP-+0
9+-+-+-+-0
xiiiiiiiiy
49...Bd6
Here, Marin writes: ‘Morozevich’ plan is hard to understand. The best chance for survival was blocking the
white pawns on light squares with 49...g5 when White still needs to prove a winning plan. See the comment
on the next move.’
50.Kd3
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-vl-mkpzp0
9zp-+-+-+-0
9-+-+L+PzP0
9+P+K+-+-0
9P+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
Now, we leave the game and take Marin’s following suggestion as the starting point of our analysis. At this
point, Black could again put up a blockade on the dark squares with:
50...g5?!
In the game, 50...Bc5 was played, and it looks like Black could have saved the game: 51.f4 (thus, Carlsen
prevents the fixing of the structure with ...g6-g5, but he can’t prevent Black from fixing the pawns on the
other colour. It looks like Morozevich had estimated well how he had to handle this endgame, and he made
a mistake only much later) 51...h5! (now, Morozevich fixes the pawns on the colour that is favourable for
him, and this could have saved the game for him) 52.g5+ Kg7 (with such a passive king, Black seems to be
beyond salvation, but the practice is different!) 53.Kc4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+p+0
9zp-vl-+-zPp0
9-+K+LzP-zP0
9+P+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
53...Bd6 (this doesn’t spoil anything, but the alternative 53...Bf2 came strongly into consideration since the
bishop cannot let go of a5 in any circumstance: 54.Kb5 Be1 55.a3 Kf7 56.b4 axb4 57.a4 (White’s best
chance is the distant a-pawn) 57...Bxh4 58.Bd5+ (58.a5 Bf2 59.Kxb4 h4 60.Kc4 h3 61.Kd3=) 58...Ke7
59.a5 Bf2 60.Kxb4 h4.
XIIIIIIIIY
9-+-+-+-+0
9+-+-mk-+-0
9-+-+-+p+0
9zP-+L+-zP-0
9-mK-+-zP-zp0
9+-+-+-+-0
9-+-+-vl-+0
9+-+-+-+-0
xiiiiiiiiy
Since Black has now created a passed pawn as well, he is able to save the game. A possible continuation is
61.Bg2 Ke6 62.Be4 Be3 63.a6 (63.Bxg6 Bxf4=) 63...Kf7 64.Kb5 h3 65.Kc6 h2 66.Kb7 Ke6 67.a7 Bxa7
68.Kxa7 h1=Q 69.Bxh1 Kf5, and the last of White’s pawns leave the board) 54.Kb5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-vl-+p+0
9zpK+-+-zPp0
9-+-+LzP-zP0
9+P+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
54...Bxf4?? (it is only at this point that Morozevich throws away his drawing chances. As mentioned above,
Black shouldn’t hand the a-pawn over to White without a fight: 54...Bb4.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+p+0
9zpK+-+-zPp0
9-vl-+LzP-zP0
9+P+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, the play runs along similar lines as we saw in the above-mentioned variation. Black’s scheme is clear:
- the bishop keeps clinging to the a5-pawn. Because of this, White has nothing better than trading off a
pawn to create a passed pawn.
- as soon as White has a passed pawn on the queenside, Black’s bishop picks up the h4-pawn and then
heads back to the g1-a7 diagonal.
- in the long run, Black will have to give his bishop for White’s passed pawn.
- in the meantime, his pawn marches to h2, after which he can play ...Ke6.
- at the right moment, Black lures White’s bishop away with ...h1=Q and then he plays ...Ke6-f5, conquering
both of White’s passed pawns.
55.Ka4 Bd2 56.Bd5 Bc3 57.a3 Be1 58.Kb5 Kf8 59.b4 axb4 60.a4 Bxh4 61.a5 Ke7 62.Kxb4 Be1+ 63.Kb5
h4 64.a6 Bf2 65.Kc6 h3 66.Be4 h2 67.Kb7 Ke6 68.a7 Bxa7 69.Kxa7
XIIIIIIIIY
9-+-+-+-+0
9mK-+-+-+-0
9-+-+k+p+0
9+-+-+-zP-0
9-+-+LzP-+0
9+-+-+-+-0
9-+-+-+-zp0
9+-+-+-+-0
xiiiiiiiiy
69...h1=Q 70.Bxh1 Kf5 – mission accomplished!) 55.Kxa5. Now, with two passed pawns on the queenside,
the win is a piece of cake for Carlsen: 55...Bg3 56.Kb5 Bxh4 57.a4 Bxg5 58.a5.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+p+0
9zPK+-+-vlp0
9-+-+L+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
A race has started between White’s queenside pawns and Black’s kingside pawns, but Black is far behind...
58...Kf6 59.a6 Be3 60.Kc6 g5 61.b4 Ke5
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9P+K+-+-+0
9+-+-mk-zpp0
9-zP-+L+-+0
9+-+-vl-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
62.b5 (he can even give up the Be4! There was also nothing wrong with 62.Bh1) 62...Kxe4 63.b6 g4 64.a7
g3 65.a8=Q Kf3 66.b7 Bf4 67.Qf8 Ke4 68.Qe8+, and Black resigned; if 68...Kf3, 69.Qxh5+ and further
resistance is futile. A marvellous endgame performance by Carlsen.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-vl-mk-zp0
9zp-+-+-zp-0
9-+-+L+PzP0
9+P+K+-+-0
9P+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
Marin: ‘It turns out that White is winning, but certain technical effort is needed still.’ Next, the Romanian
GM gives a plausible sequence of moves in which White has to be quite inventive.
Since Marin found an ingenious winning method, I am happy to present it. Alas, however, I have to conclude
that his assessment that ...g6-g5 is Black’s best chance to draw is not correct. Precisely the fixing of the
pawns (as happened in the game with ...h6-h5) turns out to be the right method to achieve the draw.
Nevertheless, I am impressed by Marin’s discoveries at the time. Today, in 2022, we have much stronger
engines at our disposal, even with artificial intelligence, allowing us to establish with almost certainty that
the Russian could have drawn this ending, despite a number of factors indicating otherwise.
51.h5!
The only way for White to try to win. The pawn on h5 plays an important role in the winning plan. If Black’s
king walks to the queenside to defend against the coming white passed pawn, White, in turn, can move his
king to the kingside to take the h6-pawn from the board and create a second passed pawn. It is essential
that, with time, ...Bf8 can be answered with Kf7 followed by Kg7, after which the h6-pawn falls.
It is clear that White needs to have deep strategic insight to win here!
51...Ke5 52.f3 Bb4 53.Kc4 Ba3
If Black hurries with his king to the queenside too early with 53...Kd6?!,
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-zp0
9zp-+-+-zpP0
9-vlK+L+P+0
9+P+-+P+-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
White can also create a second passed pawn with the violent break 54.f4!. An incredible concept. White
sacrifices two pawns to create a distant passed pawn on the h-file. After that, he will also have to create a
second passed pawn on the queenside to force his opponent to take care of that pawn. The following run of
play, indicated by Marin, shows us how White can win, but it needs at least one more study-like idea!
54...gxf4 (54...Bd2 55.f5, and White wins because he can create two distant passed pawns) 55.g5 hxg5 56.h6
Ba3.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-zP0
9zp-+-+-zp-0
9-+K+Lzp-+0
9vlP+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Black’s bishop has to move to the long diagonal, but now White comes up with his winning idea: 57.b4!!
Bb2 (the point is 57...axb4 58.Kb3!, after which the h-pawn promotes unhindered!) 58.bxa5 (now, White
has three pawns on the edge!) 58...Kc7 (this is obligatory – 58...g4 loses to 59.a6 Kc7 60.a7, and the pawn
promotes) 59.Bf3 (now, White needs to stop the black pawns) 59...Kb8 60.Kd5 (White’s plan is clear: the
king is on its way to g5 to pick up that pawn, and after that, the monarch also has to eliminate the f4-pawn)
60...Ka7 61.Ke6 Ka6 62.Kf5 Kxa5 63.Kxg5 Kb4.
Another important moment.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-mK-0
9-mk-+-zp-+0
9+-+-+L+-0
9Pvl-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
How can White make progress? Black is threatening 64...Ka3 followed by 65...Kxa2, so Bd5 will be
necessary to save the pawn, but that means that Black’s f-pawn will become a nuisance. 64.Kg4! (Black’s
passed pawn has to be kept under control since otherwise the a2-pawn will fall; certainly not 64.Kxf4?
Bc1+ 65.Kf5 Bxh6=) 64...Ka3 65.Bd5 Be5 66.h7 Kb4 67.Kf5 (this is essential; Black’s bishop cannot
protect the f4-pawn and stop the h7-pawn at the same time) 67...Bh8 68.Kxf4.
XIIIIIIIIY
9-+-+-+-vl0
9+-+-+-+P0
9-+-+-+-+0
9+-+L+-+-0
9-mk-+-mK-+0
9+-+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
And so White is definitively two pawns up, and he will also conquer the black bishop on h8. Please note that
it would still be a draw if White had the bishop of the wrong colour here! After conquering the black bishop
for the h-pawn, he will be left with the a-pawn and therefore also with the good bishop.
54.Kb5 Bb4 55.Ka4 Bd2 56.a3
This is the way to set the pawn majority in motion.
56...Kd6 57.Kb5 Be1 58.b4 axb4 59.axb4 Bd2 60.Kc4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-zp0
9+-+-+-zpP0
9-zPK+L+P+0
9+-+-+P+-0
9-+-vl-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black’s king stays on d6 as long as possible. Apart from blocking the white king’s passage to the kingside, on
d6 he keeps the choice between controlling either the b-pawn or the kingside. This means that Black has to
make a move with his bishop, and then the question is whether it has to stay on the c1-h6 diagonal or on
the e1-a5 diagonal.
60...Bc1
After 60...Be1? we see again the familiar breakthrough: 61.f4! gxf4 62.g5 hxg5 63.h6. Here the game is even
over immediately since the bishop cannot reach the long diagonal.
After the text move, White plays the clever
61.Bd5!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-zp0
9+-+L+-zpP0
9-zPK+-+P+0
9+-+-+P+-0
9-+-+-+-+0
9+-vl-+-+-0
xiiiiiiiiy
depriving Black’s king of many squares and in the meantime enabling Kc4-d4-e4 by vacating the e4-square.
61...Bf4
After 61...Bb2, 62.b5 wins; the pawn reaches b6 (and thereby also b7) after which White has free play.
The logical 61...Be3 also fails to 62.Kd3,
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-mk-+-zp0
9+-+L+-zpP0
9-zP-+-+P+0
9+-+KvlP+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
when the king can move to the juicy e4-square with gain of tempo, after which the road is open for
penetration on the king’s wing.
62.Kd4 Bd2 63.b5 Ba5 64.Ke4 Bb6 65.Ba2
He doesn’t even need the b5-pawn anymore.
65...Kc5 66.Kf5 Kxb5 67.Kg6
And White conquers the h6-pawn and wins.
Chapter 20
Rook endings
§ 20.1 Introduction
As a secondary school student, I met another pupil who was two years older and who turned out to be a
good chess player. This student, Huub van Dongen, went on to become one of my best friends. He was busy
setting up a school chess team and was looking for other chess-playing students. Thus, we came into
contact, and it wasn’t long before he also dragged me along to the local chess club. Since we shared several
hobbies, it wasn’t surprising that soon we were playing hundreds of blitz games with each other. Huub also
introduced me into the world of the endgame. With several well-known books, we tried to familiarize
ourselves with the principles of the endgame. Of course, there were many ups and downs, but eventually
we familiarized ourselves with a number of rules. There was one remark by Huub I will never forget – we
kept repeating it as a kind of mantra whenever we would stumble upon something amazing: ‘Rook
endgames are difficult, that’s because these rooks move so straight!’ Alas, my good friend passed away in
2012 at 54 – totally unexpectedly, due to heart failure. But when I am studying endgames, I sometimes feel
that he is close by, and I can hear him proclaiming his typical metaphors.

§ 20.2 Rook behind the pawn

During our quest through the various endgames, we soon found out that a good knowledge of rook endings
was indispensable for any self-respecting club player. Statistics once established that of all the endgame
types in the databases, approximately 50% are rook endings. This means that focusing on this endgame
type will certainly pay dividends. Let’s start with one of the most popular slogans about the rook endgame:

Rule of thumb
The rook belongs behind a passed pawn – both behind its own passed pawn and behind an
enemy passed pawn.

195
Constructed position
XIIIIIIIIY
9-+-+-+-+0
9+kzp-+-+-0
9-+-+-+-+0
9+-+-zp-zP-0
9-+K+-+-+0
9+-zP-+-+-0
9-+-+-+-tR0
9+-+-+r+-0
xiiiiiiiiy
White to move
White immediately puts the rook behind his own passed pawn, before the opponent puts his rook there:
1.Rg2!
And this is immediately winning for White. Why? Because thanks to the rook, White can push the pawn
forward, forcing Black’s rook to stand in front of it. Then, White has an active rook (which can move up and
down along the g-file), while Black’s rook is completely passive, and eventually Black will end up in
zugzwang. Things could develop as follows:
1...Kc6 2.g6 Rf4+ 3.Kd3 Rf8 4.g7 Rg8 5.Ke4 Kd6 6.Rg6+ Kd7 7.Kxe5 Ke7 8.Rg4 Kf7 9.Kd5
XIIIIIIIIY
9-+-+-+r+0
9+-zp-+kzP-0
9-+-+-+-+0
9+-+K+-+-0
9-+-+-+R+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black cannot go for the pawn ending, but he can switch the positions of his rook and king:
9...Rd8+
This is the standard defence in principle, but it doesn’t work here.
10.Kc6 Kg8 11.Kxc7
And White wins with his two extra pawns.
Now, we look at the same position with Black to move.

196
Constructed position
XIIIIIIIIY
9-+-+-+-+0
9+kzp-+-+-0
9-+-+-+-+0
9+-+-zp-zP-0
9-+K+-+-+0
9+-zP-+-+-0
9-+-+-+-tR0
9+-+-+r+-0
xiiiiiiiiy
Black to move
Black achieves an easy draw here by putting his rook behind White’s passed pawn. The intermediate check
is useful:
1...Rf4+
The direct 1...Rg1 is also sufficient after 2.Rh5 Rg4+ 3.Kd5 e4.

1...Rf4+ 2.Kd5 Rg4 3.Rh5

White’s rook is extremely passive.
3...e4
Black’s pawn provides enough counterplay.
4.Ke5 e3 5.Kf5 e2
White’s king has reached his own passed pawn.
6.Rh1 Rg2
 XIIIIIIIIY
9-+-+-+-+0
9+kzp-+-+-0
9-+-+-+-+0
9+-+-+KzP-0
9-+-+-+-+0
9+-zP-+-+-0
9-+-+p+r+0
9+-+-+-+R0
xiiiiiiiiy
Now, White has to watch out for a bit: Black is even threatening to win with 7...Rf2+ followed by 8...Rf1.
7.Re1 Kc6 8.Kf4
White has to run in order not to lose. It’s a draw now, since the two passed pawns (g5 and e2) will both be
captured. White loses if he makes a final winning attempt: 8.Kf6?? Kd5 9.g6 Kc4 10.g7 Kd3 11.Kf7 Kd2–+.
Black’s defence didn’t work in the first position, but the defender’s technique of putting the king in front of
the enemy pawn is a very important one:

197
Constructed position
XIIIIIIIIY
9-+-+-+-+0
9+-zp-mk-+-0
9-+-+-+P+0
9+-+-zp-+-0
9-+K+-+-+0
9+-zP-+-+-0
9-+-+-+R+0
9+-+-+r+-0
xiiiiiiiiy
Black to move
We take the same schematic position, with one important change: Black’s king is on e7 instead of b7. This
makes a world of difference: Black is now able to get his king in front of White’s dangerous passed pawn.
1...Kf8! 2.g7+ Kg8
After this manoeuvre, White’s rook cannot achieve much. Even worse: in principle, White’s king cannot
beat Black’s rook. A plausible line runs as follows:
3.Kd5 Re1 4.Kc6 e4 5.Kxc7 e3 6.c4 Rc1
 XIIIIIIIIY
9-+-+-+k+0
9+-mK-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
9-+P+-+-+0
9+-+-zp-+-0
9-+-+-+R+0
9+-tr-+-+-0
xiiiiiiiiy
7.Re2
And it’s high time for White to settle for a draw... he shouldn’t think that he can still try for a win here, since
for example 7.Rg4?? loses to 7...e2.
7...Rxc4+ 8.Kd6 Re4 9.Kd5 Re7 10.Kd4
With a draw.
We formulate this technique as follows:

Technique
If the opponent has his rook behind his passed pawn, the defender has to try to place his king
in front of the passed pawn, so as to be able to use his rook elsewhere on the board.
The reason for this is easy to understand. In rook endings, the king is an excellent blocking piece, ‘reducing’
the enemy rook to the defence of its pawn, and limiting its activity considerably. On the other hand, the
defender’s rook can suddenly play freely against the enemy king, and this will lead to an entirely different
kind of struggle.
Here is a nice practical example in which White nicely manages to exploit his surplus of tempi:

198
Yury Balashov 2570
Fridrik Olafsson 2570
Moscow 1971 (15)
XIIIIIIIIY
9-+-+-+-+0
9zp-+-+-+-0
9-+-+-+-+0
9+-+k+-zP-0
9-+p+p+-+0
9zPr+-+-+-0
9-zPK+-+R+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
White has managed to get his rook behind the passed pawn. Since the black king is slightly unfortunately
placed, it cannot creep in front of the pawn, so he will have to put his rook in front of it at a certain moment
– which means his death sentence:
50...e3
Also, 50...Rb7 51.g6 Rg7 52.Kd2 Kd4 doesn’t help.
 XIIIIIIIIY
9-+-+-+-+0
9zp-+-+-tr-0
9-+-+-+P+0
9+-+-+-+-0
9-+pmkp+-+0
9zP-+-+-+-0
9-zP-mK-+R+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Here, we see a beautiful example of the power of a rook behind a passed pawn: 53.Rg4! (White’s rook can
move freely, and so Black is slowly driven back) 53...Ke5 54.Ke3 Kf5 55.Rf4+, with a position similar to the
game.

50...e3 51.g6 Rb8

When you are losing, it’s not a bad idea to try your luck sometimes: Black could have tried 51...e2, when he
draws after 52.Rxe2 Rg3. However, White wins with 52.g7 e1=Q 53.g8=Q+, queening with check, after
which his attack decides, for example: 53...Kd4 54.Rg4+ Ke3 55.Qg5+ Kf2 56.Qh4+ Kf1 57.Rf4+, and the
rest is clear.
52.g7 Rg8 53.Kd1
XIIIIIIIIY
9-+-+-+r+0
9zp-+-+-zP-0
9-+-+-+-+0
9+-+k+-+-0
9-+p+-+-+0
9zP-+-zp-+-0
9-zP-+-+R+0
9+-+K+-+-0
xiiiiiiiiy
53...Ke4 54.Ke2 Kf4 55.Rg1!
White has a good amount of waiting moves, but Black doesn’t.
55...Ke4
Black could also make some pawn moves here, but White will calmly wait until these are exhausted: 55...a6
56.a4 a5 57.Rg2 Ke4 58.Rg3, and White wins.
56.Rg3 Kf4
This is the moment for White’s rook to leave its ideal placement:
57.Rf3+ Kg5 58.Kxe3
 XIIIIIIIIY
9-+-+-+r+0
9zp-+-+-zP-0
9-+-+-+-+0
9+-+-+-mk-0
9-+p+-+-+0
9zP-+-mKR+-0
9-zP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black resigned. His problem is that any pawn ending will be lost for him. This means that he cannot
eliminate White’s g-pawn yet. A possible continuation is 58...Kg6 59.Rg3+ Kf7 60.Kd4 Rd8+ (the only way
to get his king in front of White’s passed pawn) 61.Kxc4 Kg8. Now the king has reached g8, but Black’s fate
is sealed with two pawns less: 62.b4 Re8 63.Kb5 Re6 64.a4 Rf6 65.Rg4 Re6 66.Kc5 Rf6 67.b5 Re6 68.a5 etc.

§ 20.3 The ‘shoulder budge’ in rook endings

Just as in pawn endings, the ‘shoulder budge’, reminiscent of rugby and American football, can be a very
useful technique also in rook endings.
We can find the following diagram in various endgame books, since it provides a nice picture of a ‘shoulder
budge’ between the two kings.
We have already described the required Technique in the chapter on pawn endings: the enemy king is
held off.
White has just played 70.b5-b6, threatening to win Black’s rook. Now the question is: what is Black’s best
reply?

199
Alexander Alekhine
Efim Bogoljubow
GER/NED Wch m 1929 (19)
XIIIIIIIIY
9-+-tr-+-+0
9+-+-+-+-0
9-zPK+-zp-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+R+-+-+-0
xiiiiiiiiy
Black to move
70...Kg4??
A blunder that has found its place in chess history books.
A player of today would doubtlessly find 70...Ke4! here. The struggle of rook vs pawn, which will shortly
appear on the board, can only be drawn if Black’s king blocks the path of its opposite number to the
promotion square of Black’s f-pawn. A possible continuation is 71.b7 f5 72.b8=Q Rxb8 73.Rxb8 f4 74.Re8+
 XIIIIIIIIY
9-+-+R+-+0
9+-+-+-+-0
9-+K+-+-+0
9+-+-+-+-0
9-+-+kzp-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
74...Kd4! (the ‘shoulder budge’! There is no way for White’s king to get to the black pawn – it will be much
too late at any rate. Certainly not 74...Kf3??, which would still go wrong for Black: 75.Kd5 Kg2 76.Ke4 f3
77.Rg8+ Kf2 78.Rf8; however, 74...Kd3 is a good alternative) 75.Kd6 f3 76.Rf8 Ke3 (76...Ke4 is also good!)
77.Ke5 f2, and White can’t prevent Black from promoting.

70...Kg4?? 71.b7 f5
71...Rb8 72.Kc7 Rh8 73.b8=Q Rxb8 74.Rxb8 doesn’t make any difference:
XIIIIIIIIY
9-tR-+-+-+0
9+-mK-+-+-0
9-+-+-zp-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
White’s rook will now also get behind the pawn, securing the win: 74...f5 75.Kd6 f4 76.Ke5 f3 77.Ke4 f2
(77...Kg3 78.Rb3) 78.Rf8 Kg3 79.Ke3, and White wins the pawn just in time.
72.b8=Q Rxb8 73.Rxb8 f4 74.Kd5
The king hurries towards the pawn and does not meet its opposite number this time!
74...f3 75.Ke4 f2 76.Rf8 Kg3 77.Ke3
And Black resigned.

This endgame in practice

‘Shoulder budges’ mainly occur in rook vs pawn endings. The following position is from a game against one
of my former club mates with whom I have had quite a few scuffles over the years. Now, he is a strong
International Master with two grandmaster norms, and he once had nearly 2500 Elo. However, a busy
career and family life prevented him from actually obtaining the grandmaster title.
In a complicated ending, we both made several mistakes.

200
Frans Cuijpers
Herman Grooten
Dieren 1988 (5)
 XIIIIIIIIY
9R+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-zp-+-0
9-+-mk-+-+0
9+-+-+-mK-0
9ptr-+-+PzP0
9+-+-+-+-0
xiiiiiiiiy
White to move
After a long game, both sides have two passed pawns – the one on a2 has already reached the penultimate
rank. Clearly, this is a race, and it looks as if White will be too late.
63.h4
On the face of it, White seems to have sufficient counterplay with his h-pawn, but now Black sets his second
passed pawn in motion:
63...e4 64.Ra4+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9R+-mkp+-zP0
9+-+-+-mK-0
9ptr-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
64...Kd5?
An ugly mistake that should have cost Black half a point. With hindsight, the move is unnatural as well.
However, it was motivated by the idea that I would have to be fighting with my rook vs two white pawns
later on. Then, the king would have to stop the two passed pawns, and for that purpose d5 is the right
square!
However, supporting the e-pawn with 64...Kd3 would have secured the win: 65.Ra3+ Kd2 66.h5 (66.Ra8
e3 67.Rd8+ Kc1, and White can’t prevent one of the two pawns promoting: 68.Ra8 e2 69.Kf2 Kd1 70.Rd8+
Rd2) 66...e3, and one of Black’s pawns gets through.
65.h5?
Fortunately for me, Frans goes wrong too. Here, the ‘push pawn’ adage does not apply, even though the
move seems perfectly logical.
With 65.Kf4!, a kind of ‘shoulder budge’ in the direction of the colleague king, White could have secured the
draw as he keeps the e-pawn under control: 65...Rxg2 (65...Rf2+ 66.Ke3 Rxg2 67.h5 is a draw as well) 66.h5
Rf2+ 67.Kg3! (not 67.Ke3? due to the notorious X-ray check 67...Rf3+ 68.Ke2 Rh3, and the threat of
69...Rh1! wins the game for Black) 67...e3 68.h6, and now White’s h-pawn has become a force to be
reckoned with: draw!
65...e3!
‘Push pawn’ is the right strategy for Black, though! The pawn advances as quickly as possible.
66.h6 e2! 67.Kf2
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+k+-+-0
9R+-+-+-+0
9+-+-+-+-0
9ptr-+pmKP+0
9+-+-+-+-0
xiiiiiiiiy
67...e1=Q+!
Had White missed this tactical shot? Now, Black wins the rook – the endgame of rook vs pawn(s) is winning,
and Black’s 64...Kd5 move is justified after all.
68.Kxe1 Rb1+ 69.Kf2 a1=Q 70.Rxa1 Rxa1 71.Kf3 Rh1 72.Kf4
And White resigned at the same time, since after 72...Rxh6 73.g4 Ke6, the king arrives in front of the pawn,
with an easy win.

§ 20.4 Checking distance

Like the queen and bishop, the rook is the most effective at a large distance from the enemy king. The closer
it gets to the enemy king, the more it will be hindered in its movements. The question is, of course, where
the border lies – where is the rook still effective, and where is it really hindered? We will investigate this
within the framework of the well-known concept of ‘checking distance’ in rook endings. We will chiefly look
at the correlation between the file on which the enemy pawn is standing (supported by its king) and the file
on which the rook stands.

Rule of thumb
If the ‘checking distance’ between the rook and the pawn is three squares (more accurately,
files), the enemy king can be successfully harassed with side checks. If the distance is two files
or less, the side with the pawn can put an end to the checks and advance the pawn successfully.
Let’s have a look at the ‘border cases’ of this principle.

201
Checking Distance 1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9r+-mKP+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
White to move
We see that the ‘checking distance’ between the rook and the pawn is three squares (or three files). At this
distance, the side that gives checks can save itself and stop the advance of the pawn as follows:
1.Kd7 Ra7+ 2.Kc6
In principle, the king wants to move in the direction of the rook to walk out of the checks, but White doesn’t
want to lose his pawn. So, after
2...Ra6+
it has to move back:
3.Kd5 Ra5+ 4.Kd4 Ra4+ 5.Kc5 Ra5+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9-+-+P+-+0
9tr-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
If White finally decides to move towards the enemy rook with
6.Kb4
there follows
6...Re5
and the pawn falls, for example:
7.Rf7+ Kg6 8.Re7 Kf6 ½-½

202
Checking Distance 2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9r+KzP-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
White to move
We have moved all the pieces (except Black’s rook) one file to the left, and now we see that the ‘checking
distance’ between the rook and the pawn is only two squares/files. This makes the difference between a
draw and a loss: White can walk out of the checks without losing sight of his pawn:
1.Kc7 Ra7+ 2.Kb6 Ra2
Now, White wins due to the vertical cut-off, by going over to the famous Lucena position (see below) in
which White wins by ‘building a bridge’.
The difference mainly lies in the following line: 2...Rd7 3.Kc6 Ra7 4.d7, and the pawn reaches the back rank:
 XIIIIIIIIY
9-+-+-+-+0
9tr-+P+k+-0
9-+K+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
analysis diagram
4...Ra8 5.Kb7 Rh8 6.Kc7, and d7-d8=Q can’t be prevented.
3.d7 Rb2+ 4.Kc7 Rc2+ 5.Kd8
For the winning method in this position, we refer you to Paragraph 20.6 below.

§ 20.5 ‘Doing nothing’ (Philidor)

Most club players know the ‘Philidor position’. François-André Danican Philidor analysed the following
famous position already in 1777. The rook on the sixth rank prevents the enemy king from entering this
rank. As soon as the king tries to use the pawn as a shelter on the sixth rank, the rook dives down to the
first rank to treat the enemy king to a ‘shower of checks’.

203
Philidor position 1
XIIIIIIIIY
9-+-+k+-+0
9+-+-+-+R0
9-tr-+-+-+0
9+-+-+K+-0
9-+-+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White attempts to unsaddle the opponent with
1.e5.
However, now Black organizes his defence as follows: he keeps his king close to the promotion square and
from here on makes only ‘non-moves’ with the king on the back rank. In other words, he awaits events.
Many chess players find it difficult to ‘do nothing’. They always have the idea that they should be ‘doing
something’, but here this would only lead Black into trouble.
1...Ra6
Certainly not 1...Rb1?. After this move, it is still a draw, but Black is making things hard for himself. After
2.Kf6, he has to find 2...Re1!, and what happens next can be found under the heading ‘Short and long side’
(Paragraph 20.9 below).
So, this is Philidor’s discovery:
 XIIIIIIIIY
9-+-+k+-+0
9+-+-+-+R0
9r+-+-+-+0
9+-+-zPK+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The rook oscillates between the a6-b6-c6 squares, waiting until White pushes the pawn.
2.e6
Of course, White can also make rook moves on the seventh rank, but that doesn’t change anything. Only if
White pushes the pawn, should Black react adequately, since then White threatens Kf5-f6, with an
immediate win.
2...Ra1!
XIIIIIIIIY
9-+-+k+-+0
9+-+-+-+R0
9-+-+P+-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
The right method! White’s king is going to be attacked from the rear and can no longer find shelter against
the coming checks.
3.Kf6
Interposing the rook in reply to a check with 3.Rh4 and Rf4 is also useless, since the resulting pawn ending
is a dead draw.
3...Rf1+ 4.Ke5 Re1+ 5.Kd6 Rd1+
With a draw.

§ 20.6 Building a bridge (Lucena)

If things go wrong for the defender, a position may arise as in the following diagram. The winning method
was named after the Spaniard Luis Ramirez de Lucena, even though this position does not feature in the
book he wrote on chess, Repeticion de Amores e Arte de Axedrez (in 1497).

Lucena position 1
Characteristics with White to move
 XIIIIIIIIY
9-+-+-mK-+0
9+-+k+P+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+r+0
9+-+-tR-+-0
xiiiiiiiiy
We can specify a few characteristics of this ‘Lucena position’:
• The pawn has reached the penultimate rank, with its own king standing in front of the pawn.
• The defender’s king is on one side of the pawn, keeping the attacking side’s king from vacating the
promotion square of its pawn.
• The defender’s rook covers the file on the other side of the pawn, preventing the escape of the stronger
side’s king on that side.
To win, the stronger side uses the technique of ‘building a bridge’:

Technique: ‘building a bridge’

In the diagrammed position, the stronger side can only win by:
• driving the enemy king at least one more file away from the pawn
• playing his own rook to a certain rank from where it can be used as a ‘shield’ against enemy
checks
• moving the king to the rear of its own pawn in such a way that the rook can be placed in
between. This is what we call ‘building a bridge’.

204
Lucena position 1
XIIIIIIIIY
9-+-+-mK-+0
9+-+k+P+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+r+0
9+-+-tR-+-0
xiiiiiiiiy
White to move
White would like to promote his pawn, but how can he manage that? Black’s rook cuts off the g-file for the
king, and Black’s king is keeping the door locked on the other side. Of course, there is something White can
do about the latter:
1.Rd1+
With this check, Black’s king is driven off and space is created for White’s king.
1...Kc7
This is the most normal reply, after which we will see the technique of ‘building a bridge’ in practice.
There is one alternative that loses immediately: 1...Ke6? 2.Ke8, and White queens on the next move.
The king can also move towards the centre, and this leads to a slightly different picture: 1...Kc6 2.Rd4
(White sticks to the same technique as presented in the main line. There are various other winning
methods, but we will leave them aside to avoid confusion) 2...Kc5 3.Rd8!. This is the simplest method.
White maintains the vertical cut-off of the black king and protects the promotion square f8. This means that
the king will soon be able to appear into the open and zigzag in the direction of the black rook, as follows:
3...Kc6 4.Ke7 Re2+ 5.Kf6 Rf2+ 6.Kg6 (of course, White has to protect the pawn!) 6...Rg2+ 7.Kh5 (now, the
‘zigzagging’ can start) 7...Rh2+ 8.Kg4 Rg2+ 9.Kh3 Rf2 10.f8=Q.
2.Rd4
The ‘safest’ way is to play the rook to the fourth rank, to be able to interpose it later on.
After 2.Rd5, the rook gets harassed by 2...Kc6.
At this moment, it doesn’t make sense for White to move the king out of its hiding-place. The ‘checking
distance’ between Black’s rook and the king is large enough: 2.Ke7 Re2+ 3.Kf6 Rf2+ 4.Ke6 Re2+ 5.Kf5
Rf2+, and White’s best option is to put the king back on f8 and then start all over again.
2...Rg1
Black doesn’t have anything much better than this waiting move.
3.Ke7
And now His Majesty emerges in the daylight!
3...Re1+ 4.Kf6 Rf1+ 5.Ke6
Of course, the king shouldn’t leave the pawn to its own devices.
5...Re1+
Black can also play a waiting move, since 6.f8=Q is not a threat at this moment: 5...Rf2, and with 6.Rd5
(6.Re4 Kd8)
XIIIIIIIIY
9-+-+-+-+0
9+-mk-+P+-0
9-+-+K+-+0
9+-+R+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-tr-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
the rook can move up another rank to build a bridge on f5: 6...Re2+ 7.Re5 Rf2 8.Rf5 Re2+ 9.Kf6, and the
pawn walks on unhindered.
6.Kf5 Rf1+ 7.Rf4
Completing the bridge! In this final position, Black’s rook cannot accomplish anything and his king is too far
away as well. 1-0
If in the previous Lucena position we switch the positions of Black’s king and rook, moving the rook to the
‘long side’, the ‘checking distance’ is sufficient for a draw.

Lucena position 2
Characteristics with Black to move
 XIIIIIIIIY
9-+-+K+-+0
9+-+-zP-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
Since the situation here is different than in the previous two diagrams, we will formulate the characteristics
of this new situation:
• The pawn has reached the penultimate rank, with its own king standing in front of the pawn.
• The defender’s rook is far removed from the pawn (preferably on the edge of the board) to harass the
enemy king with checks from the side.
• This only makes sense if the door on the other side is locked due to the presence of the defender’s king.
• The weaker side can defend successfully with so-called ‘side checks’ if the ‘checking distance’ is minimally
three lines (between the file on which the rook is standing and the file on which the pawn is standing).

Technique: ‘Side checks’

In the rook + pawn vs rook endgame, the weaker side can sometimes defend by playing the
rook to the side to harass the enemy king from there with checks. A condition for success is
that the ‘checking distance’ is at least three files (see also Vancura below).

205
Lucena position 2
XIIIIIIIIY
9-+-+K+-+0
9+-+-zP-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
Black to move
1...Ra8+ 2.Kd7 Ra7+ 3.Ke6
Both 3.Kc6? Rxe7 and 3.Kd6 Ra6+ 4.Kd5 Ra5+ 5.Kc4 (in view of 5...Re5) don’t make any sense; in both
cases, White loses his pawn.
3...Ra6+ 4.Ke5
 XIIIIIIIIY
9-+-+-+-+0
9+-+-zP-mk-0
9r+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
Thus, White can make one more attempt to confuse his opponent.
4...Ra5+ 5.Kd4 Ra4+ 6.Kc5 Ra5+!
This check is still necessary! If Black puts his rook behind the pawn with 6...Re4??, he will lose after all:
7.Kd6 Rd4+ 8.Ke5 Rd2 9.Rf8! (White’s rook protects the promotion square, thus forcing the win) 9...Re2+
10.Kd6 Rd2+ 11.Kc5 (the king walks into the open field) 11...Rc2+ 12.Kb4 Rb2+ 13.Kc3+–.
7.Kb6 Re5
And the draw is secured.

Again, we make a small mutation in the previous starting position. This time Black’s king is on g8 instead of
g7. This minuscule difference again turns the draw into a loss. The king is so unfortunately placed that
Black can’t avoid losing – provided that White plays well, of course. The solution is still a bit surprising:

206
Lucena position 3
(exception)
XIIIIIIIIY
9-+-+K+k+0
9+-+-zP-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
Black to move
Now that the king is on g8, you might think that at some point White can play Rf8+ to reach the back rank
with tempo and win. However, it turns out that this is not the right idea. White wins in an entirely different
manner that is very instructive!
1...Ra8+
Black is more or less forced to give side checks. If he wants to remove his king from the unfortunate g8-
square with 1...Kg7, he gives the move to White. The winning method after this will be shown in the next
example.
2.Kd7 Ra7+ 3.Ke6
You might think that 3.Kd6 Ra6+ 4.Kd5 Ra5+ 5.Kc4 is the right method, since Black’s rook cannot get
behind the pawn due to 6.Rf8+. But Black defends with 5...Ra8, after which White can switch to the other
winning method with 6.Kc5!.
3...Ra6+
 XIIIIIIIIY
9-+-+-+k+0
9+-+-zP-+-0
9r+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
An important moment. What does White want here, actually? His second idea is to cross the f-file with his
king, on its way to the g-file, to interpose the rook if Black gives check. This ‘crossing’ of the f-file has to be
done with care.
4.Ke5!
4.Kf5? is a draw after 4...Kf7 5.Re1 Ke8.
4...Ra5+ 5.Kf6
Now the king can cross; Black doesn’t have the defence with 5...Kf7 here.
5...Ra6+ 6.Kg5 Ra5+
XIIIIIIIIY
9-+-+-+k+0
9+-+-zP-+-0
9-+-+-+-+0
9tr-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
7.Kg6!
This move is important too. Once more, White’s king dominates Black’s.
7...Ra6+ 8.Rf6
The rook can be interposed – again, a bridge has been built!
8...Ra8 9.Rd6
This ends it all quickly.
In position 3, we saw a variation in which Black gave the move to his opponent by playing ...Kg8-g7. If it’s
White’s move, he wins fairly easily:

207
Lucena position 4
 XIIIIIIIIY
9-+-+K+-+0
9+-+-zP-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
White to move
1.Rg1+!
Again, Black’s king is driven away from the pawn, allowing White to build another bridge – on the other
side of the pawn this time:
1...Kh7
1...Kf6 2.Kf8.
2.Rg4
I is also interesting to show another winning method: putting the rook behind the pawn and then walking
into the open field with the king: 2.Re1!?.
XIIIIIIIIY
9-+-+K+-+0
9+-+-zP-+k0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
analysis diagram
Rook behind the passed pawn!
A) 2...Rd2. Black tries to prevent White’s king from escaping via d7, but now it emerges on the other side:
3.Kf7 Rf2+ 4.Ke6!, and the job is done;
B) Side checks are no use since the pawn is protected by its own rook. If 2...Ra8+, White wins with 3.Kd7
Ra7+ 4.Kc6 Ra6+ (4...Ra8 5.e8=Q) 5.Kb7;
C) Also, moving back with the king, to hold off White’s king, doesn’t help now: 2...Kg7 3.Kd7 Rd2+ 4.Kc6
Rc2+ 5.Kb5 – the checks will be exhausted soon, and of course 5...Rc8 doesn’t help in view of 6.e8=Q.
2...Rd2 3.Kf7 Rf2+ 4.Ke6 Re2+ 5.Kf6 Rf2+ 6.Ke5 Re2+ 7.Re4
Again – the bridge. White wins!
We finish with a study by Grigoriev from 1937 in which Black can give side checks, but his ‘checking
distance’ is too short. We will see that the win is not trivial here.

208
Lucena position 5
Study by Nikolay Grigoriev
Shakhmaty v SSSR 1937
 XIIIIIIIIY
9-+-+K+-+0
9+-+-zP-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-tr-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
Black to move
Unfortunately for Black, his rook is not on the a-file, but on the b-file. Even though he realizes that the
‘checking distance’ is too short, he starts by giving checks anyway:
1...Rb8+
Spending a tempo to put the rook on the a-file with 1...Ra2 is too slow; White wins in the same way we have
seen in the previous position.
2.Kd7 Rb7+ 3.Kd6 Rb8
The only move, parrying the 4.e8=Q threat. The new check 3...Rb6+ would immediately lose to 4.Kc7 Re6
5.Kd7.
4.Kc7 Ra8
Of course, 4...Re8? fails to 5.Kd7 followed by 6.e8=Q.
XIIIIIIIIY
9r+-+-+-+0
9+-mK-zP-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
Finally, the rook has reached the a-file, and Black is threatening to play 6...Ra7+ with the well-known
drawing mechanism. At this moment, White has only one winning move:
5.Ra1!
The a-file is so important that White offers his rook.
5...Rh8
Not the best move (see the complications in the queen vs rook ending), but this shows why the game is
decided here.
After 5...Rxa1 6.e8=Q, the queen vs rook ending is a theoretical win. If the rook can join its own king, it is
still quite a job for White to win this, but he can do it with good play. Since this ending occurs so rarely in
practice, I will discuss it very briefly here. The Tablebases tell us that White can win in 48 moves from this
position. I will give a variation to give an indication of the winning method: 6...Rc1+ 7.Kd6 Rd1+ 8.Kc5
Rc1+ 9.Kd4 Rd1+ 10.Kc3 Rf1 11.Kd3 Rf4 12.Qe5+ Rf6 13.Ke4 Kf7 14.Qc7+ Ke6 15.Qd8 Rf7 16.Qd5+ Kf6
17.Kf4 Kg6+ 18.Kg4 Kh6 19.Qg5+ Kh7 20.Kh5 Kh8 21.Qe5+ Kh7 22.Kg5 Ra7 23.Qe4+ Kg8 24.Kf6 Rg7.
 XIIIIIIIIY
9-+-+-+k+0
9+-+-+-tr-0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+Q+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
From here on, the win is rather simple. See the frame fragment on the next page for an analysis of the
position White can aim for.
6.Kd7 Kf7 7.Rf1+ Kg7 8.e8=Q 1-0
Although the queen vs rook ending does not occur very often in practice, it does play an important role
with liquidations. Therefore, it is important to highlight one position that a club player can aim for if he
knows that the winning method is not very complicated from there (see frame on the next page).

209
Queen vs rook
Basic position
XIIIIIIIIY
9-+-+Q+-+0
9+-+-+-trk0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This position shows us what the stronger side is aiming for. This type of position can be reached on any
side of the board, mirrored over the horizontal and/or vertical axis.
The winning method is exactly the same; only, of course, in the analysis below terms like ‘the sixth rank’
must be interpreted differently.
The winning method is roughly as follows. Driving Black’s king to the edge of the board is not very difficult.
Next, it is important that White’s king occupies the sixth rank and the queen is posted behind it at a knight’s
jump distance (in this case, the e8-square, or the h5-square with the black king on g8). Since in this ideal
set-up, the rook cannot give any checks, it has to move to a square at a greater distance from its own king.
This means that it is not protected, and in all cases, White can collect the rook with a few clever checks.
In the current position, White having the move is an extra complication, since he wants to reach this
position with Black to move. With a few ‘giant steps’ by the queen, White can achieve this:
1.Qe4+ Kg8
1...Kh6?? 2.Qh4#; 1...Kh8 2.Qh1+! Kg8 (2...Rh7 3.Qa8#) 3.Qh5, and we have created our zugzwang
situation. The rook has to move away from its own king and will be lost in a few moves, similar to the main
line.
2.Qa8+ Kh7
White has achieved his goal by using the corners of the board!
3.Qe8!
XIIIIIIIIY
9-+-+Q+-+0
9+-+-+-trk0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The rook has to move into the open field and will become a victim of a double attack by the queen there.
The rook has various squares available, but it can be captured everywhere.
3...Ra7
Also after 3...Rg1, the queen manoeuvres via the centre: 4.Qe4+ Kg8 (4...Kh8 is met by 5.Qa8+ Kh7 (5...Rg8
6.Qh1#) 6.Qa7+, with a double attack) 5.Qa8+ Kh7 6.Qa7+, and again the rook goes into the box; 3...Rb7
4.Qe4+; 3...Rc7 4.Qe4+ Kg8 5.Qg2+ Kh7 6.Qh2+.
4.Qe4+
In most cases, it is wise to first give a check with the queen in the centre, and from there move to the edge
of the board.
4...Kg8
Or 4...Kh8 5.Qh1+ Kg8 6.Qg1+, and the rook is lost.
5.Qd5+!
Another centre square, although by now there were also other roads leading to Rome.
5...Kh7 6.Qh1+ Kg8 7.Qg1+
And White wins.

§ 20.7 Cutting off the king (vertically/horizontally)

The motif of cutting off the enemy king is a continuation of the motif in the previous example. Actually, the
rook prevents the enemy monarch from approaching an important pawn, enabling its own king to support
it. Here, Huub van Dongen’s mantra from our introduction is especially appropriate: ‘Rook endgames are so
difficult, these rooks are moving so straight...’! We will see that the choice between a vertical and a
horizontal cut-off of the enemy king is anything but easy.
Let’s construct a few examples that illustrate both methods. Without much effort, we will be able to distil
a few principles from the variations given.

• Vertical cut

210
Constructed position
 XIIIIIIIIY
9-+-+r+-+0
9+-+k+-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
White to move
Black’s king is cut off vertically here, but Black has just played ...Re8 to break the cut-off. If White’s rook has
to abandon the cut-off, there follows 1...Ke6, saving the half point. But White has an excellent continuation
that yields him a full point after all:
1.Re5!
Maintaining the vertical cut-off! The pawn ending after 1.Rxe8 Kxe8 2.Kf5 Kf7 is a draw.

1.Re5! Rxe5
There is actually no choice. After a move like 1...Rf8, Black’s king remains cut off, and White can use his own
king to push his pawn: 2.Kf5 Kd6 3.Re7, with the idea Kg6 followed by Kg7. If Black then defends with
3...Rg8, White can still botch it up badly. There are even two possible mistakes that cost him half a point:
A) 4.f7? Rg5+! 5.Kxg5 Kxe7 6.Kg6 Kf8 7.Kf6 stalemate;
B) 4.Rg7 Rf8 5.Ra7! – the correct way to win; there is no remedy to the threat of 6.f7 and 7.Kf6; White could
have made a second ugly mistake here by playing 5.f7? too early, since after 5...Ke7 6.Kg6 Ra8, it’s a draw!
2.Kxe5
Here, the resulting pawn ending is winning for White, as we know from the theory of key squares.
2...Kd8
2...Ke8 3.Ke6 Kf8 4.f7 Kg7 5.Ke7 is straightforward. The text is a last try to tempt White into making a
mistake.
XIIIIIIIIY
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-zP-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
3.Kd6!
The only winning move, though not very hard to find. Obviously, not 3.Ke6?? since 3...Ke8 4.f7+ Kf8 5.Kf6
is a draw.
3...Ke8 4.Ke6 Kf8 5.f7 Kg7 6.Ke7
And White wins.

This endgame in practice

This is always easier said than done, but sometimes strong players succeed in demonstrating excellent
endgame technique in a practical game.
211
Maxime Vachier-Lagrave 2719
Ian Nepomniachtchi 2717
Biel 2013 (7)
XIIIIIIIIY
9-+-+-+-+0
9+k+-+-+-0
9-+-+-+r+0
9+-+-+R+-0
9-+-zp-zP-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
59.Rd5!
Not only winning the black pawn but, equally important: White cuts off Black’s king from its own pawn
vertically.
59...Kc7
Black comes up with a few clever ploys. If White pushes his own pawn too quickly, Black can sometimes go
for a pawn ending with ...Rd6. From here onwards, both players have to know exactly what the key squares
are in any possible pawn ending.
Also after 59...Kc6 60.Rxd4 Kc5 61.Rd1, White maintains the vertical cut-off with his rook.
60.Rxd4
White not only wins Black’s pawn, but covers his own pawn at the same time, so that he will be able to use
his king shortly.
60...Rg8
This is Black’s strategy in principle. He puts his rook on the back rank to harass White’s king with checks as
soon as it leaves its hiding-place behind the pawn. At the same time, Black can always see if he can
neutralize the vertical cut-off by inviting a pawn ending with ...Rd8.
XIIIIIIIIY
9-+-+-+r+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-tR-zP-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
61.Ke4!
This is the principled way to win this ending. There are more wins, like moving the rook ‘down’ to d1. But
it’s also easy for White to go wrong there.
If White thinks he can push the pawn forward now, he will be rudely awakened, since after 61.f5?
 XIIIIIIIIY
9-+-+-+r+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+P+-0
9-+-tR-+-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Black can go into the pawn ending: 61...Rd8! 62.Re4 (62.Rxd8 Kxd8 63.Kf4 Ke7 64.Kg5 Kf7 65.f6 Kf8 is a
draw). This looks like the only method to try something, but after 62...Kd7 63.Kf4 Re8!, it also amounts to
nothing; Black holds the draw.
By playing 61.Rd1, we get a position on the board that illustrates the ‘rule of five’, which we will return to in
more detail below.
61...Kc6
Since the pawn ending is hopeless now, Black tries to move his king in the direction of the white rook to
‘harass’ it until it leaves the d4-square.
62.Ke5 Re8+
If 62...Kc5, White’s best reply is 63.Rd6, after which he can always meet a check by interposing the rook.
63.Kf6
XIIIIIIIIY
9-+-+r+-+0
9+-+-+-+-0
9-+k+-mK-+0
9+-+-+-+-0
9-+-tR-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, we see the importance of the rook on d4: it keeps the f4-pawn covered – otherwise, a check on f8
would drive White’s king back again. The ‘checking distance’ between the rook and the pawn is three
squares, but since the pawn is protected, it is quite a different story.
63...Kc5
63...Rf8+ 64.Ke7, and White has drastically improved the position of his king. There is still an essential line
here: after 64...Kc5, 65.Ra4 is forced. Thus, White must abandon the vertical cut-off, but the dominant
position of his king guarantees the win after 65...Rb8 66.f5 Rb7+ 67.Ke6 (the shoulder budge!) 67...Rb6+
68.Ke5 Rb8 69.f6, and the pawn decides.
64.Rd7
Obviously, White keeps the vertical cut-off intact, but on the seventh rank, the rook can also defend against
the checks. By coincidence, we now have a horizontal cut-off of the black king as well.
 XIIIIIIIIY
9-+-+r+-+0
9+-+R+-+-0
9-+-+-mK-+0
9+-mk-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
64...Kc6 65.Re7 Rf8+ 66.Ke5
Now, Black’s king also gets a ‘shoulder budge’, and the pawn can no longer be stopped, and so
Nepomniachtchi resigned.
By the way, there was nothing wrong with 66.Rf7 either.

• Horizontal cut

212
Constructed position
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-zPr+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+R+-+-0
xiiiiiiiiy
White to move
Of course, the ancient rule of ‘rook behind the passed pawn’ applies here as well:
1.Rf1
But with this move, White abandons his vertical cut-off, so it looks as if Black’s king will be in time to
eliminate the pawn.
1...Rg8 2.f7 Rf8
 XIIIIIIIIY
9-+-+-tr-+0
9+-+-+P+-0
9-+-+-+-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
This is an important moment. White has to work with the so-called ‘horizontal cut-off’ of the enemy king,
which means that there is only one good move:
3.Rf6!
If White continues 3.Ke4, there follows 3...Kd6, and since 4.Kf5 is not possible, White cannot prevent
4...Ke7, conquering the pawn.
XIIIIIIIIY
9-+-+-tr-+0
9+-+-+P+-0
9-+-+-tR-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The text prevents 3...Kd6, giving White time to move his king toward the pawn.
3...Kd5 4.Kf4!
With this move, White prevents Black from driving the f6-rook away from its dominant post with 4...Ke5.
Another advantage of 3.Rf6 is that White’s king can steal alongside the rear of the rook with Kf4-f5-g6 to
support the pawn and win with Kg7. Black no longer has any useful moves and can resign.

213
Constructed position
XIIIIIIIIY
9r+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+k+-mK0
9tR-+-+-+-0
xiiiiiiiiy
White to move
True, White has managed to put his rook behind the pawn, but his king is very far away from the pawn,
making the win look virtually impossible. Still, thanks to the technique of the horizontal cut-off, combined
with a few other clever shots, White manages to overcome his opponent in an elegant way:
1.Ra3!
This third-rank cut-off is essential. Black’s king has to lose time to walk to b2, before it can make its way
towards the white pawn.
1...Kd2 2.Kg3
This gives White the time to ‘cross the ocean’ with his own king.
2...Kc2 3.Kf4 Kb2
Now, the horizontal cut-off has to be abandoned, but White has already accomplished a lot.
4.Ra6 Kc3
XIIIIIIIIY
9r+-+-+-+0
9zP-+-+-+-0
9R+-+-+-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
There follows a race between both kings towards a7:
5.Ke5 Kc4 6.Kd6 Kb5
Since Black now gains a tempo on White’s rook, it seems as if he wins the race, but appearances are
deceptive.
7.Ra1 Kb6
XIIIIIIIIY
9r+-+-+-+0
9zP-+-+-+-0
9-mk-mK-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
With this shoulder budge, Black’s king prevents its opposite number from reaching the pawn, but now the
technique of domination works again:
8.Rb1+! Ka6
If 8...Kxa7, 9.Kc7 wins a rook by domination.
 XIIIIIIIIY
9r+-+-+-+0
9zP-+-+-+-0
9k+-mK-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+R+-+-+-0
xiiiiiiiiy
9.Kc7!
With this trick, White strikes home. Not 9.Kc6? Rc8+! 10.Kd7 Rh8, and Black draws.
9...Rxa7+
He is more or less obliged to take the pawn, but that costs a rook:
10.Kc6 Ka5 11.Ra1+ Kb4 12.Rxa7 1-0

§ 20.8 The rule of five

In fact, we can hardly call a rule a rule if there are many exceptions – which is the case with this rule.
Nevertheless, it offers both the stronger and the weaker side something to go by, since the rule provides
contours within which we can discern a clear technique that occurs in many rook endings. In positions in
which the pawn is protected by its own king, while the enemy king is cut off vertically, if the enemy rook is
standing in front of the pawn (at a sufficient ‘checking distance’), the rule of thumb is as follows:

Rule of thumb: the rule of five

Add up the number of the rank on which the pawn is standing with the number of files by
which the king is cut off from the defender. If the sum is more than five, it is a win in most
cases. If not, it is normally a draw.
The winning method in the following diagram (next page) still takes quite a lot of work. White will have to
apply several of the techniques we have been discussing. Averbakh once indicated the typical winning
method, but I have added quite a few variations and comments to his analysis:

214
Yury Averbakh – 1
XIIIIIIIIY
9-+-tr-+-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
9-+-zP-+-+0
9+-+K+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
White to move
Before showing the winning method, we have to establish the ‘checking distance’ between Black’s rook and
the pawn. The distance is three ranks (squares), and usually this is sufficient for the lone rook side to
prevent the pawn from moving forward with the support of its king. So, White’s king needs help from its
own rook, which is well-positioned on the back rank also for this reason, as it can be placed behind the
pawn later.
Before we start, we will first make a count. White’s pawn is on the fourth rank (4) and Black’s king is cut off
by two files (2). The sum (4+2=6, i.e. >5) tells us that the position should be winning – and it is.
1.Kc4
To win, White’s king has to emerge from its shelter. It will be subjected to a shower of checks.
1...Rc8+ 2.Kb5
Putting the king in front of its own pawn is of no use, therefore it moves to the adjacent file.
2...Rd8
This is the move on which Black’s defence rests. He is going to alternate checks with attacks on the pawn.
3.Kc5
Before White’s own rook can be brought into action, his king has to penetrate as ‘deep’ as possible into the
enemy ranks.
3...Rc8+
Black has to continue giving checks, as otherwise the pawn would take one step forward.
4.Kb6! Rd8
XIIIIIIIIY
9-+-tr-+-+0
9+-+-+-+-0
9-mK-+-+k+0
9+-+-+-+-0
9-+-zP-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
Now that the king can no longer be used, White has to use his rook. The fact that he abandons the vertical
cut-off is of course a concession, but White still manages to force the win with a clever little trick:
5.Rd1 Kf7
Black’s king rushes to the promotion square.
6.Kc7
No use is the pawn push 6.d5? in view of 6...Ke7, and White can no longer win – not even after 7.Kc7 on
account of 7...Rd7+ 8.Kc6 Rd6+ 9.Kc5 Kd7, and Black has managed to place his king in front of the pawn.
XIIIIIIIIY
9-+-tr-+-+0
9+-mK-+k+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-zP-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+R+-+-0
xiiiiiiiiy
The text move is essential to the winning process. From here, the defender can still cause all kinds of
trouble by moving the rook to the edge. White has to take care that the enemy king cannot walk to the
promotion square.
6...Rd5
Black keeps his rook in front of the pawn, but this doesn’t cause a lot of problems for his opponent.
A) 6...Ke7?? loses the rook to 7.Re1+;
B) 6...Ra8 doesn’t work either in view of 7.d5. The ‘checking distance’ from the side between the king and
rook is too small, so White can duck out of the checks as follows: 7...Ra7+ 8.Kb6 Rd7 9.Kc6.
XIIIIIIIIY
9-+-+-+-+0
9+-+r+k+-0
9-+K+-+-+0
9+-+P+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+R+-+-0
xiiiiiiiiy
analysis diagram
Another important moment, as White’s win is hanging by a thread:
B1) If 9...Ra7, 10.Rh1! wins (White has to prevent Black’s king from stealing in front of the pawn at any
price; not 10.d6?? Ke8, with a draw, i.e. 11.Rh1 Ra6+ 12.Kd5 Ra2!=) 10...Ra6+ 11.Kc7 Ra7+ 12.Kb6 Ra2.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+-0
9-mK-+-+-+0
9+-+P+-+-0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+-+R0
xiiiiiiiiy
analysis diagram
And now that Black’s rook is no longer in front of the pawn, White again has to use a vertical cut-off to
secure the full point: 13.Re1! Rc2 14.d6 Ra2 15.d7 Rd2 16.Kc7 Rc2+ 17.Kd8, and we have ended up in the
Lucena position, in which White can win by building a bridge;
B2) 9...Ke7 10.d6+ Kd8.
XIIIIIIIIY
9-+-mk-+-+0
9+-+r+-+-0
9-+KzP-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+R+-+-0
xiiiiiiiiy
analysis diagram
Black’s king has landed on the promotion square, and his rook prevents the further push of the pawn. Isn’t
this what Black was aiming for with his defence? Yes, you would say so, but his pieces are so tightly knit
now that White needs only one more move to win: 11.Rh1!.
7.Kc6 Rd8 8.d5
Now that the pawn has moved up a square, the rest is no longer too difficult. The ‘checking distance’ has
become shorter, which makes things a lot easier for White.
8...Ke7 9.d6+
Behold the power of the rook behind the passed pawn.
9...Ke6 10.d7
And White wins, as he also would with 10.Re1+.
If we move all the pieces two files to the left, the above winning method doesn’t work, since White’s king
now doesn’t have the file to the left of the a-file available; this is a simple draw.

215
Yury Averbakh – 2
XIIIIIIIIY
9-tr-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mk-+-0
9-zP-+-+-+0
9+K+-+-+-0
9-+-+-+-+0
9+-+R+-+-0
xiiiiiiiiy
White to move
1.Rd7
This is an entirely different attempt which, however, doesn’t amount to anything either.
The manoeuvring method with the king doesn’t work here. We also see now why the ‘checking distance’
between the rook and the pawn is long enough for the defender to hold the draw. Please note that with
Black’s rook on b7, the outcome would be entirely different!
1.Ka4 Ra8+ 2.Kb5 Rb8+ 3.Ka5 Ra8+ 4.Kb6 Rb8+, and White doesn’t get any further. He cannot attack the
rook since then the pawn will fall, and there is nothing for it but to return to his hiding place behind the
pawn.

1.Rd7 Ke6 2.Ra7

White uses the technique of the ‘horizontal cut’, but the pawn has not advanced far enough to create any
serious winning chances.
Also, the ‘clever’ 2.Rd4 (to protect the pawn with the rook from the side) doesn’t bring White any benefit
after 2...Ke5 3.Kc3 Rh8 4.b5 Rb8 5.Rh4 Kd6 6.Kb4 Kc7, and the king gets in front of the pawn again.
2...Kd6 3.Ka4 Kc6 4.Ka5
 XIIIIIIIIY
9-tr-+-+-+0
9tR-+-+-+-0
9-+k+-+-+0
9mK-+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
4...Rb5+!
An essential check, securing the half point.
5.Ka4 Rb8 ½-½
Certainly not 5...Kb6?? since after 6.Rb7+ Kxb7 7.Kxb5, White controls the key squares in the resulting
pawn ending and wins.
So, that a-file is playing tricks on White. What happens if the pawn has advanced one rank further (fifth
rank=5) and the enemy king is cut off by one file (=1)? Doesn’t the ‘rule of five’ (5+1=6 >5) tell us that
White should win? No – because of the liquidation to a pawn ending!

216
Yury Averbakh – 3
XIIIIIIIIY
9-tr-+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+P+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
Black to move
Not 1...Rb7?? 2.Ka5, and White wins, but with

1...Rc8!
Black breaks the cut-off, since the resulting pawn ending is a draw. Here, we see that the position of Black’s
king does matter!
So, if the position of the defending king mattered, let’s put Black’s king on d6 in the previous position, with
Black to move. Then, the position is lost again!

217
Yury Averbakh – 4
 XIIIIIIIIY
9-tr-+-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+P+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
Black to move
Black can’t break the cut-off yet, since his king is unfortunately placed.
1...Kd7
Now, he is just too late with his ...Rc8 move.
XIIIIIIIIY
9-tr-+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+P+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tR-+-+-0
xiiiiiiiiy
2.Ka5!
And since the pawn ending after 2...Rc8 is now winning, and the ‘checking distance’ is too short, the win
isn’t difficult, e.g.:
2...Ra8+
2...Rc8 3.Rxc8 Kxc8 4.Ka6 Kb8 5.Kb6!+–.
3.Kb6 Rb8+ 4.Ka6 Ra8+ 5.Kb7
And it’s game over.
Let’s return for a moment to the Averbakh No. 3 diagram and now put White’s rook on c5 instead of c1,
with Black to move. What is the situation then?

218
Yury Averbakh – 5
 XIIIIIIIIY
9-tr-+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+PtR-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Small differences sometimes have great consequences. Also here, the position of the rook determines the
difference between a win and a draw:
1...Rc8
This was Black’s method in the corresponding example with the rook on c1 – but here it doesn’t work.
Also after 1...Ra8 2.b6 Kd6 3.Kb5 Rb8 4.Rc7 Ra8 5.Rc6+ Kd7 6.b7 Rh8 7.Kb6, White will achieve the win.

If 1...Kd6, White wins with, for example, 2.Rc1 Kd7, and now Black threatens to play 3...Rc8 again, but, as
we have seen in the previous example, White is on time: 3.Ka5!, and since the ‘checking distance’ is too
short, and the pawn ending after 3...Rc8 is losing for Black, White wins also here.

1...Rc8 2.b6!
XIIIIIIIIY
9-+r+-+-+0
9+-+k+-+-0
9-zP-+-+-+0
9+-tR-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
As now the resulting pawn ending is a win:
2...Rxc5
Clearly, if 2...Rb8, 3.Kb5 is an easy win.
3.Kxc5 Kd8
3.Kxc5 Kc8 4.Kc6 Kb8 5.b7 Ka7 6.Kc7.
4.Kd6! Kc8 5.Kc6 Kb8 6.b7 Ka7 7.Kc7
And White wins.
The funny thing is that I published these two corresponding positions before in Schakend Nederland, as the
Dutch federation’s magazine was called at the time, and already indicated the difference between the rook
on c1 and c5 there.

This endgame in practice

There is another story to the two above-mentioned positions. As it turns out, the second position could
have played an important role in a club match between De Variant from Breda and Zukertort from
Amstelveen in the 1993/94 season of the Dutch club competition, where the score was 4½-4½. The white
player, Erik Knoppert, played for the home team of De Variant, and here is what happened:
219
Erik Knoppert
Piet Peelen
Netherlands tt 1993/94 (1)
XIIIIIIIIY
9-+-+-+-+0
9+r+-+-+-0
9-+-+k+-+0
9+P+R+-+-0
9-+-mK-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This is a difficult ending that looks promising for White, but there is a discrepancy between ‘feeling’ and
‘knowing’, as we will see:
1.Rh5?
This can’t be the solution. White not only abandons the vertical cut-off, but he doesn’t even get a chance to
create a horizontal cut-off either. This means that Black’s king can simply walk to the ideal square in front
of the white pawn:
1...Kd6 2.Rh6+ Kc7 3.Kc5 Kb8
And here the game was already agreed drawn. A possible continuation is 3...Kb8, which leads to a
theoretical draw after 4.b6, for example: 4...Rg7 5.Kc6 Rg8, and as we know, with a b-pawn White can’t
achieve anything here.

The problem is that with the rook on d5, White’s king cannot move to the queenside since after 1.Kc4?
Rc7+ 2.Rc5, Black plays 2...Kd6!, with an immediate draw.

Also, not 1.Rd8? in view of 1...Rd7+, and the pawn ending is a draw: 2.Rxd7 Kxd7 3.Kd5 Kc7 etc.

The remarkable 1.Rc5! is the only move that wins, although this doesn’t seem obvious at first sight. The
idea is to get near to the pawn with the king. After that, His Majesty wants to help his pawn reach the other
side – via the a-file, if possible. During the analysis (with several members of the grandmaster guild who
had also played in this match), this option soon appeared on the board, and the story goes that Knoppert
had also seriously considered this move.
Here, 1...Kd7 is the best chance to achieve something (not good is 1...Kd6 in view of 2.Rc6+ Kd7 3.Kc5, and
White wins easily – provided that he meets 3...Rc7 with 4.Kb6!): 2.Kc4.
 XIIIIIIIIY
9-+-+-+-+0
9+r+k+-+-0
9-+-+-+-+0
9+PtR-+-+-0
9-+K+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
White’s king has managed to get close to his pawn. According to the rule of five, the game is decided (pawn
on the fifth rank plus a one-file cut-off, combined with a very unfortunate placement of Black’s rook). We do
have to investigate a few options:
A) 2...Kd6 still requires a lot of accuracy. Here, 3.Rc6+! is the only correct continuation (3.Kb4? is an
immediate draw after 3...Rc7 4.Rh5 Rc1 5.Ka5 – by the way, after 5.Rh7, the horizontal cut-off, Black would
have to find 5...Ra1!; White’s king can’t be allowed to get in front of the pawn via a5 – 5...Kc7 6.Rh7+ Kb8
7.Kb6 Rc8) 3...Kd7 4.Kc5, and we have already seen this position in another variation;
B) 2...Rc7 is met by 3.b6! – this liquidation to a pawn ending is quite familiar to us! 3...Rxc5+ (3...Rc6 4.Kb5
Rxc5+ 5.Kxc5+–) 4.Kxc5 Kd8 5.Kd6!+–;
C) 2...Rb8 is the best attempt to threaten to enter a pawn ending: 3.Kb4 (the most logical move; 3.Kd5 also
wins according to the Tablebases, but we will leave that aside) 3...Rc8 (neither does 3...Kd6 help; White
seizes the win with 4.Rc1 Kd7 5.Ka5!, and Black has no defence left).
XIIIIIIIIY
9-+r+-+-+0
9+-+k+-+-0
9-+-+-+-+0
9+PtR-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Here, knowledge is required! This position was quickly found by the (grand-)masters present, and of
course they suggested the liquidation to the pawn ending that we know with 4.b6! Rxc5 5.Kxc5 Kd8 6.Kd6
Kc8 7.Kc6 Kb8 8.b7 Ka7 9.Kc7, and the job is done.
Since White didn’t play the endgame correctly, the match ended in a 5-5 tie... I couldn’t fail to point out that
chess articles are meant to be read! This fragment is partly based on analysis by Piet Peelen.
As we have seen more often in this book, we often find it hard to apply well-known principles in the right
way. What makes it so difficult is that at a given moment, a certain principle has to be preferred above
another. Sometimes, the nuances are not easy to discern in an actual game, not even by strong
grandmasters. That doesn’t alter the fact that it pays to keep studying valid techniques, for one day the
moment will come when you will know exactly what you have to do. And precisely then – often when both
players have little time on the clock, and your opponent may be groping in the dark – your knowledge
might make the difference.
The following endgame provides a nice test with this material.
220
Evgeny Postny 2670
Benjamin Bok 2594
London 2015 (5)
XIIIIIIIIY
9-+-+-+-+0
9+-+-tr-+-0
9-mk-+-+-+0
9+-+-+-tR-0
9-+-+-+K+0
9+-+-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Victory is for the taking for White – or at least that is what the Tablebases say.
54.Rd5?!
Postny plays a perfectly logical move: he wants to establish a vertical cut-off as quickly as possible. But
after this move, the win takes more time: 43 moves, no less.
In earlier examples, we saw how important it is to push the pawn as soon as possible. As long as Black’s
king cannot be placed in front of the pawn, and White doesn’t have to fear a pawn ending, this is a more
important factor than the cut-off: 54.f4! would have won in 35 moves!

54.Rd5?! Rg7+
54...Kc6 was more tenacious.
55.Kh4?!
The Tablebases indicate that White can force mate in 53 moves here. It is important that a piece is captured
along the way, since otherwise Black could claim a draw due to the 50-move rule.
55.Rg5 was necessary, but that was, of course, not a logical move with regard to the vertical cut-off.
55...Rf7 56.Rd2
The pawn is still on its initial square, while the idea is to push it forward as soon as possible.
56...Kc6 57.Kg5 Rg7+ 58.Kf6 Rg2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+k+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-tR-zPr+0
9+-+-+-+-0
xiiiiiiiiy
We have seen a few inaccuracies on both sides, but now White’s win is secured – in 51 moves. Black’s last
move prevents the advance of the pawn, so it is quite reasonable from a human perspective.
59.Ke5
Threatening 60.Rd6+ and 61.f4.
59...Kc7 60.Ke4 Kc6 61.Ke3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-tR-zPr+0
9+-+-+-+-0
xiiiiiiiiy
61...Rg7
He puts his rook on the seventh rank because sometimes he wants to work with ...Rd7.
62.Ke4?!
Making things rather difficult for himself...
Here, White should have made at least one step forward with his pawn: 62.f3! was a win in 47 moves.
White’s king gets in front of the pawn in the pawn ending here, i.e. 62...Rd7 63.Rxd7 Kxd7 64.Kf4! Ke6
65.Kg5, and with the key squares in mind, we know that White is winning.
However, certainly not 62.f4? since the pawn ending after 62...Rd7 63.Rxd7 Kxd7 64.Ke4 Ke6 is indeed
drawn.
62...Rg2!
The best defence.
63.Kf5
A strange move at first sight.
63...Kc7 64.Ke6?!
XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-tR-zPr+0
9+-+-+-+-0
xiiiiiiiiy
This was how White thought he could deal with the problem: Rd7+ followed by Rf7, and the pawn can
safely advance. But Benjamin is not to be fooled:
64...Rg6+! 65.Ke7 Rg7+ 66.Kf6 Rg2
So, all his manoeuvres haven’t done White any good!
67.Ke5 Kc6 68.Kf4
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+-+-0
9-+-tR-zPr+0
9+-+-+-+-0
xiiiiiiiiy
Just checking – from here the win takes 60 moves!!
68...Kc7?!
Here, 68...Rg8! was needed: win in 59 moves...
69.Kf3?!
This wins in 61 moves. 69.Ke3!, as he played earlier, would have won much more quickly: 49 moves.
69...Rg8 70.Ke3 Re8+ 71.Kf3 Kc6
XIIIIIIIIY
9-+-+r+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-tR-zP-+0
9+-+-+-+-0
xiiiiiiiiy
72.Rd4?
Finally, a full question mark! Now it is definitely a draw. The only two winning moves were 72.Kg4 and
72.Kg3.
72...Kc5!
An excellent riposte.
73.Re4 Ra8 74.Kg4 Rg8+ 75.Kh4 Kd5!
Slowly, Black is gaining a foothold, and the cut-off is broken.
76.Re7 Rf8 77.Kg3 Rg8+ 78.Kf3 Rf8+ 79.Ke3 Rf6
XIIIIIIIIY
9-+-+-+-+0
9+-+-tR-+-0
9-+-+-tr-+0
9+-+k+-+-0
9-+-+-+-+0
9+-+-mK-+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
Black is threatening to go over to a pawn ending at some point – but at the right moment!
80.Re8 Kd6
He could have made a mistake with 80...Re6+?? 81.Rxe6 Kxe6 82.Ke4! Kf6 83.Kf4, and White wins, as
every club player knows!
81.f3!?
Finally, this pawn move. White can still allow a pawn ending, but it doesn’t bring him much.
81...Kd5!
Certainly not 81...Re6+?? 82.Rxe6+ Kxe6 83.Ke4! Kf6 84.Kf4, and again White wins!
82.Re4 Kd6 83.Rf4 Rg6 84.Re4 Rf6 85.Kf2 Re6!
But now he can! The draw was immediately agreed since after 86.Rxe6+ Kxe6 87.Kg3 Kf5!, there is
nothing more in it for White.
Funny – I assume that on the first move, the average patzer would have done better than the GM. The
patzer, not hampered by any knowledge, would naturally play 54.f4...
Unexpectedly, while I was working on this book, a remarkable case occurred involving the new chess star
of 2021, Alireza Firouzja. He was facing another young and talented player, Kirill Shevchenko from Ukraine.
In the defence of both players, we should mention that this fragment is taken from a rapid game and both
players had little time left. Nevertheless, the commentator on chess24.com, Peter Leko, made mincemeat of
the endgame treatment by these young grandmasters. Among other comments, Leko said that he could
dream the finesses of this endgame already when he was 12...

221
Alireza Firouzja 2804
Kirill Shevchenko 2655
Warsaw rapid 2021 (9.12)
XIIIIIIIIY
9r+-+-+-+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+R+K0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The notorious ending has come on the board. The rule of five tells us that this might be a draw (fourth rank
+ one-file cut-off is only five), but it is White’s move! This makes all the difference, since his king is already
well placed. We know that Black’s defence is based on the checking distance between Black’s rook and the
pawn.
Question: What should White play to force the win?
65.Rf1?
Firouzja goes wrong, spoiling half a point. Actually, there is only one correct winning idea here – the logical
65.Kh6!, with which White makes it impossible for Black to continue giving checks: 65...Rg8 (65...Rh8+
makes no sense after 66.Kg7) 66.Rf4 Ke5 67.Ra4
 XIIIIIIIIY
9-+-+-+r+0
9+-+-+-+-0
9-+-+-+-mK0
9+-+-mk-+-0
9R+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
and Black cannot prevent the pawn from marching on. An important tactic is that 67...Kf6 is met with
68.g5+! Rxg5 69.Ra6, winning.

65.Rf1? Rh8+ 66.Kg5 Rg8+ 67.Kh4 Rh8+ 68.Kg3

XIIIIIIIIY
9-+-+-+-tr0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
9-+-+-+P+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
The king has to go back, and this position is a draw, but Black has to keep making decisions on every move.
68...Rg8 69.Rf5 Ra8
Shevchenko has found a good way to play ‘waiting moves’. 69...Rh8 was another logical continuation.
70.Kf4 Rb8 71.Kg3 Ra8 72.Rf4
The moment White protects the pawn with his rook, Black has to start paying attention. Nevertheless, there
is still no problem for him.
72...Rg8 73.Kh4
XIIIIIIIIY
9-+-+-+r+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
9-+-+-tRPmK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
73...Rh8+?
This check costs half a point – at least, it should have. There was only one defence: 73...Ke5!=. It is essential
to ‘put the question to’ White’s rook on f4. After all, if the rook is protecting the pawn, White has reasonable
freedom to manoeuvre his king in front of the pawn without it getting lost. It’s all very intricate. After
74.Rf5+ Ke6 75.Kh5 Rh8+, Black now has to start giving checks: 76.Kg6 Rg8+ 77.Kh5 Rh8+ 78.Kg5 Rg8+
79.Kh4 Rh8+ 80.Kg3, and White can make no progress.
74.Kg5 Rg8+
XIIIIIIIIY
9-+-+-+r+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-mK-0
9-+-+-tRP+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now that White is covering his pawn, he can move towards Black’s rook with his king.
75.Kh5?
But this chance is wasted on White. He could have won with 75.Kh6! – a position we have seen above, e.g.
75...Ke5 76.Ra4.
75...Ke5!
The Ukraine player grabs his chance to force a new drawn position.
76.Rf5+ Ke6 77.Kh4 Rh8+ 78.Kg3 Ra8
The players have landed in a position that has been on the board before.
79.Rf4 Rg8 80.Kh4
XIIIIIIIIY
9-+-+-+r+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
9-+-+-tRPmK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
80...Rh8+??
But apparently, the above-mentioned defence was over Black’s head, and now Firouzja finally shows that
he knows how to collect the full point from here.
Only 80...Ke5! was sufficient for a half point.
81.Kg5 Rg8+ 82.Kh6!
Yes, he found the winning idea!
82...Ke5 83.Ra4 Rh8+ 84.Kg7 Rh4
An ultimate attempt – in case White hurries and plays g4-g5??.
85.Ra5+ Ke6 86.g5
Now, the pawn reaches the seventh rank unhindered. The young Franco-Iranian commands this textbook
ending.
86...Rg4 87.g6 Rg2 88.Ra6+ Ke7 89.Kg8 Rh2 90.g7 Rh4 91.Ra1 Rg4
After 91...Rh2, we could have seen White build a bridge.
92.Re1+ Kd6 93.Rh1!
And Black resigned.
There is a lot more to be said about these endgames, but we will limit ourselves to one more theoretical
example in which the pawn is a bit further back, and the win is a little more complicated.

222
Yury Averbakh – 6
XIIIIIIIIY
9-+r+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zP-+-+-0
9-+K+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
White to move
First, we make the sum: third rank (3), two-file cut-off (2) makes 3+2=5, so with this rule in mind, Black
should be able to draw.
Unfortunately for him, his king is very poorly placed on f7. Transferring to a pawn ending with ...Re8 is not
on the cards (yet), and since the king is too far removed from the third rank, White has an instructive
method to force the pawn forward anyway, without Black being able to prevent it. Again, of course, we start
with:
1.Kb3 Rb8+ 2.Ka4 Rc8 3.Kb4 Rb8+ 4.Ka5
As ‘deep’ as possible, but the ‘checking distance’ is a lot longer than we have seen in earlier example.
4...Rc8
XIIIIIIIIY
9-+r+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9mK-+-+-+-0
9-+-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
Black persists in the strategy of alternating checks with attacks on the pawn. If White now opts for the
‘standard’ method with the rook behind the passed pawn, he will be in for a nasty surprise: 5.Rc1 Ke6 6.c4
Kd6 7.Kb6 Rb8+ 8.Ka7 Kc7, and the king has managed to get in front of the pawn, holding the draw.
5.Re3!
This time, the rook has to be placed at the side of the pawn first, to maintain the vertical cut-off. Now, if
Black’s king had stood a little ‘higher’ on the f-file, it could have harassed White’s rook earlier.
This is what makes these endgames so difficult and so beautiful at the same time. At one moment, the king
has to be on the seventh rank to break the cut-off with a rook move (provided the resulting pawn ending is
a draw) and the next moment it has to stand somewhere in the middle of the board to be able to harass the
enemy rook that is protecting its pawn from the side. Assessing such themes during a game is hard to do,
but at the same time, it’s a nice challenge.
5...Kf6 6.Kb5
This gives White the time to work with his king again. With the text move, he threatens 7.c4.
6...Rb8+
And so, Black again has to harass him with checks.
7.Kc6
XIIIIIIIIY
9-tr-+-+-+0
9+-+-+-+-0
9-+K+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zP-tR-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Here, we see how important it is that the rook protects the pawn from e3, since now White’s king has free
play along the b-, c- and d-files.
7...Rc8+
Consistently advancing the king in an attempt to drive the rook from e3 with 7...Kf5 would now fail to 8.c4!
Kf4 9.Re1 Rc8+ 10.Kb5 Rb8+ 11.Ka6 Rc8 12.Rc1, and again we have a well-known winning position for
White.
8.Kd6 Kf5
Another check doesn’t help: 8...Rd8+ 9.Kc7 Rd5 10.Kc6 Rd8 11.c4, and again White has succeeded in
reaching the fourth rank with his pawn. With the two-file cut-off, the win is secured.
XIIIIIIIIY
9-+r+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-zP-tR-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black continues on his path, putting White on the spot. How should he continue? There is only one winning
move!
9.Re5+!
Marvellous – the rook is used in a different way than in the system we introduced above.
9...Kf6 10.Rc5
Thus, the attack on White’s pawn is parried, and at the same time Black’s rook is driven off its ideal square.
Thus, White succeeds in advancing his pawn one step, which was the ultimate goal of the entire operation.
10...Rd8+ 11.Kc7 Rd3 12.c4
 XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-mk-+0
9+-tR-+-+-0
9-+P+-+-+0
9+-+r+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
12...Ke6
Black is making things as difficult as possible for his opponent by moving his king closer.
Another idea was to keep ‘hanging onto’ the pawn with 12...Rd4. The winning line is not trivial here, but
White secures the victory with 13.Kb6 Ke6 (13...Ke7 14.Rc7+ Kd8 15.c5 leads to a winning position as
well) 14.Rc7!.
XIIIIIIIIY
9-+-+-+-+0
9+-tR-+-+-0
9-mK-+k+-+0
9+-+-+-+-0
9-+Ptr-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Combining the vertical cut-off with the horizontal cut-off. 14...Rd1 15.c5 Rb1+ 16.Kc6 Ra1 17.Rb7, and
Black is unable to prevent the pawn being shepherded to the other side.
13.Rh5
A nice moment to play the rook to the side.
13...Rd7+
XIIIIIIIIY
9-+-+-+-+0
9+-mKr+-+-0
9-+-+k+-+0
9+-+-+-+R0
9-+P+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Since c4-c5-c6 was threatened, Black has to keep harassing the opponent.
14.Kc8! Rg7
After 14...Rd6, 15.c5 is the right choice (of course, not the transfer to a pawn ending with 15.Rh6+ Ke5
16.Rxd6?? Kxd6 17.Kb7 Kc5, as this is a draw) 15...Rc6+ 16.Kb7, and Black loses his rook.
Averbakh also gave 14...Ra7 15.c5 as winning for White here.
Now, White has to find the best way to continue.
15.c5
It is great to see the double function of White’s rook on h5. Not only does it support the pawn’s march, but
it also prevents Black’s king from getting in front of the pawn.
15...Rg8+ 16.Kb7 Rg7+ 17.Kb6 Rg1 18.c6 Rb1+ 19.Kc7
XIIIIIIIIY
9-+-+-+-+0
9+-mK-+-+-0
9-+P+k+-+0
9+-+-+-+R0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
Now, White works towards the Lucena position, with a theoretical win.
I can very well imagine that such a winning method is not reserved for everyone, and it doesn’t have to be.
Still, it will always help to study such fragments, preferably together with a sparring partner. This can be
done by simulating a game situation. In the starting position, you set the clock on 20 minutes, for instance,
and you play out the position against each other – also with colours reversed – and write the moves on a
score sheet. Next, you study the first part of the winning method, until a new position arises that can be
tried in a second playing session, and so on. This way, you can mix business with pleasure.

§ 20.9 Short and long side

In almost any endgame book, the theme of the ‘short and long side’ in rook endings is discussed. Let’s first
define what we mean with the ‘short and long side’. We can show this in the following picture (see diagram
next page):

Rule of thumb: ‘short and long side’

If we draw an imaginary vertical line through the middle (between the d- and e-files), then if
the pawn is on the right side of this line, we call that side of the pawn the ‘short side’. With
that, we have automatically defined the ‘long side’ too, which is the side where the enemy rook
wants to be. To prevent the threatened mate, Black has to make a move with his king. In
principle, the defender’s king has to move to the short side, to allow his own rook to move to
the long side later, where it will be able to harass the enemy king with side checks.
We mentioned already that the defender must have a keen eye for this principle, to enable him to make the
right decisions also in time trouble. One of my former pupils, grandmaster Robin Swinkels, was once given
a tough lesson in endgame technique by GM Michael Roiz. The game was played in Gibraltar 2007, a
wonderful tournament where I was playing too. I remember seeing Robin slogging away in this ending, and
we exchanged a few words during the game – just before the diagram position was reached. I asked him
how he was doing, and he replied ‘it should be drawable, but I’ll probably lose.’ This had nothing to do with
a lack of confidence, but rather with a lack of time.
After a number of exchanges, our rook + pawn vs rook ending appeared on the board, in which the
principle of the ‘short and long side’ became acute. I had once, very summarily, discussed this endgame
with him during a training session, but I reproached myself that I probably hadn’t explained the principles
well enough at the time, seeing the moves Robin was putting on the board. Even though it was a draw
throughout, Robin deviated from the scheme I sketched above, making things so difficult for himself that –
under time pressure, of course; there were no increments yet – he went wrong eventually.

223
Robin Swinkels 2428
Michael Roiz 2605
Gibraltar 2007 (8)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+R0
9r+-+-+-+0
9+-+-+-+-0
9-+-zp-+-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+K+-+-0
xiiiiiiiiy
White to move
Black has just played 41...Kc3, and now, with little time on the clock, Robin has to try to limit the damage in
this endgame.
42.Rc7+ Kd3 43.Kc1
The short side.
43...Rh6 44.Rg7
This is still just about possible...
My advice would always be to follow the principle of putting the rook behind the passed pawn with 44.Rd7
Rh1+ 45.Kb2, after which, for example, 45...Ke3 can be met with 46.Kc2.
44...Rh1+ 45.Kb2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-tR-0
9-+-+-+-+0
9+-+-+-+-0
9-+-zp-+-+0
9+-+k+-+-0
9-mK-+-+-+0
9+-+-+-+r0
xiiiiiiiiy
45...Re1
In reply to 45...Ke2, 46.Rg2+! is the only move! 46...Kf3 47.Rg8, and Black is unable to get the pawn on d3
and the king on d2; for instance, 47...d3 48.Kc3 is an immediate draw.
46.Rh7
White can even choose whether to put the rook behind the passed pawn or on the long side – in both cases,
it’s a draw.
Another thematic move here is 46.Rd7.
46...Kd2 47.Rh4 d3 48.Rh2+ Re2 49.Rh1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+p+-+-0
9-mK-mkr+-+0
9+-+-+-+R0
xiiiiiiiiy
The so-called ‘back-rank defence’. Now, it is essential for White to keep his rook on h1 for as long as
possible, controlling both the h-file and the back rank.
49...Rg2 50.Kb3 Rf2 51.Kb2 Re2 52.Kb3 Re1
This is the moment when White’s rook has to leave the back rank, but it still remains a draw.
53.Rh2+ Ke3
After 53...Kd1, White could have forced the draw with 54.Kc3 Re3 55.Rd2+.
Now, however, White makes a horrible mistake in time trouble:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+K+pmk-+-0
9-+-+-+-tR0
9+-+-tr-+-0
xiiiiiiiiy
54.Kc3??
Immediately fatal.
With 54.Rh3+, he could have obtained the well-earned draw. Black cannot make progress.
54...Rc1+ 55.Kb2 d2 0-1
The pawn reaches the other side.
This was a tough lesson not only for the player, but for the trainer too. Today, Swinkels is a successful
grandmaster, and I assume that he doesn’t need any extra training anymore to handle such endgames
adequately in the future.
We still need to take a look at the details from the first diagram of this example. Fasten your seatbelts –
now, you are going to see an ingenious treatment of this endgame!

224
 XIIIIIIIIY
9-+-+k+-+0
9tR-+-+-+-0
9-+-+K+-+0
9+-+-zP-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tr-+-0
xiiiiiiiiy
Black to move
As we have seen a few times already, the defender’s rook needs to stand at the longest possible distance
from the file on which the pawn is standing. Let’s see how this could work. At this moment, White’s rook
still occupies the a-file; if Black’s rook has to move to the b-file, the ‘checking distance’ is too short.
1...Kf8! 2.Ra8+ Kg7
The defending king has been driven from the promotion square. Now, White’s ultimate goal is to get his
pawn on e6 and his king on e7. This set-up is winning if White can maintain his rook on a8, controlling not
only the a-file but also the back rank. We will see in several other examples why this is so important.
At this moment, White can make three winning attempts, and against each attempt Black has to know
exactly how to defend.
Winning attempt No. 1:
3.Kd6
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9-+-mK-+-+0
9+-+-zP-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tr-+-0
xiiiiiiiiy
Now, 4.e6 followed by 5.Ke7 is threatened. Black repulses this threat easily:
3...Kf7
Keeping the e6-square under direct control, and so White has nothing better than to return to the starting
position.
Here, 3...Rb1? would lose. The winning method will be discussed below in fragment No. 3.
4.Ra7+ Ke8
Black always has to strive to keep his king on, or close to, the promotion square.
5.Ke6
History repeats itself.
5...Kf8 6.Ra8+ Kg7
Here is winning attempt No. 2:
7.Re8
 XIIIIIIIIY
9-+-+R+-+0
9+-+-+-mk-0
9-+-+K+-+0
9+-+-zP-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tr-+-0
xiiiiiiiiy
Now, White puts his rook on e8, exerting extra control on the e6-square. Suppose he could now play Kd7,
then Black would not have the reply ...Kf7? since e5-e6+ follows, driving away the king. Then, White
returns his rook to the a-file and puts his king on e7, achieving his winning position.
7...Ra1!
The defender’s rook can, and must, move to the a-file now to be able to start its checks from the long side.
Since now the ‘checking distance’ is long enough, he can keep the position within drawing margins in this
way.
8.Kd7
Winning attempt No. 3 is 8.Rc8 – White could also have played this instead of 7.Re8, and 7/8.Rd8 amounts
to the same thing too. The idea is to interpose the rook somewhere during the side checks. This is still a
draw as long as Black’s rook stays on the a-file, but it involves some risk and sometimes leads to
(unnecessary) defeat, as we will see in the next fragment. That is why I always advise to adhere to the
principles in such positions, i.e. back behind the pawn with the rook! 8...Re1!. Now that White’s rook is no
longer on e8, he doesn’t threaten to prepare the march of the pawn with Kd6, since Black will play ...Kf7 as
soon as e5-e6 becomes an actual threat. Therefore, White can make no progress and will have to settle for a
draw.
8...Ra7+ 9.Kc6 Ra6+ 10.Kb7 Kf7!
The easiest way to secure the draw.
In the diagram below, we see why the a-file is so important.

225
XIIIIIIIIY
9-tr-+-+-+0
9tR-+-mK-mk-0
9-+-+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
This position is winning for White, but the winning method is not easy. Nevertheless, we can make
ourselves familiar with it if we take another look at the principles. Let’s go through things systematically,
and indicate which goals White is aiming for.

Characteristics:
- White occupies the a-file with his rook.
- Black’s rook is standing on the unfavourable – for him – b-file, meaning that the ‘checking distance’ is too
short in principle.
- Black’s rook is on the back rank, which is favourable for him in some cases. The reason is that if White
plays, for example, Ke7-d7 followed by e6-e7, Black immediately forces a draw with ...Kf7; White can make
no progress in that case.
- White’s king is standing in front of its own pawn, and will have to leave this post one day to be able to set
the pawn in motion.
- Black’s king is ready to approach the pawn if White’s king leaves its post.

Winning method:
White wants to manoeuvre his rook at the right moment to place it behind his pawn. He will prefer to do
this when Black’s king is on g6. Black will want to keep his rook on the back rank as long as possible (in
view of the above-mentioned draw position), since with White’s rook on e1, behind the e6-pawn, and
Black’s king on g6, Black’s king cannot get in front of the pawn. If the king is on g7, he plays ...Kf8, and
White accomplishes nothing.
Let’s have a look at the translation of this discourse into a plausible sequence of moves:
1.Kd6+! Kf6
If Black plays 1...Kg6 voluntarily, White immediately replies 2.Ra1!, transferring to the winning method in
the main line.
If the king moves to the back rank, it gets cut off horizontally, and White can go over to a winning position
in which his dominant king plays the starring role. For example: 1...Kf8 2.Kd7! Kg8 (in principle, 2...Kg7 is
better, but that would mean a transposition to the main line) 3.Ra1! (threatening 4.Rg1+ to drive the king
further away from the pawn) 3...Rb7+ 4.Kc6 Rb8 5.Rf1 Ra8 6.e7 Kg7 7.Ra1! Rb8 (now, it would be best for
Black to go over to the queen vs rook ending – we have discussed this before: 7...Rxa1 8.e8=Q) 8.Rd1
(threatening 9.Rd8) 8...Ra8 (there is no sensible alternative; 8...Kf7 is also met by 9.Rd8) 9.Rd8, securing
the win.
XIIIIIIIIY
9-tr-+-+-+0
9tR-+-+-+-0
9-+-mKPmk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
2.Kd7!
White starts some tempo play.
2...Kg7
And now White completes the triangulation:
3.Ke7!!
 XIIIIIIIIY
9-tr-+-+-+0
9tR-+-mK-mk-0
9-+-+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
We have seen this position before, but White has managed to arrange it so that now Black has to move!
3...Kg6
Finally, Black is forced to play the king to the ‘disastrous’ square g6.
If Black’s rook leaves the back rank with 3...Rb1, the win can be attained as follows: 4.Ra8!.
XIIIIIIIIY
9R+-+-+-+0
9+-+-mK-mk-0
9-+-+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
analysis diagram
As we established before, this ‘magical’ a8-square is highly important. White’s rook is fulfilling a double
function here:
- It keeps occupying the a-file so that the ‘checking distance’ for Black’s rook is still too short.
- It controls the back rank (including the promotion square), making the threat e6-e7 topical at some
moment.
Also, Black misses his ‘back-rank defence’, and so he is running out of defensive options.
4...Rb7+ 5.Kd6 Rb6+ 6.Kd7 Rb7+ 7.Kc6 Re7 8.Kd6 Rb7 (8...Kf6 also loses, to 9.Rf8+) 9.e7, and the pawn
marches on.
If Black refuses to play 3...Kg6 and prefers to keep his rook on the back rank, the latter is so close to the
white king that White can now work with the intended rook manoeuvre: 3...Rc8
XIIIIIIIIY
9-+r+-+-+0
9tR-+-mK-mk-0
9-+-+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
4.Ra1! Rc2 5.Kd7 Rd2+ 6.Ke8 Rb2 7.e7 (the pawn has reached the seventh rank, and White is getting
‘visions of Lucena’!) 7...Rb8+ 8.Kd7, and since the checking distance is too short for the black rook, he will
be forced to give up all resistance. With the rook on a1, the defence with 8...Kf7 no longer works in view of
9.e8=Q+ Rxe8 10.Rf1+, and White collects the rook.
4.Ra1!
XIIIIIIIIY
9-tr-+-+-+0
9+-+-mK-+-0
9-+-+P+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
This is the right moment to reroute the rook.
4...Rb7+
Black is making things as difficult as possible for his opponent by luring the king away from its pawn with
side checks.
5.Kd6
I’m opting for the ‘methodical way’ to finish this endgame properly. Objectively, 5.Kd8!? is a faster way to
force the win, based on a tactical trick. After 5...Rb8+, the ‘trick’ is that 5...Kf6 fails to 6.e7 Rb8+ 7.Kc7 Rh8
8.Kd7 Kf7 9.e8=Q+ Rxe8 10.Rf1+ and White wins, as we have seen before.
5...Rb6+
Black keeps on annoying White.
6.Kd7 Rb7+ 7.Kc6 Rb2
Only now does the rook dive down; 7...Re7 8.Kd6.
8.Re1!
According to the motto ‘rook behind the passed pawn’. 8.Rf1? would be a big mistake on account of 8...Ra2!,
and Black occupies the crucial a-file!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+K+P+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-tr-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
Here, the subtle difference between the black king standing on g6 or g7 becomes clear. With the king on g7,
8...Kf8 would follow here, and White wouldn’t be able to win. But with the king on g6, Black does not have
this defence. He can only work with checks again, but they will soon run out.
8...Rc2+ 9.Kd7
White shouldn’t make the mistake of moving towards the black rook, since the rook would retreat to the
back rank at a certain point to defend against e6-e7.
9...Rd2+ 10.Ke8 Ra2 11.e7
White’s play is very accurate. Since his rook supports the pawn, his king can now move into the open field
to escape the checks.
11...Ra8+ 12.Kd7 Ra7+ 13.Kc6 Ra6+ 14.Kb7
And the promotion of the pawn can no longer be prevented: 1-0

This endgame in practice

How difficult it is to play such theoretical endgames in practice is shown by the following game.

226
Zhu Chen 2480
Mark Taimanov 2455
Roquebrune 1998 (8)
XIIIIIIIIY
9-+-+-+R+0
9+-+-+k+-0
9-+-mK-+-+0
9+-+P+-+-0
9-+-tr-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
After an epic battle, the diagram position arose. Although Black’s king is on the wrong side of the pawn, the
position is still just within the drawing margins. But at the end of a long game, both players were in time
trouble, and the further course of the game demonstrates that strong players like these two can also make
elementary mistakes sometimes.
In my trainings, I have used this fragment regularly to allow my pupils to ‘dabble’ with this subject matter.
We can all blame grandmasters for making a mess of things sometimes (on certain livestreams on the
internet, people can talk, since at the bottom of the screen the engine is running, or the bar at the left side of
the screen will signal when one of the players blunders...), but I would like to challenge you all to take the
chessboard and analyse the move sequence given below for yourselves. Indicate the turning points in the
game, give exclamation marks to strong moves and question marks to errors – the moments when one of
the players spoils a half point.
Below you can find a few instructions for analysing in a proper way. Please compare your analysis with the
one given at the end, and find out how well you have stored the principles of the endgame in your memory!
Maybe you will also be an expert on this endgame afterwards – if not, the experience may make you a little
milder for players featuring on live boards...

Analysing games
How do you analyse a game?
1) First play through the game superficially.
2) Play through the game again, and try to indicate the turning points. Ask yourself constantly who is better.
Provide question marks and exclamation marks.
3) Try to indicate at the critical points (where you have put a question mark) what should have been played.
Look for other candidate moves, and give a few variations.
4) Try to formulate for yourself (in words) what went on in this game.
5) Play through the game for the last time, and check your analysis.
Now, analyse the following fragment (starting from the diagram above) with the help of the above
instructions.
Zhu Chen – Taimanov
1.Rd8 Ra4
2.Rc8 Rd4
3.Rd8 Ra4
4.Rb8 Rd4
5.Rc8 Rd1
6.Rc2 Ke8
7.Kc6 Ke7
8.Re2+ Kd8
9.Rh2 Rc1+
10.Kd6 Kc8
11.Rh8+ Kb7
12.Kd7 Rc7+
13.Ke6 Rg7
14.Re8 Rg5
15.d6 Rg6
16.Kd7 Rg1
17.Re7 Rg8
18.Rh7 Rf8
19.Ke7 Rf1
20.d7 Re1+
21.Kd8 Kc6
22.Kc8
1–0

So now, it is time to present the worked-out analysis:

227
Zhu Chen 2480
Mark Taimanov 2455
Roquebrune 1998 (8)
XIIIIIIIIY
9-+-+-+R+0
9+-+-+k+-0
9-+-mK-+-+0
9+-+P+-+-0
9-+-tr-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
In principle, this is a draw, even though Black’s king is on the long side.
77.Rd8 Ra4!
The only correct defence. If White manages to get the pawn on the sixth rank without any struggle, Black
will be hopelessly lost. With good defence, White can make no progress, but obviously Zhu Chen is going to
try.
78.Rc8
After 78.Kc7 Ra7+ 79.Kb6, Black forces the draw with 79...Ke7.
78...Rd4!
Again, Taimanov finds the only move that draws! So, this is a good start. He can now meet 79.Kc6 with
79...Ke7, when the pawn won’t reach the sixth rank.
79.Rd8 Ra4 80.Rb8
XIIIIIIIIY
9-tR-+-+-+0
9+-+-+k+-0
9-+-mK-+-+0
9+-+P+-+-0
9r+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
80...Rd4!
The only move, as said.
81.Rc8
A trap, since now Black’s rook cannot give check from the side as White can interpose the rook; 81.Kc6
Ke7.
81...Rd1
But so far, Taimanov shows that he knows the ropes.
82.Rc2
An enveloping manoeuvre with the rook, forcing, again, an absolutely only move:
82...Ke8! 83.Kc6
Now, 84.d6 is threatened, but Black is on his guard:
83...Ke7
83...Kd8, 83...Rd3, 83...Rd4 and 83...Ra1 were all OK as well.
84.Re2+ Kd8!
Again, forced.
85.Rh2
XIIIIIIIIY
9-+-mk-+-+0
9+-+-+-+-0
9-+K+-+-+0
9+-+P+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-tR0
9+-+r+-+-0
xiiiiiiiiy
The rook emerges on the other side.
85...Rc1+?
The first signs of fatigue (or uncertainty?) are starting to show. The rook should remain behind the pawn
for as long as possible, so e.g. 85...Rd3! would have kept the position within drawing margins. Another
mistake would have been, for example, 85...Rg1?, when White wins after 86.Rh8+ Ke7 87.d6+ Ke6 88.Re8+
Kf7 89.d7.
85...Rd3 kept Black afloat, e.g. 86.Rh8+ (86.d6 Rc3+ 87.Kd5 Rd3+ 88.Ke5 Re3+ 89.Kd4 Re1) 86...Ke7, and
now 87.d6+ is not possible.
86.Kd6
Now, it is an elementary win for White since Black’s rook is no longer standing behind the pawn. White can
reach a position with Kd7 and the pawn on d6.
86...Kc8
Moving to the short side at any rate – but as we know by now, White’s rook occupies the important h-file,
and therefore the long side is not long enough for Black’s rook!
87.Rh8+ Kb7
XIIIIIIIIY
9-+-+-+-tR0
9+k+-+-+-0
9-+-mK-+-+0
9+-+P+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tr-+-+-0
xiiiiiiiiy
Thus far, Zhu Chen is playing excellently. True, Black’s king has ended up on the right side of the pawn, but
since his rook is not behind the passed pawn, White can now make actual progress:
88.Kd7! Rc7+
The most tenacious defence, since if Black allows 89.d6, the win would no longer be difficult for White.
89.Ke6 Rg7
After 89...Rc1, White can win, for instance, with 90.d6 Re1+ 91.Kd7 Rd1 92.Ke7 Re1+ 93.Kd8 Rd1 94.d7
(Lucena!) 94...Rd2 95.Rh1 Re2 96.Rb1+ Ka7 97.Rb4 (building a bridge!) 97...Re1 98.Kc7 Rc1+ 99.Kd6 Rd1+
100.Kc6 Rd2 101.Rb5 Rd1 102.Rd5 Rc1+ 103.Kd6.
At this point, too, White has only one move to guarantee the win.
90.Re8?
Inexplicable. As it turns out, Zhu Chen does not master this endgame either. The rook was ideally placed on
h8, controlling the back rank and, more importantly, the h-file, preventing Black’s rook from giving checks
from a longer distance. Of course, White now threatens to force a pawn ending with 91.Re7+, but the
Russian grandmaster doesn’t fall for that one.
White could have hauled in the loot only with 90.d6!. It is essential to establish the above-mentioned set-up.
A possible continuation then is 90...Rg6+ 91.Ke7 Rg7+ 92.Kf6
XIIIIIIIIY
9-+-+-+-tR0
9+k+-+-tr-0
9-+-zP-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
92...Rd7 (after 92...Rg1, we see why the h8-square is so important for White’s rook: now, the pawn queens
after 93.d7 since the d8-square is protected) 93.Ke6 Rg7 94.d7, and the pawn marches on.
90...Rg5?
Obvious was 90...Rg6+! 91.Kd7 Rh6!, and White can’t make progress.
The text move is also hard to understand. We know the preferences of the rook, which likes to be placed at
a maximum distance from the enemy king. Black’s move does not meet this desire, and therefore it is bad –
even losing.
XIIIIIIIIY
9-+-+R+-+0
9+k+-+-+-0
9-+-+K+-+0
9+-+P+-tr-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
91.d6!
But this is first-rate! White is again winning.
91...Rg6+
XIIIIIIIIY
9-+-+R+-+0
9+k+-+-+-0
9-+-zPK+r+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
92.Kd7?
Another strange decision, throwing away the win once more.
Obviously, 92.Ke7! was indicated, threatening 93.d7, against which there is no defence: 92...Rg7+ (92...Kc6
93.Rc8+ Kb7 94.d7) 93.Kf6 Rd7 (93...Rh7 94.Re7+) 94.Ke6 Rh7 95.Re7+, and the curtain can be drawn.
The Tablebases indicate that 92.Kd5 also wins for White.
92...Rg1?
Another blunder of the first order. Black could have attained a draw with 92...Rh6! (again, the only move
here!). After 93.Rf8 Rh7+ 94.Ke6 Rh6+ 95.Ke7 Rh7+ 96.Rf7 Rh8,
XIIIIIIIIY
9-+-+-+-tr0
9+k+-mKR+-0
9-+-zP-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Black’s rook has gained possession of both the h-file and the back rank. As we know, White can’t make any
progress here, especially since 97.d7 Kc7 yields nothing.
93.Re7?
As seen, she could have forced the win with 93.Rh8! because the checking distance for Black’s rook is too
short. A mirrored position in which we explain the winning method from then on can be found in diagram 3
of Section 19.9.
XIIIIIIIIY
9-+-+-+-+0
9+k+KtR-+-0
9-+-zP-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-tr-0
xiiiiiiiiy
If you have no knowledge of this ending, then the text move is normal. However, a thematic win could have
been obtained by, again, occupying the long side with the rook. By the way, there are various other ways to
win this endgame, but I will not consider those. What matters is that we apply our knowledge in the correct
way in a practical game, and carry out the required manoeuvres as accurately as possible.
93...Rg8?
Taimanov too goes wrong. Apparently the concept of ‘checking distance’ did not ring any bells with either
of the two players.
Of course, the h-file had to be occupied again with 93...Rh1!=.
94.Rh7!
Aha! One way or another, Zhu Chen is suddenly on the right track.
94...Rf8
Black could have made things much more difficult for his opponent with 94...Kb6! – this requires some
endgame knowledge.
95.Ke7?!
This move doesn’t throw away the win, but it doesn’t show any signs of systematic play.
After 95.Rh1 Rf2 96.Rb1+ Ka7 97.Kd8 Rd2 98.d7, we would have transposed to the Lucena position, and
we know how the bridge can be built here.
95...Rf1
Losing simply. Black could have defended more tenaciously with 95...Rg8, and now the absolute best way
for White to repair the situation is 96.Kd7 (not 96.Rh1?, since Black can then force a draw with 96...Rg7+
97.Ke8! Rg8+ 98.Kf7 Rg2 99.Rd1, and Black’s king creeps in front of the pawn: 99...Kc8!=).
 XIIIIIIIIY
9-+-+-+r+0
9+k+K+-+R0
9-+-zP-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Now, we are just waiting for 96...Kb6, after which White can play her rook to d1. The question is: how can
White make progress if Black plays 96...Rf8 again? Now, she can play 97.Rh1! Rf7+ 98.Ke8 Rf2 99.d7 Re2+
100.Kd8, transferring to the Lucena method. Not too difficult, it seems.
96.d7
This is the fastest path to victory, but there were more roads leading to Rome, for example: 96.Ke8+ Kc6
97.d7 Re1+ 98.Re7 Rh1, and now 99.Re6+!, and White wins (watch out! – 99.d8=Q?? Rh8+ 100.Kf7 Rxd8=).
96...Re1+ 97.Kd8
We’ve reached the Lucena position!
XIIIIIIIIY
9-+-mK-+-+0
9+k+P+-+R0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tr-+-0
xiiiiiiiiy
97...Kc6
Not the most stubborn defence, but the game was lost anyway.
After a waiting move by the rook, White could demonstrate the art of ‘building a bridge’, e.g. 97...Rg1
98.Rh4 Rb1 99.Re4 Rh1 100.Rb4+ Ka6 101.Kc7 Rc1+ 102.Kd6 Rd1+ 103.Kc6 Rc1+ 104.Kd5 Rd1+
105.Rd4, and the job is done.
98.Kc8 1-0
Quite probably, both players were in extreme time trouble during this ending, but nevertheless we saw that
neither of the players commanded the ‘automatisms’ required to handle this endgame. This should
encourage the club player: if even players this strong can go wrong, you could do at least as well as they do
with some systematic knowledge and skill. Just as long as you don’t get into time trouble!

§ 20.10 Hiding-places for the king

An endgame that occurs quite frequently is the one in which the rook is not behind its own pawn (or to the
side of it), but in front of it. If we consider that both sides’ rooks should strive to stand behind the pawn, we
may assume that the enemy rook will be able to keep the pawn under control by moving behind it. Many
players then push their pawn to the seventh rank, not realizing that this reduces the activity of their rook. It
is not very well placed on the back rank, unless there is some tactic yielding a sudden win. Let’s put this
quite common endgame type under the microscope.
• Shelter No. 1: king hides at the edge

228
Rook in front of the pawn – 1
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+K+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
Black easily holds here. His rook stays on the a-file and the king, if necessary, oscillates between the g7- and
h7-squares, where it is sheltered against checks by White’s rook.
1.Kf3
White’s only winning plan is to walk to b6 with his king, to protect the pawn. Then, White threatens to
remove the rook from the promotion square and win. However, at the moment White’s king lands on b6,
Black gives check and then returns to the a-file. Let’s see how this works:
1...Ra2
Black waits, keeping his rook, which is confined to the a-file, behind the pawn. If necessary, he can make
tempo moves with his king to h7 and back to g7.
Certainly not 1...Kf7?? in view of 2.Rh8! Rxa7, and White wins thanks to the well-known X-ray check 3.Rh7+.
Also, any move with the king to a rank other than the second is fatal: 1...Kf6?? 2.Rf8+ followed by 3.a8=Q.
2.Ke4 Ra1 3.Kd5 Ra2 4.Kc6 Ra1 5.Kb6
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-mk-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
This is the moment when Black has to react, as now White threatens to move his rook.
5...Rb1+ 6.Kc6 Ra1
White can make no progress: ½-½

229
Rook in front of the pawn – 2
 XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-+K+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
Also in this position, White cannot win, not even with two extra pawns. There is nothing he can accomplish
with his extra g-pawn:
1.Kf3 Ra2 2.Kg4 Ra1 3.Kh5 Ra2 4.g4 Ra1 5.Rc8
5.g5 Ra4 6.g6 Ra5+ 7.Kg4 Ra1 is not a real winning attempt either; Black keeps pursuing his policy of
‘doing nothing’.
5...Rxa7 6.Rc6 Rb7 7.g5 Ra7
And it’s a draw. There is no way to blow up the black king’s shelter.
In the previous examples, we saw that the defending king could not be driven out of its shelter. This is also
the case if White has an h-pawn, since Black’s king can hide on h7 and then ‘do nothing’ with his rook.
But what is the situation if White’s pawn is not a g- or an h-pawn? With an f-pawn, the position is winning
for White thanks to the X-ray check.

230
Rook in front of the pawn – 3
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-mk-0
9-+-+-+-+0
9+-+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+K+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
In this example, the situation is different.
1.f6+! Kf7 2.Rh8 Rxa7 3.Rh7+ 1-0
So what it comes down to is that the black king’s shelter on g7/h7 is not effective against a white f-pawn.
Now that we know this, we can see if we can win the next position, where the winning method requires a
different way of thinking that may be a revelation for some club players: schematic thinking!

Technique: schematic thinking

Especially in the endgame, it is important to think in ‘schemes’ rather than ‘moves’ – i.e., you
have to determine what the ideal squares are for your pieces and pawns and then find out how
you can get them there. We will expand further on this way of thinking in Part IV.
In the position given on the following page, White can use this schematic way of thinking to determine what
will be the ideal place for his king.

231
Rook in front of the pawn – 4
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-mk-0
9-+-+-+-+0
9+-+-+p+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+K+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
To win, White will have to attack Black’s pawn on f5. This can be done on either of three squares: g5, e5 or
e6. The question is: which of these is the best square for the king?
1.Kf2
The king embarks on a long journey. Many players try 1.Kg3 Ra2 2.Kh4 Ra3 3.Kg5, but are then rudely
awakened by 3...Ra5; White can make no progress here.

1.Kf2 Ra3 2.Ke2 Kh7

Black always has tempo moves by oscillating with his king between the g7- and h7-squares.
3.Kd2 Kg7 4.Kc2 Kh7 5.Kb2 Ra6
The rook remains tied to the a-file.
6.Kb3 Kg7 7.Kb4 Ra1 8.Kc5
Crossing the fifth rank.
8...Ra2 9.Kd6
As we know, 9.Kb6 makes no sense, owing to 9...Rb2+.
9...Ra1
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-mk-0
9-+-mK-+-+0
9+-+-+p+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
Black plays a clever move. Now, it’s White’s turn to outwit him.
10.Ke5!
The right way to put Black in zugzwang. Not 10.Ke6?! in view of 10...Ra5, and White can make no progress.
10...Ra5+ 11.Ke6
Now, the same position arises with the other side to move. Black has only one move left:
11...Kh7
But this is only relative, since now Black is manoeuvred into a real zugzwang:
12.Kf6!
Black has to let go of the pawn, after which the win is easy.
• Shelter No. 2: own pawn
It may happen that the defending king is not able to reach the ‘safe’ seventh rank, and then the enemy rook
can execute the sentence. But sometimes there are other factors coming to the defender’s aid, like for
instance an own pawn protecting the king against the check.

232
Rook in front of the pawn – 5
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-+-0
9-+-+-+p+0
9+-+-+-mk-0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-+K+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
Black’s king has not managed to get to the g7- or h7-squares, but his g-pawn protects him from checks from
the rear and the king has found a safe shelter on g5.
1.Kf3 Ra2
And there is nothing to be gained for White: ½-½
This doesn’t always work...

233
Rook in front of the pawn – 6
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-+-0
9-+-+-+p+0
9+-+-+-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+KzP0
9tr-+-+-+-0
xiiiiiiiiy
White to move
It looks as if Black’s king again enjoys shelter from its g-pawn, but the shelter is blown up with a nice
tactical shot:
1.h4+! Kg4
Of course, 1...Kh6 fails to 2.Rh8+.
2.h5!
This is what it’s about!
2...Ra2+
An ultimate attempt to escape from his predicament.
After either 2...gxh5 or 2...Kxh5, Black will be hit by a check from the rear, after which White’s pawn
promotes.
3.Kf1!
The king has to run as quickly as it can; not 3.Kg1 since then Black draws with the creative 3...g5 4.h6 Kg3
5.Kf1
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-+-0
9-+-+-+-zP0
9+-+-+-zp-0
9-+-+-+-+0
9+-+-+-mk-0
9r+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
analysis diagram
5...Rf2+! (an ingenious way to keep both of White’s passed pawns under control) 6.Ke1 Rf7. Black’s rook
keeps clinging to the a7-pawn from the side. Because of this, White still cannot move his rook, and in the
meantime he can’t use his other pawn either.
3...g5
If 3...Kg5, 4.hxg6 or 4.h6 and White wins; 3...gxh5 4.Rg8+ Kf3 5.a8=Q+.
4.h6
And White wins.
And sometimes, there is an optical illusion that plays tricks on us.

234
Rook in front of the pawn – 7
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-zp-0
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-zPK+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
Again, Black’s king is covered from the rear by a pawn – on g7 instead of g6 this time. Can White blow up
the shelter here?
1.f4+ Kg6
1...Kg4 also leads to a draw, e.g. 2.Kf2 (2.f5 Ra5 3.f6 Ra2+ 4.Kg1 Kg3 5.Kf1 Kf3 6.Ke1 gxf6) 2...Ra3 3.Ke2
Ra1 – Black keeps ‘doing nothing’ with ...Ra1 and ...Ra2, and the moment White’s king moves to f7 to take
the g7-pawn, Black takes on f4 with his king.
2.f5+ Kh7
Black has nothing to fear. Even the very risky-looking 2...Kf7 can do him no harm! After 3.f6, Black coolly
holds the draw by ‘doing nothing’: 3...Kg6 (certainly not 3...gxf6?? 4.Rh8); 2...Kg5 3.f6 Kg6.
3.Kf3 Ra3+ 4.Ke4 Ra4+ 5.Ke5 Ra1
White doesn’t get any further: ½-½

• Shelter No. 3: own rook

What exactly is intuition? If we consult Wikipedia (for what it’s worth), we find the following: in psychology,
intuition is described as implicit inspiration as a result of certain trains of thought and observations – this
in contrast to conscious or explicit knowing and observing.
In his famous thesis Thought and Choice in Chess, Dutch professor Adriaan de Groot claimed that stronger
players distinguish themselves from weaker ones by ‘chess knowledge’. Intuition, for example, originates
from experience or earlier discoveries. By studying many examples, as in this book, a person develops his
intuition (= inner knowledge). Many chess players have good intuition for finding the right moves.

235
Rook in front of the pawn – 8
XIIIIIIIIY
9R+-+-+-+0
9zP-+-tr-+-0
9-+-+k+-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
The situation has changed slightly. This time, Black’s rook is not behind White’s passed pawn; it keeps the
pawn under cover from the side. This is very important, as otherwise White would move his rook and
promote. Black’s king has a few more issues than normally. There are no safe squares on the seventh rank,
since White can move his rook with check and then promote. The queen vs rook ending is a win, as we have
seen.
Black seems to be in zugzwang. Any rook move loses, and which king move should he play?
1...Kf6+!
The only right idea! It seems paradoxical that Black’s king has to stand as far away as possible from the
enemy pawn, but if we think logically, and remember the principle of checking distance, the conclusion is
that, of course, the rook prefers to be as far removed from the enemy king as possible – and the rook has to
follow its king here. From a longer distance, it is much easier for the rook to give side checks to White’s king,
which is going to move to the queenside.
1...Kd6+? looks logical, and this is the move that was suggested by almost every player I showed this
position to. It’s in our nature to move in the direction from which the danger is coming: the pawn on a7. But
it is precisely this logical move that loses: 2.Kd4 Ke6 (the king returns to its original shelter) 3.Kc5 Ke5!.
 XIIIIIIIIY
9R+-+-+-+0
9zP-+-tr-+-0
9-+-+-+-+0
9+-mK-mk-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This sensible move comes with the idea of being able to give check after 4.Kb6. However, White can win
anyway with 4.Kc6! (here, 4.Kb6 does not win immediately; after 4...Re6+, the king has to go back), and
Black has no good moves left: 4...Ke4. The king moves forward in front of the rook to make room for checks.
However, this fails to 5.Kd6! which poses an insoluble problem to Black. The rook has to stay on its post on
e7 for two reasons: one, it has to keep the a7-pawn under cover, and two, it protects its king from checks
from the rear. However, after the text move, Black has to resign.
The alternative idea of 4...Re6+ is also refuted: 5.Kd7 Rd6+ 6.Kc7, and White obtains a winning position.
The rook no longer has the pawn in sight, and at the same time, the protection against the Re8+ check is
also lacking at this moment. Of course, 6...Re6 doesn’t help in view of 7.Rd8, and White wins.
So, if Black wants to keep the crucial d6-square under control, he must play 4...Ke6. But with that move, he
deprives his own rook of the necessary side check, and after 5.Kb6 Ke5 6.Rc8, he is counted out as well.

1...Kf6+! 2.Kd5 Rf7

Slowly, Black’s king and rook shuffle towards the other side of the board.
3.Kc6
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+r+-0
9-+K+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White’s king hurries to b6 to protect the pawn and free his rook. This means that there is only one move for
Black:
3...Kf5!
Making room for his rook to save himself with the familiar side checks.
4.Kb6 Rf6+
This is essential – otherwise, White’s rook could escape from its awkward position.
5.Kb5
 XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-+-0
9-+-+-tr-+0
9+K+-+k+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
5...Rf7!
And immediately back to the seventh rank. The rook has to keep clinging to the pawn; White can make no
progress.
6.Kc5 Kf4
In a funny rhythm, Black’s king is moving up and down in front of its own rook, which in turn is nailed to
the crucial f7-square.
7.Kd6
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+r+-0
9-+-mK-+-+0
9+-+-+-+-0
9-+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This is the final important moment. As we have seen before, White now threatens 8.Ke6, after which Black
could resign. However, Black prevents this with
7...Kf5!
and now again, White needs a lot of time to walk to b6.
8.Kc6 Kf4 9.Kb6 Rf6+
With a draw.
So here, we see how important it is for the defender’s rook to be as far removed from the enemy king as
possible. A beautiful sequence!
In the above example, we found out that the apparently ‘logical’ decision to move the king in the direction
of the enemy pawn had a totally averse effect. In fact, the king had to move to the ‘wrong side’, following a
different logic: that the rook has to be as far removed from the enemy king as possible to be able to give
side checks.
Now, it’s becoming clear that there are shelters in many different shapes and sizes. In the previous example,
Black’s rook kept the enemy pawn under cover from the side, and at the same time protected its own king
against checks from the rear by standing on the same file.

• Shelter No. 4: the enemy king

Another kind of shelter is seen in the next position. Here, Black’s king seeks shelter ‘in the shadow’ of its
opposite number. Unsurprisingly, this does not end well. We will soon see the importance of tactical tricks
in this scenario.
236
Rook in front of the pawn – 9
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+K+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+k0
xiiiiiiiiy
Black to move
Black’s rook has reached its familiar spot behind the a-pawn; however, Black’s king has no shelter against
checks from the rear, other than from its opposite number. So, the following move is forced:
1...Kg2!?
Hiding behind White’s king, who now embarks on a long hike to try to shake off his colleague.
2.Kf4
White tries to enable a check by the rook on every move; Black’s king is forced to follow White’s.
2...Kf2 3.Ke4 Ke2 4.Kd4 Kd2 5.Kc5!
It will soon become clear why White opts for this square.
5...Kc3
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-mk-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
Now, we see a gorgeous motif worth remembering:
6.Rc8!!
It might seem that 6.Kb6 is the solution here, but that doesn’t bring White any direct benefits after 6...Rb1+
7.Kc6 Ra1.
6...Rxa7 7.Kb6+
Black’s rook is lost due to this discovered check. A nice trick – hopefully, it will yield the reader a full point
one day!

§ 20.11 Side checks (Vancura)

In the previous series of positions with the rook in front of the pawn, the pawn had advanced to the
penultimate rank. We also saw that the rook is desperate to get away from its position in front of the pawn
to enable the pawn to promote. We concluded already that the rook is not well placed in front of its own
pawn in most cases, and that its position needs to be improved. Now, we will take a look at positions in
which the pawn has been advanced to the sixth rank only. So, to make things perfectly clear, these are rook
+ pawn vs rook endings in which:
• the pawn is a rook’s pawn, and not further advanced than the sixth rank.
• the stronger side’s rook stands in front of the pawn.
Let’s look at a schematic diagram.

237
Vancura – 1
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
White has left the a7-square vacant, to use it as a shelter for his king against checks from the rear. If White’s
king can reach a7 after a check on the b-file by Black’s rook, White will finally be able to move his rook and
nurse the pawn to the other side:
1.Kc3 Rc1+ 2.Kb4 Rb1+ 3.Kc5 Rc1+ 4.Kb6 Rb1+ 5.Ka7
XIIIIIIIIY
9R+-+-+-+0
9mK-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
Now that White has managed this, he can play his rook via b8 to b6, after which the king can step aside to
grant the pawn free passage. There is nothing Black can do against this, and so it is clear that this is a rather
simple winning plan for White unless Black can come up with something special early on.
The above would mean that White should never advance his pawn to a7, and he should keep this square
available for his king. This way, he would win in all cases – or...?

238
Vancura – 2
 XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-tr-+0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
It was Josef Vancura who made a clever discovery in this ending type: the defending rook doesn’t have to
remain behind the pawn forever – it is better to move it to the side. We will call this ‘clinging’. Only if the
pawn moves forward to a7, does Black have to put the rook behind it again.
The scheme looks approximately as follows: White’s king strives to reach b5 to protect the pawn, to then
move his rook out of the corner. If he can manage this, he will win. But Black’s rook can ‘cling’ to the pawn
from the side as long as it stands on a6. Only when White’s king appears on b5, does Black have to drive it
away. We can see fairly quickly that the ‘shelter’ on a7 is not a real shelter in this case, since White’s king
can be harassed by checks from the side.
Please note that the black rook will prefer to give these checks from a long distance, to avoid problems with
the checking distance. Black’s king should not be in the way of its rook, and should therefore stay in its own
shelter on the g7- and h7-squares.
Therefore, White tries a trick:
1.Ke5
Suddenly threatening 2.Rg8+!.
1...Rb6
So, Black’s rook has to move. It has to keep the a6-pawn under cover from the side, and if White plays a6-a7,
the rook has to be able to get behind the pawn with ...Ra6.
2.Kd5
The king hurries towards the pawn.
It is always important to investigate how the defender should react to a check by White’s rook: 2.Ra7+. In
this case, the king can move to the back rank, even though this means that it is cut off on the seventh rank:
2...Kg8 (2...Kg6 is also possible) 3.Kd5 Rf6.
XIIIIIIIIY
9-+-+-+k+0
9tR-+-+-+-0
9P+-+-tr-+0
9+-+K+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
It is important to be able to use all the squares on a file for side checks. This the reason why Black plays the
rook to the f-file and not to the g- or h-files, where Black’s king would be in the way. 4.Kc5. Now, Black can
choose whether to play a king move or to start with the side checks already: 4...Kh8 5.Kb5 Rf5+ 6.Kb6
Rf6+ 7.Kb7 Rf7+ 8.Ka8 Rf8+ etc.
If 2.a7, the rook has to get behind the pawn immediately: 2...Ra6 (‘clinging’ to the pawn from the side with
2...Rb7?? loses immediately to 3.Rg8+ Kxg8 4.a8=Q+).
2...Rf6
Now, the rook can safely return. As we mentioned in the above analysis, it is important to put the rook on a
file where it can keep harassing White’s king on both the sixth, seventh and eighth ranks. So, the scheme is
to keep the king on g7 (or h7) and keep the rook on the f- (or g-)file, so that the three back ranks remain
available.
Certainly not 2...Kf7?? since this is met by 3.a7 Ra6 4.Rh8, and suddenly White wins thanks to the X-ray
check.
3.Kc5 Rf5+
3...Kh7 4.Kb5 Rf5+ is fine too.
4.Kb6 Rf6+ 5.Ka7 Rf7+ 6.Kb8 Rf8+ 7.Kb7 Rf7+
And thanks to the Vancura method, Black holds the draw.
So, the idea of placing the defending rook to the side to enable side checks was invented by a certain Josef
Vancura. His father had the same first name, but he didn’t occupy himself with endgame studies, Since Josef
jr. passed away at a young age, Josef sr. later – in 1924 – published his son’s studies.
I came across this remarkable defensive system in Averbakh’s endgame book series, where he gave the
following position:

239
Vancura – 3
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+{+-+0
9+-+-+{}{0
9-+-+-}{}0
9+-+-}{}{0
9-+-}{}K}0
9tr-+-+-+-0
xiiiiiiiiy
Black to move
This mysterious diagram dazzled me when I first saw it. It is claimed that if White’s king stands on one of
the marked squares, with Black to move, the position is a draw (thanks to Vancura, among others). If the
king is not on one of these squares, White always wins! It is very difficult to understand this quickly. For
example, what is the difference between the d2- and c2-squares? With the king on the first, it’s a draw, on
the second it’s a win!
Of course, it would take us much too far to check all the squares one by one on ‘correctness’, but to give an
indication of the defensive strategy we will discuss the current position with the king on g2. Here, Black
reaches Vancura’s defensive set-up with the following manoeuvre:
1...Ra5!
The rook stays behind the pawn for a while, but it will try to move to the side with gain of tempo, i.e., after
2.Kf3
Black gives a check:
2...Rf5+ 3.Ke4 Rf6
‘Clinging’ to White’s pawn. It is also essential that the rook can immediately move back to the a-file after a6-
a7.
4.Kd5 Kh7
Black makes waiting moves.
5.Kc5 Rg6
The rook moves as far away as possible.
6.Kb5
Now that the king protects the pawn, Black has to react:
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-+k0
9P+-+-+r+0
9+K+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
6...Rg5+! 7.Kb6 Rg6+ 8.Ka7 Rg7+
And White can’t make any progress: ½-½
Incidentally, Averbakh and later also Dvoretsky ascribed this position to Peter Romanovsky.
Now that we know part of Black’s strategy, it is also important to see if we can analyse the issue further.
How can we visualize the various strategies from the above position? Let’s make a list of some of these
strategies:
For White:
• The rook is most unfavourably placed in front of the pawn, and therefore White will want to improve its
position as soon as he gets the chance.
• The king wants to get close to the pawn as soon as possible, and, if possible, will use the shelter on a7.
For Black:
• The rook can take care of the pawn in two ways:
1) behind the pawn
2) from the side
• The king also has two options:
1) it can remain in its ‘shelter’ on g7 and h7, so as not to fall victim to an X-ray check or a check from the
rear
2) under certain circumstances, it has to approach the pawn as quickly as possible, to cut off the enemy
king in case of a race
The main strategy for the defender will be to move his rook to the side as a useful improvement. We will
see that in the above position, White can put his opponent’s king in a spot by working with Ra8-a7+, when
the king has to choose between the squares f6, g6 and g8.
I have tried to chart these difficult choices for the defender in the table below. Of course, in every
instance it is important where White’s king is standing in relation to Black’s rook, and also who is to move.
With the help of this table, I have analysed various positions, as can be found on the next page.
240
Vancura – 4
Kc2/Ra1, Black to move
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+K+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
Black to move
The king is now just outside the zone, so White should be able to win this.
It is clear that a cut-off along the third rank with 1...Ra3 is totally useless since after 2.Kb2, the rook has to
remain on the a-file. After a move like 2...Rf3 (intending 3...Rf6), White wins immediately with 3.Rb8!, and
the pawn walks on: 3...Rf7 4.Rb7.

Also after 1...Ra5, Black will just fail to get the Vancura position: 2.Kb3 Rb5+ 3.Ka4! Rb6.
 XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9Ptr-+-+-+0
9+-+-+-+-0
9K+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This ought to be the idea, and if the rook were on f6 here, the draw would have been a fact. But after 4.Ka5
Rf6, White’s king protects the a5-pawn in time, and so he can move his rook and win; for example, 5.Rc8
Rf5+ 6.Kb6 Rf6+ 7.Rc6, and the advance of the pawn can no longer be prevented.

1...Rh1
This is indicated by the Tablebases as one of the most tenacious defences. But Black cannot prevent White
from collecting the full point:
2.Ra7+!
Without this zwischenschach, it doesn’t work. After, for example, 2.Kb3 Rh6, Black has already attained the
Vancura position.
2...Kf6
The king has to move ‘upward’ – moving ‘downward’ loses almost always, and certainly here: 2...Kf8 3.Rb7
followed by a6-a7, and the cut-off along the seventh rank is fatal for Black.
3.Kb3!
Now, White moves in the direction of the pawn with his king. It’s a draw after a move like 3.Rb7? Ke6 4.a7
Ra1 5.Kb2 Ra6 6.Kb3 Kd6 7.Kb4 Kc6.
3...Rb1+ 4.Kc3
Funnily enough, according to the Tablebases the win would take a little longer after 4.Kc4.
XIIIIIIIIY
9-+-+-+-+0
9tR-+-+-+-0
9P+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
4...Ra1
Black would prefer to have the rook on the side, but this loses: 4...Rb6 5.Rh7 (threatening with an X-ray
check!) 5...Kg6 6.a7 Ra6 7.Rb7,
 XIIIIIIIIY
9-+-+-+-+0
9zPR+-+-+-0
9r+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
and here we see that White’s king can very quickly move upward to support the pawn. The black king’s run
to the other side comes much too late.
Now that the rook is behind the pawn again, White vacates the a7-square again:
5.Ra8 Ra4
This is also a clear example where the king cannot move in the direction of the pawn: after 5...Ke6, White
plays 6.a7!, when either a check from the rear or an X-ray check finishes Black off.
6.Kb3 Ra1 7.Kb4 Ra2 8.Kc5 Ra1 9.Kb6
Now that the king reaches the shelter on a7 on the next move, without Black’s rook having been able to
reach the side, White wins.

241
Vancura – 5
Kc2/Ra3, Black to move
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Black can just hold by a thread here:
1...Rh3!
We might ask whether 1...Rf3? isn’t better, to immediately put the rook on the f-file, where it wants to be
anyway. But this move loses due to White’s well-known stratagem: 2.Ra7+ Kf6. Since White’s rook will now
also take up a position on the side, Black’s king has to move towards the pawn as quickly as possible. But
it’s too late, since White can occupy the h-file with his rook: 3.Rh7 Ra3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+R0
9P+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
4.Kb2 (a tactical trick!) 4...Ra5 (4...Rxa6?? 5.Rh6+), and now White has a thematic win. He puts the rook
behind the pawn with 5.Rh3!, and then:
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9P+-+-mk-+0
9tr-+-+-+-0
9-+-+-+-+0
9+-+-+-+R0
9-mK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
5...Ke6 6.Ra3! Rb5+ 7.Kc3 Rb8 8.a7 Ra8 9.Kb4 Kd6 10.Kb5 Kc7 11.Ka6, and the rest is easy. Black’s king
will be driven away from the pawn with a check, after which White’s king can force the promotion.

1...Rh3! 2.Kb2
If White wants to improve the position of his rook with 2.Rb8, Black’s rook has to be placed behind the
pawn immediately: 2...Ra3 3.Rb6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mk-0
9PtR-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Again, now we get a kind of race between the two kings who both hurry towards the pawn. Black is in time
to ward off the most important dangers, as becomes clear after 3...Kf7 4.Kb2 Ra5 5.Kb3 Ke7 6.Kb4 Ra1
7.Kb5 Kd7 8.Rb7+ Kc8 9.Rh7 Kb8, with a draw.
At this moment, 2.Ra7+ makes little sense after 2...Kf6, for two reasons:
1) White cannot play his rook to the h-file;
2) his king has to cross the third rank first, to get to the pawn, and this is only possible if White threatens to
push the pawn.
2...Rh6
Now, the rook is again ‘clinging’ to the pawn from the side.
3.Kb3 Rf6
And we have a Vancura situation on the board: ½-½

242
Vancura – 6
Kc2/Ra1, White to move
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+K+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
Of course, White’s king moves towards the pawn as quickly as possible:
1.Kb3! Rf1
Black tries to ‘cling’ to the pawn from the side. After 1...Kf6, Black no longer has a decent way to defend,
since White can simply carry out his basic plan with 2.Kb4.
2.Ra7+ Kf6
With 2.Ra7+ Kg6, Black could at any rate prevent the enveloping manoeuvre of the rook by 3.Rh7. The
downside to this move is that the king is very far removed from the pawn. White wins as follows: 3.Rb7 Ra1
(3...Rf6 4.a7 Ra6 5.Kb4 followed by Kb4-b5, threatening to build a bridge on a6 with Rb6+, winning easily)
4.a7 Kf6 (Black tries to walk towards the pawn anyway, but he is much too late) 5.Kb4 Ke6 6.Kc5, and
with this shoulder budge, White prevents Black’s king from approaching the pawn, while at the same time
supporting the pawn with his own king.
XIIIIIIIIY
9-+-+-+-+0
9tR-+-+-+-0
9P+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+K+-+-+-0
9-+-+-+-+0
9+-+-+r+-0
xiiiiiiiiy
3.Rh7!
It is this enveloping manoeuvre that brings White the win.
3...Ra1 4.Rh2!
Again, White makes use of the X-ray check.
4...Rb1+ 5.Rb2 Ra1 6.Ra2 Rb1+ 7.Ka4 Rb8 8.a7 Ra8 9.Kb5
And Black is busted:
1-0

243
Vancura – 7
Kc2/Ra3, White to move
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
1.Kb2
White starts to hunt Black’s rook, thus gaining time to move the king forward.
1...Ra5
Moving to the side with 1...Rf3 loses immediately to 2.Rb8, since the rook cannot move backwards now.
Also, he cannot prevent the pawn directly reaching the other side in view of 2...Rf7 3.Rb7.
2.Kb3 Kf6
Black quickly tries to bring his king closer, but he is too late. ‘Clinging’ from the side with 2...Rb5+ 3.Ka4
Rb6 is also too late: 4.Ka5 Rf6 5.Rc8 wins, as we have seen before.
3.Kb4 Ra1 4.Kb5
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-+-0
9P+-+-mk-+0
9+K+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
White has managed to protect the pawn already, and is thus able to free his rook. Black cannot manage to
reach the side without any trouble.
4...Rb1+
4...Re1, for example (to threaten with checks from the side), can be refuted in various ways. The simplest is
5.a7 since White threatens with a check from the rear, and since the defender’s rook is no longer on a1,
5...Kg7 is met by 6.Rb8, and 7.a8=Q can’t be prevented.
5.Kc5 Ra1 6.Kb6 Rb1+ 7.Ka7
And White wins.

244
Vancura – 8
Kd2/Ra1, Black to move
 XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9tr-+-+-+-0
xiiiiiiiiy
Black to move
The question is, of course, why the position with the king on d2 can be held to a draw by Black. I will give a
plausible sequence of moves to show why:
1...Ra3
From here on, I will give the moves that look logical for us human players, and the assessments which are
confirmed by the Tablebases.
The Tablebases also give 1...Rb1, which is not illogical. The rook is immediately threatening to cling to the
side of the pawn with 2...Rb6. True, Black allows White’s rook to escape from a8, but that can’t harm him
after 2.Rc8 Ra1 3.Rc6 Kf7 4.Kc2 Ke8 5.Kb2 Ra5 6.Kb3 Kd7 7.Rh6 Kc7, and it is clear that White cannot
win this.

1...Ra3 2.Ra7+
This is the best attempt also in this situation.
2...Kf6 3.Kc2 Ke6 4.Kb2 Ra5 5.Kb3 Kd6 6.Kb4 Ra1
XIIIIIIIIY
9-+-+-+-+0
9tR-+-+-+-0
9P+-mk-+-+0
9+-+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
7.Rh7
A final, sneaky attempt. If 7.Kb5, Black will start with his ‘shower of checks’ from the first rank, starting
with 7...Rb1+, and White can’t achieve anything.
7...Kc6
Obviously, Black’s king wants to move closer to the pawn. After 7...Rxa6??, Vancura would turn in his grave...
8.Rh6+.
8.a7 Kb6
And the pawn falls – draw!

245
Vancura – 9
Kd2/Ra3, Black to move
 XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
We don’t have to dwell on this position for too long. Black can immediately place the rook to the side of the
pawn:
1...Rb3! 2.Kc2
Of course, after 2.a7, the rook has to return to the a-file with 2...Ra3.
2...Rb6
And again, the Vancura motif is on the board! ½-½

246
Vancura – 10
Kd2/Ra1, White to move
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9tr-+-+-+-0
xiiiiiiiiy
White to move
White’s king immediately steps out of the ‘draw area’ and so White can win, although it takes some effort:
1.Kc3 Kf7
An interesting and also important attempt to save the game at any given move is 1...Rb1, intending to ‘cling’
to the pawn via the b6-square. Here, White has three ways to win, and I have selected the continuation that
is the most logical for a human player: 2.Ra7+ (in the other examples, we have already seen the idea behind
this nasty little check), and now:
A) Since the Rh7 idea, working with the X-ray check, is so essential here, the question arises whether Black
shouldn’t play 2...Kg6. However, the king is too far away for Black to be able to thwart White’s following
action: 3.Rb7
 XIIIIIIIIY
9-+-+-+-+0
9+R+-+-+-0
9P+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
analysis diagram
3...Ra1 (the Tablebases give 3...Rc1+ – which doesn’t look very logical – as slightly more tenacious, but Black
also loses after 4.Kb2 Rc6 5.a7 Ra6 6.Kb3 Kf6 7.Kb4 Ke6 8.Kb5 Ra1 9.Kc6) 4.a7 Kf6 5.Kc4 Ke5 (5...Ke6
makes no difference: 6.Kc5!+–) 6.Kc5, and White wins thanks to the well-known ‘shoulder budge’;
B) 2...Kf6
XIIIIIIIIY
9-+-+-+-+0
9tR-+-+-+-0
9P+-+-mk-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
analysis diagram
3.Rh7! Ra1 4.Kb4 (there are other good moves, but this is the most logical) 4...Kg6 (Black excludes the X-
ray check and forces White’s rook to move in front of the pawn again; if 4...Ke6 5.Kb5, and White’s king
reaches the shelter in front of the pawn) 5.Ra7 Rb1+ 6.Kc5.
XIIIIIIIIY
9-+-+-+-+0
9tR-+-+-+-0
9P+-+-+k+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
analysis diagram
This moment is essential for understanding this endgame. With the black king on g6, suddenly other
problems arise that need to be fathomed by both players.
B1) 6...Ra1 7.Ra8, and the manoeuvre Kc5-b6-a7 followed by the removal of the rook from the corner
cannot be prevented. Black hasn’t managed to get his rook to the side of the pawn;
B2) I was wondering how things would end if Black continues giving checks. If White’s king flees to a8, the
black rook will post itself behind the pawn again. White’s strategy is to bring the king to the back rank and
then interpose the rook: 6...Rc1+ 7.Kb6 Rb1+ 8.Kc7 Rc1+ 9.Kb8.
XIIIIIIIIY
9-mK-+-+-+0
9tR-+-+-+-0
9P+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-tr-+-+-0
xiiiiiiiiy
analysis diagram
Again, an important moment. 9...Rb1+? is senseless here, since 10.Rb7 wins immediately. But the black rook
has two ways to ‘cling to’ the pawn, and then how can White make progress? 9...Rc6 (if the black rook
moves behind the pawn now (9...Ra1), White has to create a shelter on a7 again; the Tablebases claim that
10.Kc8 Ra3 11.Ra8 is the most efficient win here) 10.Kb7 Rf6.
XIIIIIIIIY
9-+-+-+-+0
9tRK+-+-+-0
9P+-+-trk+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
The rook moves to the side to introduce Vancura ideas, but as long as Black’s king is not on g7 yet, he can’t
manage this: 11.Ra8 (the only way to make progress; White now threatens to improve his rook. If only
Black’s king had been on g7 or h7, this position would be a draw!) 11...Rf7+ (again, Black starts an offensive
with checks – this time, from the side. It’s quite special the way White manoeuvres with his king to escape
the checks, while still remaining inside the winning zone!) 12.Kb6 Rf6+ 13.Kb5.
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-+-0
9P+-+-trk+0
9+K+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
The king keeps protecting b5 for the moment, thereby threatening 14.Rc8. From here onwards, White’s
task is to keep the king inside the winning zone, in case Black suddenly plays ...Kg7: 13...Rf5+ 14.Kc4
(finally, the king moves in the direction of the enemy rook) 14...Rf4+ (from here on, we have to examine at
every move whether – and how! – White can win if Black’s rook moves behind the pawn again. Here, after
14...Ra5, White wins with 15.Kb4 Ra1 16.Kb5, and White will be able to improve his rook after fleeing to a7
with his king) 15.Kd5 Rf5+ (what happens after 15...Ra4 ? Quite simple – the king reaches its pawn nicely
in time: 16.Kc5 Ra1 17.Kb6+–) 16.Ke4! Ra5 (this time, the rook can’t be placed on the side with 16...Rf6,
since White can now exploit the unfortunate position of Black’s king with 17.a7 Ra6 18.Rg8+).
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-+-0
9P+-+-+k+0
9tr-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Out of necessity, the rook has ended up behind the pawn again. Please note that Black’s king is also not on
the desired g7-square yet. And since White’s king is inside the winning zone, we know that he can win as
follows: 17.Kd4 Kg7 (other moves fail as well) 18.Kc4 Ra1 19.Kb5 – with this move, the a6-pawn is
protected, and White can finally liberate his rook.
2.Kb4
As we saw also in the previous examples, the king moves towards the pawn as quickly as possible. Black’s
rook doesn’t manage to get to the side of the pawn now, e.g.:
2...Re1
Both Black’s king and his rook are wrong-footed here.
XIIIIIIIIY
9R+-+-+-+0
9+-+-+k+-0
9P+-+-+-+0
9+-+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tr-+-0
xiiiiiiiiy
White can finish the game efficiently now:
3.a7! Ra1
3...Kg7 also loses in view of 4.Rb8, followed by 5.a8=Q.
4.Rh8 1-0
Again, the X-ray check decides:

247
Vancura – 11
Kd2/Ra3, White to move
 XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
With the black rook on a3, White cannot win. This is not so strange, since the king has to make a detour to
reach the pawn. The attempt
1.Kc2
is met with
1...Rh3!
and since we have studied the rest of the game in position Vancura 5, we know this is a draw.

Technique: the ‘Vancura principle’

The Vancura principle is the defending method of moving the rook to the side, ‘clinging’ to the
stronger side’s pawn, and ready to give side checks. This doesn’t allow the enemy king to find a
place to hide and escape the draw.

Conclusions
If we take the scheme with a white rook’s pawn on a6 and a white rook on a8, with the black rook behind
the pawn (on a1 or a3), we can establish the following rules:
• By keeping the pawn on a6 instead of putting it on a7, White creates a shelter for his king, enabling it to
walk out of the checks from the rear by Black’s rook. After this, White can improve the position of his rook,
build a bridge, and nurse the pawn to the other side.
• The Vancura principle is based on giving checks from the side, so that the ‘shelter’ on a7 is no longer a real
shelter for the white king.
• The defending side will try to ‘cling’ to the pawn from the side with his rook. Thus, he prevents the white
rook from improving its position, and at the same time White’s king loses its shelter.
• White sometimes has the idea Ra8-a7+, to force the black king to make a decision. If the king moves to f6
(in the direction of the pawn), White sometimes has an ‘enveloping manoeuvre’ with the rook (Ra7-h7),
using the X-ray check motif.

This endgame in practice

During the period I was working on this book, this theme occurred in an actual game between two players
who are friends and teammates, during a knock-out match to qualify for the Dutch Championship in 2021.
Almost-grandmaster, IM Thomas Beerdsen could have made a draw against the fresh grandmaster Max
Warmerdam by using the Vancura principle, but he didn’t manage to unearth it. Chess journalists duly
jumped on this missed opportunity.

248
Max Warmerdam 2562
Thomas Beerdsen 2454
Hoogeveen ch-NED 2021 (4)
 XIIIIIIIIY
9R+-+-+-+0
9+-+-+-+-0
9P+-+-mk-+0
9+-+-+-zp-0
9-+-+-+-+0
9+-+-+PmK-0
9r+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White started with
45.f4
to trade off the pawns, and then try to reach the shelter on a7 with his king.
45...Ra4??
Black misfires horribly. He could have held the draw in two ways:
A) 45...gxf4+ 46.Kxf4 Ra4+ 47.Ke3 Kg7 48.Kd3
XIIIIIIIIY
9R+-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+K+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
48...Rb4! (this is the equalizer! As Vancura prescribed, the rook prepares to hang on to the side of the pawn.
If necessary, it can move back behind the pawn immediately. After 49.Kc3 Rb6 50.Kc4 Rf6, we have the
famous Vancura position;
B) The 45...Ra3+ check would have saved Black as well, since now after 46.Kg4 gxf4 47.Kxf4 Ra4+ 48.Ke3
Kg7, it is a draw in the same way as in Variation A.
46.fxg5+
The only winning continuation!
46...Kg6
Black’s problem is that he can’t take back the pawn with 46...Kxg5? in view of 47.a7, and a check by White’s
rook cannot be prevented.
47.Kf3 Kg7 48.Ke3 Rc4
Now, he tries to get his rook to the side of the pawn. But since White has an extra pawn now, it is an
entirely different story.
49.Rb8!
 XIIIIIIIIY
9-tR-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-+-zP-0
9-+r+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This is the difference with the position without the g-pawn! White improves the placement of his rook, after
which his king walks to the queenside.
49...Ra4 50.Rb6 Ra5 51.g6 Rd5 52.Ke4 Rd1 53.Ke5 Rd2 54.Rd6
And Beerdsen surrendered. There could have followed 54...Ra2 55.Kd5 Ra1 56.Kc5 Ra2 57.Kb6 Rb2+
58.Kc7 Ra2 59.Kb7 Rb2+ 60.Rb6+–.
The winner of this game, the young grandmaster Max Warmerdam, advanced to the four-player event
where the Dutch Championship was decided, and won the title convincingly!
When I once presented the defensive technique described above to a strong group of junior players, I
invented on the spot a way to introduce Vancura’s idea to them in an expressive way. I told them that if
White’s king has to take a walk all over the board, it will be harassed by ‘a shower of checks’ from the rear.
As a pedestrian, what do you do if you’re walking in the rain, without an umbrella, looking for shelter? You
run to the next bus stop. ‘Square a7 is the bus stop,’ I told the group, ‘and that is where the king seeks
shelter from the rain.’
But is there a way for the defender to make it rain inside the shelter, i.e. the bus stop? That’s very simple –
the rain has to come from the other side – the opposite side of the board, in our case. And that is exactly
what Vancura’s invention implies: he makes it rain inside the bus stop as well! When I met one of my
former pupils twenty years later, I noticed how effective this metaphor had been – he was still talking about
the bus stop...

Van Wely and the bus stop

Of course, it is nice if you can save a game using this principle. And it is even nicer if you can do this against
a former pupil who has grown to be a much stronger player than yourself on paper: grandmaster Loek van
Wely. We regularly visited each other in the day. After he had become very strong very quickly, Loek, of
course, still remained welcome at my home, and there were still a few areas in chess in which I could help
him. But I soon told him there were much stronger players who could be of more help to him. Time passed
by, and although we still had contact now and then, it was clear that he was taking a run up towards the
world’s top. As Loek became an ‘established player’, my visits to him became less frequent.
Sometimes, our meetings were quite curious. One day, when I was passing through, I was standing
downstairs and pushed the doorbell of his Tilburg flat. His voice sounded through the intercom: ‘Who’s
there?’ When I mentioned my name, he replied, ‘What are you here for?’ I knew Loek wasn’t being
unfriendly, and realized he might not want to be disturbed since he was probably unravelling some chess
variation. ‘I guess you’re busy looking at some opening line.’ ‘Yes, I am,’ he said. ‘If you’re not here for the
Botvinnik Variation, you’d better go!’
As it happened, I hadn’t come for the Botvinnik Variation, but I was curious to know what he had found. So
he let me in, and for a while we looked at some positions he was examining. I asked him casually whether it
was wise to work on opening theory only: ‘Wouldn’t it be good to work on rook endings for a bit, for
example?’
He knew that this was one of my hobby-horses, but he brushed the subject aside. When I saw that I wasn’t
going to make a very constructive contribution to the theory of the Botvinnik Variation, I left.
We crossed swords on the board a few times, and while I was able to beat him in his younger years, I
started suffering defeats as he was getting stronger. After he had become a grandmaster – by writing a good
many tournaments to his name, among others – it was clear that I didn’t stand much chance against him
anymore.
By coincidence, we met at the board in an early round of the Lost Boys Tournament in Antwerp, 1996. Loek
had been a grandmaster for a while already, and had just passed the Elo border of 2600 – a magical border
at the time. With 2605, he had a good 200 points more than me. Even though initially the game didn’t go
badly for me, he still managed to gradually drive me into a corner, and my defeat seemed unavoidable. But I
kept on resisting stubbornly, and reached a rook ending that looked hopeless at first glance.

249
Loek van Wely
Herman Grooten
Antwerp 1996 (2)
XIIIIIIIIY
9-+-+-+k+0
9tR-+-+-+p0
9-+-+-+p+0
9zP-+-+-zP-0
9-+-+-+-tr0
9+-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
Material is even, but nevertheless, White seems to be on the verge of winning. Black’s king is cut off on the
back rank, the distant a-pawn can advance quickly, and Black’s pawn majority is fixed.
But there is one factor that keeps White’s win from being evident: White’s rook is wrongly placed in front of
its own pawn. The optimal position for it is behind the pawn, while it would also be better placed at the side.
So, Black has to prevent any improvement of White’s rook at all times.
51.a6
A logical move, although there was an important alternative.
White has no time to improve his rook with 51.Rb7 since Black can also create a passed pawn with 51...Ra4
52.Rb5 Kg7 53.Kc3 h6!. This enables him to develop just enough counterplay to save the half point:
54.Kb3 (54.Rb7+ Kg8 55.gxh6 Rxa5 is a dead draw as well) 54...Ra1, and Black draws; for example,
55.gxh6+ Kxh6 56.Kb4 g5 57.Rb6+ Kh5 58.a6 Kg4 59.Kb5 Kf3 60.Rf6+ Ke3 61.Kb6 g4, and at a certain
moment Black can sacrifice his rook for White’s a-pawn and force the draw with his g-pawn.

51.a6 Ra4
Now, it’s high time for the rook to stand behind White’s passed pawn, otherwise White would win
immediately with 52.Rb7 followed by 53.a7, and Black will have to sacrifice his rook for the pawn.
52.Kc3
 XIIIIIIIIY
9-+-+-+k+0
9tR-+-+-+p0
9P+-+-+p+0
9+-+-+-zP-0
9r+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White’s plan is clear: he hurries ‘upward’ with his king to try to protect the pawn with it, free his rook and
then promote the pawn. Now, I conceived a diabolic defence against that plan:
52...Kh8!
Initiating counterplay in an ingenious way. We will soon see why the king has to stand here.
53.Kb3 Ra1 54.Kb4
XIIIIIIIIY
9-+-+-+-mk0
9tR-+-+-+p0
9P+-+-+p+0
9+-+-+-zP-0
9-mK-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
White continues his plan.
54...h6!!
Along with 54...h5, the only way to create a passed pawn of my own.
55.gxh6 g5!
This is why the king had to be on h8; otherwise, 56.Rg7+ and 57.a7 would have followed here, with an
immediate win for White.
56.Kb5
Now, White’s king protects his passed pawn, and he threatens to liberate his rook from its awkward
position. Black has nothing left except giving checks.
White won’t get anywhere after a move like 56.Ra8+, to create space for his own king on a7. After, for
example, 56...Kh7 57.Kb5 g4 58.Rd8 g3 59.Rd2 Kxh6 60.Rg2 Ra3 61.Kb6 Kg5 62.Kb7 Kf4, it is clear that a
draw is inevitable.
56...Rb1+
 XIIIIIIIIY
9-+-+-+-mk0
9tR-+-+-+-0
9P+-+-+-zP0
9+K+-+-zp-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+r+-+-+-0
xiiiiiiiiy
57.Kc4
In principle, this is a step in the wrong direction, since Black can now immediately force the draw by
making use of what we know as the Vancura position. Van Wely was very angry that he’d let himself be
fooled, but closer inspection teaches us that, curiously enough, there is no win for White anyway.
At the time this game was played, such positions could not be properly analysed with an engine; it would
have given a far too high evaluation for White here because the engine couldn’t ‘see’ that certain positions
are theoretical draws. The modern Tablebases provide the definite answer: it’s a draw!
A) White’s most consistent try is 57.Kc5!?, which appears to win at first sight: 57...Ra1, and now:
A1) 58.Kd4 (first, the king has to stop Black’s g-pawn, but he cannot prevent Black’s rook from moving to
the side of White’s a-pawn) 58...g4 59.Ke3 g3 60.Kf3 Ra5!. This is the essence of Black’s defence. After
61.Kxg3 Rg5+ 62.Kf4 Rg6, Black’s rook ‘clings’ to White’s a-pawn. The moment White defends the a-pawn
with his king, he gets a check slapped on him, and so he can’t make any progress. Only if White plays Ra7-
a8+ and a6-a7, does the black rook return to the a-file, thereby forcing a draw as well;
A2) Also 58.Kd5?! doesn’t help White one bit after, for instance, 58...Ra4!, and the g-pawn saves the half-
point. Black can also do this with 58...g4!, since then 59.Ke4 g3 60.Kf3 can be met by 60...Ra5! 61.Kg2
(61.Kxg3 Rg5+ 62.Kf4 Rg6 63.Ra8+ Kh7 64.a7 Ra6 is also a draw, even though Deep Fritz 11 gives no less
than +5.41!) 61...Ra1 62.Kxg3 Rg1+ 63.Kf4 Rg6, and Black is in time to start ‘clinging’ from the side – this
position too is a draw.
B) The other winning attempt with 57.Kc6 doesn’t yield anything either: 57...Rc1+ 58.Kb7 Rb1+ 59.Kc8
Ra1 60.Ra8 g4 61.Kb7+ Kh7 62.Rc8 Kxh6 63.a7.
Now, Black has to watch out.
XIIIIIIIIY
9-+R+-+-+0
9zPK+-+-+-0
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
analysis diagram
White is threatening to build a bridge. 63...Rxa7+ 64.Kxa7 Kg5, with a draw. 63...Kg5 was also good, but
63...g3?? would lose to 64.Rc6+ Kg5 (64...Kh5 makes no difference) 65.Ra6 Rxa6 66.Kxa6 g2 67.a8=Q
g1=Q 68.Qg8+.
57...Rb6!
Black shows that he knows his classics – it can be useful to give trainings sometimes... the rook ‘clings’ to
White’s a-pawn, and thus Black secures the draw. There is one more snake in the grass though...
58.Kc5 Rf6!
Of course, it is essential that Black can give checks from the side.
Here, 58...Rxh6?? would have been a gross blunder, since White can protect the pawn with 59.Kb5 without
Black being able to give check, so White liberates his rook and wins, for example: 59...g4 60.Rd7 Rh5+
61.Kc4 Ra5 62.a7 g3 63.Rd8+ Kg7 64.a8=Q Rxa8 65.Rxa8, and it’s game over.
59.Kb5
Another attempt is 59.Ra8+ Kh7 60.Kb5 (60.a7 Ra6 – this is the moment for the rook to return to the a-file,
since now 61.Rh8+ was threatened – 61.Kb5 Ra1 62.Kb6 Rb1+, and White cannot achieve anything), but
also here White can’t improve his rook: 60...Rf5+ 61.Kb6 Rf6+ 62.Ka7 Rf7+, and Black keeps giving checks
from the side!
Now, Black has to pay attention. White protects the pawn with his king, so the side checks start:
59...Rf5+!
That is why the rook had to move to the f-file.
60.Kb6 Rf6+ 61.Kb7 Rf7+ 62.Ka8 Rf8+ 63.Kb7 Rf7+
And since there is no safe shelter for his king, Loek agreed to the draw. He left the playing room in a hurry
and didn’t stay to analyse the game. Loek never shunned a post-mortem, but I could imagine his
disappointment. Of course, he knew that he was a clearly better player than me, but apparently my
comment (‘Wouldn’t it be good to work on rook endings for a bit?’) was still ringing in his head. And then, if
you think you have a won game and you can’t convert it, the grapes are sour. But as we have seen, this rook
ending couldn’t be won!
Not long afterwards, I got a telephone call. To my surprise, it was Loek. Was the proposal I had made at the
time still valid? Obviously I understood what he was hinting at, and we made an appointment right away.
He wanted to stop by to analyse some rook endings – especially endings I had thoroughly examined.
There was a position in which I had put an estimated 30 hours of analytical work. It seemed to me that this
position had hardly any secrets for me. I knew that Loek was a fast learner, but I was flabbergasted by the
many ideas he managed to come up with in that position. We went on until deep into the night, and I
wanted to go to bed. Loek had missed the last train long ago already, and at the time he didn’t yet have the
car that became his favourite toy. When in the following morning I saw that a brilliant move had been
played on the board, I knew he hadn’t been sitting still!
When I was in the ChessBase studio in Hamburg in 2021 to record a couple of DVDs, I was approached to
make a video recording with endgame specialist GM Karsten Müller for his famous ‘Endgame Magic Show’. I
was surprised that he had been looking for endgames played by me in the Mega Database, and had picked
out a few nice ones, which he compared with the existing theory, and with a few examples from his own
practice. Our dialogue grew quite animated, and when I mentioned this endgame against Loek, and told him
the anecdote, Karsten was helpless with laughter... (this instalment of Müller’s show can be found under the
following link:
https://en.chessbase.com/post/karsten-mueller-s-endgame-magic-158-with-Herman-Grooten).

§ 20.12 Zugzwang
We have encountered the concept of zugzwang several times in other endgame types. The situation can get
some sadistic traits when a player cannot make a single move and then his opponent makes a waiting move.

250
Friedrich Sämisch
Aron Nimzowitsch
Copenhagen 1923 (7)
 XIIIIIIIIY
9-+-+-+k+0
9+-+q+-zpp0
9p+-vlp+-+0
9+-+p+r+-0
9-zp-zPp+-+0
9+-+lwQ-zPP0
9PzP-vL-trL+0
9+N+-tR-tRK0
xiiiiiiiiy
Black to move
Probably with satanic pleasure, Nimzowitsch played
25...h6!!
and Black resigned, since even with an extra piece he was in zugzwang.
Let’s take a look at a few rook endings in which the obligation to move is fatal for one of the sides.

251
Constructed position
XIIIIIIIIY
9-+-+-+-tr0
9+-+-+-+P0
9-+-+-+-tR0
9+-+-+k+-0
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White’s rook is standing behind the passed pawn, reducing the activity of Black’s rook to zero. Now, we
have to do the same for the black king:
1.Kh5!
And any king move by Black is met by 2.Kg6, winning.

252
Study by Savory & Lee
British Chess Magazine 1989
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+k+r0
9R+-+-+-+0
9+-+-+K+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
So little material on the board, and still things can go wrong:
1.h6
Clearly, both Black’s king and his rook have very little activity.
1...Kg8
This is more or less forced. After 1...Rh8, White wins by domination, a theme we have seen before: 2.Ra7+
Kg8 3.Kg6.
2.Ra8+
The right continuation. Here, 2.Kg6? looks immediately decisive, but in reality this lets the win slip. Black
escapes with a stalemate trick: 2...Rg7+! 3.hxg7 stalemate! (since after 3.Kf6 Rg1, the danger for Black has
disappeared).
2...Kf7
XIIIIIIIIY
9R+-+-+-+0
9+-+-+k+r0
9-+-+-+-zP0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, White works with zugzwang:
3.Kg5!
Black has no decent move left, and will lose material. 3...Ke6 is met by 4.Kg6 Rb7 5.h7, and White wins.

253
Constructed position
 XIIIIIIIIY
9-+-+-+K+0
9+-+-+-tR-0
9r+-+-+P+0
9+-+-+-+-0
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
White has only one way to win:
1.Rh7+ Kg5 2.g7
Now, White’s pieces seem to be very clumsily placed, but he manages to disentangle in some mysterious
way:
2...Ra8+
Forced.
The apparently attractive 2...Kg6? loses to 3.Kh8!, when Black cannot prevent 4.g8=Q.
3.Kf7
White’s king and rook are still not looking great.
3...Ra7+
However, Black doesn’t have much choice, since of course 4.g8=Q was threatened.
4.Ke6 Ra6+ 5.Kd5
White’s king tries to walk out of the checks, at the same time forcing the opponent to take up an
unfortunate position with his rook.
5...Rg6
White has a strong reply to this move.
5...Ra8 would fail to 6.Rh8, and giving check with 5...Ra5+ makes no sense either after 6.Kc4 Ra4+ 7.Kb5.
6.Ke5!
Mutual zugzwang! If it were White’s move, he wouldn’t be able to make any progress; Black is condemned
to play a move that loses.
6...Kg4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zPR0
9-+-+-+r+0
9+-+-mK-+-0
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
7.Rh1!
This retreat brings about Black’s downfall.
7...Kf3
The king has to leave the g-file as soon as possible. Now, 7...Kg5 didn’t work in view of 8.g8=Q! Rxg8
9.Rg1+, and wins.
8.Rf1+
But now the white rook emerges on the other side of the pawn.
8...Kg2
8...Ke3 is no solution either; White hauls in the loot with 9.Rf7 followed by Kf5.
9.Rf7 Kg3
The black king has driven off too far, and White exploits this as follows:
10.Kf5 Rg4 11.Kf6
And now that the king can support the pawn, its promotion can no longer be prevented: 1-0

This endgame in practice

With heavily reduced material and all the pawns on one wing, the endgame is often drawn. But in the
following case, a younger Magnus Carlsen received an endgame lesson.

254
Magnus Carlsen 2776
Levon Aronian 2750
Linares 2009 (8)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+r+k0
9-+-+-+-zP0
9+-+-+pmK-0
9-+-+-+-+0
9+-+-+-tR-0
xiiiiiiiiy
White to move
This position should be within drawing margins, but White has to make a decision here: 84.Kf2 or 84.Rf1.
What would you play?
84.Rf1??
A horrible and fatal mistake. Presumably, Magnus lost his concentration after such a long and exhausting
game.
The only move to hold the draw was 84.Kf2!, which should nearly always be the strategy if the opponent
has a rook behind the passed pawn. The defending side should put his king in front of the pawn to give his
rook a ‘free hand’. For example: 84...Rf6 85.Rh1 Kg4.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-trp+0
9+-+-+-+-0
9-+-+-+kzP0
9+-+-+p+-0
9-+-+-mK-+0
9+-+-+-+R0
xiiiiiiiiy
analysis diagram
Now, White has to pay attention. Passive defence loses nearly always: 86.Rg1+! (active defence usually pays
in rook endings; 86.Rh2?? Ra6 87.Rh1 Ra2+, and either White is driven to the edge, or he has to allow the f-
pawn to promote) 86...Kxh4 87.Rh1+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-trp+0
9+-+-+-+-0
9-+-+-+-mk0
9+-+-+p+-0
9-+-+-mK-+0
9+-+-+-+R0
xiiiiiiiiy
analysis diagram
87...Kg4 (or 87...Kg5 88.Rg1+ Kh6 89.Rh1+ Kg7 90.Rh3, and White picks up the f3-pawn and will hold the
pawn ending to a draw: 90...Kf7 91.Rxf3 Rxf3+ 92.Kxf3 Kf6 93.Kf4) 88.Rg1+ Kf4 89.Rg3 Ke4 90.Rg4+ Ke5
91.Rg5+ Ke6 92.Rg3.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+ktrp+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+ptR-0
9-+-+-mK-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Again, the f3-pawn is going to be lost, and Black has no chance to enter a winning pawn ending. White has
to meet 92...Ke5!? with 93.Rg5+, with a move repetition – certainly not 93.Rxf3?? in view of 93...Rxf3+
94.Kxf3 Kf5 95.Kg3 Kg5, and this is winning for Black!

84.Rf1?? 84...f2!
This wonderful move puts White in zugzwang. His rook cannot move, and any king move deteriorates the
situation. Therefore, his king is going to be driven back.
85.Kh3 Rf3+ 86.Kg2 Kg4!
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+kzP0
9+-+-+r+-0
9-+-+-zpK+0
9+-+-+R+-0
xiiiiiiiiy
Very clever play by Aronian.
87.h5
Desperation, but this is also a venomous attempt. The pawn ending after 87.Rxf2 Rxf2+ 88.Kxf2 Kxh4
89.Kg2 Kg4 would again have been winning for Aronian!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+P0
9-+-+-+k+0
9+-+-+r+-0
9-+-+-zpK+0
9+-+-+R+-0
xiiiiiiiiy
87...Rg3+!
Aronian doesn’t fall for such cheap tricks. After 87...gxh5 88.Rxf2 Rxf2+ 89.Kxf2, the pawn ending with a
pawn on the edge is a draw.
88.Kh2
Or 88.Kxf2 Rf3+ 89.Kg2 Rxf1 90.Kxf1 Kxh5, and Black wins.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+P0
9-+-+-+k+0
9+-+-+-tr-0
9-+-+-zp-mK0
9+-+-+R+-0
xiiiiiiiiy
White hasn’t had enough yet.
88...Kf3!
Of course, Black puts his cards on the f-pawn.
89.Ra1
89.hxg6 loses immediately: 89...Rxg6, and mate cannot be avoided. 89.h6 Rg5 leads to the same result.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+ktr-0
9-+-+-zp-mK0
9tR-+-+-+-0
xiiiiiiiiy
A final attempt to escape.
89...Rg2+!
Again, very accurate.
90.Kh1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+k+-0
9-+-+-zpr+0
9tR-+-+-+K0
xiiiiiiiiy
If 90.Kh3, Black forces the promotion of the pawn with 90...Rg1.
90...Rg5! 91.Ra3+ Kf4 92.Ra4+ Kg3
The king crosses over to the other side.
93.Ra3+ Kh4
We can see the depth of Aronian’s vision if we take a look at the possible continuation: 94.Ra1 (94.Ra4+ is
met by 94...Kh3 95.Ra3+ Rg3 96.Rxg3+ Kxg3, and it’s not stalemate since White still has that cursed pawn
on h5, which Black kept alive on purpose!) 94...Rg1+!? (a professional liquidation; objectively, 94...Kh3 is
the right move that wins in the fastest way: 95.hxg6 Rg2 (threatening mate) 96.Ra3+ Rg3 97.Ra1 Re3, and
there is no remedy against 98...Re1+ and mate) 95.Rxg1 fxg1=Q+ 96.Kxg1 Kxh5, and here again, the theory
of the key squares tells us that Black is winning.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-trP0
9-+-+-+-mk0
9tR-+-+-+-0
9-+-+-zp-+0
9+-+-+-+K0
xiiiiiiiiy
Now, the king has manoeuvred to the h-file, so Black is able to interpose his rook to end the checks. Carlsen
finally resigned here.
§ 20.13 The rook’s pawn
We have encountered several endgames with a rook’s pawn in the above examples. Some knowledge is
essential, to recognize when one of the players can simplify into a winning position.
First we look at a frequently occurring type of endgame in which the stronger side’s king is tied to the edge.
Any possible escape depends on the position of the enemy king.
The following position is a well-known case.

255
Constructed position
XIIIIIIIIY
9-+-+-+-mK0
9+-mk-+-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+r+0
9+-+R+-+-0
xiiiiiiiiy
White to move
With White’s king locked up in the corner by the black rook, White has only one method to try to win this
ending: the cut-off along the g-file has to be broken. This means that White has to try to get his rook on g8. I
was taught in my time that the defending black king has to be cut off by at least four files – counted from
the g-file – to prevent it from being in the way of its opposite number. The win is achieved as follows:
1.Rf1 Kd7 2.Rf8
XIIIIIIIIY
9-+-+-tR-mK0
9+-+k+-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+r+0
9+-+-+-+-0
xiiiiiiiiy
Here, there are two variations that the stronger side has to know:
2...Ke6
Looking at the winning method in the other variation, this is the most difficult defence that has to be
navigated. After 2...Ke7 3.Rg8 Rh2, White’s king can finally emerge from the corner: 4.Kg7 (here, 4.Rg7+? is
nothing for White on account of 4...Kf8=).
 XIIIIIIIIY
9-+-+-+R+0
9+-+-mk-mKP0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-tr0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Here, Black has nothing better than giving checks. But now the white king walks into the open field, since
his rook already controls the promotion square: 4...Rg2+ 5.Kh6 Rh2+ 6.Kg6 (he still has to defend the
pawn here!) 6...Rg2+ 7.Kf5, and now the king is let loose: 7...Rf2+ 8.Ke4 Re2+ 9.Kf3 – the pawn promotes
and the game is decided.
3.Rg8 Rh2 4.Kg7 Rg2+
XIIIIIIIIY
9-+-+-+R+0
9+-+-+-mKP0
9-+-+k+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+r+0
9+-+-+-+-0
xiiiiiiiiy
This is almost the same position as the above one, only on this square the black king is blocking the escape
route to f5, so White has to go about it differently.
5.Kf8
5.Kh6 makes no sense now due to 5...Rh2+ 6.Kg6 Rg2+, and it is clear that the white king can’t get away
without losing the pawn.
5...Rf2+ 6.Ke8 Ra2
The final convulsion. White cannot promote to queen now on pain of checkmate.
7.Rg6+ Kf5
XIIIIIIIIY
9-+-+K+-+0
9+-+-+-+P0
9-+-+-+R+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
8.Rf6+!
The best way to win, though White has to know how to win the queen vs rook ending.
8...Kg5
The above-mentioned ending would be easier for White here, since the rook is not in the vicinity of its own
king, and can be collected with the correct series of checks, for example: 8...Kxf6 9.h8=Q+ Kf5 10.Qf8+
Ke4 11.Qb4+ Ke5 12.Qc5+ Kf6 13.Qe7+, and on the next move the rook will be lost to a double attack by
the queen on f7 or e6.
9.Rf8!
The fastest way to win.
9...Ra8+ 10.Kf7 Ra7+ 11.Kg8 1-0
Although rook’s pawns come with a different type of strategy, we will limit ourselves to one other
instructive case:

256
Constructed position
XIIIIIIIIY
9r+K+-+-+0
9+-+R+-+P0
9-+-+-+-+0
9+-mk-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
The first thing we notice is that White is in check, so his reply isn’t very hard to find. But before we continue,
we may notice that White’s rook is excellently placed. Black’s king is cut off not only vertically, but also
horizontally, with regard to the pawn. This aspect brings the win closer for White.
1.Kc7 Rh8
Giving a new check with 1...Ra7+ doesn’t make any sense at all. After 2.Kd8 Ra8+ 3.Ke7, Black is worse off
than in the main line.
2.Rg7!
Curiously, this is the only way for White to decide the battle in his favour. It does voluntarily give up the
vertical cut.
After, for example, 2.Kb7?, Black holds the draw, e.g. 2...Re8! 3.Rg7 Kd6!, shouldering the white king away
from the battle scene, which in this case is the side where the pawn is standing. 4.Rg8 is met by 4...Re7+
followed by ...Rxh7.
2...Kd5 3.Kd7 Ke5 4.Ke7 Kf5 5.Kf7
XIIIIIIIIY
9-+-+-+-tr0
9+-+-+KtRP0
9-+-+-+-+0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This is in fact a reciprocal zugzwang situation.
5...Ra8
Again, the only move to at least keep the struggle going. A king move like 5...Ke5 is met with 6.Rg5+ Kf4
7.Rh5 Kg4 8.Rh1, when 9.Kg7 can’t be prevented.

If it were White’s move, he wouldn’t have anything much better than 6.Ke7, but then it is a draw after
6...Ra8 7.Kf7 (or 7.Kd6 Ra6+ 8.Kd5, and Black’s rook has managed to get behind the pawn; moreover,
White cannot prevent ...Kf6 followed by ...Kg6) 7...Ra7+ 8.Kg8 Ra8+, and White can’t achieve anything.
Now, what is the best way to continue for White?

6.Rg1!
Since Black was ready to give a kind of ‘perpetual check’ with 6...Ra7+, White’s rook had to move. So what is
more obvious than taking it all the way down? The threat is both 7.Kg7 and 7.Rh1, so Black has little else
but giving checks.
Incidentally, not 6.Rg8?, which loses the pawn to 6...Ra7+.

6.Rg1! Ra7+ 7.Kg8 Ra8+ 8.Kg7 Ra7+ 9.Kh6 Ra6+

XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+P0
9r+-+-+-mK0
9+-+-+k+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-tR-0
xiiiiiiiiy
10.Kh5!
This is the way to escape the side checks; 11.h8=Q is threatened.
10...Ra2
Black’s only chance. For the moment, the promotion with 11.h8=Q is prevented, due to 11...Rh2#.
10...Ra8 fails to 11.Rg8 Ra1 12.Rf8+! (certainly not 12.h8=Q?? in view of 12...Rh1#) 12...Ke6 13.h8=Q, and
White wins.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+k+K0
9-+-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+-tR-0
xiiiiiiiiy
11.Rg4!
The fastest road to victory, although by this time there were several ways (for example, with 11.Rf1+). This
rook move enables White to build a bridge on h4 (11...Rh2+ is met by 12.Rh4), and at the same time keeps
the option of working with Rg8.
11...Ra8 12.Rg8
The advantage of this move is, as we have seen before, that the rook covers the promotion square. It will
now become clear why this is convenient.
12...Ra1
Again, Black is threatening mate, but now White prevents this mate in the same way as in the above
variation:
13.Rf8+ Ke6 14.h8=Q
And White wins.

§ 20.14 Tactics – exercises

When I first studied rook endings, it struck me how difficult it was sometimes to master various
stereotypical manoeuvres like the ones that passed the review in the previous section. For me, this was also
a motivation to start classifying them by theme, to be able to familiarize club players with the techniques as
well. Rook endings contain a lot of tactics, a feature that makes our game extra interesting. Therefore, I will
present a collection of different tactical themes in the next few pages.
In your chess life, there has to be someone who lights the fire, who awakens your interest in this area.
For me, the influence of one specific teammate was very important. I am talking about IM Peter Scheeren,
who was the strongest player at our club for a long time. He studied electrical engineering, and got his Ph.D.
in this discipline. During our many blitz nights held at the home of one of our fellow club members, Jules
Welling, we always set ourselves the task of analysing one recently played game at the beginning of the
evening. Since these were games by the world’s top players, endgames would also come on the board –
including, of course, rook endings. All the attentive listeners at the time (a generation of players, five of
whom went on to become International Masters) noticed how excellent Scheeren’s knowledge was in this
area. He was able to teach us the tricks of the trade along the way, and this is one of the reasons why I
‘automatically’ learned a lot from these instructive sessions. I don’t remember much of the blitz game after
that, which used to go on until deep into the night, but this may be due to the amount of alcoholic beverages
consumed...
In the selection below, I have opted for the themes that are the most attractive and also the most frequent
in rook endings. In my opinion, club players can profit from getting familiarized with the, sometimes
hidden, motifs in them. A good awareness of such themes may be a key to success, and being able to bring
such motifs to the surface in a game tends to give a player a lot of satisfaction.
Have fun solving these exercises and working out the variations!

1 – Checkmate
XIIIIIIIIY
9-+R+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-zp-+-0
9-+-tr-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9mk-+-+-+-0
xiiiiiiiiy
White to play and win
2 – Checkmate
XIIIIIIIIY
9-+-+-+-+0
9tr-+-+P+-0
9p+-+-+-+0
9zP-+-+K+-0
9-+-+-+-mk0
9+-+-+-+p0
9-+-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

3 – Checkmate
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+P+rmk0
9+-+-+-+-0
9-tR-+K+-+0
9zp-+-+-zp-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

4 – Checkmate
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
9-+-mK-+-+0
9tr-+-+-+-0
9P+-tR-+-+0
9mk-+-+-+-0
xiiiiiiiiy
White to play and win
5 – Checkmate
XIIIIIIIIY
9-+-+-+-+0
9+-+-+pzp-0
9-+-+P+-+0
9+-+-+-+-0
9-+-+K+-mk0
9+-+-+-+-0
9-+-+-+R+0
9+-+r+-+-0
xiiiiiiiiy
White to play and win

6 – Checkmate
XIIIIIIIIY
9-+-+-+-+0
9+KtR-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9Pmk-+-+-+0
9+-+-+-+-0
9rzP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

7 – Checkmate
XIIIIIIIIY
9-+-+-+-mk0
9+-+pmK-+-0
9-+-zp-tRP+0
9+-+-+-trp0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

8 – Checkmate
 XIIIIIIIIY
9-+-+-+k+0
9+-+-mK-+-0
9-+-+-+-+0
9tr-+-+p+-0
9-+-+P+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
White to play and win

9 – Stalemate
XIIIIIIIIY
9r+-+-+-mk0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9p+-+-+-zp0
9tR-+-+-+-0
xiiiiiiiiy
White to play and draw

10 – Stalemate
XIIIIIIIIY
9-+-+-+-+0
9+-+k+-+-0
9-+-zp-+-+0
9+-+-+-+R0
9-+-mK-+-sn0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and draw

11 – Stalemate
 XIIIIIIIIY
9-+-+-+-tr0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+Rzp0
9+k+-+-+-0
9-+p+-+-+0
9+-mK-+-+-0
xiiiiiiiiy
White to play and draw

12 – Stalemate
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+P+-+-+0
9+-+-+-+-0
9pmk-+-+-+0
9+-+rzP-+-0
9-mK-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
White to play and draw

13 – Passed pawn(s)
XIIIIIIIIY
9-+-+-+-+0
9mK-mk-+-zP-0
9-+r+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-tR0
9+-+-+-+-0
xiiiiiiiiy
Black to play – White wins

14 – Passed pawn(s)
 XIIIIIIIIY
9-+-+-mK-mk0
9+-+-+-+p0
9-+-+-+-tR0
9+-+-+-zPP0
9-+-tr-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

15 – Passed pawn(s)
XIIIIIIIIY
9-+r+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+-+k0
9-+p+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
White to play and win

16 – Passed pawn(s)
XIIIIIIIIY
9-+r+-+-+0
9+-zpp+-+-0
9-+PzP-+-+0
9+-+-+-+-0
9-+-+R+-+0
9+-+-+-+-0
9-mk-+-+-+0
9+-+K+-+-0
xiiiiiiiiy
White to play and win

17 – Liquidation
 XIIIIIIIIY
9-+-tR-+-+0
9+-+P+K+-0
9-+-mkr+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

18 – Liquidation
XIIIIIIIIY
9-+-+-tr-mk0
9+-+-+ptR-0
9-+-+-mKP+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

19 – Liquidation
XIIIIIIIIY
9r+-+-+-+0
9+-+K+-+-0
9P+-zp-+-+0
9+-+-+-+-0
9-+-+-+R+0
9mkP+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

20 – Liquidation
 XIIIIIIIIY
9-+k+K+-+0
9zp-+r+-+-0
9P+-+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
White to play and win

21 – Interception/bridge
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+rmk-0
9R+-+-+-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

22 – Interception/bridge
XIIIIIIIIY
9-+-+k+-+0
9+-+-+-mK-0
9-zP-+-+-+0
9+-+-+-+-0
9-zp-+-+-+0
9+-+-+-+-0
9r+-+-+-+0
9+-+-+-tR-0
xiiiiiiiiy
White to play and win

23 – Interception/bridge
 XIIIIIIIIY
9-mK-+-+-+0
9zp-+-+-+-0
9-+-+-+P+0
9+-+-+-+-0
9-+-+k+-+0
9+-+-+-+-0
9r+-+-+P+0
9+-+-+R+-0
xiiiiiiiiy
White to play and win

24 – Interception/bridge
XIIIIIIIIY
9-+-+-+-+0
9+-zP-tR-+-0
9-+-+-+p+0
9+-+K+-mkp0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+P+0
9+-tr-+-+-0
xiiiiiiiiy
White to play and win

25 – Exploiting poor position of king/rook

XIIIIIIIIY
9R+-+-+-+0
9+-+-zp-zp-0
9P+-tr-+-+0
9mk-mK-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

26 – Exploiting poor position of king/rook

 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-zp-0
9-+-mkp+r+0
9+-+-tR-zP-0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
White to play and win

27 – Exploiting poor position of king/rook

XIIIIIIIIY
9-+-+-+-+0
9+-+-zp-+-0
9r+-+-+p+0
9mk-+KzP-+-0
9-+-+-+-+0
9+R+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

28 – Exploiting poor position of king/rook

XIIIIIIIIY
9-+-+K+-tR0
9+-zp-+-+-0
9-+-mk-+r+0
9+-+P+-+-0
9-zP-+-+p+0
9+-+-+-+-0
9-+P+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

29 – Exploiting poor position of king/rook

 XIIIIIIIIY
9-+-+-+-mk0
9+p+-+p+-0
9P+-+rzP-zp0
9+-+K+-+-0
9-zp-+-+-+0
9+-+-+-+-0
9-zP-+-+-+0
9+-+-+R+-0
xiiiiiiiiy
White to play and win

30 – Exploiting poor position of king/rook

XIIIIIIIIY
9-+-+-+-tR0
9+-+p+-+-0
9-+-mK-+-+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+r+-+k+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

31 – Exploiting poor position of king/rook

XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mkP0
9-+K+-+-+0
9zp-+-trp+-0
9-+-+p+-+0
9+-+-+-+-0
9-+-+-+-+0
9+R+-+-+-0
xiiiiiiiiy
White to play and win

32 – Exploiting poor position of king/rook

 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+p+-0
9-+-+-+-+0
9+-+R+-+-0
9P+-+-+-+0
9+k+K+N+r0
xiiiiiiiiy
White to play and win

33 – Mix
XIIIIIIIIY
9R+-+-+-+0
9zP-+-+-+-0
9-+-mK-+-+0
9+-+-+-+-0
9-+-mk-+-+0
9+-+-+-+-0
9-+-+-+-+0
9tr-+-+-+-0
xiiiiiiiiy
White to play and win

34 – Mix
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+Pmk-+-+-0
9-tr-+-+-+0
9+-mKR+-+-0
xiiiiiiiiy
White to play and win

35 – Mix
 XIIIIIIIIY
9-+-+-mk-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-tr-0
9-+-+-+ptR0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

36 – Mix
XIIIIIIIIY
9-+-+-+-+0
9+-tr-+kzp-0
9-+-+-+-+0
9+-+-+-zPP0
9R+-+-+K+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

37 – Mix
XIIIIIIIIY
9-+-+-+k+0
9zp-zP-+-+-0
9P+r+-+-+0
9+-+-mK-+-0
9-+p+-+-+0
9+-+-+-+-0
9-+R+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

38 – Mix
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+R+-+-+-0
9-+-+-+-+0
9mKptr-+-+-0
9-+k+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and draw

39 – Mix
XIIIIIIIIY
9-+-+r+-mk0
9+-+-+-+-0
9R+-+-+-+0
9+-+-zP-+-0
9-+-+-+-mK0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

40 – Mix
XIIIIIIIIY
9-+-+-tR-+0
9+-+-+P+-0
9-zp-+-+-+0
9+k+-+r+-0
9-+-+-+-+0
9+-+-+-+-0
9PmK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

41 – Mix
 XIIIIIIIIY
9-+R+-+-+0
9+-+-+-+-0
9-+P+-+r+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
White to play and win

42 – Mix
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-tR-+-+pzP0
9+-+K+-+-0
9k+-+-zp-+0
9+-+-+r+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

43 – Mix
XIIIIIIIIY
9-+-mk-+-tr0
9+-+-tRp+-0
9-+-zPp+-zp0
9+-+P+-+-0
9-+-+-+-+0
9+K+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

44 – Mix
 XIIIIIIIIY
9-+-+-tr-+0
9tR-+Pzp-mk-0
9-+-+-+-+0
9+-+P+p+p0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

§ 20.15 Tactics – solutions

1
Study Hans Fahrni & Hans Keemink
Het Eindspel 1928

This position should ring a bell, since Black’s king in the corner may fall victim to certain mate motifs.
1.Kc2!
Threatening to give checkmate, starting with 2.Ra8+.
1...Ra4 2.Kb3!
With this double attack, White eliminates all counterplay. He is threatening both 3.Rc1# and 3.Kxa4.
We know this motif from the famous Saavedra endgame.

2
Study by Gyula Neukomm
Magyar Sakk Lap 1906

With Black’s king caught on the edge, as so often, a mating motif is possible:
1.f8=R!
White has to promote to a rook; queening would allow Black to escape with stalemate: 1.f8=Q? Rf7+
2.Qxf7.

1.f8=R! Rh7 2.Kg6

The double threat of 3.Rf4# and 3.Kxh7 ends the game.
2...Rc7 3.Rf4#

3
Study by Vladislav Bunka
Sachove Umenie 1970

White combines the power of his passed pawn with mating patterns, thanks to the vulnerable position
of Black’s king on the edge of the board:
1.e7 Re6+ 2.Kf5 Rxe7
If 2...g2, White plays the strong 3.Rb6! (not 3.Kxe6? g1=Q 4.e8=Q Qe1+, and Black holds the draw)
3...Rxb6 (if 3...g1=Q 4.Rxe6+ Kg7 5.e8=Q Qf2+, White still has to escape the checks, which he manages with
6.Ke4 Qe2+ 7.Kd4 Qd2+ 8.Kc5 Qc3+ 9.Kb6, and soon White will be able to interpose a piece) 4.e8=Q
g1=Q 5.Qh8#.
3.Kf6 1-0
 The double threat of 4.Rh4# and 5.Kxe7 decides.

4
Study by Ladislav Prokes
Sach 1941

White is two pawns up, but they are both about to fall. Therefore, White has to play very inventively to
get more than a draw here:
1.Kc4
Not 1.Rd3? Rxa6 (a mistake would be 1...Rxa2? 2.Rd1+ Kb2 3.Rd2+ Kb3 4.Rxa2 Kxa2 5.a7) 2.a3 Kb2,
and the final pawn falls.

1.Kc4 Rxa6 2.a4!

A wonderful concept!
2...Rxa4+
If Black doesn’t take the pawn, he will lose technically, for example: 2...Ra8 3.Kb5 Rb8+ 4.Kc6 Ra8
5.Rd4 Kb2 6.Kb7 Ra5 7.Kb6 Ra8 8.a5, and White wins.
3.Kb3 Ra8 4.Rd1#

5
Study by Ladislav Prokes
Sach 1940

White has a dangerous pawn, and Black’s king on the edge is another theme. But how does White
combine these themes to win?
1.e7!
1.exf7 Rf1 is a draw: 2.Rxg7 Kh5 3.Ke5 Kh6, and the pawn falls.

1.e7! Re1+
Forced.
2.Kf4!
Suddenly, White threatens checkmate with 3.Rh2#.
2...Kh3
Black neutralizes the mate threat, but cannot prevent White from winning by building a bridge. After
2...Kh5 too, White can suddenly build a bridge with 3.Rg5+ Kh6 4.Re5.
Also essential is 2...g5+, which is best answered by 3.Kf3, keeping the mating pattern with Rh2# intact:
3...g4+ (3...Kh3 4.Re2) 4.Rxg4+ (4.Kf4? g3) 4...Kh5 5.Re4, and again White wins thanks to a bridge.
If 2...Rxe7, 3.Rh2#.
3.Rg3+ Kh2 4.Re3 1-0
The bridge has been established.

6
Study by Ladislav Prokes
Zemedeske Noviny 1964

Two extra pawns, but they are both hanging, and a draw seems to be the most probable result.
1.Kc6 Rxa4
Clearly, 1...Rxb2 fails to 2.Rb7+ Kc3 3.Rxb2 Kxb2 4.a5.
2.Rb7+ Ka5
With a king on the edge, mate patterns are looming. The king couldn’t go to c4 in view of 3.b3+.
3.b4+
‘Alas’, 3.b3! wins too... 3...Ra3 4.Rb8!, winning the rook.
3...Rxb4 4.Ra7#

7
Study by Josef Hasek
Ceskoslovensky Sach 1953

King in the corner – this is mostly a signal that we can play for mate.
1.Kf8
Threatening to use his foremost g-pawn.
1...h4
After 1...Rxg2 2.Rf5, checkmate can’t be prevented, as also in the variation 1...d5 since after 2.g7+ Rxg7
3.Rh6+ Rh7 4.Rg6, there is no decent way to parry the threat of 5.Rg8#.
2.Rf4 1-0
There is nothing to be done against the double threat of 3.Rh4+ and mate, or 3.g7+ etc.

8
Study by Alexey Selezniev
Tidskrift for Schack 1922

Both pawns threaten to disappear from the board, but White has a beautiful finesse that allows him to
keep his pawn:
1.Rg1+
1.exf5? is a draw due to the side checks: 1...Ra7+ 2.Kf6 Ra6+ 3.Kg5. This is the only way to walk out of
the checks, but now follows 3...Kf7, reaching the Philidor position.

1.Rg1+ Kh7 2.e5!

This is the move that brings White the win.
2...Rxe5+
2...Ra7+ 3.Kf6 Ra6+ 4.e6 leads to a losing endgame for Black.
3.Kf7
Threatening 4.Rh1#. His own pawn on f5 causes Black’s downfall.
3...Kh6 4.Kf6
With the double threat of 5.Rh1# and 5.Kxe5: 1-0

9
Study by Helmut Pruscha
Schach-Echo 1975

With two pawns less, and with Black having the rook behind one of his passed pawns, the situation
looks hopeless for White. For starters, Black threatens 1...h1=Q+.
1.Kg2 Kg7
1...h1=Q+ makes no sense due to 2.Kxh1 Kg7 3.Kg2!, and White’s king is just in time to cut off Black’s
king: 3...Kf6 4.Kf3 Ke5 5.Ke3 Kd5 6.Kd3 Kc5 7.Kc3, with a draw.
2.Kh1!
Certainly not 2.Kxh2? Kf6!, and the black king ends up on b2, winning.
2...Kf6
And now, White uncorks his idea:
3.Rxa2! Rxa2
Stalemate!
10
Study by Hans Fahrni
Das Endspiel im Schach 1917

One knight and a pawn less – time to resign?

1.Rxh4!
No! Sometimes, there are tricks you just have to remember.
1...Ra4+ 2.Kd5 Rxh4
Stalemate!

11
Study by Henri Weenink
Algemeen Handelsblad 1918

1.Rd4!
The only way to save the game.
1...h3
After 1...Kc3 2.Rxh4, it’s stalemate, or otherwise completely drawn. After 1...Rh7 2.Rd7, we see the
theme from the main line.
2.Rd8!
White starts a stunning chase of Black’s rook.
2...Rh7
2...Rxd8 is stalemate!
3.Rd7 Rh6
3...Rxd7 would lead to the same result.
4.Rd6
The black rook can’t get away.
4...Rxd6
Stalemate!

12
Study by Harold van der Heijden
& Yochanan Afek
Josten JT Internet 2002

This famous pair of endgame study composers concocted this study together.
White to play and win – this is certainly not trivial!
1.Rxa4+!
A surprising rook sacrifice. It looks logical to put the rook behind the passed pawn with 1.Rc1?, but
after 1...Rd8 2.c7 (another attempt is 2.e4, but after 2...Rc8 3.e5 Kb5 4.e6, Black can eliminate the c-pawn
also here with 4...Rxc6 5.Re1 Rc8 6.e7 Re8, after which the king walks towards the e-pawn and captures it)
2...Rc8 3.e4 Kb5 4.e5 Kb6 5.e6, the c7-pawn will simply be lost, after which it is a draw.

The direct 1.c7? doesn’t bring White anything either: 1...Rb3+ 2.Ka2 (if the king goes to the c-file
(2.Kc2), Black immediately collects the pawn with 2...Rc3+ followed by 3...Rxc7) 2...Ra3+.

1.Rxa4+! Kxa4 2.c7

 XIIIIIIIIY
9-+-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9k+-+-+-+0
9+-+rzP-+-0
9-mK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
It seems that the black rook can’t stop the promotion of the c-pawn. The d8-square is off-limits, and
the rook can’t get behind the pawn either, since the white king controls the c3-square. So – time to resign?
2...Rb3+
An important intermediate check. 2...Rd2+ is strongly met by 3.Kc1! (certainly not 3.Kc3? in view of
3...Rd5! (the temptation here is 3...Rd1?, when White wins with 4.Kb2! – 4.Kc2? Rd5!, threatening with
stalemate – 4...Rd2+ 5.Kc1!)
XIIIIIIIIY
9-+-+-+-+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+r+-+-0
9k+-+-+-+0
9+-mK-zP-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
4.c8=R! (this underpromotion to a rook yields White a winning rook ending; not 4.c8=Q? Rc5+ 5.Qxc5,
with stalemate) 4...Kb5! 5.e4, but now Black can liquidate into a pawn ending that is a dead draw with
5...Rc5+ 6.Rxc5+ Kxc5) 3...Rd5 4.c8=Q!. Now, White can promote to a queen, since there is no stalemate
anymore; 4.c8=R? is nonsense: 4...Rd3 5.e4 Re3 6.Rc4+ Kb5, and White can no longer win this.
3.Kc2 Rb5!
Now, Black has his stalemate trick, but White doesn’t fall for it:
4.c8=R!
Again, this underpromotion to a rook, after which the remaining endgame needs to be evaluated.
4.c8=Q? Rc5+ 5.Qxc5 stalemate!
4...Rd5
A last try. Also other moves, like 4...Kb4, don’t save Black: 5.Kd3 (5.e4? Rc5+ 6.Rxc5 Kxc5) 5...Rb7 6.e4
Rd7+ 7.Ke3, and since the black king is cut off vertically and the checking distance is much too short, Black
cannot prevent the advance of the pawn.
 XIIIIIIIIY
9-+R+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+r+-+-0
9k+-+-+-+0
9+-+-zP-+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
5.Rb8!
Cutting off the black king by one more file, after which the position can’t be saved. Incidentally, this
final move is in quite nice accordance with the ‘rule of five’.
Tempting was 5.Kc3?, but then Black plays 5...Kb5!, with the threat 6...Rc5+, after which the pawn
ending is a draw. This means that White has to abandon the cut-off and can no longer win.

13
Study by Nikolay Minev
Source unknown, 1980

White has a good passed pawn, but his king is on the edge and will never be able to support the pawn.
So, the latter seems to be doomed:
1...Rg6
This looks sufficient for a draw. However, White still has some tactics that bring him the win.
It’s easy to see that after 1...Rc1?, White can simply queen (2.g8=Q), since after 2...Ra1+, he can defend
with 3.Ra2.

1...Rc3 looks more clever for this reason, but White still wins with 2.Rc2! Rxc2 3.g8=Q, and the queen
covers the a2-square!

1...Rg6 2.Rh7!
Now, White threatens to give a discovered check with 3.g8=Q+; not 2.Rb2 Rg1 3.Ka6 Kc6=.
2...Kd6
After 2...Kd8 3.Rh8+ Ke7, White can also promote to queen.
3.Rh6
Thus, White forces the promotion of the pawn. The queen vs rook ending is winning, though the
technique is not so simple, as we saw in the analysis at the end of section 20.6.

14
Axel Akerblom
Mora Tidning 1923

A king in the corner is a bad omen.

1.g6!
Threatening mate in one.
1...Rd7
Thus, Black keeps everything under control, parrying both 2.Rxh7+ and 2.g7+. So what can White do
now?
After 1...Rd8+ 2.Ke7 Rd5 3.Rxh7+ Kg8 4.Kf6, the resulting endgame is winning for White.
2.Rxh7+!
 A magnificent idea!
2...Rxh7
XIIIIIIIIY
9-+-+-mK-mk0
9+-+-+-+r0
9-+-+-+P+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
3.h6!!
The point! One of the white pawns will get through – two connected passed pawns on the sixth rank
are mostly stronger than a rook.
3...Rxh6 4.g7+ Kh7 5.g8=Q#

15
G.van Nieuwkuijk
1930

Both players have their rook behind the passed pawn, so the position seems to be dynamically
balanced. However, as it turns out this is not the case:
1.Rh1+!
A nasty in-between check! 1.Rc1 Ra8 is a draw.

1.Rh1+! Kg2 2.Rc1

Only now does White’s rook move in front of the enemy pawn. The difference becomes clear after
2...Ra8
2...Kh3 doesn’t help either due to 3.Rxc2
3.Rxc2+.
The pawn is taken with check, and after that White saves his own pawn and wins. This analysis was
based on a game Van Nieuwkuijk- Pijpers, Alkmaar 1930.

16
composer & date unknown

This is what is called a ‘nerve block’ in the Netherlands, with four pawns that can all capture each
other. Can White do business with this?
1.Rb4+!
Only so! Not 1.cxd7? Rd8 2.dxc7 Rxd7+, and both of White’s pawns perish.

1.dxc7 is no good either: 1...Rxc7, and it’s a draw.

1.Rb4+! Kc3 2.Rb8

This rook sacrifice works wonders!
2...Rxb8 3.dxc7
Thanks to the rook sacrifice, Black’s rook cannot immediately take a pawn, which gives White time to
get his other pawn to the seventh rank as well.
3...Rc8
 3...Rb1+ 4.Ke2 Rb2+ 5.Ke3 doesn’t help either, of course.
4.cxd7 1-0

17
Constructed position

White’s pawn threatens to be lost after 1...Re7+ and 2...Rxd7.

1.Re8!
With this strong move, White puts an end to all the drawing chances for Black. Apart from vacating the
d8-square to threaten promotion, White also threatens to take the black rook.
1...Kxd7 2.Rxe6 1-0

18
Study by Alexey Selezniev
Endspiele 1940

It’s almost a draw... or is there still something in it for White?

1.Rh7+ Kg8
And now?
2.g7!
The cat’s out of the bag:
2...Kxh7 3.gxf8=R 1-0
Black is mated on the next move. Certainly not 3.gxf8=Q? stalemate!

19
Study by Ladislav Prokes
Suomen Shakki 1951

White has an extra pawn, but both his pawns are hanging, so the draw seems to be within reach for
Black.
1.Kc6!
The king intervenes.
1.Ra4+? makes no sense due to 1...Kxb3 2.Ra5 (2.Ra1 d5!) 2...d5 3.Kc7 d4, and Black draws easily in
the rook vs pawn ending that will soon arise.

Also, 1.b4? fails to 1...Rxa6 2.b5 (2.Kc7 Ra4) 2...Rb6, with a draw.

1.Kc6! Rxa6+
Now, 1...Kxb3 doesn’t work on account of 2.Kb7 Rd8 3.Rd4!? (3.a7 also suffices) 3...Kc3, and here
White has the important zwischenzug 4.Kc7! Ra8 5.Rxd6, and the white pawn decides.
2.Kb5
Exposing the awkward positions of Black’s king and rook. We see here that the d6-pawn is also badly
in the way of Black’s rook.
2...Ra8 3.Ra4+ Rxa4 4.bxa4
The remaining pawn ending is a simple win, for example:
4...d5 5.a5 d4
5...Kb3 6.Kc5.
6.Kc4 1-0

20
Study by Albert van Tets
Ndaba 1983
 White is a pawn up, but in principle that isn’t sufficient to win. Something special needs to be done.
Tempting is 1.b6, but Black can defend with, for instance, 1...Rd2!? 2.bxa7 (2.b7+ Kb8 3.Rc1 Re2+
4.Kf7 Re7+, and Black has a ‘crazy rook’) 2...Rd8+ 3.Ke7 Rd7+ 4.Ke6 Rxa7, with a draw.

1.Rc1+
This forces
1...Rc7
but now that Black’s rook is pinned, White has a gorgeous move:
2.b6!! Rxc1
If 2...axb6, 3.a7! wins.
3.bxa7 1-0
The foremost pawn will promote; there is no way for Black to stop it.

21
Study by Josef Kling
& Bernard Horwitz
Chess Studies 1851

So little material, and yet there is a win for White:

1.h7
This is forced, otherwise the pawn will be lost.
1...Rf8
The only move, but now White has:
2.Rg4+! Kh6 3.Rg8
Using the g8-square to support the promotion of the pawn. White wins.

22
Study by Jindrich Fritz
Prace 1953

If a pawn has almost reached the back rank, there is sometimes a trick in the position.
1.b7 Ra7
This pinning move is the only way for Black to prevent promotion. But this gives White a beautiful idea:
2.Re1+ Kd8 3.Re7!
A pretty move with multiple intentions. Firstly, the pawn is protected. Secondly, it is also ‘un-pinned’,
and at the same time the black king is deprived of the c7-square. Thus, promotion can only be prevented by
either giving up the rook or allowing the queen vs rook ending:
3...Kxe7 4.b8=Q
And White wins.

23
Study by Alois Wotawa
Deutsche Schachzeitung 1936

‘Doubled pawns are worth nothing’, is often claimed. But this is a quite different case! What can White
do to force promotion?
1.Rf4+!
A wonderful rook sacrifice.
1.g7? Rxg2 2.Rf7 a5 yields nothing.

1.Rg1 Ra6 2.g7 also leads to a draw after 2...Rg6.

 1.g3 looks clever too, but this is also a draw after 1...Ra6 2.Rf4+ Ke5 3.Rg4 Re6 4.g7 Re8+ 5.Kxa7 Rg8
since after 6.Kb6 Kf5 7.Rf4+ Kg6,
XIIIIIIIIY
9-+-+-+r+0
9+-+-+-zP-0
9-mK-+-+k+0
9+-+-+-+-0
9-+-+-tR-+0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
the ending can’t be won by White.

1.Rf4+! Kxf4
Declining the rook sacrifice with 1...Kd5 loses to 2.Rg4 Re2 3.g7 Re8+ 4.Kc7! Rg8 5.Kd7, and the white
king rushes to f7 to win the black rook and then win with his second pawn: 5...a5 (5...Ke5 6.Ke7 Kf5 7.Kf7
Rb8 8.g8=Q goes wrong for Black as well) 6.Ke7 a4 7.Kf7.
After 1...Ke5 2.Rg4, the black rook can’t get in front of the pawn; 2...Rd2 is met by 3.Kc7!, and White
wins.
2.g3+
The point of the previous sacrifice. By giving up the pawn, White creates an interception of the g-file,
enabling the foremost g-pawn to walk on unhindered.
2...Kxg3 3.g7 1-0

24
Constructed position

White’s pawn has almost reached the other side, while Black still has to put in some effort to create a
passed pawn of his own. With his king on the extremely handy d5-square, White has a brilliant way to win.
With an intermediate check, he manages to place his rook between his pawn and the enemy rook, building
a bridge:
1.Re5+! Kf4
The most logical square, but Black can’t prevent White from putting his rook on an ideal square.
Moving the king backwards doesn’t bring any solace either: 1...Kf6 2.Re6+! Kf5 3.Rc6 and again,
Black’s rook cannot keep the pawn under cover: 3...Rd1+ 4.Kc5. Since Black’s rook can’t move to the back
rank, he has to keep giving checks, but White has a clever way to walk out of them: 4...Rc1+ 5.Kd6 Rd1+
6.Ke7! (White takes care that Black can never give his rook for the c-pawn) 6...Re1+ 7.Kf7, and the checks
are exhausted.
After 1...Kh6, Black has no counterplay anymore, since his king is too passive. White wins with 2.Kd6
(threatening to build a bridge on c5!) 2...Rxc7 3.Kxc7 g5 4.Kd6 h4 5.Re4 Kh5 6.Ke6 g4 7.Kf5 h3 8.g3!, and
White gives mate.
2.Re4+! Kg3 3.Rc4
There it is – the bridge!
3...Rd1+ 4.Kc6 1-0
Pawn promotion can’t be prevented.
25
Study by Jindrich Fritz
Lidova Demokracie 1961

Despite the heavily-reduced material, White can win. The placement of the black king on the edge and
White’s dangerous a-pawn give White a few arguments:
1.a7
Now, White threatens to move away the rook and promote the pawn.
1...Ra6
1...Rd7 is met by 2.Kc6 Rd6+ 3.Kc7 Ra6 4.Kb7 Rd6 5.Rc8, and the pawn decides.
If 1...Ka6, 2.Re8! Kxa7 Rxe7+ wins the black rook.
2.f5
This pawn move not only paralyses the two black pawns, but it also functions as a tempo move – Black
has hardly any move left.
2...Ka4
XIIIIIIIIY
9R+-+-+-+0
9zP-+-zp-zp-0
9r+-+-+-+0
9+-mK-+P+-0
9k+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black seems to have everything under control, but White has a quite special idea here:
3.Rd8!
With this lovely move, White introduces a special motif: shutting in the enemy rook.
3...Rxa7 4.Kb6
Suddenly, the black rook loses almost all its squares.
There is only
4...Ra5
but that fails to:
5.Rd4+ 1-0

26
Study by Jaroslav Hasek
Le Monde des Échecs 1946

It is clear that Black’s rook has no squares, but how can White exploit this?
1.Kf2!
The white king has to protect his rook first, before going on its way to h3.
1...Ke5 2.Kg2 Kd4
2...Kf6 3.Kh3 Kf5 4.Ra3; 2...Ke6 loses to 3.Kh3 Kf5 4.Ra3, and the threat of 5.Ra5+, winning the
unfortunate black rook, cannot be parried. Also, 4...e3 doesn’t help after 5.Ra5+ Kf6 6.Kxg4 e2 7.Ra1.
3.Ra3!
Now, the rook moves to the side since
3...e3
(3...Ke5 4.Kh3 Kf5 5.Ra5+ is also over and out) fails to
4.Ra4+.
 White wins.

27
Study by Wouter Mees
Schakend Nederland 1961

With the king and rook on the edge, Black is asking for trouble. Still, White has to come up with
something special to seize the full point:
1.Kc5 Ka4 2.Rb1 Re6
The rook could not move to a square on the a-file, but it ensconces itself on e6.
White now wants to try to take away all the squares from the rook, which almost works:
3.Ra1+ Kb3 4.Kd5
XIIIIIIIIY
9-+-+-+-+0
9+-+-zp-+-0
9-+-+r+p+0
9+-+KzP-+-0
9-+-+-+-+0
9+k+-+-+-0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
4...Rb6
Now the rook is lost:
5.Rb1+
And White wins.
Instead of playing 4...Rb6, Black can liberate his rook with 4...Kb2, but it won’t help him after 5.Ra7
Rb6 6.Rxe7, and the remaining endgame is winning for White. The reason is that White supports his passed
pawn with the king, while Black’s pawn has no support. A possible winning continuation is 6...g5 7.e6 g4
8.Rg7 Rb5+ 9.Kd6 Rb6+ 10.Ke5 Rb3 11.Rxg4, and White wins.

28
Study by Henri Rinck
The Chess Amateur 1922

It is easy to see what White should do first here. Black’s rook is behind its own passed pawn, so the
latter constitutes a danger.
1.Kf7!
Asking the rook what it’s going to do.
1...Rg5
The rook has little space, but it appears to escape via this square.
1...g3 2.Rd8+ Ke5 3.Re8+, and the rook can go to e1 after White takes Black’s rook.
2.Kf6 Rxd5
 XIIIIIIIIY
9-+-+-+-tR0
9+-zp-+-+-0
9-+-mk-mK-+0
9+-+r+-+-0
9-zP-+-+p+0
9+-+-+-+-0
9-+P+-+-+0
9+-+-+-+-0
xiiiiiiiiy
3.c4!
The point. The rook has nowhere to go on the fifth rank, so...
3...Rd1
3...Rd4 amounts to the same.
4.Rd8+
And White wins.

29
Study by Henri Rinck
Schweitzerische Schachzeitung 1922

A somewhat chaotic position, in which it isn’t easy to unearth the winning theme.
1.axb7
In any position, it is always important to look for moves with a forced character. If anything, this
capture forces Black to react. 1.a7 would also force Black to react, but after 1...Ra6, there is not a lot to see
for White – not even after 2.Rg1 Rxa7 3.Rg7 b6, with equality.

One of my pupils tried 1.Rg1 bxa6 2.Rg6, but after 2...h5 (certainly not 2...fxg6?? in view of 3.Kxe6, and
White wins) 3.Rh6+ Kg8, if anything, Black is better here.

1.axb7 Rb6
The rook stops the pawn from the rear, and threatens to win it at the same time.
Retreating with 1...Re8 loses in an elementary way. White can force the pawn promotion with 2.Rc1!,
since there is nothing to be found against the threat of 3.Rc8.
2.Ra1!
White tries to promote the pawn, whatever it takes. We will soon see why the rook moves along the a-
file. Another interesting move is 2.Re1; however, it leads to a draw after 2...Rxb7 3.Re7 Rb5+ 4.Kc6 Rf5
5.Rxf7 Kg8 6.Rg7+ Kf8 7.Rg6 Kf7 8.Rxh6 Rf2.
2...Rxb7
Black has to take, otherwise the pawn gets through (after 3.Ra8+).
2...Kh7 comes down to the same as the main line.
3.Ra8+ Kh7
 XIIIIIIIIY
9R+-+-+-+0
9+r+-+p+k0
9-+-+-zP-zp0
9+-+K+-+-0
9-zp-+-+-+0
9+-+-+-+-0
9-zP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
And even though Black is now a pawn up, he still loses due to the poor positioning of his rook:
4.Kc6!
The rook is caught! White controls both the a-file and the back rank, and the f7- and b4-pawns are
badly in the way of the rook. Consequently, White wins.

30
Study by Dmitry Petrov
All Russian Tourney 1964

Another position with very little material. We have already seen that rook’s pawns are often not
favourable for the side that plays for a win. Still, there is a deeply hidden winning motif here:
1.h6!
White’s best chance; the pawn has to move forward.
During a training session, 1.Kxd7 was tried, but it doesn’t yield anything if Black defends well. He has
to apply Vancura’s principle of side checks: 1...Ra2!. Then, White’s best attempt is 2.Rg8+ Kf3 3.Rg5
(Vancura applies after 3.h6 Ra6! 4.h7 Ra7+, with a draw) 3...Kf4 4.Rb5 Kg4 followed by 5...Rh2, and the
pawn is lost.

Improving the rook with 1.Rg8+ doesn’t lead to anything: 1...Kf3, and now:
A) 2.Kxd7 Rh2 3.Rh8 (3.Rg5 Kf4 4.Ra5 Kg4 is also equal) is not possible in view of 3...Kg4 4.h6 Kg5
5.Rg8+ (5.h7 Kg6=) 5...Kf5! (5...Kxh6?? 6.Rh8+) 6.Rh8 Kg6 and Black draws;
B) 2.h6 Rc6+ 3.Kxd7 Rxh6 with a draw.

1.h6! Rc6+!
Black’s best option.
A move like 1...Kh1 (which will be played later as well) is met by 2.Kxd7 (not 2.Rg8? in view of
2...Rc6+ (2...Rh2 doesn’t help on account of 3.Rg6 followed by 4.Kxd7, and the king walks to g7) 3.Kxd7
Rxh6), and Black doesn’t manage to bring his rook to the side, as after 2...Ra2 3.Rc8! White wins.
 XIIIIIIIIY
9-+-+-+-tR0
9+-+p+-+-0
9-+rmK-+-zP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
xiiiiiiiiy
2.Ke5!!
Brilliant, and the only move, as will become clear.
The pawn is ‘harmful material’: 2.Kxd7? Ra6!, and big thanks to Josef Vancura again:
XIIIIIIIIY
9-+-+-+-tR0
9+-+K+-+-0
9r+-+-+-zP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
xiiiiiiiiy
Black ‘clings’ on from the side with checks: 3.Ke7 (3.h7 Kh3!? (to the h-file!) 4.Ke7, and Black can
choose: 4...Ra7+ – side checks, or behind the pawn with 4...Rh6 5.Kf7 Rh4 6.Kg6 Rg4+ 7.Kh5 Rh4+ 8.Kg6
Rg4+, and White doesn’t get any further) 3...Kh3 4.Kf7 Kh4 5.Kg7 Ra7+, and White can make no progress,
since Black either keeps giving checks or clings on to the pawn with his rook.
We also looked at 2.Ke7, but then it’s a draw after 2...d5! (not 2...Kh3? this time in view of 3.Kf7 d5
4.Kg7 Rc7+ 5.Kg6 Rc6+ 6.Kg5,
XIIIIIIIIY
9-+-+-+-tR0
9+-+-+-+-0
9-+r+-+-zP0
9+-+p+-mK-0
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
and now Black lacks the side check on the fifth rank; since White protects the h6-pawn with his king,
White can improve his rook on the next move) 3.h7 and now Black draws thanks to his own d-pawn, as
follows: 3...Rh6 4.Rg8+ Kf3 5.h8=Q Rxh8 6.Rxh8 d4 7.Kd6 d3 8.Kc5 d2 9.Rd8 Ke2
 XIIIIIIIIY
9-+-tR-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-zpk+-+0
9+-+-+-+-0
xiiiiiiiiy
followed by promotion, when the pieces can be put back in the box.
2...Kh3
Black’s king has to go back to the h-file because otherwise Rg8+ followed by h6-h7 would come one
day.
2...d5 is nonsense in view of 3.Rg8+, and if 3...Kf3 4.h7.
3.Kf5
XIIIIIIIIY
9-+-+-+-tR0
9+-+p+-+-0
9-+r+-+-zP0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White’s king is on the way to g5 to protect the pawn, and so liberate his rook from the bad h8-square.
Black’s side checks are obstructed by the pawn on d7, so the white king will move to the seventh rank.
3...Kh4!
3...d5 at this point fails as well: 4.Kg5 (because of the pawn on d5, Black has no side check) 4...d4 5.Rd8
Rc5+ 6.Kg6 Rc6+ 7.Kg7 Rc7+ 8.Kg8
XIIIIIIIIY
9-+-tR-+K+0
9+-tr-+-+-0
9-+-+-+-zP0
9+-+-+-+-0
9-+-zp-+-+0
9+-+-+-+k0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
and Black has run out of checks. He has to give his rook for the pawn, while his king has strayed too far
from his own pawn.
The text move prevents 4.Kg5, and it looks as if Black has saved himself in the nick of time. White
cannot protect the pawn with his king, and therefore can’t improve his rook. But now follows the totally
unexpected apotheosis of this study:
4.h7 Rh6
XIIIIIIIIY
9-+-+-+-tR0
9+-+p+-+P0
9-+-+-+-tr0
9+-+-+K+-0
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The rook gets behind the pawn – with a draw, you would think, since White’s king is unable to help.
5.Re8!!
White gives up his most important trump card, the h-pawn! How is that possible?
5...Rxh7 6.Kg6!
XIIIIIIIIY
9-+-+R+-+0
9+-+p+-+r0
9-+-+-+K+0
9+-+-+-+-0
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The rook is caught; only now do we see why the d7-pawn is ‘harmful material’.
6...Rh5
The only square for the rook where it is still protected, but...
7.Re4+
Winning. A marvellous finish!

31
Study by Alexey Troitzky
Novoye Vremja 1895

Black even has two extra pawns, but there are times when he would rather do without such pawns.
White has an artistic idea to exploit the black rook’s lack of freedom of movement:
3.Rb7+ Kh8 4.Rb8+!
The back rank is very important!
4...Kxh7 5.Kd6 1-0
Catching the rook in the middle of the board!

32
Jindrich Fritz
Prace 1953

If I remember well, I once saw this study during a training Garry Kasparov gave to a great young talent,
which had been recorded on film. The solution is so special that we can’t omit showing it here.
 White is a full knight up, but he has no time to save the knight since then his last pawn will be lost. The
rook + knight vs rook ending is a theoretical draw (although Kasparov himself once won it against Judit
Polgar!).
The winning move is very hard to find since the motif is quite hidden. But with the knowledge, and,
hopefully, also the skills we’ve gained from the previous positions, we have to be able to find the move:
1.a3!!
A truly brilliant idea, based on the theme of closing in the enemy rook.
1...Rxf1+ 2.Ke2 Rf4
What can we do here?
Obviously, the rook can’t escape via the back rank with 2...Rh1 in view of 3.Rd1+ Rxd1 4.Kxd1, and the
pawn ending is winning for White.
3.Rb3+ Ka2 4.Rb4!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+p+-0
9-tR-+-tr-+0
9zP-+-+-+-0
9k+-+K+-+0
9+-+-+-+-0
xiiiiiiiiy
The gorgeous point is revealed here: there is no escape route for the black rook! And 4...Rg4 loses, due
to 5.Rxg4 fxg4 6.a4. What a fabulous endgame study!

33
Study by Metodej Fila
Ceskoslovensky Sach 1931

White’s rook is highly unfortunately placed on a8, but if he can attack Black’s king from the back (with
tempo!), he will be able to promote the pawn. However, Black’s king is hiding behind White’s king.
Can White come up with something good?
1.Kc6 Kc4 2.Rc8!
This tactical trick wins a rook.
2.Kb6 does not achieve the aim yet because of 2...Rb1+, and the king has to go back again.
2...Rxa7 3.Kb6+ Kb4 4.Kxa7 1-0
We already saw this idea when we discussed the shelters of the black king.

34
Constructed position

White liquidates into a winning pawn ending:

1.Rd3+! Kxd3 2.Kxb2
According to the theory of key squares, this ending is a win for White.
2...Kd4 3.Ka3 Kc5 4.Ka4! Kb6 5.Kb4 1-0

35
Study by St Wojcik
Szachy 1959
 A king on the edge is often a bad sign. Even though Black is a pawn up, he is in deep trouble. White
profits as follows:
1.Kf6
Attacking the rook and creating a (mate) threat.
1...Rg7
Forcing White to choose the right move order.
2.Rh8+
First, Black’s rook has to be lured to the back rank, otherwise Black could save himself with 2...Rf7+.
2...Rg8 3.Rh7
The cut-off along the seventh rank is serious – but even more serious is the fact that Black’s rook is
badly hindered in its activity by its own pawn (= ‘harmful material’).
3...Ke8
3...g3 doesn’t make any sense in view of 4.Ra7, but now follows the same move:
4.Ra7
And Black is lost:
4...Rf8+ 5.Ke6 Kd8 6.Ra8+ 1-0

36
Study by Nikolay Kopaev
Source unknown, 1957

As a rule, a rook ending with two pawns vs one is a draw. But if the pawns are far advanced, tactics
may crop up.
White can win here, but he has to be inventive:
1.g6+ Kf8
Kopaev gives this as the main line, but also after 1...Kf6 2.Rf4+ Ke6, White has to find something. That
‘something’ is 3.h6! gxh6 4.Kh5,
XIIIIIIIIY
9-+-+-+-+0
9+-tr-+-+-0
9-+-+k+Pzp0
9+-+-+-+K0
9-+-+-tR-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
and since the king is cut off vertically, and White’s king can support its own pawn with Kxh6, the win
is secured.
1...Ke7 too is met by 2.h6! gxh6 3.Kh5 Kf8 4.Kxh6, since the king will be driven off with a check on the
back rank.
2.Ra8+ Ke7 3.h6!
Here also, this breakthrough decides.
3...gxh6 4.g7 1-0

37
Study by Ladislav Prokes
Prace 1951
 White’s c7-pawn is doomed, and also Black’s rook is behind his own passed pawn. So, no trouble at all
for Black, you might think... but the truth is different.
1.Rb2
Threatening 2.Rb8+ and 3.c8=Q.
1...Rxc7 2.Rb8+
XIIIIIIIIY
9-tR-+-+k+0
9zp-tr-+-+-0
9P+-+-+-+0
9+-+-mK-+-0
9-+p+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Forcing the black king to the seventh rank.
2...Kf7
Now, White can win by using the useful supporting square b7. Other king moves lead to the same
result, e.g. 2...Kg7 3.Rb7 Rf7 4.Ke6!; even 2...Kh7 fails to 3.Rb7 Rg7 4.Kf6!.
3.Rb7 Re7+ 4.Kd6
And because of the double attack on e7, Black has to trade on b7, after which White’s pawn decides the
issue.

38
Study by Peter Krystufek
Moderne Schachrätsel 1991

White is in a very tight spot: his king is ‘jammed in’ on the edge, and Black’s king and rook are very
dominantly placed. A rook move like 1...Rc8 would yield Black the win right away.
Still, White has an ingenious way to save himself:
1.Rb4!
The only correct move. Why does the rook have to go to this square? To see what can go wrong, we
will analyse two other rook moves: 1.Rb6? Rc7 2.Re6 Ra7+ 3.Kb4 Rb7+ 4.Kc4 b2, and the pawn marches on;

or 1.Rh5? b2+, and 2...b1=Q wins immediately.

1.Rb4! Rc7
The only way to try something, but now we are going to see what White has thought up.
Black couldn’t strengthen his position without a rook move along the c-file, since after, for instance,
1...Rg3 2.Rb5, he has nothing better than to retreat, after which White repeats the position with 2...Rc3
(2...b2+ 3.Ka2= costs the pawn) 3.Rb4.
2.Rc4+! Rxc4
Stalemate!

39
Study by Jarl Ulrichsen
(& Fernando Saavedra)
EBUR 2001
 White has only one pawn left, which hardly has any support, so the draw is within reach for Black – at
least, that’s the first impression. But White now starts an interesting manoeuvre to catch his opponent in
the corner:
1.Kg5!
Not 1.Kh5? in view of 1...Kg7,

while 1.e6? doesn’t make any sense either on account of 1...Re7 followed by 2...Kg7, with a draw.

1.Kg5! Rxe5+
After 1...Kg7, the king is driven to the back rank immediately with 2.Ra7+ Kf8 3.Kf6, and we have a
theoretically winning position. There could follow 3...Rc8 4.e6 Re8 5.Rf7+ Kg8 6.Rg7+ Kh8 (6...Kf8 7.e7+)
7.e7.
2.Kg6
Threatening mate!
2...Re8 3.Kf7
Where have we seen this theme before? Saavedra again...
3...Re1 4.Rh6#

40
Josef Kling
Chess Player’s Magazine 1866

The black king has found a shelter in front of its own pawn on b6. But White can blow up this shelter:
1.a4+ Kb4 2.a5 bxa5
Other moves also lose immediately.
3.Rb8+
And White wins.

41
Study by John Nunn
Secrets of Rook Endings 1992

Here, White’s rook stands in front of its own pawn, but by threatening to pull it away with check, White
can win – so:
1.c7!
It is a draw after 1.Kh2? Rg5! (curiously, this is the only defence!) 2.Kh1 Kg3, and White cannot win.
Actually, the point is the following defence: 2.c7 Rg7! 3.Kg2 Kf4+ 4.Kf2 (slowly, Black’s king shuffles in the
direction of the pawn, cutting off its white colleague in the meantime) 4...Rf7 5.Ke2 Re7+ 6.Kd3 Ke5
followed by ...Kd6, and Black will gobble up the pawn.

1.c7! Rg7
 XIIIIIIIIY
9-+R+-+-+0
9+-zP-+-tr-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+K0
xiiiiiiiiy
2.Kh2!
A wonderful tempo move that brings White the win.
Who wouldn’t play 2.Kg2 here? But that doesn’t win after 2...Kf4+ 3.Kf2 Rf7 4.Ke2 Re7+ 5.Kd3 Ke5
and again, Black’s king can reach the pawn.
2...Rh7+
If 2...Kg5, White has to find the right way again: 3.Kh3! (after 3.Kg3?, Black again steals the half-point:
3...Kf5+ 4.Kf3 Rf7 5.Ke3 Re7+ 6.Kd4 Ke6 7.Kc5 Kd7, with a draw) 3...Rh7+ 4.Kg3 Rg7 5.Kf3 Rf7+ 6.Ke4
Kf6 7.Kd5, and White wins.
3.Kg2 Rg7 4.Kf2
Thus, White’s king gets a free passage to the queenside; it can’t be cut off by the black king.
4...Rf7+ 5.Ke3 Rg7
5...Kf5 6.Kd4 Kf6 7.Kd5+–.
6.Kd4 Kg5
XIIIIIIIIY
9-+R+-+-+0
9+-zP-+-tr-0
9-+-+-+-+0
9+-+-+-mk-0
9-+-mK-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
7.Ke5!
And the curtain falls for Black:
7...Kg6
If 7...Kg4, 8.Kf6! drives the rook from its crucial square. On f7, the rook would ‘cling’ to the c7-pawn
and at the same time prevent checks along the f-file.
This is not possible after the text move.
8.Kd6
And White wins.

42
Study by Alexey Troitzky
500 Endspielstudien 1924

White has a distant passed pawn on h6, but it seems to be controlled by Black’s rook. The fact that
Black’s king is on the edge of the board gives White an extra impulse to try for the win:
1.Kc4!
Threatening checkmate all of a sudden.
1...Ka5
1...Ka3 is met by 2.h7 Rh3 3.Rb3+! Rxb3 4.h8=Q, and with a queen against rook plus two pawns, White
is winning, even though the conversion is still quite a job.
2.Rb5+! Ka6
And now follows an interception on the h-file:
3.Rh5! gxh5 4.h7 1-0
Again, the queen will beat the rook plus two pawns.

43
Study by Aron Nimzowitsch
Deutsche Schachzeitung 1918

The Great Teacher also composed endgame studies once in a while. This one is rather comical:
1.Ra7
White threatens to win the black rook.
1...Re8
Black cannot hide the rook on h7 (1...Rh7) in view of 2.dxe6.
2.d7 Re7
Obviously, 2.d7 Rf8 doesn’t work in view of 3.Ra8+ Ke7 4.Rxf8.
3.d6!
The doubled pawn slides forward like a caterpillar.
3...Rxd7 4.Ra8#

Unfortunately for Nimzowitsch, the simple 1.dxe6! is also winning in the starting position, for example
1...fxe6 2.Ra7 Re8 3.Kc4 e5 4.Kd5 etc.

44
Study by Nikolay Minev
Shakmatna Misl 1974

White has a pawn on the seventh rank, and usually this doesn’t bode well for the opponent.
1.d6!
White tries to break open the seventh rank.
Not 1.Rc7? in view of 1...Rd8, and the danger for Black has disappeared.

1.d6! Kf6
Of course, 1...exd6?? is impossible due to 2.d8=Q+.
2.dxe7 Kxe7
Black seems to have the pawn under control now, but there is more to this position than meets the eye.
 XIIIIIIIIY
9-+-+-tr-+0
9tR-+Pmk-+-0
9-+-+-+-+0
9+-+-+p+p0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White comes up with something astonishing:
3.d8=Q+!
Giving up his last pawn, with the idea that Black’s king is now cut off along the seventh rank, which
leads to big trouble for Black.
3...Kxd8 4.Ke5
The king reports at the front.
4...f4
Black cannot do much, mainly because his rook is very passive. Without the f-pawn, it would be a draw,
since Black could save himself with 4...Rf1 here.
5.Ke6!
This technique, which we have seen before under the ‘domination’ theme, brings White victory.
5...Re8+ 6.Kd6 Kc8 7.Ra8+ Kb7 8.Rxe8 1-0
Chapter 21
Queen endings
Queen endings may be the hardest-to-fathom endgame type since it seems that here the laws are different.
This is not so strange, if we realize that both sides are left with the strongest piece. The queen is very
mobile, and able to create danger quickly, and so accidents can easily happen. A queen ending can also end
in a marathon session since the weaker side can often chase the enemy king with ‘endless’ checks. The well-
known Dutch arbiter Arthur Schuering once told me during a tournament game that he was going to try to
be the terror of the arbiter on duty: he wanted to get a queen ending on the board and then bleed it dry for
as long as possible!
It would go too far for me to present all the highly complex problems of this endgame type. That is not
the aim of this book. I mainly intend to give the club player some practical advice with the help of generally
known principles. A club player may be able to find a way, even in this endgame type if he is presented with
an arrangement of themes, rules of thumb and techniques, to know what he should pay attention to. If you
really want to know all the ins and outs, I would refer you to the specialist works in endgame literature. For
example, I have the book Fundamental Chess Endings by GM Karsten Müller & Frank Lamprecht on my
bookcase. This book describes extensively what types of problems can occur. For this endgame type, Müller
and Lamprecht base their case on John Nunn’s book, Secrets of Chess Endings. The famous trainer, Mark
Dvoretsky, also provided many handles for trying to fathom this endgame type in his much-lauded
Dvoretsky’s Endgame Manual. Marvellous literature, but I can’t shake off the impression that for the average
Joe at the chessboard, their explanations are too complicated and too extensive. Of course, it will be useful
for anyone to read this material, but for the practical player it mostly suffices to have a few ‘handles’ in his
arsenal.
I had little experience with queen endings myself, until I got involved with them in a quite special way. It
allowed me to learn a few principles that had been unknown to me.
One of the first times I met several strong sub-top players of our country was after I had qualified for the
so-called ‘preliminaries’ of the Dutch Championship, which were played in Eerbeek in 1979. In those days,
games were still adjourned after the 40-move time control. The players stayed in a hotel where they also
dined, and between the adjournment and the resumption of the game, there was a break of roughly two
hours. FM Fred van der Vliet, from The Hague, had been participating in these tournaments for years. He
was a lawyer by profession and a strong chess player, probably of master level, who qualified for the
national championship countless times, which yielded a player quite a few privileges in those days. Van der
Vliet hardly played in international events, but for fifteen years he was a pillar for the Volmac Rotterdam
team that was sponsored by chess Maecenas and multimillionaire Joop van Oosterom.
When Van der Vliet had an adjourned game in Eerbeek one day, he looked around for someone who wanted
to help him analyse the position. As a young and inquisitive, upcoming player, I volunteered to help him. I
don’t remember the exact position, but I’ve made a kind of reconstruction that illustrates approximately
what it was about.
257
Fred van der Vliet game
Eerbeek 1979 (reconstructed)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zp-0
9-+-+-+Pzp0
9+-+Q+K+P0
9k+-+-+-+0
9+-wq-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
This was more or less the type of position he’d got on the board. Both players have two fixed pawns on the
kingside. White’s king is quite close to the kingside, and wants to force an entrance there. The black king is
at the other side of the board, and with his queen alone Black cannot prevent White from picking up his
pawns.
We agreed that Fred would play his moves, and I had to try to find replies that might cause him headaches.
This approach, highlighting the ideas you should pay attention to in such positions, helped me to improve
my understanding of how such endgames should be handled.
White’s task, to penetrate with the king, turned out to be a tricky business at closer inspection. By the way,
the defence was still much more difficult...
To serve the club player even better, it seemed useful to me to show a number of techniques that occur
frequently in queen endings, as used by strong players. Whose endgames could be better analysed than
those by the current World Champion himself, Magnus Carlsen, who is famous for his formidable endgame
technique? We will present a number of fragments by Magnus in the coming sections.

§ 21.1 Queen behind the passed pawn

Just as in rook vs rook endings, the queen can be put behind its own passed pawn to help it get to the other
side. If the opponent has to put his queen in front of the pawn, the activity of his queen dwindles
enormously. Moreover, the queen will not be able to fulfil its blockading function for long, since the
attacking queen will soon be looking for a support point close to the pawn.

258
Daniil Dubov 2710
Magnus Carlsen 2847
chess24.com 2021 (13)
XIIIIIIIIY
9q+-+-+k+0
9+-+-+pzp-0
9-+-+p+-zp0
9+-+-+-+-0
9-+-wQ-+PzP0
9zp-+-zP-+K0
9-+-+P+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Black’s queen is ideally placed behind the passed pawn.
37...a2 38.Qa1
White’s queen has to move to an extremely passive square to stop the pawn.
38...Qa6 0-1
The white player had seen enough here already.
39.Kg3 Qxe2 40.Qc1 (40.Qc3 Qd1! amounts to the same) is strongly met by 40...Qd3.
XIIIIIIIIY
9-+-+-+k+0
9+-+-+pzp-0
9-+-+p+-zp0
9+-+-+-+-0
9-+-+-+PzP0
9+-+qzP-mK-0
9p+-+-+-+0
9+-wQ-+-+-0
xiiiiiiiiy
analysis diagram
This is a multifunctional move. The black queen wants to come to b1 to enable the pawn to promote, and at
the same time the b1-h7 diagonal is covered from checks, so the black king has nothing to fear. Also against
38...Qa3 followed by ...Qb3 and ...Qb1, White wouldn’t have been able to undertake anything.

Rule of thumb
The queen is well placed behind its passed pawn to help push it forward.

§ 21.2 Safety of the queen’s own king

In the previous example, we already saw how much value should be attached to the safety of the queen’s
own king. This is even more clearly visible in the example below, in which we will be able to show a second
principle as well.

259
Aryan Tari 2642
Magnus Carlsen 2855
Stavanger 2021 (7)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+pmk-0
9-+-zP-+p+0
9+-+-wq-+-0
9pzp-+-+-zp0
9zP-+QzP-+P0
9-+-+-+P+0
9+-+-+-mK-0
xiiiiiiiiy
Black to move
This sharp position requires accurate calculation. One queen ending will transfer into a new one:
45...Qa1+ 46.Kf2 bxa3 47.d7 Qf6+ 48.Ke2 a2 49.d8=Q Qxd8 50.Qxd8 a1=Q 51.Qxh4 a3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+pmk-0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-wQ0
9zp-+-zP-+P0
9-+-+K+P+0
9wq-+-+-+-0
xiiiiiiiiy
Since the black queen is ideally placed on the long diagonal, further resistance is futile, and so Tari resigned
here.

Rule of thumb
A king surrounded by its own pawn is safeguarded against checks if its own queen is on a
diagonal on which enemy checks would be possible.
Let’s take a look at two more examples of Carlsen angling for a win by using a passed pawn, and at the same
time safeguarding his king against enemy checks.
In the position to our right, material is equal, but that does not count for much in queen endings. Essential
is which of the sides has a passed pawn, and how far advanced it is. Also, the threats to both kings
determine who will come away with the full point.
In this case, it is clear that White is winning, and Carlsen wraps up efficiently:

260
Magnus Carlsen 2881
Shakhriyar Mamedyarov 2760
Shamkir 2014 (1)
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-+-0
9pzp-+Q+-zp0
9+-zp-zPK+P0
9-+P+-+P+0
9+-+-+-+-0
9-+-wq-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to move
46.Qe8+ Kg7 47.Qe7+
 XIIIIIIIIY
9-+-+-+-+0
9+-+-wQ-mk-0
9pzp-+-+-zp0
9+-zp-zPK+P0
9-+P+-+P+0
9+-+-+-+-0
9-+-wq-+-+0
9+-+-+-+-0
xiiiiiiiiy
And the Azeri GM had seen enough, since if 47...Kg8 48.Kg6!, the white king penetrates devastatingly. It
can also escape the checks after 48...Qd3+ 49.Kxh6 Qe3+ 50.g5.

261
Magnus Carlsen 2810
Leinier Dominguez Perez 2712
Wijk aan Zee 2010 (11)
XIIIIIIIIY
9-+-+q+k+0
9+-+-+-+p0
9p+-+P+p+0
9+-+-+-+-0
9p+-+P+P+0
9+-wQ-+-+-0
9-zP-+-+-zP0
9+-+-+-mK-0
xiiiiiiiiy
White to move
White has an extra pawn, but there are technical difficulties.
38.Qf6
Obviously, Magnus would like to keep his foremost e-pawn.
38...Qc8
Now the question is: how can the king walk out of the coming checks? Carlsen has mapped out a beautiful
route!
39.Kg2 Qc2+ 40.Kh3
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-+p0
9p+-+PwQp+0
9+-+-+-+-0
9p+-+P+P+0
9+-+-+-+K0
9-zPq+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
40...Qc5
Thus, Black covers the f8-square, otherwise White could have given mate starting with 41.Qf7+.
A check on the third rank makes no sense since it stops after 40...Qd3+ 41.Kh4.
40...Qb3+ 41.Kh4 Qb7 loses to 42.Qd8+ Kg7 43.Qd7+ Qxd7 44.exd7.
41.Kh4
The white king is miraculously safe here!
41...Qb4
Black has to try something: now, he wants to attack the enemy king from the rear with ...Qe1+.
42.Qf7+ Kh8
Carlsen takes no notice of his opponent’s ‘threat’.
With his last queen move, he creates space for his king and at the same time nurses his pawn one step
forward.
43.e7 Qe1+ 44.Kg5 Qe3+
After 44...Qa5+, His Majesty also escapes the animosities: 45.Qd5 (definitely not 45.Kh6?? Qd2+ 46.g5
Qxh2#).
45.Qf4!
Black resigned in view of the variation 45...Qc5+ 46.Kh6 (now, yes! Also, the pawn ending after 46.e5
Qxe7+ 47.Qf6+ Qg7 48.Qd8+ Qg8 49.Qxg8+ Kxg8 50.Kf6 Kf8 wins for White after 51.g5 since Black is put
in zugzwang) 46...Qxe7 47.Qb8+.

§ 21.3 Exploiting a passed pawn/defending against checks

What we actually saw already in the previous sections was various ways to make use of a passed pawn.
However, in many positions it takes a lot more effort to nurse a distant passed pawn forward – for example,
in positions where the kings are standing more or less in the ‘open field’.

262
Magnus Carlsen 2843
Fabiano Caruana 2822
Stavanger 2018 (1)
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-zp-0
9-+-+-+-zp0
9zP-+-+p+-0
9-zP-+-+-+0
9+-+-wQ-+P0
9-+-+-+P+0
9+-+q+K+-0
xiiiiiiiiy
White to move
White has two passed pawns, but how can he help them, or one of them, to the other side? First, he has to
put an end to the checks:
45.Kf2 Qc2+ 46.Kg3
 XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-zp-0
9-+-+-+-zp0
9zP-+-+p+-0
9-zP-+-+-+0
9+-+-wQ-mKP0
9-+q+-+P+0
9+-+-+-+-0
xiiiiiiiiy
46...g5
In principle, Black doesn’t want to weaken his king’s position, but he has little choice here since otherwise
he would lose ‘quietly’. In this way, he also generates threats against the white king.
Much worse would have been 46...Qc7+ 47.Qf4 Qc3+ 48.Kh2, and the white king has found a wonderful
shelter.
47.Qe5+ Kh7 48.Kh2!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+-zp0
9zP-+-wQpzp-0
9-zP-+-+-+0
9+-+-+-+P0
9-+q+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
In queen endings, king safety is high on the agenda!
48...f4
Therefore, Black has to try to create some draught for the white king.
49.Qd5 Qa4 50.Qf7+ Kh8
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+Q+-0
9-+-+-+-zp0
9zP-+-+-zp-0
9qzP-+-zp-+0
9+-+-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
51.Qg6!
This is a characteristic of the play of the best players in the world. Carlsen finds a way to liquidate into an
endgame in which he has an uncomplicated win. He isn’t distraught by the fact that he has to give up one of
his passed pawns for this. What matters is to get one of them to the other side.
51.Qf8+ Kh7 52.Qd6, to try to push one of the pawns with a5-a6, would allow some trouble after 52...Qc2,
with the threat 53...f3 – incidentally, White is still winning.
51...Qxb4 52.Qxh6+ Kg8 53.Qxg5+
Picking up two black pawns, mainly to dismantle the counterplay.
53...Kh7 54.Qh5+ Kg7 55.Qg5+ Kh7
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+-+0
9zP-+-+-wQ-0
9-wq-+-zp-+0
9+-+-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
56.h4
Now, he uses another passed pawn, also to create luft for his own king.
There was nothing against 56.a6, but then it would still be a lot of work to win, for example: 56...Qa3
57.Qf5+ Kh8 58.Qc8+ Kg7 59.Qc7+ Kg8 60.Qc4+ Kg7 61.Qd4+ Kg6 62.a7 Qg3+ 63.Kh1 Qe1+ 64.Qg1
Qa5 65.Kh2 Qa6 66.Qb1+ Kf6 67.Qb8.
56...Qd6 57.Qh5+ Kg7 58.Qg5+ Kh7 59.h5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+k0
9-+-wq-+-+0
9zP-+-+-wQP0
9-+-+-zp-+0
9+-+-+-+-0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
There are no normal ‘safety regulations’ for Carlsen! He simply allows a discovered check, seeing that it’s
not dangerous for him.
59...f3+ 60.g3
60.Qg3! may have been slightly more clever, after which White will have two extra pawns at least, and
Black cannot pose many threats to his king.
60...f2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+k0
9-+-wq-+-+0
9zP-+-+-wQP0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-zp-mK0
9+-+-+-+-0
xiiiiiiiiy
61.Qg6+
Liquidation into a new queen ending!
61...Kh8
This is what it’s about. Of course, the pawn ending is immediately won for White: 61...Qxg6 62.hxg6+ Kxg6
63.Kg2+–.
62.Qxd6 f1=Q
White is three pawns to the good, but as long as his king is under threat, Black has every right to play on.
63.Qh6+ Kg8
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-+-0
9-+-+-+-wQ0
9zP-+-+-+P0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-+-mK0
9+-+-+q+-0
xiiiiiiiiy
64.Qe6+
There are always booby-traps to step on: 64.Qg6+ Kh8 65.h6??, and now follows 65...Qg2+!, with stalemate
after 66.Kxg2 !
64...Kh8 65.Qe3!
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-+-+0
9zP-+-+-+P0
9-+-+-+-+0
9+-+-wQ-zP-0
9-+-+-+-mK0
9+-+-+q+-0
xiiiiiiiiy
This is instructive. With the queen on e3 and the king (soon) on g3, all the back-rank checks will be
prevented.
65...Qb5 66.Qc3+
One way or another, the pawn has to be pushed, and therefore Carlsen is still searching for the right
placement of his queen.
66...Kh7 67.g4
Apart from covering h5, the pawn vacates the g3-square for its king too.
67...Qd5
On this central square, Caruana prevents the march of the a-pawn (68.a6? Qa2+). Please note how safe
White’s king is after 67...Qe2+ 68.Kg3.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+-+0
9zP-+-+-+P0
9-+-+-+P+0
9+-wQ-+-mK-0
9-+-+q+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
All checks have been prevented! This kind of pattern is very useful to remember.
68.Qc7+ Kg8 69.Kg3
XIIIIIIIIY
9-+-+-+k+0
9+-wQ-+-+-0
9-+-+-+-+0
9zP-+q+-+P0
9-+-+-+P+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Also here, White’s queen is ideally placed on a dark square. Apart from keeping a5 protected, she also takes
care of her own king along the diagonal, who is going to have to find a safe shelter somewhere in this open
position.
69...Qe6
Now, White does have to take measures.
Both players saw that after 69...Qd3+ 70.Kh4, there is no check to the white king for the present.
70.Qd8+ Kh7 71.Qd3+ Kh8
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-+-0
9-+-+q+-+0
9zP-+-+-+P0
9-+-+-+P+0
9+-+Q+-mK-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The queen has again taken up a place on the third rank – on a light square this time, enabling the a-pawn to
make another step forward.
72.a6 Qe5+
Again, we see that Black’s back-rank checks will be quickly exhausted: 72...Qe1+ 73.Kg2.
73.Kh3
This too is a good shelter for White’s king at this juncture.
73...Qa1 74.Qd8+ Kh7 75.Qe7+
75...Kh6
This helps White to find his ideal set-up, but the alternative wasn’t attractive either: after 75...Kh8 76.g5,
mate patterns emerge, even though White’s king still has to make a long journey: 76...Qc3+ 77.Kg2 Qc6+
78.Kg3 Qc3+ (the idea behind 76.g5! is that the a6-pawn is defended by a tactic: 78...Qxa6 79.Qf6+, and
White wins) 79.Kf4 Qd2+ 80.Kf5 Qd3+ 81.Kf6 Qxa6+ 82.Qe6 Qa1+ 83.Qe5.
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-wQ-zPP0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9wq-+-+-+-0
xiiiiiiiiy
analysis diagram
And after a new check, White’s king moves, giving discovered check followed by mate.
76.Qe3+ Kh7 77.a7 1-0
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+k0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-wQ-+K0
9-+-+-+-+0
9wq-+-+-+-0
xiiiiiiiiy
Carlsen again uses the same technique to protect his king on the one hand, and push his pawn forward on
the other. At this point, Caruana lowered the flag, since he has no perpetual check and cannot prevent the
pawn’s promotion. A fantastic demonstration of endgame technique by the World Champion!
The winning line could run as follows: 77...Qh1+ 78.Kg3.
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+k0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-wQ-mK-0
9-+-+-+-+0
9+-+-+-+q0
xiiiiiiiiy
analysis diagram
Here, we see the beauty of the set-up of White’s king and queen. Black doesn’t have a single check: 78...Qb7
79.Qd3+, and now White can play the queen to b8 after either 79...Kg7 80.Qg6+ Kh8 (on 80...Kf8, White
wins with the quiet move 81.Qc2!, threatening to trade the queens with 82.Qf2+, after which it’s
immediately game over) 81.Qe8+ Kg7 82.Qb8 or 79...Kh8 80.Qd8+ Kg7 81.Qb8. With this, he prevents
any possible enemy checks and enables his pawn to promote. The new queen can be used to eliminate new
checks.
What is the instructive technique Carlsen showed us here? He very elegantly combined pushing his passed
pawn with taking up a characteristic set-up with his queen and king. By putting both on same-coloured
squares on the third rank, the back-rank checks can be controlled to a large extent. Apparently, this type of
set-up with the king and queen is standard procedure for top-level players. Let’s have another look at two
of these stereotypical set-ups from the preceding fragment:
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-+-0
9P+-+-+-+0
9+-+-wq-+P0
9-+-+-+P+0
9+-+Q+-+K0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
set-up 1
The Kh3 + Qd3 take away many squares from the enemy queen, reducing perpetual check opportunities to
a minimum. With the queen on d3, the white player also supports the pawn push from a5 to a6.

XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+k0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-wQ-mK-0
9-+-+-+-+0
9+-+-+-+q0
xiiiiiiiiy
set-up 2
The Kg3 + Qe3 again prevent the enemy queen to start an attack on the white king from the back. Can we
call this line of play an important technique? Yes, we can!

Technique
A king and queen on the third rank, both on the same colour, preferably with one square
between them, prevent the enemy queen from giving check from the back rank.
When I saw this technique used by Carlsen, I was reminded of the following brilliant fragment from a game
between two former World Champions. Here, it is the white king in cooperation with a dark-squared
bishop who succeed in warding off the enemy checks.
263
Tigran Petrosian
Max Euwe
Zurich ct 1953 (25)
XIIIIIIIIY
9-+-+-+-+0
9+p+k+lwQ-0
9-+p+p+-zp0
9+-zPq+-+-0
9-zP-+-zPPzP0
9+-+-mK-+-0
9-vL-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
The game was adjourned in this position, and during the break Petrosian searched for the right set-up of
his pieces.
It is clear that the only thing White has to watch out for is a perpetual check by Black. The way he puts a
sharp curb on the checks is downright instructive!
41...Ke8 42.Bf6 Qb3+ 43.Bc3 Qd1 44.Qh8+ Kd7 45.Qb8
XIIIIIIIIY
9-wQ-+-+-+0
9+p+k+l+-0
9-+p+p+-zp0
9+-zP-+-+-0
9-zP-+-zPPzP0
9+-vL-mK-+-0
9-+-+-+-+0
9+-+q+-+-0
xiiiiiiiiy
White threatens to win a pawn on b7 and then to penetrate.
But how can he walk out of the checks?
45...Qc1+
Essential is White’s reaction to 45...Qg1+. With 46.Kd2 Qf2+ 47.Kd1 Qf1+ 48.Kc2 Qe2+ 49.Bd2, he would
reach the same position as the one that came on the board in the game, as follows: 49...Qe4+ 50.Kb2 Qd4+
51.Bc3 Qf2+ 52.Ka3.
46.Bd2 Qg1+ 47.Kd3 Qf1+
Black can’t activate his bishop with 47...Bg6+, since it will be lost after 48.f5 exf5 49.Qd6+.
48.Kc2
Forcing the black queen to assist in the defence on a passive square.
48...Qa6
After 48...Qc4+ 49.Kb2 Qd4+ 50.Bc3 Qf2+ 51.Ka3, White has steered clear from the annoying checks and
can now start grabbing material:
 XIIIIIIIIY
9-wQ-+-+-+0
9+p+k+l+-0
9-+p+p+-zp0
9+-zP-+-+-0
9-zP-+-zPPzP0
9mK-vL-+-+-0
9-+-+-wq-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
51...e5 52.Qxb7+ Ke8 53.Qxc6+ Kf8 54.Qd6+ Kg8 55.Qd8+ Kh7 56.Qd2 Qf1 57.Bxe5 Qa6+ 58.Kb2 Qa2+
59.Kc1, and White wins.
XIIIIIIIIY
9-wQ-+-+-+0
9+p+k+l+-0
9q+p+p+-zp0
9+-zP-+-+-0
9-zP-+-zPPzP0
9+-+-+-+-0
9-+KvL-+-+0
9+-+-+-+-0
xiiiiiiiiy
49.h5!
White is still making fun of the f7-bishop by giving it as little space as possible. At the same time, he fixes
Black’s h-pawn on a dark square, to be able to comfortably exchange queens later, since in the opposite-
coloured bishops ending the h6-pawn will be easy prey for his bishop.
White is not in the least discomforted by Black’s new series of checks that now begins. He has accurately
worked out that his bishop forms a good shield against the shower of checks:
49...Qa2+ 50.Kd3 Qb1+ 51.Ke2 Qe4+ 52.Kf2 Qd4+ 53.Be3 Qxb4
After 53...Qb2+ 54.Kg3, there is an ‘echo’ of the variation on the 48th move, when White’s king escaped via
a3. Please take note of the beautiful cooperation of the king and the rook in warding off the black queen!
XIIIIIIIIY
9-wQ-+-+-+0
9+p+k+l+-0
9-+p+p+-zp0
9+-zP-+-+P0
9-wq-+-zPP+0
9+-+-vL-+-0
9-+-+-mK-+0
9+-+-+-+-0
xiiiiiiiiy
54.Qf8!
Up to here, Petrosian had analysed everything during the break. He had seen that the black queen has to go
back to f6, where it has to fulfil a passive role, giving White a free hand to finish off without any risk.
54...Qb2+ 55.Kg3 Qf6 56.Qd6+
Petrosian’s play was always characterized by ‘safety first’. He could also have liquidated to a queen ending
with 56.Bd4 Qxd4 57.Qxf7+ Kd8 58.Qg8+ Ke7 59.Qh7+ Ke8 60.Qxh6, which should be winning since
White’s h-pawn has advanced quite far already.
We will give the rest of the game without comments since it no longer fits in with the theme in this section.
56...Kc8 57.Bd4 Qd8 58.Qxd8+ Kxd8 59.Bg7 Kc7 60.Bxh6 b6 61.cxb6+ Kxb6 62.Kh4 1-0
62...c5 63.g5 c4 64.g6 Bg8 65.Bg7 Kc5 66.h6+–.
A true masterpiece by Petrosian!

§ 21.4 Centralization
What I have learned from analysing queen endings with strong players is a kind of slogan that is still
ringing in my ears... let’s turn it into a rule of thumb:

Rule of thumb
Centralize your queen as much as possible, on a square where it protects as many of your own
pawns as possible, but especially your most distant pawn. Then, send your king into the ‘open
field’ in search of a shelter against the enemy checks. During this king march, work with the
threat of liquidation into a winning pawn ending.
The following fragment shows how such a technique can be used in practice. Semi-central squares like c4
or f4 (for the black queen) can also be regarded as central ones, where a queen is often very well placed.

264
Hou Yifan 2603
Magnus Carlsen 2861
Wijk aan Zee 2013 (9)
XIIIIIIIIY
9-mk-+-+-+0
9+-zp-+-+-0
9-zp-+-+-+0
9zp-+-+-+-0
9-+-wq-+-+0
9+-+-+-+-0
9-+-+QzPP+0
9+-+-+K+-0
xiiiiiiiiy
White to move
In such positions, the material balance is usually less important than how far a pawn has advanced. In some
cases, material can play a role, as we will see later. With the queen nicely centralized on d4, Carlsen can be
satisfied already. However, Hou Yifan now sets one of her pawns in motion to create counterplay:
45.g4 Qd5
Carlsen decides to put his queen on another central square, even threatening to give mate in one!
Also worthy of attention was 45...Qf4!?, another ‘centralizing’ move. The queen keeps White’s g-pawn
under control, and prevents White from using her other pawn as well. At the same time, he introduces the
possibility of ...a5-a4.
46.Qe8+
White cannot push another pawn here, since Black will then liquidate into a winning pawn ending. This is
based on the idea that a lone king can’t stop three connected passed pawns, whereas it can stop two;
Black’s king can simply block White’s pawns. An illustration of this principle was given by GM Mihail Marin,
who analysed this game for Mega Database. White could try 46.f4 here, with the possible continuation
46...Qh1+ 47.Kf2 Qh2+ 48.Kf3 Qxe2+ 49.Kxe2 Kc8.
 XIIIIIIIIY
9-+k+-+-+0
9+-zp-+-+-0
9-zp-+-+-+0
9zp-+-+-+-0
9-+-+-zPP+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
And here, it turns out that the three black pawns can put White’s king in zugzwang, whereas two pawns
can’t manage that with the black king! For example: 50.Kd3 Kd7 51.Kc4 c5 52.Kb5 Ke7 53.Ka4 c4 54.Ka3
b5 55.Kb2 a4 etc.
46...Kb7 47.Qa4
XIIIIIIIIY
9-+-+-+-+0
9+kzp-+-+-0
9-zp-+-+-+0
9zp-+q+-+-0
9Q+-+-+P+0
9+-+-+-+-0
9-+-+-zP-+0
9+-+-+K+-0
xiiiiiiiiy
She tries to stop Black’s a-pawn, but Carlsen isn’t caught out that easily.
47...b5! 48.Qxa5 Qd1+ 49.Qe1 Qxg4 50.Qe5 Qc4+ 51.Kg2 Qc6+
With this ‘semi-centralizing’ move, he safeguards his own king, and now he can push his b-pawn forward.
52.Kf1 b4 53.f4 b3 54.f5
XIIIIIIIIY
9-+-+-+-+0
9+kzp-+-+-0
9-+q+-+-+0
9+-+-wQP+-0
9-+-+-+-+0
9+p+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
On the face of it, the black machine is faltering here. The advance ...b3-b2 is not possible now, while White’s
centralized queen (apart from stopping 54...b2) enables f5-f6. What should Black do?
54...Ka6!
As ever, Carlsen’s technique is impeccable. He threatens 55...Qb5+ to liquidate to a winning pawn ending –
especially since ...b1=Q will be check!
55.Qa1+
She tries to sow confusion. If 55.f6,
 XIIIIIIIIY
9-+-+-+-+0
9+-zp-+-+-0
9k+q+-zP-+0
9+-+-wQ-+-0
9-+-+-+-+0
9+p+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
analysis diagram
Black has the very strong 55...b2!, using a well-known tactical motif in queen endings (55...Qb5+ 56.Qxb5+
Kxb5 57.f7 b2 58.f8=Q b1=Q+ leads to an endgame in which the b-pawn will bring the win, as is confirmed
by the Tablebases, but of course this is a difficult job that would take a long time; so, if possible, the black
player will try to find something else): 56.Qxb2 Qh1+ 57.Ke2 Qg2+ (the X-ray check!) 58.Kd3 Qxb2.
55...Kb6
The most convenient square for the king. Here, 55...Kb5 also won, since White has no good continuation.
After 56.Qe5+ Qc5 57.Qe8+ c6 58.Qb8+ Kc4 59.Qf4+ Qd4, it’s game over.
56.Qh8
56.Qd4+ Qc5 57.Qf6+ Kb7 backfires more quickly in view of 58.Qb2 Qxf5+ 59.Ke1 Qe4+, transferring to a
winning pawn ending.
XIIIIIIIIY
9-+-+-+-wQ0
9+-zp-+-+-0
9-mkq+-+-+0
9+-+-+P+-0
9-+-+-+-+0
9+p+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
xiiiiiiiiy
56...Qc1+
Thematic was 56...Qb5+ to get behind the own passed pawn: 57.Kf2 b2 58.Qb8+ Ka5 59.Qa8+ Kb4
60.Qe4+ Qc4 61.Qb1 Kc3 62.f6 Qf4+ 63.Kg2 Qc1, and wins.
57.Kg2 Qc2+ 58.Kh1
58.Kg3 leads to the same kind of finish as in the game: 58...b2.
58...b2 59.Qb8+
 XIIIIIIIIY
9-wQ-+-+-+0
9+-zp-+-+-0
9-mk-+-+-+0
9+-+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
9-zpq+-+-+0
9+-+-+-+K0
xiiiiiiiiy
Now, Black has to find a way to walk out of the checks.
59...Ka5 60.Qa7+ Kb4 61.Qb7+ Ka3 62.Qf3+
62.Qa6+ Qa4 63.Qd3+ Ka2 64.Qd5+ Ka1 65.Qe5 c5! is similar to the game.
62...Qb3 63.Qa8+ Qa4 64.Qf3+ Ka2 65.Qd5+ Ka1 66.Qe5 c5! 0-1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-zp-wQP+-0
9q+-+-+-+0
9+-+-+-+-0
9-zp-+-+-+0
9mk-+-+-+K0
xiiiiiiiiy
The intention behind Black’s move is to finish with a ‘super-centralization’: 67...Qd4, lifting the pin on the
long diagonal and thus enabling ...b1=Q.

§ 21.5 Queen plus pawn vs queen

This may be the most tedious subject in this book, but I will try to make it a little more accessible for club
players. What is better than connecting it with an anecdote?
In 1981, I was part of a Dutch students’ team that participated in the Under-26 World Team Championship
in Graz, Austria. Since we were competing in the top regions, we got the opportunity to cross swords with
the strongest teams in the world – among them the Soviet Union with the then 18-year-old Garry Kasparov
on first board. Our first-board player, GM John van der Wiel, wasn’t doing badly at all, and against Kasparov
he even reached a winning position in the far middlegame, which he spoiled in time trouble.
In this tournament, games were adjourned, which meant that the players would resume their game on that
same evening. If the game wasn’t finished by then, a new session would start in the morning before the new
round – and, possibly, continue that night. You understand – this was what happened in our match with the
United States.
The U.S. has always been a strong chess country, partly due to the great number of Russian immigrants.
Unfortunately, we ended up trailing 2½-½, with one adjourned game, between John van der Wiel and John
Fedorowicz.
Our countryman had thoroughly analysed a certain variation in the Rauzer Sicilian at home, and had
concluded that he could achieve an advantage despite the wholesale simplifications in the position. After 40
moves, a rook ending had arisen that was promising for him. This ended in an exciting race in which White
(our John) had to give up his rook for Black’s dangerous passed pawn. In the meantime, White was pushing
two pawns, for which Black also had to give up his rook.
Eventually, a queen ending appeared on the board in which White had an extra a-pawn. Van der Wiel had
already joked to us that this position had been reached by force straight from the opening – and if he
managed to win, the Rauzer would have been refuted! This game was continued during the following days,
and took no less than 143 moves in the end. I forgot how many sessions this was, but at any rate, it was a
battle of attrition, and many participants were making jokes about it...
I do know that the Americans were quite bent on drawing this game. It was more or less a matter of
prestige to see whether the endgame could be held or not. Van der Wiel told us that Timman had won a
similar type of ending against Gligoric relatively easily one year earlier:

265
Svetozar Gligoric 2565
Jan Timman 2600
Bugojno 1980 (4)
XIIIIIIIIY
9-+Q+-+-+0
9+-+K+-+-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-+p+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
60...g1=Q
This endgame is a draw according to the Tablebases.
61.Qc3 Qg3 62.Qc1 Qg4+ 63.Kc7 Qg7+ 64.Kb6 Qf6+ 65.Ka7
Here, White’s choice was limited already, but Gligoric shows that he knows the ropes.
65...Qe5 66.Qg1 Kh3 67.Qh1+ Kg4 68.Qd1+ Kf4 69.Qf1+ Kg5 70.Qg2+ Kf6 71.Qf3+ Ke7 72.Qa3+
Qd6 73.Qf3 Qd4+
XIIIIIIIIY
9-+-+-+-+0
9mK-+-mk-+-0
9-+-+-+-+0
9+-+-+-+p0
9-+-wq-+-+0
9+-+-+Q+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Always a good advice: centralize your queen as much as possible, protect the pawn, and then start working
with the king.
74.Kb7 h4 75.Qh5 Kd6 76.Qg6+ Kc5 77.Qc6+ Kb4 78.Qg2 Qc4 79.Kb6 Qc5+ 80.Kb7 Qd4 81.Kc6
Qc3+ 82.Kd7 Qd3+ 83.Ke6 h3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-mk-+-+-+0
9+-+q+-+p0
9-+-+-+Q+0
9+-+-+-+-0
xiiiiiiiiy
It looks as if Black is slowly making progress, but the position is still a draw. However, according to the
Tablebases, at this moment White has only two drawing continuations: the text move and 84.Qf2 are
correct, all other moves lose.
84.Qh2 Kc3
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mkq+-+p0
9-+-+-+-wQ0
9+-+-+-+-0
xiiiiiiiiy
And here, we see how thankless the task can be for the defender. The attacker can do what he wants, but
the defender has a very difficult job. Gligoric didn’t manage to find one of the three correct continuations
here. These were all king moves, i.e. 85.Ke5, 85.Kf6 and 85.Kf7.
85.Qe5+?
Also, the task can be daunting for the stronger side. The Tablebases indicate that here 85...Kd2 is the only
correct move.
85...Kc2? 86.Qh2+?
Please don’t ask me why the Tablebases give 86.Qf4! as the only move to hold the game, but such is the cold
conclusion of modern chess technique.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+q+-+p0
9-+k+-+-wQ0
9+-+-+-+-0
xiiiiiiiiy
86...Kd1!
Now, Timman finds the only winning move!
87.Qh1+ Kd2 88.Kf6
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+q+-+p0
9-+-mk-+-+0
9+-+-+-+Q0
xiiiiiiiiy
At this moment, Black has a choice between 88...Qd6+, 88...Qc3+ and 88...Qg3 to gain the full point. Can we
blame Timman for not being able to calculate this? The move he plays is perfectly logical for a human
player:
88...Qe3? 89.Kf5?
89.Qh2+ would have saved White.
89...Qf2+ 90.Ke5 Qe1+ 0-1
However, there were more precedents. By coincidence, we found out that the Americans went much
further in their analyses than we could have surmised. They even ‘sacrificed’ a player on the team to get to
the bottom of the endgame in his hotel room. This was Ron Henley, who turned out to possess more than
exceptional analytical skills. Later, he was even approached by Karpov to help him in his team of seconds
during a World Championship match. Henley did his homework very well: he found out where the black
king should stand in this type of endgame. We shouldn’t forget that in those days there weren’t any (strong)
computer chess programs, let alone endgame databases like the Tablebases of today, which indicate exactly
which position is winning and which is drawn in endgames with seven pieces or less. At the time, little was
known about this, and no rules had been laid down.
What I gathered from it is that with a white a-pawn, Black’s king is best posted on the h2- or h3-squares.
This is connected to White’s familiar winning method: centralizing his own queen (while keeping the pawn
protected) and then trying to escape the checks with a long king walk. The idea is that White can end the
long series of checks by threatening to trade queens at a certain point. We saw this principle nicely
illustrated in Timman’s game. Henley had found out more than this. Since the Americans also knew that
Van der Wiel was going to give it a serious try, they allowed the pawn ‘voluntarily’ to march from a2 to a7 –
otherwise, the game would have taken much longer still. At the moment the pawn emerged on a7,
Fedorowicz’s king had to be on the right square, and then he would have to give the right checks. This
required great accuracy, as is shown in the analysis below. Henley had told their first-board player how it
had to be done. So, after the regular playing round, Fedorowicz had to have coffee with Henley several
times to learn how to defend this endgame.
When playing through the final part of this queen ending, I consulted the Tablebases, and I have to take my
hat off to what the Americans accomplished at the time. What Henley did was important, but I have even
more respect for Fedorowicz, who had to find it all in a practical game. The slightest inaccuracy would have
been fatal for him (if we assume that Van der Wiel would have found the win then...), but according to the
Tablebases, Fedorowicz found the extremely narrow path to the draw. I have added exclamation marks at
the moments where he found the only correct move, and there are quite a lot of them!
I had also become fascinated by the endgame, and I tried to help Van der Wiel as well as I could – not only
in analysis, but also with moral support, by accompanying him to the tournament hall after breakfast, when
the two gentlemen started their marathon sessions in the early hours. I can remember that one morning I
saw Fedorowicz stumbling through the corridor, big cup of coffee in hand, moaning, in his thick American
accent: ‘I still have to defend this f*** endgame, tooooooo early in the morning...’

266
John van der Wiel
John Fedorowicz
Graz Wch tt U26 1981 (3)
XIIIIIIIIY
9-+-+-+-tr0
9zPKzP-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9P+-+p+-+0
9+-+-+-+k0
xiiiiiiiiy
White to move
63.c8=Q
The only winning chance. After the promotion with the other pawn, all pieces leave the board: 63.a8=Q
Rxa8 64.Kxa8 e1=Q 65.c8=Q Qa5+ 66.Kb8 Qxa2.

63.c8=Q Rxc8 64.Kxc8 e1=Q 65.a8=Q+

White queens with check, and has also maintained his extra pawn. This is the beginning of a queen ending
with a passed a-pawn in which White seems to have chances, also because he can centralize his queen right
away.
65...Kh2 66.Qd5 Qc3+ 67.Kd7 Qg7+ 68.Kc6 Qc3+ 69.Kb5 Qb2+ 70.Kc5 Qa3+ 71.Kd4 Qa7+ 72.Kd3
Qa3+ 73.Ke2 Qe7+ 74.Kd1
XIIIIIIIIY
9-+-+-+-+0
9+-+-wq-+-0
9-+-+-+-+0
9+-+Q+-+-0
9-+-+-+-+0
9+-+-+-+-0
9P+-+-+-mk0
9+-+K+-+-0
xiiiiiiiiy
The checks are exhausted for now.
74...Qa3 75.Qc4 Qd6+ 76.Kc2 Qg6+ 77.Qd3 Qc6+ 78.Kd1 Qa8 79.a3
Finally, White has managed to take a step with the pawn.
79...Kg2 80.Kc2 Kf2 81.Qd4+ Kf3 82.a4 Qe4+
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9P+-wQq+-+0
9+-+-+k+-0
9-+K+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black can play this with impunity: after the queen trade, the pawn ending would be a draw.
83.Kc3 Qc6+ 84.Qc4 Qf6+ 85.Kb3 Qb6+ 86.Qb5 Qe3+ 87.Kc4 Qe4+ 88.Kc5 Qe5+ 89.Kc6 Qe6+
90.Kc7 Qe7+ 91.Qd7 Qe5+ 92.Kb7 Qb2+ 93.Qb5 Qg7+ 94.Ka6 Qg6+ 95.Ka7 Qg1+ 96.Qb6 Qa1
97.a5
Another step forward.
97...Ke2 98.Qe6+ Kf1 99.Qf5+ Ke1 100.Kb6 Qb2+ 101.Kc6 Qc3+ 102.Kd6 Qb4+ 103.Qc5 Qb8+
104.Ke7 Qb7+ 105.Kf6 Qf3+ 106.Qf5 Qa8 107.Qe6+ Kf1 108.a6
XIIIIIIIIY
9q+-+-+-+0
9+-+-+-+-0
9P+-+QmK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+k+-0
xiiiiiiiiy
And, just like that, the pawn can move to the sixth rank as well. Our hopes were rising.
108...Kg1 109.Qb6+
XIIIIIIIIY
9q+-+-+-+0
9+-+-+-+-0
9PwQ-+-mK-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-mk-0
xiiiiiiiiy
109...Kh2!
This is absolutely the only move! All other moves lose; after, for instance, 109...Kf1, White has precisely one
winning move: the mysterious 110.Kg7!.
110.Qc7+ Kg1 111.Qc5+ Kh2 112.a7
And now, it is even on the seventh rank. But we see that Black’s king has taken up the right position, and
now follows a ‘shower’ of checks from which White’s king cannot find shelter. A glance at the Tablebases
teaches us that Henley had indeed analysed excellently: the position is a draw.
112...Qh8+
The funny thing is that according to the endgame databases, Black has only one, two or three good moves
every time. This goes for almost all the moves that follow from here.
113.Kg5 Qg7+ 114.Kf5 Qh7+ 115.Kf4 Qf7+ 116.Ke3 Qb3+ 117.Ke4 Qb7+!
XIIIIIIIIY
9-+-+-+-+0
9zPq+-+-+-0
9-+-+-+-+0
9+-wQ-+-+-0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-mk0
9+-+-+-+-0
xiiiiiiiiy
This is absolutely the only move. Fedorowicz knows exactly what he’s doing! Also further on, he manages to
find the very narrow path to the draw.
The longer I am watching those tables, the more formidable I find the American’s achievement.
118.Kd3 Qa6+ 119.Ke3 Qe6+ 120.Kd4 Qg4+!
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-wQ-+-+-0
9-+-mK-+q+0
9+-+-+-+-0
9-+-+-+-mk0
9+-+-+-+-0
xiiiiiiiiy
Again, the only move.
121.Kc3 Qf3+! 122.Kd2 Qg2+ 123.Ke1 Qe4+! 124.Kd1 Qf3+! 125.Kc1 Qh1+ 126.Kb2 Qb7+!
127.Ka3 Qf3+! 128.Kb4 Qe4+! 129.Kb5 Qe2+ 130.Kc6 Qe8+ 131.Kb6 Qe6+! 132.Kb7 Qe4+
133.Kc7 Qh7+! 134.Kc6 Qe4+! 135.Qd5 Qa4+! 136.Kb7 Qb4+! 137.Kc7 Qe7+ 138.Qd7 Qc5+!
139.Kb8 Qf8+ 140.Qc8 Qb4+! 141.Ka8 Kg1 142.Qc1+ Kf2
XIIIIIIIIY
9K+-+-+-+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-wq-+-+-+0
9+-+-+-+-0
9-+-+-mk-+0
9+-wQ-+-+-0
xiiiiiiiiy
Here, Van der Wiel had seen enough. Always a humorist, he finished with a joke:
143.Qb2+ Qxb2 ½-½
We have advanced quite a few years since then. The modern grandmaster has powerful computers at his
disposal, he has access to gigantic opening databases, chess programs with artificial intelligence, and the
Tablebases have grown out to a database in which all endgames with seven pieces or less have been
analysed to the end. If we consider the fact that some opening lines have been analysed to the 40th move,
and that we have a database with endgames of which we know the result, it is not so strange that the well-
known Dutch chess commentator Lex Jongsma once sighed that in future grandmasters would have to play
only 20 moves per game: the rest is already settled!
Let’s see how the World No. 1 hauls in the win in our decade:

267
Magnus Carlsen 2862
Nils Grandelius 2663
Wijk aan Zee 2021 (9)
XIIIIIIIIY
9-+-+Q+-+0
9+-+-+-zpk0
9-+-+-+-zp0
9+-+-+P+-0
9-+-zp-+-+0
9+-+-+-zP-0
9-+P+-+KzP0
9+-+q+-+-0
xiiiiiiiiy
White to move
There is a lot to be done here to start with, but Carlsen knows like no other how to go about things.
50.Qg6+! Kh8
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-zp-0
9-+-+-+Qzp0
9+-+-+P+-0
9-+-zp-+-+0
9+-+-+-zP-0
9-+P+-+KzP0
9+-+q+-+-0
xiiiiiiiiy
51.f6! Qe2+ 52.Kh3 Qe6+
52...Qf1+ makes no sense, as now White marches with his king:
A) 53.Kh4 would probably be played by a flesh-and-blood grandmaster. Black has no checks left. The
computer thinks that after 53...gxf6,
 XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-+-0
9-+-+-zpQzp0
9+-+-+-+-0
9-+-zp-+-mK0
9+-+-+-zP-0
9-+P+-+-zP0
9+-+-+q+-0
xiiiiiiiiy
analysis diagram
54.h3! is by far the best move. Actually, I think this is an insane move, and a human player wouldn’t easily
think of it (again, 54.Qxh6+ looks more logical: 54...Kg8 55.Qg6+ Kf8, but indeed, the win is not exactly
trivial here). 54...Qf2 55.Kh5 Qf3+ 56.Qg4 Qe3 (56...Qd5+ 57.Kxh6) 57.Kg6 indeed gives White a winning
position; curiously, here again, White positions his queen well and then starts walking with the king;
B) 53.Kg4! Qe2+ 54.Kf4 Qd2+ 55.Ke5 – through all the lines, a stereotypical winning method.
53.Kh4 gxf6
The pawn ending is hopeless: 53...Qxf6+ 54.Qxf6 gxf6 55.Kg4 Kg7 56.Kf5, and White picks up the d4-pawn.
54.Qxh6+ Kg8
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-+-0
9-+-+qzp-wQ0
9+-+-+-+-0
9-+-zp-+-mK0
9+-+-+-zP-0
9-+P+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
55.Qg6+! Kf8 56.g4
Carlsen can be trusted to convert such an endgame neatly.
56...Qe5 57.h3 Ke7 58.Qf5
This is typical when you have a better position: most of the time, you can liquidate into a pawn ending,
while your opponent has to avoid it in most cases. This allows Carlsen to keep improving his queen.
58...Qe3
XIIIIIIIIY
9-+-+-+-+0
9+-+-mk-+-0
9-+-+-zp-+0
9+-+-+Q+-0
9-+-zp-+PmK0
9+-+-wq-+P0
9-+P+-+-+0
9+-+-+-+-0
xiiiiiiiiy
59.Kh5!
Well found.
59...Qxh3+ 60.Kg6 Kd6 61.Kxf6 Qe3 62.g5 Qe7+ 63.Kg6 Qe8+ 64.Kg7 Qe7+ 65.Kg8
Now, each check is answered by a queen trade, and so Grandelius threw in the towel.

§ 21.6 Tactics – Exercises

There are countless tactical motifs in chess, and we have seen many in this book. I have made a small
selection of the most frequent tactical themes in queen endings, which the reader can study with the help of
exercises.

• Domination
We have encountered this theme in the queen or rook vs pawn endings. Sometimes, you can push the
enemy king together with the queen to the edge of the board, limiting the enemy queen severely in her
mobility. This is a rewarding theme for endgame study composers!

1
XIIIIIIIIY
9-+-+-+k+0
9+-+p+-+-0
9-+-+-wQ-+0
9+q+-+pmKp0
9-+-+-+-+0
9+-+-+-+-0
9P+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

2
XIIIIIIIIY
9-+-+-+-+0
9zp-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-zP-+-+-zp0
9zP-+-+K+-0
9-+-+-+-mk0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
3
XIIIIIIIIY
9-+-wQ-+-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+-+0
9+-+-+-+-0
9-zpq+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

4
XIIIIIIIIY
9-+-+-+-+0
9+-zp-+-+-0
9-+-+p+-+0
9wq-zp-zpp+-0
9-+-+-mk-mK0
9+P+-+-+-0
9-+-zP-+Q+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

5
XIIIIIIIIY
9-+k+-+-+0
9+-zp-+-+q0
9p+P+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-wQ-+-+p0
9-+-mK-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

6
XIIIIIIIIY
9-+-+-+-wq0
9+-wQ-+-+-0
9-+-+-mkp+0
9+-+-+-+-0
9-+-+-+P+0
9+-+p+K+-0
9-+-zP-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
7
XIIIIIIIIY
9-+-+-+-+0
9zPQzpk+K+-0
9-+-+-+-+0
9+-wq-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

8
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+K0
9-+-+Q+-+0
9+-+-+-+k0
9-+-+-+-zp0
9+-+-+-+p0
9-+-+-zPqzP0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

9
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpk0
9-+-+-+q+0
9+-+-+-+p0
9-+-+-+-zP0
9+-+-+-+-0
9-+Q+-+-+0
9+K+-+-+-0
xiiiiiiiiy
White to play and draw

10
 XIIIIIIIIY
9-+-+-+-+0
9zp-+-+-zpk0
9-+-+-+q+0
9zp-+-+-+p0
9P+-+-+-zP0
9+-+-+-+-0
9-mKQ+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and draw

11
XIIIIIIIIY
9-+-+-+-+0
9zp-zp-+-+-0
9kzP-+-+Q+0
9zP-+-+-+-0
9-+-+-+-+0
9+-+-+-+q0
9-+-+-+-+0
9+K+-+-+-0
xiiiiiiiiy
White to play and win

12
XIIIIIIIIY
9-+-+-+-+0
9mk-+K+-+-0
9qzp-zp-+-+0
9+-+-+p+-0
9-zp-+-zp-zp0
9+P+-+-+-0
9-zP-+-+Q+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
13
XIIIIIIIIY
9-+-+-+Q+0
9+-+K+-+-0
9-+-+-+-zp0
9+-mk-+-+-0
9p+-+-+-+0
9+-+-zP-+-0
9-+-+-+-+0
9+q+-+-+-0
xiiiiiiiiy
White to play and win

14
XIIIIIIIIY
9-+-+-+-+0
9+-zp-+p+-0
9q+P+p+-+0
9zp-+-+-+-0
9-+-+Q+-+0
9+-mK-+-+-0
9-+P+-+-+0
9mk-+-+-+-0
xiiiiiiiiy
White to play and win

15
XIIIIIIIIY
9-mk-+-+-+0
9+-+Q+-+-0
9p+-+-+-+0
9zPp+-+-+-0
9qzP-+-+-+0
9+-+-+-+-0
9-mK-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

16
 XIIIIIIIIY
9-+-+-+-+0
9+p+-+qzp-0
9-+p+-zp-+0
9+-zP-+P+p0
9-+-+-+-zP0
9+-+-+-+k0
9-mK-+-zP-+0
9+-+Q+-+-0
xiiiiiiiiy
White to play and win

17
XIIIIIIIIY
9-+-+-+-+0
9+-wQ-+-+P0
9-+-+-+-wq0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-+-zp-+0
9+-+-+-+k0
xiiiiiiiiy
White to play and win

18
XIIIIIIIIY
9-+-+-+-+0
9+-+-+P+-0
9-+-+-+-+0
9wQ-+-+-+-0
9K+-+-+-+0
9+-+-+-+-0
9-+-+-+qzp0
9+-+-+-+k0
xiiiiiiiiy
White to play and win
19
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-zp-+k+-+0
9+P+-+-+-0
9-zp-mK-+-zp0
9+-+-+-+P0
9-+-+-+-wQ0
9+-+-wq-+-0
xiiiiiiiiy
White to play and draw

20
XIIIIIIIIY
9-+-+-+-+0
9+pmkP+-+-0
9-+-+q+-+0
9+-zp-+-+-0
9-zp-+-+-+0
9+-+-+-zP-0
9-zP-mK-+Q+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

21
XIIIIIIIIY
9-+-+-+Q+0
9+-+K+-+-0
9-+-+-+-zp0
9+-mk-+-+-0
9-+-+-zp-+0
9+-zP-+-+-0
9-+-+-+-+0
9wq-+-+-+-0
xiiiiiiiiy
White to play and win

22
 XIIIIIIIIY
9-mK-mk-+-+0
9+-+-+-wQ-0
9-+-+-+-+0
9+-+-+-zpq0
9-+-+-+-zp0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

23
XIIIIIIIIY
9-+-+-+-+0
9+-+p+-+-0
9p+-zp-+-+0
9zp-+Q+-+-0
9-+-+-+-+0
9mKp+-+-+-0
9-+q+P+-+0
9mk-+-+-+-0
xiiiiiiiiy
White to play and win

24
XIIIIIIIIY
9-+-+-+-+0
9+-+-+K+-0
9Q+-+-+p+0
9+-+-mk-+-0
9-+-zp-+-+0
9+-+-+p+-0
9-+-+-+-zP0
9+-wq-+-+-0
xiiiiiiiiy
White to play and win
25
XIIIIIIIIY
9-+-+-+KwQ0
9+-+-wq-+P0
9-+-+-+k+0
9+-+-+-zp-0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and draw

26
XIIIIIIIIY
9Q+-+K+k+0
9+-+-+-zp-0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+q0
xiiiiiiiiy
White to play and win

27
XIIIIIIIIY
9-+-+-wq-+0
9+Q+-+-zp-0
9-+-mkp+p+0
9+-+-+-zP-0
9p+-+P+-+0
9zP-+K+-+p0
9-+-+-+-zP0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

28
 XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+-0
9-mk-+-+-+0
9+-+-+-wQ-0
9-+p+-+-+0
9+-+-mK-+-0
9-+-+-+-+0
9+q+-+-+-0
xiiiiiiiiy
White to play and win

29
XIIIIIIIIY
9-+-+-wQ-+0
9+-zp-+-+-0
9-zpP+p+-+0
9+q+-mk-+-0
9-zp-+-+-+0
9+P+P+-+-0
9-+-+-+K+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

30
XIIIIIIIIY
9-+-+-+-mk0
9+-zp-wQ-+-0
9-+-+-+-+0
9+-+p+K+-0
9-+-+p+-+0
9+pwq-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win
31
XIIIIIIIIY
9-+Q+-+-+0
9+-+-+-zpk0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-zpP+-0
9-+-+-+-+0
9+-+-wq-+-0
xiiiiiiiiy
White to play and win
32
XIIIIIIIIY
9-+-+-wq-+0
9wQ-+-+-+-0
9-+-mkp+-+0
9+-+-+-+-0
9-+-+P+-+0
9+p+K+-+-0
9-zP-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White to play and win

§ 21.7 Tactics – Solutions

1
Study by Alexey Troitzky
Study 1930

With the king at the edge, mate patterns emerge. But for this, help is always needed, in this case from
White’s own king, to push Black’s king into an even tighter spot.
1.Kg6
Threatening mate in one, so Black has only one defence:
1...Qc6
This leads to a pawn ending involving a pawn race.
2.Qxc6 dxc6 3.a4 f4 4.a5 f3 5.a6 f2 6.a7 f1=Q
Black is even the first to reach the other side, but...
7.a8=Q+ Qf8
 XIIIIIIIIY
9Q+-+-wqk+0
9+-+-+-+-0
9-+p+-+K+0
9+-+-+-+p0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
With the queen and king jammed in on the back rank, there are various mate patterns here. The win
still takes a little time, but it is highly elegant:
8.Qa2+ Kh8 9.Qb2+ Kg8 10.Qb3+ Kh8 11.Qc3+ Kg8 12.Qc4+ Kh8 13.Qd4+
With a zigzag movement, White’s queen slowly approaches.
13...Kg8
XIIIIIIIIY
9-+-+-wqk+0
9+-+-+-+-0
9-+p+-+K+0
9+-+-+-+p0
9-+-wQ-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
14.Qd7!
Depriving the black queen of all its useful squares. The domination is complete, and mate can no longer
be prevented.

2
Study by V.Zikov
Roycroft Jubilee Tourney 1978

Again, a pawn ending going over into a queen ending. White’s move order requires a delicate touch:
1.Kf2!
With 1.a4?, he would spoil the win: 1...h3 2.Kf2 a5 3.bxa5 Kh1 4.a6 h2, and White can’t win.

1.Kf2! h3 2.b5!
The right order. Again, 2.a4? is an error due to 2...a5.
2...Kh1
Now, 2...a6 is met by 3.b6 a5 4.b7 a4 5.b8=Q+, and Black gets mated; and also after 2...a5 3.b6 Kh1
4.b7 a4 5.b8=Q, Black ends up on the wrong foot: 5...h2 6.Qb7#.
3.Kf1!
Another clever move!
3...h2
Again, 3...Kh2 4.a4 Kh1 5.a5 Kh2 6.b6 a6 7.b7 Kh1 8.b8=Q wins for White.
 XIIIIIIIIY
9-+-+-+-+0
9zp-+-+-+-0
9-+-+-+-+0
9+P+-+-+-0
9-+-+-+-+0
9zP-+-+-+-0
9-+-+-+-zp0
9+-+-+K+k0
xiiiiiiiiy
Black is trying to stalemate himself.
4.b6!
White has an ingenious plan to prevent stalemate.
4...a5
After 4...axb6 5.a4 b5 6.a5 b4 7.a6 b3 8.a7 b2 9.a8=Q#, White is right on time!
5.b7 a4 6.Ke2!
Now, White lets the enemy king escape so as to enter a winning queen ending.
6...Kg2 7.b8=Q h1=Q 8.Qb7+ Kg1 9.Qb6+
The queen starts the ‘zigzag’ movement again.
9...Kh2 10.Qd6+ Kg2 11.Qd5+ Kh2 12.Qh5+ Kg2 13.Qg4+
In most cases, this is the desired square for the queen.
13...Kh2!
Domination!
14.Kf2
And White wins.

• Discovered check
Sometimes, in a long, forced variation, you have to look ‘behind the horizon’. If both queens end up on
the same diagonal, it may be possible to move a pawn that is ‘coincidentally’ standing in between away with
check (i.e., with tempo gain!). This way, the enemy queen falls victim to an extremely nasty discovered
check!

3
Study by Henri Rinck
Basler Nachrichten 1938

It is hardly imaginable that White can win this, but the composer has weaved a diabolical trick into the
position that Black can’t evade:
1.Qd5+! Kh4
1...Kg4 is met by 2.Qg2+, transposing to the main line.
2.Qh1+!
It’s highly unlikely for such a check to decide a game.
2...Kg4
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+k+0
9+-+-+-+-0
9-zpq+-zP-+0
9+-+-+-+Q0
xiiiiiiiiy
3.Qg2+!
This queen check reveals White’s wonderful idea.
3...Kh4
The first point is that Black’s king cannot escape to the centre: 3...Kf4 4.Qg3+ Kf5 (or 4...Ke4 5.Qg6+),
since White wins the queen with the X-ray check 5.Qg6+.
4.Qh2+
Both queens are now on the second rank, with a white pawn in between...
4...Kg5
4...Kg4 5.f3+.
5.f4+
That’s the idea! With this discovered attack, White wins he queen. A beautiful study!

4
Study by Leonid Kubbel
Smena 1929

Black is quite a few pawns to the good, but his king is in danger.
1.Qe2!
A quiet move that brings White the win. Other attempts lead nowhere, e.g.:
A) 1.Qf1+? Ke4 2.Qc4+ Kf3 3.Qd3+ Kf2, and White doesn’t even have perpetual check;

B) 1.Qg3+? Ke4 2.Qe3+ Kd5 3.Qd3+ Kc6, and the king escapes.

1.Qe2! e4 2.Qe1!
Another silent move. Suddenly, mate is threatened with 3.Qg3#.
2...Ke5
The only move to parry the mate; 2...Kf3 3.Qf1#.
3.d4+!
Here is the discovered check: 1-0

• X-ray check
Sometimes, it pays to chase the enemy king until it ends up on the same file, rank or diagonal as its
queen. This is the signal for an X-ray check. This is also a rewarding theme in studies: the solver can rack
his brains to find the forced move sequence leading to the ultimate goal!

5
Study by Vasile Tacu
L’Échiquier de Paris 1948
 White has only one pawn against Black’s four. But it is, among other things, this pawn on c6 that causes
problems for the black king, and Black’s pawns hinder their own queen in its activity:
1.Qf6!
Not only depriving Black’s queen of a number of squares, but also threatening a sudden mate with
2.Qf8#.
1...Qg8
The king can’t escape with 1...Kb8 because of a beautiful enveloping manoeuvre with the queen:
2.Qb2+ Kc8 3.Qb7+ Kd8 4.Qb8+ Ke7 5.Qxc7+, and White wins the queen.
2.Qf5+
An extremely nasty little check.
2...Kb8
Forced (2...Kd8 3.Qd7#), but now White wins the queen with the same technique as above:
3.Qb1+ Kc8 4.Qb7+ Kd8 5.Qb8+ Ke7 6.Qxg8 1-0

6
Study by Frantisek Prokop
The Chess Amateur 1928

An interesting case in which White has to avoid a couple of pitfalls.

1.Qd6+
Who wouldn’t start with 1.Qc3+ and after 1...Kg5 take Black’s queen with 2.Qxh8, only to discover
that Black is stalemated!

1.Qd6+ Kf7
Importantly, the pawn ending after 1...Kg7 2.Qe5+ Kg8 3.Qxh8+ Kxh8 4.Ke3 wins for White.
Now, 1...Kg5 is a mistake in view of 2.Qf4+ Kh4 3.g5+, and White either wins the queen or gives mate.
2.Qd7+ Kf6
As said, Black cannot go into the pawn ending: 2...Kf8 3.Qd8+ Kg7 4.Qxh8+ Kxh8 5.Ke3.
3.Qd4+
Now that the queen is on d4, White gives a check on the long diagonal anyway.
3...Kg5 4.Qf4+
This is the check that has to do the job. We had already established that 4.Qxh8? is stalemate; but also
4.Qe3+? Kh4 (4...Kf6? 5.Qd4+) 5.Qf2+ Kg5 yields nothing (5...Kh3? 6.Qg3#).
4...Kh4 5.g5+! 1-0
We have already seen this position. As said, not 5.Qg3+? Kg5, and preferably not 5.Qh2+? Kg5,
although White can repair this with 6.Qf4+.

• Checkmate
We have seen many times that you can play for checkmate also in endgames. With a strong piece like
the queen, this is not a problem, of course. One mate pattern is even prettier than the other... it’s difficult to
choose from among the many available examples.

7
Study by Ernest Pogosyants
Shakhmaty v SSSR 1976

White is on the verge of promoting, but will he be winning with a second queen? No; it’s easy to see
that Black will have perpetual check by moving the queen along the fifth rank: 1.a8=Q Qf5+ 2.Kg7 Qg5+
3.Kh7 Qh5+. So, White has to come up with something special:

1.Qc8+! Kxc8
The alternatives lose more quickly:
 1...Kd6 2.Qe6#; 1...Kc6 2.a8=Q+.
2.a8=Q+ Kd7
But also now, Black gets caught in a mating net.
3.Qe8+ Kd6 4.Qe6#
A lovely epaulette mate (the escape squares for the king are blocked by its own pieces).

8
Study by Valeriu Petrovici
Revista de Romana de Sah 1982

With a king on the edge, it’s all hands on deck...

1.Qf5+
The right check to cause the black queen to be in the way of her own king. 1.Qh6+ is no use in view of
1...Kg4;

neither is 1.Qf7+ on account of 1...Kg4.

1.Qf5+ Qg5 2.Qf7+ Kg4 3.f3#

A checkmate with a pawn is always kind of aesthetic.

• Stalemate
If a king is driven into a tight spot, the enemy pieces also often deprive it of most of its escape squares.
Chances are that it will be mated, but sometimes there is salvation! When the king has so few squares, in
some cases, an escape by stalemate is possible.

9
Study by Bruno Schurig
Source unknown

If you are a pawn down, it is always quite a tough job to keep the game within drawing margins. In this
case, White has an elegant way out:
1.Ka1! Kh6
1...Qxc2 is stalemate!
2.Qd2+ Kh7 3.Qc2
And Black doesn’t get any further: ½-½

10
Study by Emanuel Lasker
Lehrbuch des Schachspiels 1926

In this study, the situation seems analogous to the previous one from the 19th century.
However, there is a small difference...
1.Ka3!
The right square for the king.
White would lose after 1.Ka1? due to the tempo move 1...a6!, and White can’t maintain the stalemate
motif.

1.Ka3! Kh6
1...Qxc2 is stalemate too!
2.Qc1+ Kh7 3.Qc2
The same position, and now Black has to play his trump card – the tempo move with the a-pawn:
3...a6
 Now, it seems as if White has landed in a precarious situation. If he can’t maintain the pin along the b1-
h7 diagonal, Black will have excellent winning chances. But White has a curious reply up his sleeve:
4.Qb1!
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpk0
9p+-+-+q+0
9zp-+-+-+p0
9P+-+-+-zP0
9mK-+-+-+-0
9-+-+-+-+0
9+Q+-+-+-0
xiiiiiiiiy
Delightful! With this new stalemate idea, White saves the game!
½-½

• Zugzwang
Zugzwang is always a difficult theme. Therefore, it is good to get a grip on the quintessential feature of
this concept. It all revolves around the ‘obligation to move’. If it is our move, we always want to undertake
all kinds of things, but in some positions there is nothing constructive we can do. Even worse, each move
we play has enormous drawbacks. And so, sometimes it pays to give the move to our opponent, if any move
he makes will cost him the game. There are many beautiful examples on this theme.

11
Study by Ladislav Prokes
Schweitzerische Arbeiter Zeitung 1951

Pawns on the same wing, otherwise the board is empty, and Black’s queen is about to have a field day –
what is there to see?
1.b7+!
A fantastic pawn sacrifice that has to be accepted. Not 1.bxc7+ in view of 1...Kb7, and White has
nothing.

1.b7+! Kxb7
After 1...Kxa5, White promotes and he will be able to stop the checks with two queens.
2.a6+
That was the idea.
2...Kb8 3.Qe8+ Qc8 4.Qb5+ Ka8 5.Qc6+ Kb8
 XIIIIIIIIY
9-mkq+-+-+0
9zp-zp-+-+-0
9P+Q+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+K+-+-+-0
xiiiiiiiiy
All forced, but what now?
6.Kc1
A marvellous quiet move that brings about the zugzwang. This is the only square for the king where it
can’t be checked: 1-0

12
Study by Georg Bernhardt
Deutsche Schachzeitung 1961

We have some experience with kings on the edge. We can cause trouble for them by pushing them
even further into a tight spot:
1.Kc7
Not only parrying 1...Qb7+, but also creating a mate threat on b7. For one thing, this keeps the queen
on a6 under control; it cannot move because of mate.
1...h3
And so, Black is sentenced to making pawn moves. This is the most sensible one, since now White has
only one good answer:
2.Qh1!
The only correct move.
2.Qf3? is strongly met by 2...b5!, threatening 3...Qb6+. So ,White has to give check on the g1-a7
diagonal, but after ...Ka8, he would like to penetrate on the back rank via an open file or diagonal, and that
isn’t possible here.
The direct 2.Qd5? is met with 2...h2! 3.Qg2 and here again, Black has the strong 3...b5!, untangling his
pieces.
2...h2
All other pawn moves are useless:
A) 2...b5 runs into 3.Qg1+ Ka8 4.Qg8+ Ka7 5.Qb8;
B) After 2...f3 3.Qxf3 b5, now the queen enters via the open e-file: 4.Qe3+ Ka8 5.Qe8+ Ka7 6.Qb8#.
3.Qd5! 1-0
Now, yes, and it is over for Black. Let’s have a look at a few more variations.
The following manoeuvre proves why the queen is so strong on d5: 3...b5 (here, 3...h1=Q is no use in
view of 4.Qxh1 b5 5.Qg1+ Ka8 6.Qg8+ Ka7 7.Qb8#) 4.Qd4+ Ka8. From the centre, the queen uses the long
diagonal to reach the back rank: 5.Qh8+ Ka7 6.Qb8#.

• Liquidation
This book has seen many examples of the importance of having an eye open for opportunities to
transfer to an advantageous endgame. In queen endings, such opportunities often exist in the possibility of
trading queens and moving on to a pawn ending. Pawn endings often lead to new queen endings in turn,
which – contrary to the first version – may be winning. Good calculation of such move sequences, but
especially good assessment of the newly arisen ending, are essential tools for winning games. Along the
way, there will be many dangers, temptations, and tricks to avoid. We will see a few of those in the next two
fragments.

13
Study by Milenko Dukic
Schach-Echo 1985

White starts with a very clever check:

1.Qc8+!
Other moves lead to nothing.
1...Kd5
It’s clear that Black loses the queen if his king goes to the b-file: 1...Kb5 2.Qb8++–; 1...Kb4 2.Qb8++–.
2.Qc6+
Here, we see the first function of the e3-pawn. The king has only one square:
2...Ke5 3.Qe6#
And this lovely mate picture, in the middle of the board, underlines the part played by the little pawn!

14
Study by Leonid Kubbel
Shakhmaty v SSSR 1931

White is even two pawns down, but still there is a hidden and highly original path to the win:
1.Qa4+
First, the black king is driven towards the centre. We will discover below why this is done.
1...Kb1 2.Qb3+ Kc1 3.Qb2+!
The correct continuation.
Here, White’s idea reveals itself: he would like to liquidate to a pawn ending with 3.Qb7, but this
wasn’t the right moment to play this move, since the black queen could escape with 3...Qf1 and could then
start a series of checks.
3...Kd1 4.Qb1+
Another check is needed to put Black’s queen ‘under lock and key’. So, again, not 3...Kd1 4.Qb7?, due to
4...Qf1.
4...Ke2 5.Qb7!
XIIIIIIIIY
9-+-+-+-+0
9+Qzp-+p+-0
9q+P+p+-+0
9zp-+-+-+-0
9-+-+-+-+0
9+-mK-+-+-0
9-+P+k+-+0
9+-+-+-+-0
xiiiiiiiiy
Thus, Black’s fate is sealed. He has to liquidate to a lost pawn ending since his queen has no squares: 1-
0

A) to a pawn ending
Liquidation to a pawn ending is, of course, one of the most important weapons a player can have.
15
Study by Bernard Horwitz
The Chess Monthly 1882

The winning path is reasonably obvious, even though it requires quite a bit of dexterousness. White’s
plan is to conquer the a6-pawn with tempo and then force a trade of queens.
1.Qd8+ Kb7 2.Qb6+ Kc8 3.Qxa6+ Kd7
Now, White has to put in some effort to get his queen to b3.
After 3...Kd8, it is rather simple: 4.Qd6+ Kc8 5.Qe6+ Kd8 6.Qb3.
4.Qa7+ Kc8
Moves like 4...Ke6 lose more quickly: 5.Qe3+ Kd6 6.Qb3.
5.Qc5+ Kb8 6.Qf8+
With some enveloping moves, White prepares the intended manoeuvre.
6...Kb7 7.Qf7+ Kb8 8.Qb3 1-0
The remaining pawn ending is hopeless for Black.

16
Leonid Kubbel
Leningradskaya Pravda 1936

In this position, the winning plan isn’t entirely clear at first glance. Therefore, White centralizes his
queen first with two forcing moves, before finding out what’s in it for him:
1.Qh1+
Also attractive is the quiet move 1.Qg1?, to threaten mate, but Black can parry this with 1...Qc7.

1.Qh1+ Kg4 2.Qe4+ Kh3 3.Qe6!

With pawns on c5 and f5, there are liquidations to a pawn ending in the position, and since Black’s
queen isn’t very mobile, this seems to be the obvious way.
3...Qc7
XIIIIIIIIY
9-+-+-+-+0
9+pwq-+-zp-0
9-+p+Qzp-+0
9+-zP-+P+p0
9-+-+-+-zP0
9+-+-+-+k0
9-mK-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
3...Qf8 leads to the same winning idea:
4.Qd6!
With this move, White crowns his concept. Black has a choice between a prospectless pawn ending and
getting mated.
4...Qf7
4...Qxd6 5.cxd6.
5.Qg3#
When both sides have a pawn that is about to promote, sometimes it pays to sacrifice a queen so as to gain
a tempo for a promotion that enables the new queen to enter the struggle, as happens in the following
study:
 B) to a new queen ending
It is a little more complicated if a liquidation leads to a pawn ending, in which a pawn race leads to a
new queen ending. What is the situation after that?

17
Study by Velimir Kalandadze
Chavchavadze Memorial Tourney 1987

White starts with a cute move:

1.Qc1+! Qxc1 2.h8=Q+ Kg1
XIIIIIIIIY
9-+-+-+-wQ0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+K+-0
9-+-+-zp-+0
9+-wq-+-mk-0
xiiiiiiiiy
Actually, White has sacrificed a pawn to move his queen to h8, while luring away the enemy queen.
And now, suddenly it turns out that White has checkmate:
3.Qg7+ Kf1 4.Qg2+ Ke1 5.Qxf2+ Kd1 6.Qe2#
A combinatory motif can’t always be discovered so easily. Therefore, it can be sensible to not ferret out
such a motif endlessly, but first examine a series of forced moves (in your head), after which sometimes the
theme occurs as if by itself! See the next study:

18
Otar Tabidze
A Chess Philosophy of Otar Tabidze 1999

1.Qe1+ Qg1 2.Qe4+ Qg2 3.Qb1+ Qg1 4.Qb7+ Qg2

XIIIIIIIIY
9-+-+-+-+0
9+Q+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
9K+-+-+-+0
9+-+-+-+-0
9-+-+-+qzp0
9+-+-+-+k0
xiiiiiiiiy
With a few ‘long queen moves’, White has succeeded in putting his queen on b7.
Black had no choice, and since his queen is a defender of her king, she can now be lured away with a
gorgeous move:
5.f8=Q!!
Finding such a motif would make any player happy!
5...Qxb7 6.Qf1#
 • Miscellaneous
It’s time to practice all the above-mentioned themes without telling you which theme is applicable.
Unfortunately – or should we say fortunately? – when you play a game there is no-one behind you tapping
you on the shoulder and whispering: ‘Pay attention! You can win by domination!’ This series also contains
one more theme that I have deliberately omitted in the instruction part. The reader can find out for himself
what is this new motif he should pay attention to!

19
Study by Jonathan Tate
Springaren 2011

White’s king is vulnerable, his queen is passive in the corner, and he is a pawn down: it doesn’t look
good. But the darkest hour is just before the dawn:
1.Qe2+! Qxe2
Stalemate!

20
Study by B.Corcos
New Statesman (and Nation) 1953

With a pawn on the seventh rank, often tactical shots are available:
1.Qxb7+! Kxb7
And now, an underpromotion:
2.d8=N+
And the resulting endgame is an easy win.

21
Study by Paul Heuäcker
Schach Beobachter Bayreuth 1937

Sometimes, you can play for mate on an ‘empty’ board. Other tactical themes, like the X-ray check, can
help.
1.Qc8+
The right check.
1...Kd5
1...Kb6 is strongly met by 2.Qb8+ Kc5 (the a-file is always taboo: 2...Ka6 3.Qa8+) 3.Qb4+ Kd5
4.Qd4#, and Black is mated in the middle of the board.
Or 1...Kb5 2.Qc6+ – here, we see one of the functions of the c3-pawn, as now the king has to go to the
a-file: 2...Ka5 3.Qa8+.
2.Qc6+ Ke5 3.Qe6#
Another gorgeous mate picture: the flight square f4 has been blocked by his own pawn.

22
Study by Tigran Gorgiev
Shakhmaty (Riga) 1928

We have seen before that pieces on the edge may get stuck in a jam. Here, Black loses the queen in an
artistic way:
1.g4!
Black’s queen has extremely few squares, and this will be his downfall.
1...Qxg4
 The queen is out, but immediately lands on a fatal diagonal.
On 1...Qe8, White constructs a charming mate picture with 2.Qc7#.
2.Qf8+ Kd7
XIIIIIIIIY
9-mK-+-wQ-+0
9+-+k+-+-0
9-+-+-+-+0
9+-+-+-zp-0
9-+-+-+qzp0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
3.Qc8+ 1-0

23
Study by Frantisek Prokop
Ceskoslovensky Sach 1938

A somewhat strange position in which Black has no fewer than four extra pawns, though two sets of
them are doubled. White wins thanks to the poor placement of Black’s king, and, presently, also of the
enemy queen:
1.Qh1+ Qb1 2.Qh8+ b2
XIIIIIIIIY
9-+-+-+-wQ0
9+-+p+-+-0
9p+-zp-+-+0
9zp-+-+-+-0
9-+-+-+-+0
9mK-+-+-+-0
9-zp-+P+-+0
9mkq+-+-+-0
xiiiiiiiiy
3.Qg8!
A quiet move, signing Black’s death warrant! The queen has to stay on b1 to keep the a2-square
covered, and she can’t give check either, since the d3-square is controlled by White’s pawn. Now, White
only has to wait for Black’s pawn moves to run out. The composer stopped here, but I will give a few more
moves: 3...a4 4.Qd5 a5 5.Qc4 d5 6.Qxd5 d6 7.Qg8 (here, 7.e4 wins too) 7...d5 8.Qxd5, and now it is really
over!

24
Study by Alexey Troitzky
500 Endspielstudien 1924

Sometimes, there are pawns in the way of a king that is being hunted and is desperately searching for
an escape.
1.Qf6+
 Also attractive is 1.Qe6+, but then the black monarch escapes to safety: 1...Kf4 2.Qf6+ Kg4 3.Qe6+
(not 3.Qxg6+? Qg5!, and it’s Black who wins), and after both 3...Kf4 and 3...Kh4, the game ends in a draw.

1.Qf6+ Ke4
The unenviable d4-pawn is fatal for Black after 1...Kd5 2.Qe6+ Kc5 3.Qc8+, winning the queen.
2.Qxg6+ Kd5
Retreating doesn’t help either: 2...Ke5 3.Qe6+ Kf4 4.Qh6+ – this theme always plays a role, of course.
3.Qe6+ Kc5 4.Qc8+
And the curtain falls.

25
Study by Alexander Kuznetsov
Shakhmaty (Riga) 1967

We’ve seen that if a player’s king and queen are squeezed into a corner, he is likely to suffer an
abominable fate. Fortunately, there are exceptions. Here, White saves himself from a precarious situation
with a pretty motif:
1.Qf6+! Qxf6
Neither does Black win after 1...Kxf6 2.h8=Q+ Kg6 3.Qh5+=.
2.h8=N+! Kh6 3.Nf7+ Kg6 4.Nh8+ ½-½
Black’s king cannot walk out of the checks. A curious save!

26
Study by Pietro Rossi
L’Italia Scacchistica 1975

White is ready for a devastating discovered check by his king, but unfortunately there is one problem
that needs solving: his queen is hanging! And since White can’t take Black’s queen (1.Qxh1?) on pain of
stalemate, he has to search for new ideas to win this battle.
1.Qb8!
The only move that wins, which is far from easy to see.
A) 1.Qxh1? stalemate!;

B) Also, 1.Qc8? Qh3 2.Qd8 Qe6+ 3.Qe7 Qxe7+ 4.Kxe7;

C) And 1.Qd8? Qb7 does not work: Black draws in all cases.

1.Qb8! Qh2
Black has to keep attacking White’s queen due to the stalemate motif.
The 1...Qe4+?? check would lose immediately to 2.Kd7+.
2.Qd8!
Again, the only move that wins for White.
Once more, not 2.Qc8? in view of 2...Qh3, with another draw!
2...Qh4
 XIIIIIIIIY
9-+-wQK+k+0
9+-+-+-zp-0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-wq0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
There were no useful alternatives. But now we see White’s hidden, brilliant idea:
3.Qd5+! Kh8 4.Qa8!
The queen returns to the corner where it started, but the difference is that now Black’s king is on h8
instead of g8. This means that the stalemate is gone, and so Black cannot prevent a discovered check by
White’s king: 1-0

27
Study by Gennady Zgerskye
(after Prokop)
Sachova Skladba 2006

When we try to find our bearings in a position, we will be inclined to follow our first hunch. Often, this
is also the right move, but sometimes there is a snake in the grass that has to be tracked down before we
can draw the right conclusion.
1.e5+!
Only with this pawn sacrifice can White draw in the net.
Black has a nasty swindle after 1.Qb4+?: 1...Ke5! 2.Qd4+ (2.Qxf8 is stalemate!) 2...Kf4 3.Qf2+ Ke5,
and White can make no progress since 4.Qxf8 leads to stalemate again.

1.e5+! Kxe5 2.Qe4+

First, look for moves to which your opponent only has one answer, and calculate those. With this in
mind, 2.Qb5+ also came into consideration, but then Black escapes with 2...Kf4.
2...Kd6 3.Qb4+
Finally, White can win the queen.
3...Ke5 4.Qxf8 1-0
There is no more stalemate.

28
Study by A.P.Grin
Shakhmaty v SSSR 1976

We saw in an earlier fragment that it sometimes pays to sacrifice your queen to get a new queen by
promotion. But it takes some care:
1.Qa5+!
Not 1.a8=Q? in view of 1...Qc1+, collecting the Qg5.

1.Qa5+! Kxa5 2.a8=R+!

The only winning move! Here, queening would lead to a surprising stalemate after 2.a8=Q+? Kb4
3.Qb8+ Kc3 4.Qxb1.
2...Kb4
 2...Kb6 3.Rb8+.
3.Rb8+ Kc3 4.Rxb1 1-0
Now, there is no stalemate.

29
Study by Claude Sansom
Endgame Studies by Sansom 1990

At first glance, it looks as if White plays to win the black queen with an X-ray check:
1.Qh8+
Nothing is gained by 1.Qg7+ due to 1...Kf5.
XIIIIIIIIY
9-+-+-+-wQ0
9+-zp-+-+-0
9-zpP+p+-+0
9+q+-mk-+-0
9-zp-+-+-+0
9+P+P+-+-0
9-+-+-+K+0
9+-+-+-+-0
xiiiiiiiiy
Now, Black has many king moves, and all of them have their drawbacks.
When giving checks, White has to consider that he is allowed to use the queen vertically as well as
horizontally.

1.Qh8+ Kf4
After 1...Kd6 2.Qd8+, Black again has two options, both of which are losing:
A) 2...Kc5 3.d4+ Kxc6 4.Qe8+; 2...Ke5 3.Qg5+ – this horizontal check wins the queen;
B) 2...Kxc6 3.Qe8+ Kc5 4.d4+ and here too, the queen is lost.
If 1...Kf5 2.Qh5+.
1...Kd5 transposes to previous lines after 2.Qd8+ Kxc6 3.Qe8+ Kc5 4.d4+.
2.Qh4+ Ke3
Of course, 2...Ke5 (or 2...Kf5) is again met by 3.Qg5+ (or 3.Qh5+).
3.Qf2+
It is this check that forces Black to put his king on a fatal diagonal:
3...Kxd3 4.Qf1+ 1-0

30
Study by Emil von Feyerfeil
Österreichische Schachzeitung 1872

Black has many pawns, which are in the way of his queen. White hopes to exploit this.
1.Kg6 Qc6+ 2.Kf7
Since mate is threatened, Black has only one defence:
2...Kh7!
2...Qh6 fails to 3.Qe8+ Kh7 4.Qg8#.
3.Qh4+
Now, Black’s queen is forced to the edge anyway.
3...Qh6
 XIIIIIIIIY
9-+-+-+-+0
9+-zp-+K+k0
9-+-+-+-wq0
9+-+p+-+-0
9-+-+p+-wQ0
9+p+-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
4.Qg4!
And then, of course, White can work with domination! Black cannot prevent mate: 1-0

31
Study by Alexey Sochnev
Postsjakk 1986

How often have we seen king and queen weaving a mating net? But this fragment is a true piece of art:
1.Kf7
White threatens both 2.Qf5+ and 2.Qh3+, followed by mate, and 2.Qg8+ too is a tough threat since
after 2...Kh6 3.Qxg7+, White will have a support point on g4 for the queen and can therefore chase the king
back to h6 and give mate with Qg6#.
So, there are certain squares Black has to cover with his queen.
1...Qg3
The most tenacious defence, but this move doesn’t ‘feel’ right...
1...Qh4 fails to 2.Qg8+ Kh6 3.Qxg7+ Kh5 4.Qg6#; and fleeing with 1...Kh6 doesn’t work either:
2.Qh3+ Kg5 3.Qg4+ Kh6 4.Qg6, also ending in mate.
2.Qg8+ Kh6 3.Qh8+ Kg5 4.Qxg7+
XIIIIIIIIY
9-+-+-+-+0
9+-+-+KwQ-0
9-+-+-+-+0
9+-+-+-mk-0
9-+-+-+-+0
9+-+-zpPwq-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
4...Kf4
Or 4...Kh4 5.Qh6#.
5.Qf6#
A quite unique epaulette mate. The escape squares e3 and g3 have been blocked by Black’s own pieces,
while the miraculous pawn on f3 covers two more flight squares.
Mr Prokop has composed various studies that look alike but always have different solutions, often highly
original ones. See the following example:
32
Study by Frantisek Prokop
The Chess Amateur 1928

The winning path here is a culmination of ideas the Czech composer had found in his earlier
investigations.
1.Qb6+
Unfortunately, there is a ‘cook’ here: as was discovered later, 1.e5+! also wins.

1.Qb6+ Kd7 2.Qb7+ Kd6 3.Qb4+

We know these tricks from Prokop’s earlier compositions.
3...Ke5 4.Qd4+
White can’t win the queen with 4.Qxf8? due to stalemate.
4...Kf4
The king has to go to the f-file.
5.Qf2+
Here, 5.e5+ is absolutely nothing for White since the board is large enough for the black king to escape.
5...Ke5 6.Qh2+!
Since now the kingside can be used to the full, White seizes this opportunity with both hands.
It would be stalemate again after 6.Qxf8?.
6...Qf4
Obviously, Black would lose the pawn ending resulting after 6...Kf6 7.Qf4+ Ke7 8.Qxf8+ Kxf8 9.Kc3,
and so the text move is forced.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+p+-+0
9+-+-mk-+-0
9-+-+Pwq-+0
9+p+K+-+-0
9-zP-+-+-wQ0
9+-+-+-+-0
xiiiiiiiiy
7.Qh8+!
Now the edges of the board are put to full use.
7...Kd6
The beauty of it is that 7...Qf6 fails to 8.Qb8#, a splendid mate picture!
8.Qb8+!
And now it’s an X-ray check that forces Black to resign.
 Part IV
Curious endgames
Chapter 22
Introduction
The link of theory with practice... how often do we hear chess players sigh that it’s not easy to turn
something you have learned into success on the board. For your author too, chess life hasn’t been a bed of
roses. It is only after many ups and downs that you will finally succeed now and then in applying a certain
theme, motif or concept into practice in the correct way. It’ll cost you blood, sweat and tears before you get
to the point when you can say: ‘Yes, I’ve managed to convert years of hard labour into success!’
What many people, including myself, tend to forget, is that everything you do in chess will bring you
success at unexpected moments. Apart from the fact that there are whole crowds of people who derive
pleasure from our royal game in one way or another, even for success-starved tournament sharks it’s
always useful to immerse themselves in endgames that may never even come on the board.
Think of the example of the rook ending that transfers into the theoretical queen vs rook ending. If
you’ve never had a look at this, because you never saw the sense of it, it may well be that you avoid
liquidations into such an endgame even if it is the most convincing way to win. But if you store the
elementary principles in your memory, you will play the rook ending with great self-confidence when the
time comes. As always, fear is a bad counsellor.

268
Herman Grooten 2387
Semen Dvoirys 2553
Hoogeveen 2006 (8)
XIIIIIIIIY
9-+-+-mk-+0
9+-+-+p+p0
9-+-+-zp-+0
9+-+-+-zp-0
9-mK-zP-+P+0
9+-+r+-+P0
9P+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
White to move
Just before the 40th move, I have acquired a winning position, and the win comes closer with the following
move:
40.a4!
Since my rook is behind the passed pawn, this pawn now moves forward as quickly as possible.
40...Rxd4+ 41.Kc5 Rd2 42.a5 Rc2+ 43.Kd4 Rd2+ 44.Ke3 Rd7 45.a6 Ra7
 XIIIIIIIIY
9-+-+-mk-+0
9tr-+-+p+p0
9P+-+-zp-+0
9+-+-+-zp-0
9-+-+-+P+0
9+-+-mK-+P0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
This position must be winning for White. Black’s rook is extremely passive, and White has an active king.
46.Ke4?
Too cautious, and simply not offering the best chances. It is questionable whether White is still winning
after this, although he still has a chance.
I could probably have forced the win with 46.Kd4. I didn’t play this because I saw the breakthrough ...f6-f5
coming. A little calculation shows that this breakthrough does not give Black enough counterplay: 46...f5
47.gxf5 Ke7 (47...h5
XIIIIIIIIY
9-+-+-mk-+0
9tr-+-+p+-0
9P+-+-+-+0
9+-+-+Pzpp0
9-+-mK-+-+0
9+-+-+-+P0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
analysis diagram
48.f6! Ke8 49.Kc5 g4 50.hxg4 hxg4 51.Kb6 Ra8 52.Kb7 Rd8 53.a7+–) 48.Kc5 Ra8 49.Kb6 h5 50.a7 Kf6
XIIIIIIIIY
9r+-+-+-+0
9zP-+-+p+-0
9-mK-+-mk-+0
9+-+-+Pzpp0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
analysis diagram
51.Ra5! g4 52.h4! g3 53.Ra3 (53.Kb7 g2 54.Ra1 Re8 55.Rg1 Kxf5 56.Rxg2 Re7+ 57.Ka6 Re6+ 58.Ka5 Re5+
59.Kb4 Re4+ 60.Kb3 Re3+ 61.Kb2 Re8 62.Rg5+ Ke4 63.Rxh5 also wins, but this is all rather complicated).
46...Kg7 47.Kd4
Anyway, but now Black has improved his position.
47...f5 48.gxf5 Kf6
 XIIIIIIIIY
9-+-+-+-+0
9tr-+-+p+p0
9P+-+-mk-+0
9+-+-+Pzp-0
9-+-mK-+-+0
9+-+-+-+P0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
49.Ra5?
Another ‘half-move’. The f5-pawn is totally uninteresting; White simply has to shepherd his pawn to the
other side as quickly as possible, forcing Black to sacrifice his rook. Then, the question is whether White
will be back in time to stop Black’s last pawn.
Obviously, 49.Kc5! was the right continuation. Actually, the win is quite smooth here, with only logical
moves. I will give a stereotypical example: 49...Kxf5 50.Kb6 Ra8 51.Kb7 Rg8 52.a7 h5 53.a8=Q Rxa8
54.Rxa8.
XIIIIIIIIY
9R+-+-+-+0
9+K+-+p+-0
9-+-+-+-+0
9+-+-+kzpp0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This is typically the kind of ending in which the rook is superior to Black’s pawns. The rook will be ideally
placed behind the pawns, while White’s king will return to the ‘new’ battle scene as quickly as possible.
What we don’t really see very clearly here is that the white pawn on h3 may play an important role:
54...Ke4 (Black is already trying to keep the enemy king at bay; 54...g4? 55.h4! – maintaining this pawn is
crucial! – 55...g3 56.Rg8 Kf4 57.Kc6 f5 58.Kd5 Kf3 59.Rg5 f4 60.Ke5 g2 61.Rxg2 Kxg2 62.Kxf4 and White
wins) 55.Rh8 h4 56.Kc6 f5 57.Kc5
XIIIIIIIIY
9-+-+-+-tR0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+pzp-0
9-+-+k+-zp0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
A) 57...f4 58.Kc4 Ke3 59.Kc3 f3.
XIIIIIIIIY
9-+-+-+-tR0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-zp-0
9-+-+-+-zp0
9+-mK-mkp+P0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
We know this type of position from the rook vs pawn chapter. Without the other three pawns, White’s win
is elementary. It turns out that the winning method for White is similar to the one in the above endgame:
60.Re8+ Kf2 61.Kd2 Kg3 62.Rg8 (62.Ke3 is also possible) 62...f2 63.Ke2 Kxh3 64.Rxg5, and the job is
done;
B) Black can also try 57...g4, after which White has only one good move: 58.Rxh4!. It is very important to
keep the two pawns, h3 and f5, on the board for a little longer (the position is a draw after 58.hxg4? fxg4
59.Rxh4 Kf3, and Black’s g-pawn cannot be conquered). After 58...Kf3,
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-mK-+p+-0
9-+-+-+ptR0
9+-+-+k+P0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
White now replies 59.Kd4!, and here the difference becomes clear. Now, 59...g3 (59...Kg3 doesn’t save
Black because of 60.hxg4! Kxh4 61.gxf5+–) is no use to Black since after 60.Rh8 g2 61.Rg8, the white pawn
is still on the board. The rook is given up for Black’s g-pawn, White’s king takes care of Black’s f-pawn, and
White’s h-pawn marches on victoriously! Here is an illustrative line: 61...f4 62.h4 Kf2 63.Ke4 f3 64.h5
g1=Q 65.Rxg1 Kxg1 66.Kxf3.
49...h5
 XIIIIIIIIY
9-+-+-+-+0
9tr-+-+p+-0
9P+-+-mk-+0
9tR-+-+Pzpp0
9-+-mK-+-+0
9+-+-+-+P0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
50.Ke4
Perhaps, during the game I was under the impression that my king could now ‘safely’ cross the fifth rank
with 50.Kc5 since 50...Kxf5?? would be answered with the discovered check 51.Kb6+, winning the rook.
But eventually I rejected this line anyway, because it is a draw after 50...Ra8!; for example, 51.Kb6 g4 52.h4
g3 53.a7 g2 54.Ra1 Kxf5,
XIIIIIIIIY
9r+-+-+-+0
9zP-+-+p+-0
9-mK-+-+-+0
9+-+-+k+p0
9-+-+-+-zP0
9+-+-+-+-0
9-+-+-+p+0
9tR-+-+-+-0
xiiiiiiiiy
analysis diagram
and White can gain nothing here, not even after 55.Rg1 Kg4 56.Rxg2+ Kxh4 57.Kb7 Rxa7+ 58.Kxa7 Kh3
59.Rg8.
XIIIIIIIIY
9-+-+-+R+0
9mK-+-+p+-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Of course, I hadn’t seen this position in my calculations, but this is a draw, although Black can still go badly
wrong: 59...f5! (the f-pawn has to be pushed right away, before the king can enter the square of the pawn; if
59...h4??, 60.Kb6 f5 61.Kc5, and White wins) 60.Kb6 f4 61.Kc5 f3 62.Kd4 f2 63.Rf8 Kg2.
50...Re7+ 51.Kf3 Ra7 52.Kg3 Ke7
 XIIIIIIIIY
9-+-+-+-+0
9tr-+-mkp+-0
9P+-+-+-+0
9tR-+-+Pzpp0
9-+-+-+-+0
9+-+-+-mKP0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Here, I sat and looked at my position with great frustration. How clumsy had I handled it! Why hadn’t I
seized my chances? My king on g3 is much too passive for me to have any serious winning chances.
53.h4 f6 54.Ra1 Kd7 55.Kf3 g4+ 56.Ke4 Kc6
XIIIIIIIIY
9-+-+-+-+0
9tr-+-+-+-0
9P+k+-zp-+0
9+-+-+P+p0
9-+-+K+pzP0
9+-+-+-+-0
9-+-+-+-+0
9tR-+-+-+-0
xiiiiiiiiy
57.Rc1+
An attempt to confuse him.
57...Kd6!
After 57...Kb6?! 58.Kd5 Rxa6 59.Ke6 Kb5+ 60.Kf7, White gets slight winning chances, but it doesn’t seem
sufficient after a move like 60...g3 or 60...Rc6.
58.Rd1+ Kc6 59.Rc1+ Kd6 60.Ra1 Kc6 61.Kd4 Kb6 62.Kd5 Rxa6 63.Re1 Kc7
XIIIIIIIIY
9-+-+-+-+0
9+-mk-+-+-0
9r+-+-zp-+0
9+-+K+P+p0
9-+-+-+pzP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-tR-+-0
xiiiiiiiiy
With a tempo less, this is nothing for White. Even worse: he is a pawn down now...
64.Re7+ Kd8 65.Rh7 g3 66.Rxh5 g2 67.Rh8+ Ke7 68.Rg8 Ra5+ ½-½
I found out that tournament play is a tough learning school when I had the honour of twice crossing swords
with a former World Champion. In 1992, the organizers of the famous Interpolis Tournament in Tilburg
decided to introduce a new formula. Instead of the traditional twelve-player round robin, the knock-out
formula was tried. Apart from the usual company of chess stars, Dutch (sub-)top players also entered the
arena. The minimum rating was 2400, and with my 2405 Elo, I was paired with Vasily Smyslov, one of the
most amiable World Champions ever. In the first game, in which I was Black, the opening didn’t develop in
an unfavourable way for me. A Catalan transposed to a Benoni, and in the middlegame I was definitely not
worse. But after a long struggle, my usual optimism and hopes for a good result were reduced to nothing.
For the comments to this fragment, I have made grateful use of GM Igor Stohl’s analysis in ChessBase
MegaBase.

269
Vasily Smyslov 2530
Herman Grooten 2405
Tilburg 1992 (1)
XIIIIIIIIY
9-+-+-+-+0
9+-+-+pmk-0
9-+-zp-+q+0
9zP-zpP+-+l0
9-+-+-+-+0
9+-+-zP-+-0
9-+-+-zPL+0
9+-+-+QmK-0
xiiiiiiiiy
Black to move
My assumption that Smyslov, who was quite advanced in age already, would perhaps not be able to hold
out well in such a long game, was totally misplaced. The Maestro ‘sat me out’ in a long endgame he treated
expertly; an endgame lesson I wasn’t going to forget in a hurry...
After a fierce middlegame fight, a horribly difficult endgame is now on the board. At first sight, the position
doesn’t look so bad for Black, but it is worse than it seems since White’s a-pawn is an important factor.
40...Bf3 41.a6 Bxd5 42.a7
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+pmk-0
9-+-zp-+q+0
9+-zpl+-+-0
9-+-+-+-+0
9+-+-zP-+-0
9-+-+-zPL+0
9+-+-+QmK-0
xiiiiiiiiy
42...Ba8?!
During the game, I thought I didn’t have much choice here, but in fact the text move is a huge inaccuracy.
Probably, I should have used my own passed pawn. After 42...c4! 43.f3 c3 44.e4 Ba8 45.Qc4 Qf6, I haven’t
been able to discover any serious winning chances for White.
43.f3 Kf8 44.Qa1
44.Qa6 Ke7.
44...Ke7 45.e4!?
 XIIIIIIIIY
9l+-+-+-+0
9zP-+-mkp+-0
9-+-zp-+q+0
9+-zp-+-+-0
9-+-+P+-+0
9+-+-+P+-0
9-+-+-+L+0
9wQ-+-+-mK-0
xiiiiiiiiy
White has two aims:
1) He tries to play his queen to b8;
2) By creating a foothold on d5, he has a very strong plan to break the blockade by transferring his bishop
to that square at one point.
45...Qh6!?
Black attempts to disturb White’s king.
46.Qc3 f5?
A violent try to attack the foothold on e4, and also to activate my bishop. However, this turns out to be a
‘slip’.
A logical defensive try was 46...Kd7, but this is also doomed to fail since Black’s king will be very vulnerable
after, for example, 47.Qb3 or Bf1.
A good continuation was 46...Qf4.
47.exf5! Qf6 48.Qe3+
XIIIIIIIIY
9l+-+-+-+0
9zP-+-mk-+-0
9-+-zp-wq-+0
9+-zp-+P+-0
9-+-+-+-+0
9+-+-wQP+-0
9-+-+-+L+0
9+-+-+-mK-0
xiiiiiiiiy
48...Kd7?
This move deserves disapproval. Better, and also much more logical, was to steal away with the king over
the dark squares: 48...Kd8!, and for the present, a clear win for White is not in sight.
49.Kh2
Here, Stohl gave an interesting alternative: 49.f4!?, with the idea to meet 49...d5 with the strong 50.Qe5. But
Black calmly replies 49...Qg7! (49...Qa1+ 50.Kh2 Bxg2 51.Kxg2 Qxa7 52.Qe6+ Kc6 53.f6+–; 49...d5
50.Qe5), after which things are anything but clear. It looks as if White should have played 49.Bh3!, after
which he gets excellent winning chances.
 XIIIIIIIIY
9l+-+-+-+0
9zP-+k+-+-0
9-+-zp-wq-+0
9+-zp-+P+-0
9-+-+-+-+0
9+-+-wQP+-0
9-+-+-+LmK0
9+-+-+-+-0
xiiiiiiiiy
He wants to protect the f5-pawn with a tactic, but this may not have been the best move.
49...c4?!
Black also wants to put one of his passed pawns to some use, but this ugly move gives away some
important dark squares.
49...Qxf5?? 50.Bh3+–.
I could also have selected a dark square for the king here, but this leads to some problems after 49...Kc7
50.f4.
Perhaps, 49...Qh4+ 50.Bh3 Kc7 was an idea. The fact is that White still has all the chances.
50.f4!
Of course, Smyslov plays his trump card now.
50...d5
Giving up even more dark squares, and now the game cannot be held anymore.
Also if 50...Qh4+, 51.Kg1 Qg4 52.Qe6+ Kc7 (52...Kd8 53.Qxd6+ Kc8 54.Qf8++–) 53.Qe7+ Kb6 54.Qd8+
would have decided the battle.
51.Bh3
XIIIIIIIIY
9l+-+-+-+0
9zP-+k+-+-0
9-+-+-wq-+0
9+-+p+P+-0
9-+p+-zP-+0
9+-+-wQ-+L0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
51...Kc7
51...d4 is too slow as well, as is shown by 52.Qe5! Qxe5 53.fxe5 d3 54.f6+ Ke8 (54...Kc7 55.f7) 55.e6 c3
56.Bg4 c2 57.Bh5+.
Now that the pawn has been pushed forward, Black should have been consistent and pushed his c-pawn.
However, after 51...c3, he also draws a blank after 52.Qc5!, putting an end to all Black’s illusions.
52.Qc5+
Maybe not the strongest, but more than sufficient to win.
Smyslov avoided another obstacle by not playing 52.Qe5+? since then Black would have drawn even with
52...Qxe5 53.fxe5 c3, as correctly indicated by Stohl.
The engine recommends 52.Qe8.
52...Qc6
 XIIIIIIIIY
9l+-+-+-+0
9zP-mk-+-+-0
9-+q+-+-+0
9+-wQp+P+-0
9-+p+-zP-+0
9+-+-+-+L0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
53.Qxc6+
True to his style, Smyslov opts for the systematic solution.
Curiously, the sharp 53.Qe7+! would have been much more convincing, which is illustrated by the
following variation: 53...Kb6 54.f6, and the pawn promotes to a queen unhindered: 54...d4 55.Qd8+ Kxa7
56.Qa5+ Kb8 57.Qe5+ Ka7 58.f7 Qh1+ 59.Kg3 Qg1+ 60.Kh4 Qf2+ 61.Kg5 Qg3+ 62.Bg4, and it’s over.
53...Bxc6 54.f6!
XIIIIIIIIY
9-+-+-+-+0
9zP-mk-+-+-0
9-+l+-zP-+0
9+-+p+-+-0
9-+p+-zP-+0
9+-+-+-+L0
9-+-+-+-mK0
9+-+-+-+-0
xiiiiiiiiy
The two distant passed pawns taken together are an irresistible force. Every move loses for Black now.
54...Kd6
Also after 54...c3 55.Bf5, White has everything sorted out.
55.Kg3 c3 56.Bf5 c2
56...d4 is strongly met by 57.Kf2 Bd5 58.Bg6, and Black has no chance against White’s two distant passed
pawns.
57.Bxc2 Ke6
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+-0
9-+l+kzP-+0
9+-+p+-+-0
9-+-+-zP-+0
9+-+-+-mK-0
9-+L+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Thus, Black eliminates a passed pawn at least. However, the other f-pawn will be fatal for him. Now,
Smyslov finishes off very neatly:
58.Kf2 Kxf6 59.Ke3 Ke6 60.Kd4 Kd6 61.Bb3 Bb7 62.f5 Ba8 63.f6
And here I shook the hand of my illustrious opponent, since after 63.f6 63...Ke6, there would follow 64.f7
(or, more spectacularly, 64.Bxd5+ Bxd5 65.f7) 64...Kxf7 65.Bxd5++–.
After the game, I was bold enough to ask him if he wanted to analyse. But no, he was pretty tired, and he
brushed my suggestion aside with a friendly gesture. Of course, I understood; I would have to try to win the
second game. This hope too turned out to be illusory. I didn’t obtain any advantage; in the middlegame, he
was in the driver’s seat, and the more pieces that got exchanged, the more miserable my prospects became.
At a certain moment, I saw nothing better than to go for a rook ending that looked highly precarious – and
this against a great endgame specialist. However, fortunately for me, the great man made a mistake on
move 48. He allowed a hidden trick by which I managed to escape with a half point. He probably couldn’t
care less, since he had advanced to the next round!

270
Herman Grooten 2405
Vasily Smyslov 2530
Tilburg 1992 (2)
XIIIIIIIIY
9-+-+-+r+0
9+-+-+-+-0
9R+-+-+-+0
9+pzp-+K+-0
9-+-+-+P+0
9zP-+-+-+-0
9-mk-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
49...Rb8
The situation looks quite miserable for White, but after trading off one of Black’s queenside pawns, he can
sacrifice his rook for the other one and force the draw with the distant g-pawn.
50.a4! c4
50...b4 51.Rc6, and Black can’t protect the c5-pawn: 51...b3 (51...Ka3 52.Rxc5 b3 53.Rc3 Ka2 54.g5 b2
55.Rc2, with a draw) 52.Rxc5 Ka1 53.Rc1+ Ka2 54.g5 b2 55.Rc2 is a draw as well.
51.axb5 c3 52.Rc6 c2 53.g5 c1=Q 54.Rxc1 Kxc1 55.g6 Rxb5+ ½-½
Still, this second little endgame felt like a small success. Perhaps it proved that rook endgames weren’t
wasted on me after all? For a long time, I’d been feeling that this was the sum of many years’ study, analysis,
and exchanges of ideas with strong(er) players. Often, I’d felt obliged to conclude that my endgame
treatment left room for improvement.
However, in 1984 there was a moment in a tournament in Ramsgate, England, when I could say to myself:
‘Finally, finally I’ve managed to put my knowledge and skills into practice to win a drawish endgame
against a strong player.’
During winter, Ramsgate was a pretty desolate seaside resort. Therefore, a certain hotel owner, also a chess
player, had decided to upgrade his hotel to turn it into the site of a chess festival. At the same time, he had
organized a language course for students from Southern Europe who could work in the hotel service in
return. This was an ideal way to learn the language, while, at the same time, the hotel was chock full! The
tournament was part of a Grand Prix series, and nearly the entire top echelon of English players was
present. The rounds were played in the afternoon, after which the game would be adjourned if it wasn’t
finished. The tables with chessboards were transformed into dinner tables, and the chessboards
disappeared in the wings. I remember well that the Dutch master Rob Hartoch had to seal a move in a
virtually winning position. He was the only player sitting there at the chessboard, while all around him the
tables had been set for dinner. Disturbed by the noise, he was unable to think, so after some time he
decided to find his opponent and offer him a draw...
At a certain point, I too had an adjourned game, and I had learned from Hartoch’s experience: I quickly put
a move in the envelope and tried feverishly to unravel the secrets of the position during dinner with a
pocket chess set.
My opponent was the highly talented and original player Julian Hodgson, who later won the Championship
of Great Britain four times. I had some extra material, but the position seemed rather drawish. In the dining
room, I could see from a distance that Hodgson wasn’t taking the situation too seriously and was making
merry at the bar. Was he being too light-hearted about it? When after dinner the food boards had been
replaced with chessboards again, the game was resumed.
I had found a few hidden ideas and I used them to set my opponent thinking. At least that was something! I
kept presenting him with new problems, landing him in time-trouble as well. Was it really going to
happen...?

271
Julian Hodgson
Herman Grooten 2305
Ramsgate 1984 (5)
XIIIIIIIIY
9-+-+r+k+0
9+-+-+p+p0
9-+-tR-+p+0
9+-zp-+-+-0
9p+p+-+-+0
9zP-+-+-+P0
9-zP-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
Black to move
Black is ‘virtually’ two pawns up, but since he has a doubled pawn, the win is still far away. In fact, if White
manages to exchange all his queenside pawns for Black’s, the latter’s winning chances will be minimal with
3 vs 2 on one wing. I had to play one more move before the time control:
40...c3!?
With hindsight, this is not the strongest continuation, but there are enough practical chances left to play for
a win.
The most accurate would have been 40...Rb8, a move I had considered: 41.Ra6 (White’s only chance is to
eliminate Black’s potentially dangerous a-pawn; 41.Rc6? Rxb2 42.Rxc5 Rb3 43.Rxc4 Rxa3, and with two
extra pawns, the job should be doable; passive defence with 41.Rd2 loses to 41...c3 42.bxc3 Rb3 43.c4 Rc3,
and Black will be two pawns to the good) 41...Rxb2 42.Rxa4
 XIIIIIIIIY
9-+-+-+k+0
9+-+-+p+p0
9-+-+-+p+0
9+-zp-+-+-0
9R+p+-+-+0
9zP-+-+-+P0
9-tr-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
42...Rc2! (the only chance to achieve something; not so good is 42...c3 in view of 43.Rc4 c2 44.Rxc5 Kf8
45.a4 Ke7 46.a5 Ra2 47.a6, and White’s a-pawn will be traded for Black’s c-pawn, after which the ending
with 3 vs 2 on one wing is a draw; these are ideas I had seen in my analysis) 43.Ra5 Kf8! (the king has to
enter the stage quickly) 44.Rxc5 Ke7 45.a4 Kd6 46.Rc8 Kd5.
XIIIIIIIIY
9-+R+-+-+0
9+-+-+p+p0
9-+-+-+p+0
9+-+k+-+-0
9P+p+-+-+0
9+-+-+-+P0
9-+r+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Now that the king has taken over the protection of the c4-pawn, Black can deploy his rook elsewhere on the
board: 47.a5 Ra2 48.Ra8 Kc5 49.a6 Ra4 50.Kg3 Kb6.
XIIIIIIIIY
9R+-+-+-+0
9+-+-+p+p0
9Pmk-+-+p+0
9+-+-+-+-0
9r+p+-+-+0
9+-+-+-mKP0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Black’s king has crowned its long journey by controlling White’s a-pawn now. White’s king is too late to put
a spoke in the wheel. The winning line may run as follows: 51.Rb8+ Kxa6 52.Kf4 c3+ 53.Ke3 Rc4 54.Rb1
Ka5 55.Kd3 c2 56.Rc1 Kb4, and Black wins.

40...c3!? 41.bxc3 Re3

Strangely, the engine thinks that 41...Kg7 is a slightly better idea, with the following possible continuation:
42.Rd5 Re1 43.Rxc5 Ra1 44.Ra5 Rxa3 45.c4 Kf6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+p0
9-+-+-mkp+0
9tR-+-+-+-0
9p+P+-+-+0
9tr-+-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Black’s king fully participates in the battle again; on the other hand, White’s king is a passive onlooker at
the edge of the board.
XIIIIIIIIY
9-+-+-+k+0
9+-+-+p+p0
9-+-tR-+p+0
9+-zp-+-+-0
9p+-+-+-+0
9zP-zP-tr-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
The sealed move.
42.Ra6 Rxc3 43.Rxa4
XIIIIIIIIY
9-+-+-+k+0
9+-+-+p+p0
9-+-+-+p+0
9+-zp-+-+-0
9R+-+-+-+0
9zP-tr-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
If now the a3-pawn is traded for the c5-pawn, the pieces can be put back in the box – without doubt, this
was also the assessment of my opponent.
43...Kg7!
The idea is that Black’s king can come near his c-pawn, after which he can make a serious attempt to win.
44.Ra7
Presumably, 44.Ra6 was a better fighting chance, when a possible continuation is 44...h5 45.h4 (White can’t
allow Black to fix his pawns with 45...h4) 45...Kf8 46.Ra7 Ke8 47.a4
 XIIIIIIIIY
9-+-+k+-+0
9tR-+-+p+-0
9-+-+-+p+0
9+-zp-+-+p0
9P+-+-+-zP0
9+-tr-+-+-0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
47...Ra3! (a nice moment to put the rook behind White’s passed pawn) 48.g3 (protecting h4, but there are
new drawbacks to this move; White could not ‘wipe the queenside clean’ without punishment since after
48.Rc7 Rxa4 49.Rxc5 Rxh4+, Black is two pawns up again, and the endgame should be winning for him)
48...Ra2+ 49.Kg1.
XIIIIIIIIY
9-+-+k+-+0
9tR-+-+p+-0
9-+-+-+p+0
9+-zp-+-+p0
9P+-+-+-zP0
9+-+-+-zP-0
9r+-+-+-+0
9+-+-+-mK-0
xiiiiiiiiy
analysis diagram
A) Normally, 49...Kd8 could be the winning idea, to prevent 50.Rc7, but in this case Black’s winning chances
become very narrow since White’s king can get in front of the enemy pawn: 50.Kf1! c4 51.Ke1 c3 52.Kd1.
Here, 52...Rg2 is not a real winning attempt either since White pointedly replies with 53.Rxf7 Rxg3 54.a5
and now, if Black doesn’t pay attention, White’s a-pawn will even become dangerous: 54...Kc8! (54...Rg4?
55.a6! Ra4 56.a7+–) 55.a6 Kb8 56.Kc2, and the draw can’t be avoided;
B) 49...c4! (at the right moment, Black sets his pawn in motion) 50.Kf1! (again, obviously, the king hurries
to the promotion square) 50...c3! (the pawn seems to be able to walk to the other side unhindered) 51.Rc7
(now, the rook has to get behind the pawn; 51.Ke1?? loses to a tactical trick: 51...Ra1+! – covering the
promotion square! – 52.Ke2 c2, and Black queens) 51...c2 52.Ke1. This is also forced, since otherwise,
again, 52...Ra1+ would follow.
 XIIIIIIIIY
9-+-+k+-+0
9+-tR-+p+-0
9-+-+-+p+0
9+-+-+-+p0
9P+-+-+-zP0
9+-+-+-zP-0
9r+p+-+-+0
9+-+-mK-+-0
xiiiiiiiiy
analysis diagram
52...Ra1+! (super-accurate! White’s king is forced to the d2-square; after 52...Rxa4 53.Rxc2 Ra3, White can
simply protect the g3-pawn with 54.Kf2) 53.Kd2 Rxa4 (now he does take! The reason is that now Black
immediately threatens 54...Ra3, winning the g3-pawn and the game) 54.Rc3 (also forced, to prevent
54...Ra3; but now Black can keep his pawn!) 54...Ra2 55.Re3+ Kd7.
XIIIIIIIIY
9-+-+-+-+0
9+-+k+p+-0
9-+-+-+p+0
9+-+-+-+p0
9-+-+-+-zP0
9+-+-tR-zP-0
9r+pmK-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
With two points to the good, the win should be within reach for Black. However, he has to find a clever way
to create a second passed pawn, and take care that it doesn’t get lost right away. Here is a possible winning
move sequence: 56.Rf3 Ke6 57.Kc1 f6 58.Kd2 Ke5 59.Kc1 g5 60.hxg5 fxg5 61.Kd2.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-mk-zpp0
9-+-+-+-+0
9+-+-+RzP-0
9r+pmK-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
And now the pretty breakthrough 61...h4! 62.gxh4 g4 63.Rf8 g3 64.Rg8, after which the familiar X-ray motif
brings the win: 64...g2! 65.Rxg2 c1=Q+ 66.Kxc1 Rxg2.
44...Kf6 45.a4 Ke6
 XIIIIIIIIY
9-+-+-+-+0
9tR-+-+p+p0
9-+-+k+p+0
9+-zp-+-+-0
9P+-+-+-+0
9+-tr-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
46.Ra6+?
There were various ways for White to bring the draw closer.
There is the attempt to bring the king into play with 46.Kg1 Rc2 47.Kf1 c4 48.Rc7 f6 (also 48...h5 49.a5 f6
50.a6 brings Black nothing) 49.a5 (the most logical; 49.Rxh7 is possible as well: 49...Ra2 50.g4 Rxa4 51.Ke2,
and even though Black keeps practical chances, I have not been able to find a win here) 49...Kd5 50.a6 Ra2
51.a7 Kd4 52.Rxh7, and Black’s winning chances have shrunk considerably: 52...Kc3 (after 52...c3 53.Rd7+
Ke3 54.Re7+, he cannot play 54...Kd2?? in view of 55.Re2+) 53.Ke1 Kb2 54.g4 Ra1+ 55.Ke2
XIIIIIIIIY
9-+-+-+-+0
9zP-+-+-+R0
9-+-+-zpp+0
9+-+-+-+-0
9-+p+-+P+0
9+-+-+-+P0
9-mk-+K+-+0
9tr-+-+-+-0
xiiiiiiiiy
analysis diagram
55...g5! (55...c3?? 56.Rb7+ Kc1 57.h4, and White even wins with his two distant passed pawns!) 56.Rb7+
Kc2 57.Ke3 c3 58.Rc7 Ra4 59.Kf2 Kd3 60.Rd7+=.
46...Kd5! 47.a5 Ra3
XIIIIIIIIY
9-+-+-+-+0
9+-+-+p+p0
9R+-+-+p+0
9zP-zpk+-+-0
9-+-+-+-+0
9tr-+-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
48.Ra8
White wants to attack the black king from the rear, and in the meantime use his own a-pawn as a distractor.
But since Black’s king is very well placed, this little plan falls to pieces.
Now, among others, the problem for White is that here 48.Ra7 c4 49.Rxf7 fails to 49...Rxa5, and the rook
ends up behind its own passed pawn, winning for Black.
48...Kc6 49.a6 Ra5
XIIIIIIIIY
9R+-+-+-+0
9+-+-+p+p0
9P+k+-+p+0
9tr-zp-+-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
With the idea of putting the rook behind his own pawn soon.
50.Ra7!
Seizing his best chance.
Another idea was to activate his king: 50.Kg3 Kb6 51.Rh8 h5 52.Kf4 Rxa6 53.Ke5
XIIIIIIIIY
9-+-+-+-tR0
9+-+-+p+-0
9rmk-+-+p+0
9+-zp-mK-+p0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
53...Ra2! (in rook endings, it is important to always remain flexible, and to be able to change plans
according to circumstances. Now that the situation has been altered – White’s king has become very active
too! – Black does well to go on a foray with the rook. Putting the rook behind the pawn made less sense
now, since White’s king is able to get in front of Black’s passed pawn; incidentally, Black’s winning chances
still remain large after 53...Ra7 54.Kd5
XIIIIIIIIY
9-+-+-+-tR0
9tr-+-+p+-0
9-mk-+-+p+0
9+-zpK+-+p0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
54...Rd7+ 55.Kc4 Rd4+ 56.Kc3 Rf4) 54.g4 hxg4 55.hxg4 Re2+ 56.Kf6 (56.Kd5 Rd2+ 57.Kc4 Rd4+ 58.Kc3
Rxg4–+) 56...Re4 57.g5 Rf4+ 58.Ke7
 XIIIIIIIIY
9-+-+-+-tR0
9+-+-mKp+-0
9-mk-+-+p+0
9+-zp-+-zP-0
9-+-+-tr-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
58...c4–+.
50...c4
Not only using his own pawn, but now also the rook can be placed behind it in short order.
51.Rxf7 Rxa6
XIIIIIIIIY
9-+-+-+-+0
9+-+-+R+p0
9r+k+-+p+0
9+-+-+-+-0
9-+p+-+-+0
9+-+-+-+P0
9-+-+-+PmK0
9+-+-+-+-0
xiiiiiiiiy
First, White’s passed pawn has to be eliminated.
52.Rxh7
Perhaps, White could have tried to save his skin here by building a fortress with 52.Kg3 c3 53.Rf3 Ra3.
However, after 54.Kf2 (54.Rf1 Kd5!; 54.Kg4 Kb5!) 54...c2 55.Rxa3 c1=Q 56.Rf3,
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+k+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+R+P0
9-+-+-mKP+0
9+-wq-+-+-0
xiiiiiiiiy
analysis diagram
there is no fortress. Black carefully nurses a pawn to g4, depriving White’s rook of several important
squares: 56...Qd2+ 57.Kg3 g5 58.Rf2 Qe3+ 59.Rf3 Qe1+ 60.Kh2 h5 61.Rf5 g4 62.hxg4 Qh4+ 63.Kg1 hxg4,
definitively breaking all resistance.
52...c3
‘The pawn has to do the work!’ – Euwe.
53.Re7 c2 54.Re1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9r+k+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+p+-+PmK0
9+-+-tR-+-0
xiiiiiiiiy
54...Kd5!
The king quickly leaves the c-file to enable ...Rc7 – and, of course, it moves to the centre to give the enemy
king a vicious shoulder-budge!
55.Rc1 Rc6 56.Kg3 Ke4! 57.Kg4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+r+-+p+0
9+-+-+-+-0
9-+-+k+K+0
9+-+-+-+P0
9-+p+-+P+0
9+-tR-+-+-0
xiiiiiiiiy
57...Rc5!
The last accurate move. Now, Hodgson lowered the flag, since after 58.h4 Ke3 59.g3 Kd2 60.Rg1 c1=Q
61.Rxc1 Kxc1 62.Kf4 Kd2 63.g4 Kd3, he is still unable to create a passed pawn without it being lost
immediately.
After this game, my opponent looked rather crestfallen. Later, I often met him at tournaments, and it didn’t
take long for him to surpass me completely as a player. He was always a nice guy to hang around with, and
we had a lot of laughs. His unique view on the game, and especially the joy he always radiated, were
fantastic. That’s why I was baffled when I heard at a certain point that he had given up chess and quit his
active career as a player...
Chapter 23
Thinking in schemes
When I first saw the term ‘thinking in schemes’ in Alexander Kotov’s book Think like a Grandmaster,
something dawned upon me. Here, something had been put on paper that I had also tacitly gathered from
my analyses with strong players.
Kotov gives the following anecdote:
‘Once, during the third Moscow International Tournament held in 1936, several chess masters were
analysing the ending of a game. They could not find any solution, but there was more than enough of
arguing. Suddenly, into the tiny room where they were analysing the game walked Capablanca, who loved
to stroll about while waiting for his turn to move. On learning the cause of the dispute, the imposing Cuban
suddenly scattered the pieces all over the board and then showed what kind of arrangement the active side
should try to achieve. It is not a slip of the tongue: Don Jose had literally scattered the pieces without
making any moves, but simply placed the pieces to their proper positions. And then all became clear at once:
the scheme was ready and a win could be easily achieved... Later on, I have seen such a way of thinking in
the play of contemporary outstanding endgame masters, Flohr and Smyslov.’
We introduced this way of thinking already in Part III:

Technique: thinking in schemes

Especially in the endgame, it is important to think not only in ‘moves’ but in ‘schemes’ –
meaning that one should think about finding the ideal squares for one’s pieces and pawns.
What it comes down to is that a player doesn’t automatically look at specific moves or tries to calculate
variations, but instead asks himself where, logically, his pieces can be best placed a couple of moves further
down the line. This is a kind of mini-plan he should be trying to achieve, before contemplating how to
continue after that. In an enormous amount of books (including my earlier book, Chess Strategy for Club
Players), and also on the Internet, the game Capablanca-Ragozin, Moscow 1936, is mentioned as an example.
However, the idea of ‘thinking in schemes’ is used not only in the endgame, but also in the middlegame.
This was in the days when games were adjourned at a certain point. When strong computer engines made
their entry into the chess world, the moment came that the adjournment of games was abolished.
Kotov gave a long game played by himself, which was adjourned countless times and in which again and
again he had to formulate a new phase in the winning process for himself. Below is a depiction of this
process in the form of a ‘comic strip’ with diagrams as pictures, showing the various phases:
272
Alexander Kotov
Gudmundur Arnlaugsson
Amsterdam ol 1954 (6)
Phase 1
XIIIIIIIIY
9-+-wq-+r+0
9zp-+-mkp+r0
9lzp-zp-zp-zp0
9sn-zpPzpP+N0
9P+P+P+-zP0
9+-zPLtR-+K0
9Q+-+-+P+0
9+-+-+R+-0
xiiiiiiiiy
White has a small advantage since he has more space, while Black has a few weak points – f6 and h6, among
others. On the other hand, White has the traditional Nimzo-Indian weakness on c4. If White wants to win,
he will have to work hard to find an entry point into Black’s position. There are two options: via the g-file,
or via the ‘transfer point’ h5. At this moment, it isn’t clear where the knight on h5 should go to, and what
the white queen or bishop could achieve on h5.

Without wanting to think ahead too far, White formulated the first phase of his ‘multi-stage’ plan:
1) The exchange of one set of rooks on the g-file will help him with his subsequent operations on the
kingside;
2) He will try to liberate the queen from its duty to protect c4 by playing the knight to e3;
3) With the help of the queen, White can offer the exchange of the second set of rooks. Since in that case
White’s queen will be more active than Black’s, Black will probably decline the exchange of the second set
of rooks, allowing White to gain control of the g-file with the rook.

Phase 2
XIIIIIIIIY
9-+-+q+-tr0
9zp-+lmkptR-0
9-zp-zp-zp-zp0
9sn-zpPzpP+-0
9P+P+P+-zP0
9+-zP-sNQ+K0
9-+L+-+P+0
9+-+-+-+-0
xiiiiiiiiy
White has succeeded in rounding off the first phase of his plan. Now what? White has to find a way to
penetrate deeper into the black position, and so he formulates the next plan:
1) Liberate the knight from the duty of protecting the c4-pawn by playing the king to d3;
2) Next, play the knight to h5, after which the threat of Qg6 becomes acute, attacking f7 and f6.

Phase 3
 XIIIIIIIIY
9-+-mklwq-tr0
9zp-+-+ptR-0
9-zp-zp-zp-+0
9sn-zpPzpP+p0
9P+P+P+-zP0
9+-zPK+-wQ-0
9-+-+-+P+0
9+-+L+N+-0
xiiiiiiiiy
What now? White has to invent a new plan since the structure has changed somewhat. Here, suddenly
there is the threat of Bxh5 since ...Rxh5 could be met with Rg8, and then after ...Qe7, White can catch
Black’s rook. This means that the h4-pawn has to be protected, so it is advisable to play the king back to h3!
The reader can prepare to enjoy some amusing turning-points in this game, which was adjourned various
times and continued in a series of morning sessions. There was a kind of silent agreement that a member of
an Olympiad team who was playing an adjourned game in the morning, wouldn’t take part in a new match
in the afternoon. Kotov: ‘Therefore, my teammates started looking at this game with a pessimistic eye, since
they were going to have to play more games, as I was chained to the board nearly every morning.’

Phase 4
XIIIIIIIIY
9-+-mkl+R+0
9zpn+-+p+-0
9-zp-zp-zp-wq0
9+-zpPzpP+r0
9P+P+P+NzP0
9+-zP-+-wQK0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
After the previous action, Kotov writes: ‘Here I breathed a sigh of relief, and now expected quick
resignation. Black calmly took my bishop, played his queen to h6, and then after Ng4, there suddenly
followed:
84...Qg5!
Black gives up his queen in hopes of creating a fortress!’
85.Rxg5 fxg5 86.Qe3 Rxh4+ 87.Kg3 Rxg4+!! 88.Kxg4 f6
Phase 5
 XIIIIIIIIY
9-+-mkl+-+0
9zpn+-+-+-0
9-zp-zp-zp-+0
9+-zpPzpPzp-0
9P+P+P+K+0
9+-zP-wQ-+-0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
Kotov: ‘Now I looked at the board, completely bewildered. How was I going to win this position? Where
could I find a way through? My king is shut out and the queen cannot do it on her own. I sat there for about
a quarter of an hour, and then I formulated yet another plan – the fourth already! I wasn’t confused by the
pitiful glances of my teammates. My only interest was to find a win, since otherwise it would mean that I
had spoiled many “working days”.’
This was the new plan:
1) Play the queen to h8 to limit the mobility of Black’s pieces as much as possible;
2) Next, come up with a4-a5 to open lines on the queenside, even at the cost of a pawn;
3) By creating threats with the queen (both on the kingside and via the other side of the board), White will
eventually be able to drive away the bishop from e8, after which the white king can invade.

XIIIIIIIIY
9-+-snl+-+0
9+-wQ-mk-+-0
9-+-zp-zp-+0
9zp-zpPzpPzp-0
9p+P+P+K+0
9+-zP-+-+-0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
Here is the finish to this epic game, which is enormously instructive from the viewpoint of learning to make
plans in chess.
With a few ‘long’ queen moves, White has invaded the black position, and now he gives a ‘deadly’ check.
98...Bd7

The bishop is forced to leave its important post, and now White’s king has the last word. There followed:
99.Kh5 Nf7 100.g4 Nd8 101.Kg6 Nf7 102.Kg7 Nd8 103.Qxa5 Nf7 104.Qb6 Be8 105.Qb1 Nd8
106.Qh1
 XIIIIIIIIY
9-+-snl+-+0
9+-+-mk-mK-0
9-+-zp-zp-+0
9+-zpPzpPzp-0
9p+P+P+P+0
9+-zP-+-+-0
9-+-+-+-+0
9+-+-+-+Q0
xiiiiiiiiy
And since the threat of Qh1-h6xf6 can no longer be rebuffed, Black finally gave up.
A magnificent example of ‘thinking in schemes’ carried out in an actual game. Precisely the ‘verbal
rendering’ of Kotov’s ‘mini-plans’ is illustrative for the way thinking ahead should be structured. Kotov
mainly uses terms like ‘goals’ and ‘sub-goals’ to achieve a certain progress in the position. In every new
phase, he tries to accomplish something that is within the bounds of possibility, without it becoming too
abstract. Precisely this ‘verbalizing’ of the various phases in this ‘multi-stage planning’ cements how he
intends to achieve these goals.
Botvinnik once said that a player always has to be able to formulate which plan he wants to pursue. As a
trivial example, he gave the way you intend to checkmate a lone enemy king with king plus queen. ‘No mate
without a verbal rendition of this checkmating process’ is the message. Apparently, the former World
Champion considered this verbalization essential.
To see whether this could be learned, I once practised this with both a group of young junior players and a
group of adults. Both groups thought it was difficult – on the whole, the latter group did better at it than the
former group. I think this has to do with the ‘abstractness’ of the thinking lines. The ‘young ones’ (most of
them primary school kids) were hardly capable of thinking in this way. Only after the transfer from primary
to secondary school, where they are taught mathematics (i.e. abstract thinking!), did it turn out that this
process gradually went better for them.
Nevertheless, we can’t get around the fact that especially endgames should be approached in this way.
Eventually, it will be clear that this procedure enables us to determine concrete goals, and every club player
can use this to his advantage.
As it turns out, this way of thinking is harder to learn for ‘calculators’ than for ‘strategists’. Even so, there
are many roads leading to Rome, and we can learn from model examples how we can actually convert
favourable positions to a win. As it turns out, in endgames too, besides abstract thinking, variations often
have to be calculated very concretely.
I think that the following example may appeal to the reader’s imagination. We see how a favourable ending
is gradually converted into a win through various phases, with the help of concrete goals.

273
Loek van Wely 2632
Alexei Shirov 2726
Monaco rapid 1999 (1)
 XIIIIIIIIY
9-+-+-+k+0
9+-+-zpp+p0
9-+-+-vlp+0
9+-+-+-+-0
9-tR-zP-+-+0
9+-+-zPN+-0
9r+-+-zPPzP0
9+-+-mK-+-0
xiiiiiiiiy
White to move
White is a pawn up, but all the pawns are on one wing. He does have another advantage: the knight is
stronger than the bishop here, since it is helped by the fact that the pawns are not standing on two wings;
the black bishop is, as it were, ‘biting on hot air’.
34.g4!
A typical move to gain space, but also to vacate the g2-square for the king, allowing White to escape the
rook checks.
34...e6 35.Rb7
Looking at the further course of the game, the question arises whether White should have played 35.g5
here: 35...Be7 36.Rb7 Bd6 37.h4, and in this way White could have achieved the plan for which he had to go
through so much trouble later on.
35...h6 36.h4 Ra5
Black prevents g4-g5.
37.Kf1
XIIIIIIIIY
9-+-+-+k+0
9+R+-+p+-0
9-+-+pvlpzp0
9tr-+-+-+-0
9-+-zP-+PzP0
9+-+-zPN+-0
9-+-+-zP-+0
9+-+-+K+-0
xiiiiiiiiy
37...e5
Black seizes the opportunity to exchange two pawns, thus reducing the winning margin. This is always an
important strategy for the defender: the side that is behind in material should exchange pawns, and the
superior side mostly wants to exchange pieces.
38.dxe5 Bxe5 39.Kg2
After the exchange of knight for bishop, there would be a great chance that the resulting rook ending is a
draw.
39...Bf6
 Phase 1
XIIIIIIIIY
9-+-+-+k+0
9+R+-+p+-0
9-+-+-vlpzp0
9tr-+-+-+-0
9-+-+-+PzP0
9+-+-zPN+-0
9-+-+-zPK+0
9+-+-+-+-0
xiiiiiiiiy
White wants to carry through the g4-g5 push. For that purpose, the king has to go to f4. White has to make
a few preparations for this: rook to e4.
40.Rb4 Kg7 41.Kg3 Be5+
Black makes things as difficult as possible for his opponent.
42.Kh3 Bf6 43.Re4
White plays systematically: he prevents Black’s ...Be5 and now enables the manoeuvre Kh3-g3-f4, to carry
through the g4-g5 push.
43...Rc5 44.Re8
XIIIIIIIIY
9-+-+R+-+0
9+-+-+pmk-0
9-+-+-vlpzp0
9+-tr-+-+-0
9-+-+-+PzP0
9+-+-zPN+K0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
44...Rd5?!
English GM Glenn Flear, who also commented on this game, let this moment pass by unnoticed.
I think Black had a better defence here.
It is interesting to see what happens if Black ‘passes’: 44...Ra5 45.Kg3 Rb5 46.Kf4.
 XIIIIIIIIY
9-+-+R+-+0
9+-+-+pmk-0
9-+-+-vlpzp0
9+r+-+-+-0
9-+-+-mKPzP0
9+-+-zPN+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
This should be the plan, but unfortunately this position contains a little trick: 46...Bxh4! (46...Rb4+ 47.e4
Rb5 (here too, 47...Bxh4 should be examined: 48.Nxh4 g5+ 49.Kg3 gxh4+ 50.Kxh4, and even though White
has a favourable pawn formation, the position should still be within drawing margins) 48.g5 hxg5+ 49.hxg5,
and White has carried though his plan) 47.Nxh4 g5+ 48.Kg3 gxh4+ 49.Kxh4, and this rook ending should
be a draw, as we already concluded above.
45.Kg3 Bd8
Shirov transfers the bishop to c7 to harass the white king from there.
46.Kf4 Bc7+ 47.Ke4 Ra5
The rook is put on a more active square, forcing White to react in the right way.
48.Re7
Not 48.g5?, as was the plan, since after 48...hxg5 49.hxg5 Ra2, White can no longer protect the f2-pawn.
48...Ra4+ 49.Kd3 Bd8
If Black keeps giving checks with the rook, White’s king will take a stroll to g2: 49...Ra3+ 50.Ke2.
50.Rd7 Bf6
XIIIIIIIIY
9-+-+-+-+0
9+-+R+pmk-0
9-+-+-vlpzp0
9+-+-+-+-0
9r+-+-+PzP0
9+-+KzPN+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
51.g5!
White has manoeuvred his rook to a favourable square, and can now finally pursue his plan.
51...hxg5
Black is obliged to trade on g5. True, another set of pawns leaves the board now, but also the h-file is
opened. It was hard to foresee that this will cause problems later on.
52.hxg5 Ba1

Phase 2
 XIIIIIIIIY
9-+-+-+-+0
9+-+R+pmk-0
9-+-+-+p+0
9+-+-+-zP-0
9r+-+-+-+0
9+-+KzPN+-0
9-+-+-zP-+0
9vl-+-+-+-0
xiiiiiiiiy
Now, it’s three pawns vs two on one wing, but the cooperation between White’s three pieces is superior to
that of Black’s. White has to put a pawn on f4 to give the g5-pawn its natural protection.
53.Nd2
The start of White’s new plan. To enable him to play f2-f4, the knight has to move.
53...Ra3+ 54.Ke2 Ra2
Now, White has to watch out: Black is threatening 55...Bc3 since after the trade of the light pieces, it’s a
draw! In general, any piece or pawn trade is favourable for Black, except the trade of all the pieces: the
pawn ending is winning for White.
55.Kf3 Bc3 56.Ne4
XIIIIIIIIY
9-+-+-+-+0
9+-+R+pmk-0
9-+-+-+p+0
9+-+-+-zP-0
9-+-+N+-+0
9+-vl-zPK+-0
9r+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
Van Wely has seen correctly that the knight is like a king on this square! Not only does it control f6, but it
also covers the potential weakness on f2.
56...Bb4 57.Rb7 Ra4 58.Nf6
A ‘poke’: suddenly, White threatens mate in two!
58...Bc3 59.Ne4
XIIIIIIIIY
9-+-+-+-+0
9+R+-+pmk-0
9-+-+-+p+0
9+-+-+-zP-0
9r+-+N+-+0
9+-vl-zPK+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
The knight returns empty-handed. The question is: why did Van Wely play like this? He is following
Capablanca’s ancient principle of ‘Don’t hurry!’ Apart from the fact that White isn’t spoiling anything with
his move, the repetition also has a psychological effect: the opponent gets to ‘feel’ who is calling the shots in
the game. Moreover, he can also make a mistake, and this is what White is testing. Bobby Fischer too often
made use of this ‘twofold repetition’.
59...Bb4 60.Rd7
Threatening 61.Rd4, with a lethal pin.
60...Ra5
60...Ra2 61.Kg4.
61.Kg4
A good moment to finally achieve the formulated sub-goal: White is ready for f2-f4.
61...Re5 62.Rd4 Be7 63.f4

Phase 3
XIIIIIIIIY
9-+-+-+-+0
9+-+-vlpmk-0
9-+-+-+p+0
9+-+-tr-zP-0
9-+-tRNzPK+0
9+-+-zP-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
The second phase of the plan has been concluded.
Phase 3: White is going to try to push his pawn from e3 to e5, to be able to play the knight to f6, enabling
mate threats.
63...Re6 64.Kf3 Kf8 65.Ra4 Rc6 66.Ra8+ Kg7 67.Ra7 Kf8 68.Nf2
XIIIIIIIIY
9-+-+-mk-+0
9tR-+-vlp+-0
9-+r+-+p+0
9+-+-+-zP-0
9-+-+-zP-+0
9+-+-zPK+-0
9-+-+-sN-+0
9+-+-+-+-0
xiiiiiiiiy
Preparing not only Nf2-g4, but also e3-e4-e5.
68...Rc8?
Black persists in passive defence.
68...f6! seems to me to be the only way to defend successfully. As said, Black is helped with any exchange,
especially if it’s pawns leaving the board. This way, he also removes later mating patterns from the position:
69.gxf6 Bxf6. It is known that a rook ending with white pawns on e5 and f4 vs a black one on g6 is drawn in
principle; with knight vs bishop added, it looks like White has significantly better winning chances, but I
doubt if the balance has really been broken. Of course, the white player could keep on trying endlessly.
69.Ng4 Rb8 70.Ne5
White (temporarily) abandons his plan to hand out another ‘poke’. You never know if during the blitz phase
someone might fall for a knight fork trick...
70...Rc8 71.Rb7 Re8 72.Nd7+
 XIIIIIIIIY
9-+-+rmk-+0
9+R+Nvlp+-0
9-+-+-+p+0
9+-+-+-zP-0
9-+-+-zP-+0
9+-+-zPK+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Now, the march of the e-pawn becomes topical.
72...Kg7 73.e4 Ra8
Here, 73...f6?? was impossible already: 74.Nxf6, with an immediate win.
74.e5

Phase 4
XIIIIIIIIY
9r+-+-+-+0
9+R+Nvlpmk-0
9-+-+-+p+0
9+-+-zP-zP-0
9-+-+-zP-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
Phase 3 has been concluded. Phase 4 is determining the right moment for the knight to jump to f6, after
which mating threats will abound.
74...Ra6 75.Nf6
And that moment comes quickly!
75...Bf8
After the light-piece trade with 75...Bxf6 76.gxf6+, the rook ending is lost since White’s pawns have
advanced very far already: 76...Kg8 77.Rb8+ Kh7 78.Rf8 Ra7.
XIIIIIIIIY
9-+-+-tR-+0
9tr-+-+p+k0
9-+-+-zPp+0
9+-+-zP-+-0
9-+-+-zP-+0
9+-+-+K+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
White has a wonderful pawn phalanx, and he can break through as follows: 79.Kg4 (this king move is
essential, as we will soon see) 79...Rb7 80.e6 fxe6 81.Re8, and since White threatens to force a pawn ending
with 82.Re7+, Black’s rook has to leave the seventh rank, after which f6-f7 decides. Important also is that
81...Rf7 fails to 82.Kg5, when Black is in zugzwang: 82...Ra7 83.Re7+ Rxe7 84.fxe7.
76.Rb1
XIIIIIIIIY
9-+-+-vl-+0
9+-+-+pmk-0
9r+-+-sNp+0
9+-+-zP-zP-0
9-+-+-zP-+0
9+-+-+K+-0
9-+-+-+-+0
9+R+-+-+-0
xiiiiiiiiy
Now that the h-file is open, the rook threatens to invade there.
76...Ra3+ 77.Kg4 Be7 78.Rh1 Bxf6
Now Black can pack it in after all. The resulting rook ending is winning for White, since now there are
breakthroughs in the position.
79.gxf6+ Kg8
Phase 5
XIIIIIIIIY
9-+-+-+k+0
9+-+-+p+-0
9-+-+-zPp+0
9+-+-zP-+-0
9-+-+-zPK+0
9tr-+-+-+-0
9-+-+-+-+0
9+-+-+-+R0
xiiiiiiiiy
Phase 4 is over. In the remaining rook ending with the giant pawn on f6, White tries to force a pawn
breakthrough.
80.Rb1 Ra8
Shirov doesn’t allow 81.Rb8+ – we already know the consequences of having a white rook on the back rank.
81.Rb7 Re8 82.Kg5 Kf8 83.Kh6!
XIIIIIIIIY
9-+-+rmk-+0
9+R+-+p+-0
9-+-+-zPpmK0
9+-+-zP-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
White’s king surprisingly penetrates via the h6-square.
83...Rc8
83...Kg8 is met by 84.e6! Rxe6 (84...fxe6 85.f7+) 85.Rb8++–.
84.e6!
At just the right moment, White liquidates into a rook ending with a passed f-pawn.
84...fxe6 85.Kxg6
XIIIIIIIIY
9-+r+-mk-+0
9+R+-+-+-0
9-+-+pzPK+0
9+-+-+-+-0
9-+-+-zP-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
We know that if the black king is cut off along the seventh rank, White’s f-pawn will bring him the win.
85...Ra8 86.Rh7 Kg8 87.f7+ Kf8 88.Rh8+
And Black resigned.
A wonderful technical performance by Van Wely!
Chapter 24
Analysis of practical endgames
Analysing games is something I have always liked to do. It is of crucial importance for a chess player to look
critically at his own games, but I’ve always gone a little further in this respect. I was also interested in the
secrets that were hidden in a position. As a result, it could happen that I kept working out things in my head
for days on end, trying to find out what was going on. Of course, you need to test your ideas for correctness,
and that is why I’ve always liked to cooperate with other chess players – also outside tournaments.
Botvinnik’s advice to publish your analyses, so that a large audience will be able to check your findings –
which was also observed by Jan Timman later – is something I have been doing in smaller circles. For years,
I have been a club bulletin editor at the clubs where I was a member. When I qualified in 1980 as an
outsider for the Dutch Championship in Leeuwarden, 1981, a new world opened up for me. In this
fourteen-player event, I got to know the then top players of the Netherlands, and during my stay there it
became clear to me that all of them were prepared to analyse games, and even opening variations, together.
Of course, some of them kept their cards close to their chest in some cases, but it was clear to me that such
analysis sessions were regarded by most players as a welcome opportunity to exchange ideas. In our
modern age, things are quite different. The players lock themselves up in their hotel room with their
laptops, to prepare for the next game. Analysing with each other is hardly ever done anymore; why make
the competition wiser than they already are?
Personally, I think it’s a pity that there is hardly any cooperation among the players nowadays – if only
because of the social aspect. Fortunately, some like-minded players still find ways to cooperate by working
for each other as seconds. Think of the fertile cooperation between the Dutch grandmaster Jorden van
Foreest and his good friend Max Warmerdam during the Tata Steel Tournament in 2021. Jorden surprised
friend and foe by winning the tournament with a fantastic performance, ahead of World Champion Magnus
Carlsen, the then number two in the world Alireza Firouzja, and his fellow countryman Anish Giri. After the
event, Jorden commented that, apart from Max’s chess-technical capacities, the social ties with his second
had also been a great help to him. Conversely, their cooperation resulted in Warmerdam, who had just
earned the grandmaster title, passing the 2600 Elo limit, and convincingly winning the Dutch national title
at the end of the same year.
In 2010, I was approached by an amateur chess player, Kees Schrijvers, to help to build a new Dutch-
language website that was supposed to be a ‘cheeky’ counterweight to the site of the Dutch chess federation:
schaaksite.nl, for which, from the beginning, I have been providing content together with GM and former
Dutch Champion Dimitri Reinderman. Apart from writing many ‘timeless’ articles, we also report on
important recent tournaments. After ten years, Schrijvers thought it was enough, and passed on the site to
the two of us, together with webmaster Lennart Ootes. We still offer reports on topical tournaments, often
right on the same night after the round has been played. Actually, in some sense we’re mashing two
potatoes with one fork: some quickly-fabricated analyses I later extended into more thorough examinations
that I can use for training purposes. I thought it might be nice to present two such game analyses here.

274
Stefan Kindermann 2570
Loek van Wely 2655
Germany Bundesliga 1997/98 (12)
 XIIIIIIIIY
9-+-+-trk+0
9+-+-+-+p0
9R+-+-+p+0
9+p+-+-+-0
9-+-+-+-+0
9+-+PvL-+-0
9-vl-+-+PzP0
9+-+-+-mK-0
xiiiiiiiiy
Black to move
An equal number of pawns, a rook each, same-coloured bishops, and a passed pawn on both sides. Black’s is
the most distant passed pawn, and it is also one square further advanced than White’s. Even so, Black
seems to have very little, but Van Wely is able to make something of very little – in fact, a lot of ‘something’
after some time!

28...Rb8!
Of course, Black starts by placing his rook behind his passed pawn.

29.Kf2 b4 30.Ke2 b3
The pawn is pushed forward immediately, making it an important factor. Black is now ready to remove his
bishop (...Be5!) followed by ...b3-b2, with an immediate win.

31.Rb6
White is forced to exchange the dominant black rook.

31...Rxb6 32.Bxb6
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-+p0
9-vL-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+p+P+-+-0
9-vl-+K+PzP0
9+-+-+-+-0
xiiiiiiiiy
White opts for the bishop ending, since the rook ending offers him no prospects. Now, Black has two
advantages: the most distant passed pawn, and the king in the centre (though it still has to arrive there).
32...Bc1
With this move, Van Wely drives White’s king to a passive square to keep the b3-pawn under control. He
could also have started a king’s march immediately. White’s king is unable to conquer the black pawn; for
instance, 32...Kf7 33.d4? Bc3 34.Kd3 b2 35.Kc2 Ke6, and Black can navigate into a winning pawn ending
with ...Kd5, followed by ...Bxd4.
33.Kd1 Bh6 34.h3 Kf7 35.Bc5 Ke6 36.Bf2
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+k+pvl0
9+-+-+-+-0
9-+-+-+-+0
9+p+P+-+P0
9-+-+-vLP+0
9+-+K+-+-0
xiiiiiiiiy
Kindermann understands that he can’t do anything active and assumes a wait-and-see attitude. Important
in his defensive plan is that he keeps the d3-pawn for as long as possible, denying Black’s king access to the
c4- and e4-squares.
36...Kd5
The king is there.
37.Ba7 Kc6 38.Bd4 Kb5 39.Bc3
White is making things as difficult as he can for his opponent.
39...Kc5 40.Be5 Be3 41.Bg7 Kb4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-vLp0
9-+-+-+p+0
9+-+-+-+-0
9-mk-+-+-+0
9+p+Pvl-+P0
9-+-+-+P+0
9+-+K+-+-0
xiiiiiiiiy
42.Bb2
Forced, since if Black’s king can enter via a3 or c3, his b-pawn will decide the game. An illustrative line is
42.Bf6? Ka3 43.Be5 Ka2.
42...Ka4
With this move, Black intends to liquidate into a pawn ending by working with ...Ba3.
43.d4
White becomes active. Also possible seems 43.Ke2 Bc5 44.Kd2 Bb4+ 45.Kc1 Ba3 46.d4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+-+-+-0
9k+-zP-+-+0
9vlp+-+-+P0
9-vL-+-+P+0
9+-mK-+-+-0
xiiiiiiiiy
analysis diagram
46...Bd6 (46...Bxb2+ 47.Kxb2 Kb4 48.d5 Kc5 49.Kxb3 Kxd5 just fails to win by a thread) 47.Kd2 h5, and
it’s not entirely clear yet.
43...Bf4 44.Ke2
Another idea was 44.d5 Kb5 (the original idea with ...Ba3 doesn’t work now: 44...Bd6 45.Kd2 Ba3 46.Bxa3
Kxa3 47.d6, with a draw) 45.Ke2 Kc4 46.Kf3 Bd6 47.Ke4.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-vl-+p+0
9+-+P+-+-0
9-+k+K+-+0
9+p+-+-+P0
9-vL-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Even though Black has good winning chances here, it doesn’t look easy: 47...Bg3 (47...Bb4? 48.Ke5=)
48.Kf3 Bb8 49.Ke4 Bd6 50.Ba1, and I can’t see how Black can make any progress.

44...Bd6 45.Kd3 Ba3

As the white pawn hasn’t been pushed yet, Black can liquidate.

46.Kc3
Certainly not the solution, but there was no longer any defence.
46.Bc3 Kb5 47.d5 Kc5 48.Ba5 b2 49.Kc2 Kxd5 was also hopeless.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+-+-+-0
9k+-zP-+-+0
9vlpmK-+-+P0
9-vL-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
46...h5?!
This is a logical move that appears to be necessary in a possible ensuing pawn ending. But Black had a
forced win here. However, calculating such winning lines is a very difficult job, and Van Wely didn’t want to
burn his fingers.
We can’t blame Loek for not exchanging, but the pawn ending with the most distant pawn was winning for
him. He would have needed to see the following variation though: 46...Bxb2+ 47.Kxb2 Kb4 48.d5 Kc5
49.Kxb3 Kxd5 50.Kc3 Ke4 51.Kd2 Kf4 52.Ke2 Kg3 53.Kf1.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-mkP0
9-+-+-+P+0
9+-+-+K+-0
xiiiiiiiiy
analysis diagram
So far, everything is clear. But what now? It turns out that there is only one winning move, and I can fully
imagine that one can’t calculate it to the end: 53...Kh2!! (53...h5? 54.Kg1 h4 55.Kf1 Kh2 56.Kf2, and Black
doesn’t get any further) 54.Kf2.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+P0
9-+-+-mKPmk0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Here again, there is only one move: 54...g5!, and now:
A) 55.Kf3 h5 56.Kf2 g4
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+p+0
9+-+-+-+P0
9-+-+-mKPmk0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
57.h4 (his last chance, but it won’t help White; 57.hxg4 hxg4 is an elementary win) 57...Kh1! 58.Kg3
(58.Kf1 g3) 58...Kg1, and Black wins;
B) Of course, we also have to investigate the other king move, 55.Kf1, after which things aren’t very simple
either: 55...h5 56.Kf2 g4, transferring to the above lines; Black wins.
47.d5
Obviously, White sets his pawn in motion too.
47...h4
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+p+0
9+-+P+-+-0
9k+-+-+-zp0
9vlpmK-+-+P0
9-vL-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
48.Ba1
Of course, not 48.Bxa3? Kxa3 49.d6 b2 50.d7 b1=Q 51.d8=Q, and Black wins the queen thanks to a small
‘coincidence’: 51...Qc1+ 52.Kd4 Qd2+.
48...Be7
There is nothing for the bishop to do on a3 now, and therefore Black transfers it elsewhere.
49.Kd3
The king has to make way. It makes no sense to play 49.Kc4, on account of 49...Ka3 50.Kb5 b2 51.Bxb2+
Kxb2 52.Kc6 Kc3 53.Kd7 Bg5,
XIIIIIIIIY
9-+-+-+-+0
9+-+K+-+-0
9-+-+-+p+0
9+-+P+-vl-0
9-+-+-+-zp0
9+-mk-+-+P0
9-+-+-+P+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
and Black’s king is very fast on the queenside; therefore, it is useful to add a move sequence that illustrates
how Black can win here: 54.Ke6 Bf4 55.Kf6 Kd4 56.Kxg6 Kxd5 57.Kf5 Bg3 58.Kg4 Ke4, and Black’s king
picks up White’s two pawns, at the same time preventing the white king from reaching the corner: 59.Kg5
Ke3 60.Kf5 Kf2 61.Ke4 Kxg2 62.Ke3 Kxh3 63.Kf3 Kh2.
49...Ka3 50.Ke4 Bd6!
Thus, the bishop cuts off the white king’s passage through the centre. Also, Van Wely has seen sharply that
he will be left with exactly one pawn on the kingside that can’t be traded off.
51.Kf3 g5 52.Kg4 Bf4 53.Bf6
53.d6 is met by 53...Bxd6 54.Kxg5 Bg3,
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-mK-0
9-+-+-+-zp0
9mkp+-+-vlP0
9-+-+-+P+0
9vL-+-+-+-0
xiiiiiiiiy
analysis diagram
and Black wins. But we have to clarify why this position is winning. Doesn’t Black have the ‘wrong bishop’?
In other words: can’t White’s king hide in the corner on h1? Let’s see what Black has up his sleeve: 55.Kg4
b2 56.Bxb2+ Kxb2 57.Kf3 Kc2 58.Ke2 Kc1 59.Kf1 Kd1 60.Kg1 Ke1 61.Kh1 Ke2 62.Kg1 Bf2+ 63.Kh1.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zp0
9+-+-+-+P0
9-+-+kvlP+0
9+-+-+-+K0
xiiiiiiiiy
analysis diagram
The king is in the corner, and there it will be skilfully stalemated, only in such a way that he has to bring
about his own undoing with a pawn move: 63...Kf1 64.Kh2 Bg1+ 65.Kh1 Kf2! 66.g4 hxg3 67.h4 g2#.
53...b2 54.Bxb2+ Kxb2 55.d6 Bxd6 56.Kxg5 Bg3 57.Kg4 Kc2 58.Kf3 Kd2
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zp0
9+-+-+KvlP0
9-+-mk-+P+0
9+-+-+-+-0
xiiiiiiiiy
Now, Black even keeps the white king outside the corner, which is of course fine as well.
59.Ke4 Ke2 60.Kf5 Kf2
And here the German grandmaster gave up the fight. A captivating endgame!

275
Tim Grutter 2253
Eline Roebers 2173
Almelo ch-NED U18 2021 (3)
XIIIIIIIIY
9-+-+-+qmk0
9+-+n+-zpp0
9-+-+-+-+0
9+-+P+-+-0
9-+pwQ-+-zP0
9+-zP-+-zP-0
9-+-+-zPK+0
9+-+-+-+-0
xiiiiiiiiy
Black to move
This position arose in a game between the strong Dutch youth player Tim Grutter and the online Girls U14
World Champion from 2020, Eline Roebers from the Netherlands. White has been forced to give a piece, but
he is ready to gather up multiple pawns in return. At this moment, Black has to make an important decision:
play 42...Nf6 or 42...Qc8 ?
42...Nf6
During the livestream in which I commented on this Dutch Under-18 Championship event, I also looked at
42...Qc8.

analysis diagram
Black (for the moment) retains the pawn on c4, and can then try to play the king to the centre. At the same
time, Black can also try to improve her pieces. I liked this move a little better than the move in the game,
since the side that plays for the win should prefer to keep as many pawns on the board as possible. This is
an important strategy to remember: the side that is worse will, as a rule, want to exchange as many pawns
as possible, and the side that is better will want the opposite. We can also assume that the stronger side will
want to exchange as many pieces as possible, since then the predominance of an extra piece will manifest
itself more clearly.
If now 43.d6, 43...Qc6+.

42...Nf6 43.Qxc4 Nxd5!

But in this case, the winning chances for the knight’s side are probably insufficient after 43...Qxd5+
44.Qxd5 Nxd5 45.c4.
 XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-zpp0
9-+-+-+-+0
9+-+n+-+-0
9-+P+-+-zP0
9+-+-+-zP-0
9-+-+-zPK+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Sometimes, a knight can be a clumsy piece when it comes to keeping a distant passed pawn under control.
In the live broadcast, I tried to lure the pawn forward to make it more vulnerable: 45...Nb6 46.c5 Nd7 47.c6
Nb6, but also here White’s king enters the stage quickly: 48.Kf3 Kg8 49.Ke4 Kf7 50.Ke5 Ke7, and after
51.f4, Black can forget about winning, as my colleague commentator Vincent Vleeming and I concluded
together.
You might consider blocking the passed pawn with 45...Nc7, but since White’s king is able to approach so
quickly, this doesn’t yield any realistic winning chances, e.g. 46.Kf3 Kg8 47.Ke4 Kf7 48.Ke5 Ke7 49.c5,
and eventually White will try to walk into the kingside; the position is completely balanced.
Nonetheless, there is a paradoxical move here that allows Black to control the pawn: 45...Nc3!!. What is the
idea behind this curious knight move? It is a clever way to keep White’s king at bay, and sometimes the
knight can also attack it from the rear. 46.Kf3 Kg8
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-zpp0
9-+-+-+-+0
9+-+-+-+-0
9-+P+-+-zP0
9+-sn-+KzP-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
47.Kf4 (after 47.Ke3?, the black knight attacks from the rear: 47...Nd1+ 48.Ke2 Nb2 49.c5 Na4 50.c6 Kf7,
and White’s pawn will be lost without him getting any counterplay for it) 47...Nd1 (this is the basic idea;
the knight creates confusion in White’s ranks, while Black’s king takes care of the c-pawn) 48.c5 Kf7 49.f3
(bad is 49.Ke5 in view of 49...Nxf2, and Black is winning already) 49...Ke6.
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpp0
9-+-+k+-+0
9+-zP-+-+-0
9-+-+-mK-zP0
9+-+-+PzP-0
9-+-+-+-+0
9+-+n+-+-0
xiiiiiiiiy
analysis diagram
But also here, it is clear that Black’s king is much better placed than in the variations given above. This has
to be winning, for example: 50.Ke4 Nc3+ 51.Kd3 Nb5 52.Kc4 Na7 53.Kd4 Nc6+ 54.Ke4 Na5 (the knight
is jumping around like crazy to lure White’s pawn further forward, and then gobble it without White’s king
being able to penetrate the king’s wing) 55.h5 Nb3 56.c6 Nc5+ 57.Kf4 Kd6 58.Kf5 Kxc6.
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-zpp0
9-+k+-+-+0
9+-sn-+K+P0
9-+-+-+-+0
9+-+-+PzP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
And here, it turns out that White has no access to the e6-square to enter Black’s position. The following
variation is still essential: 59.h6 g6+ 60.Kf6 Kd5 61.Kg7 Ke6 62.Kxh7 Kf7 63.f4 Ne6 64.Kh8.
XIIIIIIIIY
9-+-+-+-mK0
9+-+-+k+-0
9-+-+n+pzP0
9+-+-+-+-0
9-+-+-zP-+0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
And now, a remarkable move.
64...g5!! – Black voluntarily trades off her last pawn to achieve the exceptional checkmate of the king in the
corner: 65.fxg5 Nxg5 66.g4 Kf8 67.h7 Nf7#.
 XIIIIIIIIY
9-+-+-mk-mK0
9+-+-+n+P0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+P+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
What a dénouement!

The text, at any rate, is the right continuation. Black could also exchange the queens, which, considering our
statement in the previous comment, should be the indicated continuation.
44.Qd4 Qa8
Threatening mate in two, starting with 45...Nf4+.
45.Qe4
Typically Tim, the clever tactician. Any ‘normal’ human being would play 45.Kg1 here, but also in that case
Black has ample winning chances after, for example, 45...h6. Tim is not afraid of ghosts, simply allowing a
discovered check, as he has seen that he can quickly get his king into the centre in this way. Now, Eline
thought for a long time. She felt this was a crucial moment in the game – and she was right:
45...Ne3+
Nothing is given by 45...Nf4+ since after 46.Kf3, the knight is very clumsily placed. 45...h6 was also possible,
and another idea was 45...Qc6, to play the king to the centre next.

analysis diagram
This would have been a more systematic build-up, but also in that case Black would still have quite a lot to
prove.
46.Kf3 Qxe4+ 47.Kxe4 Nd1

This was the idea, of course: an ordinary knight fork. Two of White’s pawns are hanging. And yet, what
Black is doing looks very dangerous: the knight will now stand at a great distance from White’s passed
pawn, while White’s king is also able to support that pawn. But things are not as bad as they seem. Eline has
calculated very well, and the only thing she has to do now is keep the c-pawn under control.
48.c4 Nxf2+
XIIIIIIIIY
9-+-+-+-mk0
9+-+-+-zpp0
9-+-+-+-+0
9+-+-+-+-0
9-+P+K+-zP0
9+-+-+-zP-0
9-+-+-sn-+0
9+-+-+-+-0
xiiiiiiiiy
49.Kd5
Of course, not 49.Kf3 Nd3 50.Ke4 Nb4 51.Ke5 Kg8 52.Ke6 Nd3!, and since the c-pawn cannot move
forward now, Black can further strengthen her position.
49...Kg8
The funny thing is that Black’s king is still within the square of the c-pawn, so it can simply stop the pawn.
50.c5
Trying to keep Black’s king at bay with 50.Ke6 makes no sense after 50...Ne4 when, apart from the c-pawn
being stopped, the g3-pawn is hanging as well.
XIIIIIIIIY
9-+-+-+k+0
9+-+-+-zpp0
9-+-+-+-+0
9+-zPK+-+-0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-sn-+0
9+-+-+-+-0
xiiiiiiiiy
Another critical moment in this endgame. Eline thought for a relatively long time here, until she had almost
no time left, and had to play on her increment from here on.
50...Ng4?
And unfortunately, she goes wrong.
Of course, now it was important to involve the king in the game, and this seems to be the winning idea.
During our analysis, we looked at the following possible sequence: 50...Kf7, and now:
A) 51.c6 Ke7.
XIIIIIIIIY
9-+-+-+-+0
9+-+-mk-zpp0
9-+P+-+-+0
9+-+K+-+-0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-sn-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Now that Black’s king controls the pawn, the win seems no longer difficult. But there are still a couple of
snakes in the grass, as the following complex of variations shows: 52.c7 (White’s only chance is to lure
away Black’s king and then penetrate with his own king on the kingside) 52...Kd7 53.c8=Q+ Kxc8 54.Ke6.
XIIIIIIIIY
9-+k+-+-+0
9+-+-+-zpp0
9-+-+K+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-sn-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
It took us a while to spot that here 54...h5!? wins for Black (it turns out that there are more wins here: even
54...Ne4, but we will leave this out of consideration; only later did we spot that 54...Kd8 55.Kf7 g6 56.Kg7
Ke7 57.Kxh7 Kf7 also wins pretty easily, but then you would have to see that after 58.Kh6 Kf6 59.Kh7
Ng4 60.Kg8 Nh6+ 61.Kh7 Nf5, it is no longer very difficult): 55.Kf7 Ne4 56.Kxg7.
XIIIIIIIIY
9-+k+-+-+0
9+-+-+-mK-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+n+-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
Another position that shows how difficult chess can be, and especially the endgame. Black has only one way
to win:
A1) We looked exclusively at 56...Nxg3 and couldn’t find a win here – in fact, there isn’t one. However,
White’s defence here is very much hidden, and, actually, quite brilliant! 57.Kf6! (making use of the Réti
manoeuvre, the diagonal march of the king, to save his skin! We evaluated 57.Kg6? here, but that move
loses!: 57...Kd7 58.Kg5 Ke6 (not 58...Ke7? 59.Kf4, with a draw) 59.Kf4 Nf5 60.Kg5 Ng7 61.Kg6, and here
we stopped trying, but it is precisely this position that is winning for Black!: 61...Ke5! 62.Kxg7 Kf5!
(shouldering the enemy king – again, we see how essential this technique is; instead, 62...Kf4?? would
nullify all the above!) 63.Kf7 Kg4 64.Kf6 Kxh4, and White is exactly one move short to lock in Black’s king
on the edge) 57...Kd7 58.Ke5!, and there is nothing Black can gain here, for example: 58...Ke7 59.Kf4 Nf1
60.Kg5 Ng3 61.Kf4, with move repetition;
A2) 56...Kd7!! – the king hurries as quickly as possible in the direction of the white pawns. As it turns out,
the position is a win by one tempo: 57.Kg6 Nxg3 58.Kg5 Ke6 59.Kf4 (thus, the knight is driven from g3,
but now it can protect the pawn from the rear: 59...Nf5 60.Kg5 Ng7 61.Kg6 Ke5 (here, we see why the
position of the black king was so essential) 62.Kxg7 Kf5!, and the resulting pawn ending is a win.
B) 51.Kc6 also seems to come into consideration, to take control of the promotion square with Kb7, but
after 51...Ne4 52.Kb6 Ke7 (the computer indicates that even 52...Nxg3 is possible, but a human isn’t likely
to play this– certainly not with little time) 53.c6 Kd8 54.Kb7 Nc5+ 55.Kb6 Ne6,

analysis diagram
Black is winning here. For example, 56.Kb7 h5 57.Kb8 Ke7 58.c7 Nxc7 59.Kxc7 Ke6, and the rest is easy.
51.c6 Nf6+
51...Ne3+ is no help either: 52.Kd6 Nf5+, and now:
A) Our analysis continued 53.Kd7?, but now Black draws the long end of the stick after all with 53...Kf7
54.c7 Ne7 55.c8=Q Nxc8 56.Kxc8.
XIIIIIIIIY
9-+K+-+-+0
9+-+-+kzpp0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
We knew that pawn endings can be deceptive, but as deceptive as this one... Black wins, but how?
A1) We considered 56...h5? for a while, but soon saw that this doesn’t yield anything because of 57.Kd7
Kf6 58.Ke8 g6 59.Kf8 Kf5 60.Kg7 g5 61.hxg5 Kxg5 62.Kh7, with a draw;
A2) We thought we saw a win for Black with 56...Ke6?,
XIIIIIIIIY
9-+K+-+-+0
9+-+-+-zpp0
9-+-+k+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
but at closer inspection this turns out to be incorrect: 57.Kd8! (now, we see that Black was wide off the
mark, since she hasn’t kept the door locked for the white king) 57...h5 58.Ke8 Kf5 59.Kf7 Kg4 60.Kg6!,
and the draw is secured (certainly not 60.Kxg7?? Kxg3 61.Kg6 Kxh4–+);
A3) Black wins by playing 56...Ke7!!.
Why...?
A31) Of course, White can prevent ...h7-h5 with 57.g4, but this doesn’t help him: 57...Ke6 58.Kd8 Ke5
59.Ke7 Kf4 60.g5.
Black still has to watch out here!
XIIIIIIIIY
9-+-+-+-+0
9+-+-mK-zpp0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-mk-zP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
A311) How many players wouldn’t play the logical 60...Kg4? here without thinking, giving away the half
point after all their efforts? White saves his skin again with 61.g6!!, e.g. 61...hxg6 (61...h5?? is an
unwarranted winning attempt – even worse, it loses to 62.Kf7 Kxh4 63.Kxg7 Kg3 64.Kf6 h4 65.g7 h3
66.g8=Q++–) 62.Kf7 Kxh4 63.Kxg6, with a draw!;
A312) 60...g6! is again the only winning move: 61.Kf6 Kg4 62.Kg7 Kxh4 63.Kxh7 Kxg5, and now Black is
in time to win!
A32) 57.Kc7 h5 (this is a good moment to play this move) 58.Kc6 Ke6.
 analysis diagram
White’s king has to penetrate Black’s position ‘through the back door’, but he is too late: 59.Kc7 Kf5
60.Kd7 Kg4 61.Ke6 Kxg3, and Black is easily in time.
A4) Even 56...Ke8!! yields Black a full point. How does it work? Well, for example: 57.Kc7 Ke7 58.Kc6 Ke6
59.Kc7 Kf5 60.Kd6 Kg4 61.Ke6 Kxg3 62.Kf7 Kxh4 63.Kxg7 h5.
B) 53.Ke6! Nd4+ 54.Kd7 Nb5 55.c7 Na7 56.c8=Q+ Nxc8 57.Kxc8.
XIIIIIIIIY
9-+K+-+k+0
9+-+-+-zpp0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
It looks as if Black has the win for the taking, but she may be in for an unpleasant surprise: 57...Kf7 58.Kd7
Kf6 (also, nothing is given by the characteristic move 58...h5...
XIIIIIIIIY
9-+-+-+-+0
9+-+K+kzp-0
9-+-+-+-+0
9+-+-+-+p0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
analysis diagram
but here White would have to find the narrow path to the draw with 59.Kd8! Kf6 60.Ke8 Kf5 61.Kf7 g5=)
59.Ke8 (also here, White’s king penetrates ‘along the bottom’, securing the draw) 59...h5 60.Kf8, with a
draw.
52.Ke6 Ne8 53.Ke7 Nc7
 XIIIIIIIIY
9-+-+-+k+0
9+-sn-mK-zpp0
9-+P+-+-+0
9+-+-+-+-0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
54.g4
This is a useful move anyway, as we have seen. Black will often want to fix White’s structure with ...h7-h5.
By the way, there was nothing wrong either with 54.Kd8 Na6 55.c7 Nxc7 56.Kxc7 Kf7 57.Kd6=.
54...Nb5 55.g5 Na7 56.c7 Nc8+
XIIIIIIIIY
9-+n+-+k+0
9+-zP-mK-zpp0
9-+-+-+-+0
9+-+-+-zP-0
9-+-+-+-zP0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
57.Kd7
I thought it was also possible to keep Black’s king locked in with 57.Ke8, and the engine confirms this:
57...g6 58.Kd8 Nb6 59.Ke7 Nd5+ 60.Kd7 Nxc7 61.Kxc7 Kf7 62.Kd7, and this is, again, a draw.
57...Nb6+ 58.Kc6 Nc8 59.Kd7 Na7 60.Ke7
White decides to keep Black’s king locked in its cage.
60...Nc8+
And the peace treaty was signed. A fascinating and instructive endgame!
Chapter 25
Exercises
1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+k0
9-+-+-+R+0
9+-+-+rzPP0
9-+-+-+-+0
9+-+-+-+-0
9-+-+K+-+0
9+-+-+-+-0
xiiiiiiiiy
What is your verdict on this position? Does White have a winning method, or can Black hold the draw?

2
XIIIIIIIIY
9-tR-+-+-+0
9+P+-+-+-0
9-tr-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+K+-zp-+0
9+-+-+k+-0
xiiiiiiiiy
Black can only win if he can move his king from behind the f-pawn, but this fails to a rook check followed by
promotion. Still, Black has an ingenious winning plan in this position. Which one?

3
 XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-vl-+-+-+0
9+-zp-+p+p0
9P+k+-zP-zP0
9+-+-+-zP-0
9-+-+-+-+0
9+-+KvL-+-0
xiiiiiiiiy
61...Kb3??
Black goes too far in his winning attempts. After e.g. 61...Bd8, the game would have ended in a draw.
What did Black miss?

4
XIIIIIIIIY
9-+-+-+-+0
9vl-+-+p+-0
9P+-+pmk-+0
9+L+-+-zp-0
9-+-+-+K+0
9+-+-zP-+P0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
Since Black’s bishop is tied to the a-pawn, White has good winning chances. White conceives a beautiful and
effective plan. Do you see it?
5
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+q+0
9zP-+-+-+-0
9-+-+-+K+0
9+P+-+-zp-0
9-+-+-+k+0
9+-+Q+-+-0
xiiiiiiiiy
White’s king is in check. Which way should it go?

6
XIIIIIIIIY
9-+-tr-mk-+0
9+-wq-+pzp-0
9-+P+l+-+0
9+P+-tR-+p0
9-+-+-+-zP0
9+-+-+-wQ-0
9-+-+-zPP+0
9+-+-+-mK-0
xiiiiiiiiy
Sometimes a liquidation into a winning ending takes some very special moves. Are you just as creative as
White was in this position?
7
XIIIIIIIIY
9R+-+-+-+0
9+-mk-+-+-0
9P+-+-+-+0
9+-+-+-zp-0
9-+-+-+K+0
9tr-+-+P+-0
9-+-+-+-+0
9+-+-+-+-0
xiiiiiiiiy
What is the best way for White to liquidate into a winning position? Play through the fragment and
determine the good and less good moves made by the players. Give, where appropriate, exclamation marks
and question marks, and alternatives for less good moves.
53.a7 Kb7 54.Rg8 Kxa7 55.Rxg5 Kb6 56.Rd5 Kc6 57.Rd1 Ra8 58.f4 Rg8+ 59.Kf3 Rf8 60.Rd2 Kc5
61.Rd7 Kc6 62.Re7 Kd6 63.Re5 Rg8 64.Ke4 Ra8 65.Rb5 Ke7 66.Kf5 Rf8+ 67.Ke5 Kf7 68.Rb7+ Kg6
69.Rb6+ Kg7 70.f5 Re8+ 71.Re6 Ra8 72.Re7+ Kf8 73.Kf6 Ra5 74.Re6 Ra7 75.Rc6 Rf7+ 76.Kg5 Rg7+
77.Kf4 Ra7 78.Ke5 Re7+ 79.Re6 Kf7 80.Rxe7+ ½-½
Chapter 26
Solutions
1
Veselin Topalov
Alexander Grischuk
Wijk aan Zee 2003 (13)

The answer is not very difficult. Black’s rook keeps ‘clinging’ to the g5-pawn, disallowing White from
improving the position of his rook.
65...Ra5!
Bringing White’s king closer doesn’t help either since then the rook will keep giving side checks until
the king lets go of the g5-pawn again, and then the black rook will again ‘stick’ to the g5-pawn.
Draw!

2
Constructed position

62...Rc6+!
To win, Black has to change the position of his rook first. The rook will have a double task. On the one
hand, it has to prevent White’s pawn from promoting, and on the other it has to provide coverage for its
own king from the rear. This can be accomplished by moving the rook to the side of the pawn!
63.Kb3 Rc7
XIIIIIIIIY
9-tR-+-+-+0
9+Ptr-+-+-0
9-+-+-+-+0
9+-+-+-+-0
9-+-+-+-+0
9+K+-+-+-0
9-+-+-zp-+0
9+-+-+k+-0
xiiiiiiiiy
Here, we have the winning concept! Now, the black rook is ‘clinging onto the side’ of the b-pawn, so
White’s rook cannot move on pain of losing the pawn. With this set-up, Black can play his rook to the e- or
g-files, enabling the black king to emerge safely from behind the f-pawn without getting harassed by a rook
check from the rear.
64.Kb4
White’s king rushes towards its own pawn, but it is precisely one tempo short.
64...Rg7!
Creating a ‘hiding file’ for his king.
65.Kc5
Not very clever is 65.Kb5 Kg2 66.Rf8, since Black can eliminate White’s pawn with check and can then
return: 66...Rxb7+ 67.Kc6 Rg7–+.
65...Kg2 66.Rf8
White has to prevent the f-pawn’s promotion, but this means that he loses his own crucial pawn. If he
had one more move, the king would have been able to protect the c7-pawn, and it would have been a dead
draw.
66...Rxb7
 We know this type of position. It has the characteristics of Lucena’s position (building a bridge), but in
this case it is even much simpler. Black wins, for example, as follows:
67.Rg8+ Kf1 68.Kd4 Re7 69.Kd3 Ke1
And the pawn decides.

3
Johan Goormachtigh 2270
Herman Grooten 2390
Sas van Gent 1990 (6)

62.a5! Ba7
After 62...Bc7 63.a6 Bb6, White has the same break as in the game...
XIIIIIIIIY
9-+-+-+-+0
9vl-+-+-+-0
9-+-+-+-+0
9zP-zp-+p+p0
9-+-+-zP-zP0
9+k+-+-zP-0
9-+-+-+-+0
9+-+KvL-+-0
xiiiiiiiiy
63.g4!
Yielding White two distant passed pawns, deciding the game.
63...fxg4 64.f5 c4 65.f6 Be3?
Perhaps, Black could have tried 65...c3!?, since after 66.Bxc3 Kxc3 67.f7 Bc5, White can still go wrong:
A1) 68.a6?? g3 69.f8=Q Bxf8 70.Ke2! (70.a7?? g2 71.a8=Q g1=Q+, and Black wins) 70...Bc5 71.Kf3
Kc4 72.Kxg3=;
A2) 68.Ke2!, and now the g4-pawn is under control and one of White’s pawns will march on.
B) 65...Bc5 66.f7 c3 67.Bxc3 Kxc3 68.Ke2 would transpose to the variation given above.
66.Ke2! g3 67.Kxe3 1-0

4
Anish Giri 2780
Maxime Vachier-Lagrave 2760
Bucharest 2021 (7)

45.Be8 Ke7 46.Bxf7!

Beautiful; White sacrifices his bishop to create a second distant passed pawn.
46...Kxf7 47.Kxg5
Thanks to the bishop sacrifice, he can clear away the important g5-pawn.
47...Kg7 48.h4 e5
Also after 48...Bb6 49.h5 Bd8+ 50.Kf4 Bb6 51.Ke5 Kf7 52.Kd6, White wins easily.
49.Kf5 Bb8
 XIIIIIIIIY
9-vl-+-+-+0
9+-+-+-mk-0
9P+-+-+-+0
9+-+-zpK+-0
9-+-+-+-zP0
9+-+-zP-+-0
9-+-+-zP-+0
9+-+-+-+-0
xiiiiiiiiy
50.Ke4 1-0
In the game notation, the text move is given, but probably both players put the kings on e4/d5 to
indicate that White won (as is customary with electronic boards these days).
Actually, I think that Giri would have played 50.Ke6 here, with the possible continuation 50...Kg6 51.f4
exf4 52.exf4 Ba7 53.f5+ Kg7 54.f6+ Kf8 55.h5, and this ‘improved’ pawn ending is no longer very difficult.

5
Hans Hubert Sonntag
Herman Grooten
Brussels 1992

84.Kf4??
The right move was 84.Kh4!=, although this may not be so obvious. The point is that on f4 the king is
in the way of its own queen, and at the same time it is too exposed: 84...Kh2 85.Qd2+ (the only move!)
85...g2 86.Qf4+ Kh1 87.Qf3, and Black has no way to make progress.

84.Kf4?? Kh2
Here, we see why White’s king is in the way of his queen; he doesn’t have the check on f4.
85.Qe2+
85.Qe1 Qh6+ 86.Kf3 Qh5+ 87.Kf4 g2 88.Qg3+ Kh1.
85...g2 86.a6 Kh1
XIIIIIIIIY
9-+-+-+-+0
9+-+-+-+-0
9P+-+-+q+0
9+-+-+-+-0
9-+-+-mK-+0
9+P+-+-+-0
9-+-+Q+p+0
9+-+-+-+k0
xiiiiiiiiy
87.a7
The second reason why White’s king is wrongly placed on f4 has to do with this defensive method.
Sometimes, White can prevent the further march of Black’s dangerous g-pawn by pinning it. But if he plays
87.Qf3 now, there follows a simple queen exchange by 87...Qf6+ 88.Kg3 Qxf3+ 89.Kxf3 g1=Q, and Black
wins.
87...Qf6+
An important in-between check; 87...g1=Q?? 88.a8=Q+ would have helped White to win...
88.Ke4
If White’s king goes to the g-file, Black promotes with check!
88...Qc6+ 89.Kf4 g1=Q
Now Black can safely promote.
90.Qh5+ Qh2+
And White resigned.

6
Daniil Dubov 2714
Ivan Saric 2644
Riga 2021 (6)

35.Rxe6!!
An exchange sacrifice to liquidate.
35...Qxg3
The zwischenzug 35...Rd1+ doesn’t take the sting out of the position either, as after 36.Kh2 Qxg3+
37.Kxg3, White’s two pawns on the sixth rank are stronger than the rook: 37...fxe6 38.b6 Ke7 39.b7 Rb1
40.c7.
36.fxg3 fxe6 37.c7!
The right order. Certainly not 37.b6?? in view of 37...Ke7 38.b7 (38.c7 Rd1+ 39.Kf2, and now Black’s
king controls the pawns: 39...Kd7–+) 38...Kd6–+.
37...Ke7
Black tries to transpose into a pawn ending, but this turns out to be hopeless as well.
37...Rc8 38.b6 Ke7 39.b7 Kd7 (39...Rxc7 40.b8=Q Kd6 41.Qb6+ Kd7 42.Qd4+ Ke8 43.Qe5, and this
triple attack yields White at least a pawn, after which the battle is over) 40.b8=Q Rxc7 41.Qb5+ Ke7
42.Qxh5, and White wins.
38.cxd8=Q+ Kxd8
XIIIIIIIIY
9-+-mk-+-+0
9+-+-+-zp-0
9-+-+p+-+0
9+P+-+-+p0
9-+-+-+-zP0
9+-+-+-zP-0
9-+-+-+P+0
9+-+-+-mK-0
xiiiiiiiiy
White has the most distant passed pawn and wins in an elementary way:
39.Kf2 Kc7 40.Ke3 Kb6 41.Kd4 Kxb5 42.Ke5 1-0
There could follow 42...Kc4 43.Kxe6 Kd4 44.Kf7 Ke3 45.Kxg7 Kf2 46.Kg6, and Black is (much) too
late.

7
Paul Hummel
Jan Willem van de Griendt
Netherlands tt 2021/22 (1)

White has a quite clever way to liquidate:

53.a7 Kb7 54.Rg8 Kxa7 55.Rxg5 Kb6 56.Rd5
The characteristic vertical cut-off.
 According to the Tablebases, 56.f4 wins even more quickly: 56...Kc6 57.Rg7! (horizontal cut-off!)
57...Kd6 58.Kg5 Ra1 59.f5 Rg1+ 60.Kf6 Rf1 61.Kg6 Ke5 62.Re7+ Kd6 63.f6 etc.
56...Kc6 57.Rd1
Thus, Black’s king has been cut off at a considerable distance, but still White manages not to win this:
57...Ra8
Black also takes up the theoretically known defensive position. The rook returns to the back rank to
harass the king with checks from there.
58.f4
XIIIIIIIIY
9r+-+-+-+0
9+-+-+-+-0
9-+k+-+-+0
9+-+-+-+-0
9-+-+-zPK+0
9+-+-+-+-0
9-+-+-+-+0
9+-+R+-+-0
xiiiiiiiiy
The pawn swiftly marches forward. According to the ‘Rule of five’, the position is winning now, since
the rank of the pawn (4) added with the number of lines by which the enemy king is cut off (2) makes 6.
58...Rg8+ 59.Kf3
The typical way to win such positions is as follows: 59.Kh5 Rf8 60.Kg5 Rg8+ 61.Kh6 Rf8 62.Rf1 Kd6
63.Kg7 Rf5 64.Kg6 Rf8 65.f5, and the rest is fairly straightforward.
59...Rf8 60.Rd2
This is a first indication that White doesn’t know how to win this.
60...Kc5 61.Rd7
This too indicates that White doesn’t have the necessary knowledge. There is still a chance to repair it...
61...Kc6 62.Re7?
But after this move, the win is definitively gone. White simply allows Black’s king to come one file
closer.
62.Rd1, to carry out the original plan, would still have won. But if you don’t know this, and you also
don’t see how it should be done, it’s not easy at all...
62...Kd6 63.Re5 Rg8 64.Ke4 Ra8 65.Rb5 Ke7 66.Kf5 Rf8+ 67.Ke5 Kf7 68.Rb7+ Kg6 69.Rb6+ Kg7
70.f5 Re8+ 71.Re6 Ra8 72.Re7+ Kf8 73.Kf6 Ra5 74.Re6 Ra7 75.Rc6 Rf7+ 76.Kg5 Rg7+ 77.Kf4 Ra7
78.Ke5 Re7+ 79.Re6 Kf7 80.Rxe7+ ½-½
Index of names
(numbers refer to pages)

A
Afek 350
Akerblom 352
Alahakoon 32
Alburt 164
Alekhine 251
AlphaZero 210
Arnlaugsson 438
Aronian 330-332
Averbakh 76, 223, 225, 270-271, 273-275, 282, 285, 310-311

B
Bähr 161, 166, 192-195
Balashov 250
Barbier 156
Beerdsen 321-322
Benko 42
Bernhardt 407
Boden 151
Bogoljubow 251
Böhm 91
Bok 233, 278
Botvinnik 441, 448
Browne 185
Brunia 145
Bunka 346

C
Capablanca 437, 444
Carlsen 15, 51, 55-56, 181-182, 231-232, 240-241, 243, 330, 332, 372-381, 383-384, 392-393, 448
Carlsson 152
Caruana 376, 378-379
Castro 32
Centurini 225
Colas 235
Corcos 411
Cruijff 48
Cuijpers 253

D
Dall’Ava 184
Dedrle 154
De Groot 304
De Vreugt 73
Dominguez Perez 375
Donner 61
Drtina 147
Dubov 373, 467
Dukic 407
Dvoirys 421
Dvoretsky 193, 311, 371

E
Ellenbroek 46
Enkhnar 236
Erwich 104, 223, 225
Euwe 11, 20, 42-45, 381, 437

F
Fahrni 161, 178, 206, 346, 349
Fedorowicz 386, 389, 391
Fila 364
Firouzja 231, 280-282, 448
Fischer 42-45, 444
Flear 442
Flohr 437
Floor 105
Flores 214
Foltys 205
Fontana 173
Fritz 190, 355, 357, 363

G
Genttner 205
Gheorghiu 59
Giri 23, 448, 466
Gligoric 386-387
Goormachtigh 465
Gorgiev 412
Grandelius 392-393
Grigoriev 90, 124, 137, 262
Grin 415
Grischuk 464
Grooten 9, 24, 46, 73, 105, 127, 141, 155, 201, 232, 253, 324, 421, 425, 428, 430, 465-466
Grutter 453

H
Hartoch 429
Hasek 348, 358
Henley 388-389, 391
Herbstman 234
Heuäcker 411
Hodgson 12, 429-430, 436
Hoetmer 17
Hörning 237
Horwitz 177, 355, 408, 417
Hou Yifan 383
Hummel 468
I
Imbens 17

J
Jongsma 392

K
Kalandadze 410
Kalra 236
Karpov 55, 388
Karstedt 220
Kasparov 27-29, 31, 163-164, 203, 363, 386
Keemink 206, 346
Kindermann 449-450
Kling 177, 355, 367
Knoppert 275-276
Koblenz 31
Kohlrausch 17, 158, 218
Kohser 209
Kok 219
Kopaev 365
Kortchnoi 55, 58-59, 61
Kotov 437-440
Krystufek 366
Kubbel 219, 403, 408-409
Kuipers 232
Kurt 17, 116
Kuznetsov 413

L
Lamprecht 371
Lasker,Ed. 153
Lasker,Em. 405
Lautier 239
Lee 328
Leko 280
Ljubojevic 185
Lolli 88
Louma 209
Lucena 257, 291, 296

M
Mamedjarova 127
Mamedyarov 127, 374
Mandler 175
Marin 240, 243-244, 384
Mees 174, 358
Melnichenko 188
Melnikov 187
Minev 351, 370
Mohamed 155
Moravec 135, 151
Morozevich 240-242
Muhren 91, 111
Müller 17, 181, 239, 327, 371

N
Nakamura 16, 116, 181-182
Negra 141
Nepomniachtchi 55-56, 266, 268
Neukomm 346
Nimzowitsch 203, 328, 369
Nunn 60, 368, 371

O
Olafsson 250
Ootes 449
Otten 109, 158

P
Pachman 157
Pajeken 239
Peelen 276-277
Petrosian 17, 140, 381-383
Petrov 360
Petrovici 405
Philidor 58, 255-256
Pijpers 353
Pliester 9
Pogosyants 63, 179, 203, 212-213, 404
Polgar 363
Ponziani 160
Portisch 17
Postny 278
Pribyl 203
Prokes 178, 210, 217-218, 220, 235, 347-348, 354, 366, 406
Prokop 404, 412, 414, 417
Pruscha 349

R
Radjabov 51
Ragozin 437
Reinderman 449
Réti 152, 169-170, 172
Rey Ardid 176
Richter,E. 180
Richter,F. 207
Riihimaa 211, 229
Rinck 72, 359, 402
Rödl 166
Roebers 453
Roiz 286-287
Romanovsky 311
Rooda 201
Rossi 71, 413
Rublevsky 239

S
Saavedra 156-157, 367
Sackmann 221
Sämisch 328
Sansom 415
Saric 467
Sarratt 176
Savory 328
Scheeren 337
Schmitt 181
Schrijvers 449
Schuering 371
Schurig 405
Schuurman 194-195
Selezniev 109, 348, 354
Selivanov 136
Shevchenko 280-281
Shirov 441, 443, 446
Smyslov 239, 425, 427-428, 437
So 116
Sochnev 416
Sonntag 466
Spassky 58-59
Speckmann 221
Speelman 52-54
Sprangers 23-24
Stohl 425-427
Struik 91, 111
Sunilduth Lyna 235
Swinkels 286-288

T
Tabidze 410
Tacu 403
Taimanov 293-295, 298
Tari 374
Tate 411
Tebben 216
Timman 16-17, 52-54, 60, 128, 386-389, 448
Tofighi 233
Topalov 464
Troitzky 189, 204, 206, 216, 363, 369, 401, 413

U
Ulrichsen 367

V
Vachier-Lagrave 266, 466
Vancura 260, 307-308, 310, 317, 322-323, 360-361
Van de Griendt 468
Van den Herik 52
Van der Heijden 350
Van der Vliet 372
Van der Wiel 386, 389, 391
Van Dongen 247, 265
Van Foreest 448
Van Kerkhof 104
Van Nieuwkuijk 352
Van Oosterom 372
Van Tets 354
Van Wely 12, 323-325, 441, 444, 447, 449-452
Van Wijgerden 11-12, 145
Vleeming 454
Vogt 201
Von Feyerfeil 416
Vukic 28, 31

W
Warmerdam 321-322, 448
Weenink 349
Welling,G. 157
Welling,J. 157, 337
Whitehead 58-59
Wojcik 365
Wotawa 356

Y
Yakhontov 72

Z
Zalkind 215
Zgerskye 414
Zhu Chen 293-298
Zikov 401
Zinar 172
Explanation of symbols
The chessboard
with its coordinates:

XIIIIIIIIY
8rsnlwqkvlntr0
7zppzppzppzpp0
6-+-+-+-+0
5+-+-+-+-0
4-+-+-+-+0
3+-+-+-+-0
2PzPPzPPzPPzP0
1tRNvLQmKLsNR0
xabcdefghy
K King
Q Queen
R Rook
B Bishop
N Knight

² White stands slightly better

³ Black stands slightly better
± White stands better
µ Black stands better
+– White has a decisive advantage
–+ Black has a decisive advantage
= balanced position
! good move
!! excellent move
? bad move
?? blunder
!? interesting move
?! dubious move
# mate
ch championship
zt zonal tournament
izt interzonal tournament
ct candidates tournament
tt team tournament
ol olympiad
m match
cr correspondence
Bibliography
Averbach, Jurij, Wat iedere schaker van het eindspel moet weten, Variant 2002
Awerbach, Juri, Lehrbuch der Schachendspiele Band 1 en 2, Sportverlag Berlin 1979
Beliavsky, Alexander & Adrian Mikhalchishin, Winning Endgame Strategy, Batsford 2003
Dvoretsky, Mark, Dvoretsky’s Endgame Manual, Russel Enterprises 2009
Euwe, Max/Cor van Wijgerden, Het eindspel 3 Toreneindspelen, Het Spectrum 1981
Kasparyan, Ghenrikh, Domination in 2,545 Endgame Studies, Progress Publishers 1980
Koblenz, Alexander, Schach positionell – Training mit Alexander Koblenz, Sportverlag 1991
Kotov, Alexander, Think like a Grandmaster, Batsford 2002
Krogius, Nikolaj, De wetten van het eindspel (Leer goed schaken 2), Keesing BV 1973
Müller, Karsten & Frank Lamprecht, Fundamental Chess Endgames, Gambit 2001
Staudte, Hans Hilmar & Milu Milescu, Das 1x1 des Endspiels, Beyer Verlag 2004

ChessBase MegaBase
Harold van der Heijden’s Endgame Study Database