First published in 2020 by Gloucester Publishers Limited, London.

Copyright © 2020 Neil McDonald

The right of Neil McDonald to be identified as the author of this work has
been asserted in accordance with the Copyrights, Designs and Patents Act
1988.

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, photocopying, recording or otherwise,
without prior permission of the publisher.

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

ISBN: 978 1 78194 529 2

Distributed in North America by National Book Network,

15200 NBN Way, Blue Ridge Summit, PA 17214. Ph: 717.794.3800.

Distributed in Europe by Central Books Ltd.,

Central Books Ltd, 50 Freshwater Road, Chadwell Heath, London, RM8
1RX.

All other sales enquiries should be directed to Everyman Chess.

email: info@everymanchess.com; website: www.everymanchess.com

Everyman is the registered trade mark of Random House Inc. and is used in
this work under licence from Random House Inc.

Everyman Chess Series

Commissioning editor and advisor: Byron Jacobs

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.
Printed by TJ International Limited, Padstow, Cornwall.
About the Author
Neil McDonald became a grandmaster in 1996 and a FIDE trainer in 2017.
He is a regular coach of the England Junior team at international events. Neil
has written numerous books on openings, endgames, tactics and strategy as
well as biographies of famous players. He lives in Gravesend in Kent,
England.

Also by the Author:

Break the Rules!
Catalan: Move by Move
Chess Secrets: The Giants of Power Play
Chess Secrets: The Giants of Strategy
Coach Yourself
Concise Chess Endings
Concise Chess Middlegames
Concise Chess Openings
Dutch Leningrad
French Winawer
How to Play against 1 e4
Main Line Caro Kann
Modern Defence
Play the Dutch
Positional Sacrifices
Practical Endgame Play
Rudolf Spielmann: Master of Invention
Starting Out: 1 e4
Starting Out: Queen's Gambit Declined
Starting Out: The Dutch Defence
Starting Out: The English
Starting Out: The Réti
The King’s Indian Attack: Move by Move
The Ruy Lopez: Move by Move
Contents
About the Author
Introduction

1 Improving the Activity of your Pieces

2 Stopping the Opponent Playing Good Moves
3 Full Grovel Mode
4 Punishing Faulty Freeing Moves
5 Exploiting a Hole
6 Manoeuvring Against Pawns
7 Promoting a Pawn
8 Using a Pawn as a Battering Ram
9 Sacrificing to Gain the Initiative
10 Deciding the Character of the Game in the Opening

Index of Games
Introduction
The sixth world champion, Mikhail Botvinnik, described chess as the art of
logic. Indeed, if music is the art of sound, and dance is the art of movement,
then chess, as the greatest of intellectual games, is the art that most clearly
expresses decision making.
As the title of the book suggests, chess is also a war game. Nevertheless,
the competitive pleasure we get in beating our opponent and notching up a
point cannot be separated from the satisfaction of having our ideas tested and
validated. If this wasn’t so, we wouldn’t prize a win over a skilful player
more than that over a beginner. And the pain of losing for most players far
outweighs the joy of winning – not only has our opponent triumphed over us,
but our plans have been proved wrong.
Perhaps at the moment you see chess in terms of developing moves,
attacking moves and defensive moves. The purpose of this book is to show
you these things and a lot more. Imagine you have all your pieces in play in
an equal position where there is nothing to attack and no threat to deal with.
What should you do then? You should manoeuvre and probe, stop the
opponent carrying out the advances he wishes, fortify strong points, try to
create or seize control of holes, and so on. Such play is a direct challenge to
the opponent: it isn’t neutral. It forces him to keep up, to tread a fine line
between playing with too much energy or too little. If and when he slips up,
you will get the advantage. Then you will have something to build upon.
The games have been chosen not just for their instructional value but also
because I find them aesthetically pleasing. I hope you enjoy them and pick up
some good battle plans.

Neil McDonald,
Gravesend, January 2020
Chapter One
Improving the Activity of your Pieces
In a nutshell the purpose of a plan is to add energy to your pieces and pawns
while lessening the power of your opponent’s pieces and pawns. As the third
world champion José Raul Capablanca observed, the key theme which runs
through all phases of the game is the co-ordination of the pieces. You have to
get them working together.
As our first venture into planning we’ll study examples from the
endgame. This will allow us to examine planning in its purest form.
Compared with the middlegame or opening, there is less ‘noise’ or ‘clutter’ to
distract us from following the plan in question through its stages. Once we
have grasped the basics of strategy we can then apply it to the more complex
situations that arise in other phases of the game.
We’ll begin by looking at happy instances where you have succeeded in
paralysing or entombing the opponent’s pieces. In other words, you have
drained them of their energy. Your plan is to find a precise series of moves to
assure victory.

Entombing an Enemy Piece

Game 1
And.Volokitin-S.Mamedyarov
European Cup, Eilat 2012
 Question: Can you see the move for White that persuaded a world-class
player
to resign at once?

Answer: Black has two extra pawns, but after 34 g4! he immediately gave
up. An extreme example of reducing the energy level of the enemy pieces:
the black rook is entombed on h8.
Let’s imagine that Black had played on for a bit. If he just waits, a simple
winning plan for White would be to take on b2 and then march his king up
the board, capture on a6, and queen the a-pawn. White’s pieces have almost
unopposed power to manoeuvre, though he still has to be a tiny bit careful.
Let’s consider two ways Black might try to free his rook.

Question: How should White respond to 34 ... f5 - ?

After 34 ... f5, 35 Bxf5?? would be a colossal mistake for White, as it

allows Black to extract his rook with 35 ... Kf7 and then, say, 36 ... Bb4 and
37 ... Rb8.
Answer: Instead, 35 gxf5! keeps the bind in place. Black could then try
35 ... h5, hoping to escape with 36 ... h4 and 37 ... h3, when the rook can
emerge via the h4-square. But 36 h4 puts a stop to that. (If you find it
aesthetically unpleasing that White puts his pawn on a dark square then 36
Kxb2 h4 37 h3! is also good enough.) Note that giving up the exchange with
36 ... Rh6 and 37 ... Rxg6 is entirely hopeless for Black, as after 38 fxg6 his
king and bishop remain boxed in.
Returning to the diagram, if Black tries the alternative 34 ... h5, White
has to avoid the abysmal 35 Bxh5??, as after 35 ... Rh6! (even better than the
immediate 35 ... g6) there is no good way to prevent 36 ... g6. Also 35 h3??
allows the black rook freedom after 35 ... hxg4 36 hxg4 Rh2+. The correct
response is 35 gxh5, keeping the rook imprisoned. Black gains a passed pawn
on the f-file, but it doesn’t matter. For example: 35 ... f5 36 Kxb2 f4 37 Re4
Bd6 38 Re8+ Bf8 39 Kc2 and the white king will deal with the passed pawn,
leaving Black still stuck in the bind.
Finally, if 34 ... Rh7, then 35 Bxh7+ Kxh7 36 Re6 should win for White
in the long run; but why on earth would you play like that rather than keep
the black rook in torment with 35 Kxb2 etc.
Notice how the thoughtless 34 Kxb2?? in the initial position would have
allowed Black to solve his problems with 34 ... h5 and 35 ... Rh6 etc.

Game 2
P.Nikac-P.Dukaczewski
IBCA Championship, Cagliari 2019
 It’s White to move. The diagram position is not yet an endgame, but
White is going to show us how to exchange to victory: when you have shut an
opponent’s piece out of the game, the simplest way to win is often to swap
off the other pieces to leave his army hopelessly outnumbered in an endgame.

Question: But to start with, can you see the move with which White
destroyed
the co-ordination of the black pieces?

White has a huge advantage in the diagram position. His knight

dominates the entire board and can never be dislodged. In contrast the black
horse is denied any active squares: it cannot go to b6 and, thanks to White’s
kingside pawns, is shut out of both e5 and f6. White’s bishop is also far
superior to its opposite number which is almost boxed in on f8.
I happened to watch this game live in the tournament hall at the 2019
Blind and Visually Impaired Championships. Influenced by the nasty hole in
Black’s structure on f6 and White’s mobile pawns, I entertained myself by
trying to calculate a decisive breakthrough for White. For example, 24 fxg6
hxg6 25 Bb6! (a neat shot to deflect the black knight from its defence of f6)
25 ... Nxb6 26 Nf6+ Kg7 27 Qh3 forces Black to surrender his queen to stop
mate on h6 or h7. However, 24 ... fxg6! gives Black’s king a breathing space
on f7 and lets him fight on, even if his position remains very unpleasant.
When IM Predrag Nikac made his move it was much simpler and
stronger than what I had been looking at.
Answer: 24 f6!
Plugging the hole on f6 and renouncing the opportunity to open lines of
attack with 24 fxg6. Nonetheless, this shows excellent judgment. There was
no need to bother organizing a direct onslaught on the black king with all its
attendant risks. Instead, White can win effortlessly by entombing the black
bishop on f8, and then exchanging off all the other pieces. It really is very,
very easy as the remaining moves show:
24 ... Nb8 25 Qg4 Qd7 26 Qxd7 Nxd7 27 Rxc8 Rxc8 28 Rc1 Rxc1+
In principle Black should try to keep some pieces on the board, but 28 ...
Rd8 29 Rc7 is ghastly for him.
29 Bxc1 h6 30 h4 hxg5 31 hxg5
 The multiple exchanges have accentuated the disadvantage of having an
immured bishop. We can even express it numerically: before the exchange of
queens and rooks both players had pieces (excluding pawns) worth 25 points
in total, whereas now it is down to six points (i.e. a knight and bishop each).
But let’s say Black’s imprisoned bishop is only really worth one point (rather
than three points), because it is little better than a pawn, while White’s bishop
is worth the full three points. We’ll be generous to the black knight and say it
is worth three points, the same as the white knight.
Then the numerical balance in forces has changed from 25:23 in White’s
favour before the exchanges to 6:4 after them. In other words, White’s
superiority has grown from around 8% to 33%. And this has happened not
through winning any more pieces, but simply through exchanging pieces off.
In real terms this means that there are no longer other black pieces, apart
from the knight, available to cover for the inert bishop when it comes to
defending key squares and pawns. And the knight is outnumbered by the
white pieces by 2-1. Well, it is actually worse than 2-1, as the imprisonment
of Black’s bishop means that his king is also kept out of the game. White can
win by advancing his king to capture the e4-pawn. What he should not do is
go after the a6-pawn with Nc7 too quickly, as this might allow Black to free
his bishop by ... d6-d5 (he would remain lost, but it would be criminal to give
the bishop the slightest activity).
31 ... Bh6
Evidently Dukaczewski shares my opinion of the ‘real’ value of his
bishop.
32 gxh6
Even 32 Kf2 would be good enough, heading for the e4-pawn.
32 ... Kh7 33 Kf2 1-0

A similar numerical reasoning to that at move 31 above explains the

advice: When you are a piece up, exchange pieces, not pawns. When you are
a piece down, exchange pawns, not pieces.
You can see an example of Caruana saving himself by exchanging off
Carlsen’s last pawn in Chapter Three.
My mistake when analysing the position before White’s 24th move was
trying to make the white pieces do something, when the correct plan was to
stop one of the black pieces from doing something. Incidentally, I think that
is a common fault in post-mortems and when kibitzing other players’ games.
After a game we sit down and focus on its interesting moments, maybe
investigate a sacrifice not played, or a lively episode. Given the chance to
explore spectacular opportunities, we aren’t looking for steady prophylaxis.

Breaking Resistance by Throwing a Knight into the Attack

Famous generals such as Julius Caesar and Napoleon showed impeccable
judgment when it came to choosing the right time and place to throw their
reserves into a battle. For example, the soldiers on the far left of an enemy’s
defensive line might be holding out resolutely against a frontal attack. Then,
suddenly, fresh hostile forces spring out of the cover of a forest at their back
and charge towards them.
The shock is too much for the defenders. There is panic and disorder and
they turn and run away. This might save their lives (for a time at least) but it
dooms the rest of their army by opening up a gap in their defences.
In chess strategy the entrance of a reserve piece can overwhelm what had
hitherto been a watertight defence. It becomes the proverbial straw that
breaks the camel’s back.
Game 3
V.Rasik-V.Laznicka
Czech Championship, Karlovy Vary 2004

Question: It’s White to move. How do you overwhelm the defence?

Answer: Black’s rooks are just about holding out against the white passed
pawn. But with their queen hopelessly placed to offer any help, a charging
horse will provide the extra pressure to break their resistance.
30 Nd4! Kf8
The threat was 31 Nxe6+.
31 Nc6 Kg7 32 Na7 1-0
The barrier on c8 collapses. White wins a rook (to start with!).

Game 4
F.Caruana-S.Shankland
US Championship, St. Louis 2016

Question: It is White to play. Assess the position and then suggest a

long-range plan for White.

Answer: Black’s queen and rook are tied down to the defence of their
knight. The knight itself has no safe moves going forwards or sideways,
while retreating to d8 would drop the rook and going to b8 allow Ra8 with a
fatal pin. Meanwhile, the bishop is boxed in on f7 – after Black’s next move
it can only hobble between f7 and g8.
The black pieces are therefore completely dominated. Nevertheless, they
are holding on, if only by a thread. If Caruana wants to win he has to find a
way to snip that thread. His queen, rook and bishop are all excellently placed.
What he needs now is the help of his knight. It begins a long manoeuvre right
the way to b5, when the threat is Nc7, cutting off the defence of the black
knight by the rook.
43 Nf4 Kh7 44 Ng2!
A bit of readjustment as the route to b5 will be through e3.
44 ... Bg8 45 Ne3 Bf7 46 Nc2 Bg8 47 Na3 Bf7 48 Kf2!
Black is apparently paralysed, but not quite. If 48 Nb5?, the black knight
escapes its shackles with 48 ... Ne7!. Then 49 Qxe7 Qxb5 50 Ra7 would win
the bishop, but Black has enough for perpetual check with 50 ... Qb1,
intending 51 ... Qg1+ etc. If instead 49 Bxe7, Black regains the piece with 49
... Rb8 as White’s knight will fall. Finally, if 49 Ra7 then 49 ... Nf5+ (note
the importance of this being check as otherwise f7 would fall) 50 Kh3 Bg8
and Black has much improved his chances by getting his previously terrible
knight to the f5-square.
You always have to be vigilant when you have the opponent’s pieces
under lock and key. A resourceful opponent will be waiting for just such a
slip-up as 48 Nb5?.
Caruana decides that the best way to prevent any risk of a perpetual
check, such as we saw in a variation above, is to move his king over to the
queenside. The king will also be well out of the way of ... Nf5 coming with
check.
48 ... Bg8 49 Ke3
There is no need for White to investigate lines like 49 Nb5 Qf7!? (a
heroic move) 50 Qxc8 Qf4, when 51 Qxc6? Qd2+ again gives perpetual
check.
49 ... Bf7 50 Kd2 Bg8 51 Kc2 Bf7 52 Kc1
Caruana finally has his king where he wants him.
52 ... Kg8
If 52 ... Bg8 53 Nb5 Qf7, in the style of the note to move 49, White can
just take all the pieces with 54 Qxc8 Qxf3 55 Qxc6 as his king can easily
escape perpetual by moving up the queenside.
53 Nb5
 After a rapid gallop from f4 to a3 in the space of five moves, the knight
had to wait on a3 for another six before completing the last step of its
journey. Now there’s nothing good Black can do against the threat of 54 Nc7,
blocking the rook’s defence of its knight.
53 ... Ne7 54 Na7! 1-0
As we saw above, 54 Qxe7? Qxb5 or 54 Bxe7? Rb8 (when b5 drops)
would be wrong steps for White. But now Black’s rook and knight are both
hanging and there are no more tricks.

The Entrance of the King is Decisive

In the endgame the advance of the king to a key strategic point is often of
crucial importance.

Game 5
Y.Habu-G.Kasparov
Rapid Match, Tokyo 2014
 Question: It is White’s move. Evaluate the position and decide on the
best
plans for White and Black. How should the players try to implement
them?

Answer: We might suggest:

Plan for White: Mobilize his pawns on the queenside to achieve
counterplay.
Plan for Black: Use his active king to force through his passed pawn
before White gains counterplay on the queenside.
Kasparov has a significant positional advantage. His king is well
centralized and ready to support the advance of his centre pawns. Meanwhile,
White’s king is relatively passive and his three pawns on the queenside are
held back by two black pawns with the help of their rook, making it difficult
for White to create a passed pawn.
Yoshiharu Habu (who, incidentally, is a top Shogi player as well as a
chess FIDE Master) tried to activate his pawn majority with:
41 a3
Now I suspect quite a few club players would want to play 41 ... Kf4 with
the plan of 42 ... e5 and 43 ... e4, when the black king is at hand to shepherd
the passed pawn to the queening square. But I also think they might get cold
feet about giving up the b4-pawn as occurs after 42 axb4 axb4 43 Rd4+ Kg3
44 Rxb4. White then has two connected passed pawns. What if it turns out
that he finds a way to stop the e-pawn and pushes these pawns to victory?
After the game Black might be uttering some of the saddest words ever
spoken by a chess player: “I lost a game it was impossible to lose!”
So, if Black isn’t 100% sure whether 41 ... Kf4 is winning, he might
decide to play 41 ... Rb7 to guard b4 in preparation for the king move.
However, White can then break out from the bind with 42 axb4 axb4 43 c3!,
followed by 44 cxb4 unless Black exchanges on c3. In any case the white
king will be able to help the resulting passed pawn advance and, in some
cases, oppose the advance of Black’s e-pawn.
It is one of those situations where every tempo matters. You can’t play it
safe. Kasparov trusted in his intuition and calculation and put his king on f4
straight away:
41 ... Kf4!
In Shereshevsky’s fine book Endgame Strategy there is a chapter called
“Do not hurry!” Our motto should be: do not hurry – except when you must
hurry!
42 axb4 axb4 43 Rd4+ Kg3 44 Rxb4 e5 45 Rc4
The best try, hoping for 45 ... Rxc4? 46 bxc4, when both sides will queen.
45 ... Re7!
As Reuben Fine says in Basic Chess Endings: a rook belongs behind a
passed pawn, whether it is your own pawn or the opponent’s pawn. Now
neither the white king nor the rook will be able to stop the black pawn.
46 b4
This is painfully slow compared with the fleet-footed e-pawn.
46 ... e4 47 fxe4 fxe4
 48 b5
He might have tried 48 Kc1, when 48 ... e3? 49 Kd1 blocks the pawn as
49 ... Kf2? 50 Rf4+ drives the black king away with advantage to White. If
instead 48 ... Kf2?!, White has 49 Rc5 to harass the king with 50 Rf5+, when
he is still resisting. So the most efficient way for Black to win is to utilize the
f-pawn: 48 ... f5! 49 b5 f4 50 b6 e3 51 Kd1 f3 52 gxf3 Kf2! and the e-pawn
will queen. Note the way Black has sacrificed his f-pawn to create a shelter
from checks on the f-file which confounded his king in the previous lines.
48 ... e3 0-1
White can give up his rook for the e-pawn beginning with, say, 49 Rc3.
Here’s a possible finish: 49 ... Kxg2 (breaking the pin on e3, pocketing
another pawn, and clearing the way for the f-pawn) 50 Rxe3 Rxe3 51 c4 f5
52 b6 Re8 53 c5 f4 54 c6 f3 55 c7 f2 56 b7 f1Q 57 c8Q Qb5+ 58 Kc1 Re1+
and the white king will soon be mated.
Did Kasparov calculate this line during the game? It’s possible, but right
from 41 ... Kf4! I expect he felt in his bones that he was winning. It would be
strange if, with the white king so passive and the black e-pawn so fast, White
could rely on his passed pawns to save him.
You can see how the pace at which a player is required to carry out his
plan can change during the course of an endgame. After 41 a3 Black had no
time to lose: the position compelled him to play 41 ... Kf4! or else forfeit his
advantage. But in the note to move 48, if he had played the direct 48 ... e3 or
48 ... Kf2 he would have at least complicated the win. What was required at
this point was the slow advance of the f-pawn with 48 ... f5, 49 ... f4 and 51
... f3.

Game 6
S.Karjakin-W.So
FIDE Candidates, Berlin 2018

It’s White to play. “A child could have drawn this endgame,” said Wesley
So ruefully after the game.

Question: Nonetheless, can you list the sources of danger for Black in the
 diagram position?

Answer: With severely limited material, no pawns on the queenside, and

an almost symmetrical pawn structure on the kingside it is hard to imagine
that Black won’t hold comfortably. What are White’s advantages? Let’s
compare the pieces:
Rooks: White has a rook on the seventh rank which can’t be easily
challenged. Meanwhile, the black rook is passive.
Knights: The white knight is preventing Black from activating his rook
with ... Rc8 as Ne7+ is a deadly fork. But there is more to the knight’s power
than a tactic to restrain the rook. As we shall see, the knight is excellently
placed to attack Black’s pawns on both e6 and g7 in key lines. In contrast, the
black knight is a long way from home. It will require several moves to be
brought back to the kingside and established on a decent square there.
Kings: The white king can be used to support an aggressive plan of action
by marching up the board. The unhappy black king is tied down by the white
rook and cannot leave its first rank.
Pawns: As is usually the case, when a player has the better pieces his
pawns have more energy than the opponent’s, even if the structure is almost
symmetrical. White’s e-pawn is the spearhead of his attack; Black’s e-pawn
is a liability requiring careful defence.

We can see from the assessment above that the white rook and knight are
already on optimum squares. So our task is to utilize our two reserve assets:
the king and the pawns.
27 e4!
 The first step is to gain space. White intends f2-f4 and e4-e5 to establish a
pawn on e5. Then the black e6-pawn will be fixed as a permanent target. For
example, White could play Nd4 and tie down a black piece (very likely the
rook) to the pawn’s defence. Once White gets a pawn to e5 and brings his
king forward, Black can’t rely on simplification to save him – a king and
pawn endgame could turn out to be lost, as in the note to 32 ... Nh4 below.
27 ... Nc4
A sensible move to return the knight to the rest of Black’s army. In view
of the comment above, you might wonder why Black didn’t play 27 ... e5.
This puts the pawn on a safe square (that is, defended by its fellow pawn on
f6) and, at the same time, prevents the plan of f2-f4 and e4-e5 which aims to
fix it on e6 where it has to be defended by the rook.

Question: So how should White respond to 27 ... e5 - ?

Answer: Here again we see the power of the white knight. After 27 ... e5
28 Ne7+ Kh8 (or 28 ... Kf8 29 Nf5 likewise) 29 Nf5 it takes up a tremendous
post, attacking the g7-pawn. Now 29 ... Rg8 leaves the black rook and king
entirely boxed in, while after 29 ... g6 30 Ne3 it’s bad enough that Black has
a weak pawn on f6 that can be attacked by Nd5 or Ng4, but he also has the
problem of how to get his knight to safety as the c4 retreat square has been
removed.
28 Kd3
Bringing the king forwards. If now 28 ... Ne5+ 29 Nxe5 fxe5, Black has
got rid of the pesky white knight but loses a pawn after 30 Ra5.
28 ... Nd6 29 f4

Continuing with the plan outlined above. Instead, after 29 e5 fxe5 30

Nxe5 White has given Black an isolated pawn and placed his knight on a
strong post in front of it. However, it also reduces the cramp in Black’s
position and gives his pieces a bit more breathing space. Establishing a pawn
on e5 and then advancing the king to support the knight and rook keeps more
tension. It is therefore from a practical point of view the more promising
plan. Besides, if e6 falls then White will already have a passed pawn.
 Question: Can you work out why 29 ... h5 would be a good defensive
pawn
move for Black?

29 ... Kf8
Black is impatient to free his rook and so prepares ... Rc8 without
allowing the fork on e7.
Answer: Instead, 29 ... h5! looks a good defensive move. After 30 e5 fxe5
31 fxe5 Nf5 the black knight has found a fairly stable post as it can’t be
kicked away by g2-g4 as in the game. This line shows the black pawns
haven’t lost all their vitality. Of course White doesn’t have to force the issue
with 30 e5. He could, for example, probe with 30 Ra5 h4 31 Ra7 in the hope
that the h4-pawn becomes a weakness in the future.
30 e5 fxe5 31 fxe5
White finally completes his plan of getting a pawn to e5. The e6-pawn is
now fixed.
31 ... Nf5 32 g4!
 Due to Black’s omission of 29 ... h5 his knight now has to continue its
wandering.
32 ... Nh4

Question: Was there any hope in 32 ... Ne7 - ? Try to work out the king
and
pawn endgame that arises after exchanges on e7.

Answer: We see the power of the white e-pawn in the following variation
which shows Black can’t escape from the pressure through simplification: 32
... Ne7? 33 Nxe7 Rxe7 34 Rxe7 Kxe7 35 Kd4 Kd7 36 Kc5 Kc7 (Black has
the opposition and so keeps the enemy king out for the moment, but he is
going to run out of pawn moves and then have to give way) 37 g5 h6 38 gxh6
gxh6 39 h3 h5 40 h4 Kd7 (no more pawn moves, so Black will lose the e6-
pawn) 41 Kb6 Kd8 42 Kc6 Ke7 43 Kc7 Ke8 44 Kd6 Kf7 45 Kd7 Kf8 46
Kxe6 and White wins.
33 Kc4
Because there are so few pieces White has to make every unit work as
hard as possible. Now there are sinister ideas such as putting the king on d7,
in order to smother the black rook, or playing Kd6 and Nd4 to take on e6.
(Imagine if the king was already teleported to d6 with all the other pieces on
the same squares; then after 1 Nd4 the black rook would jump at the chance
to have revenge on the white knight with 1 ... Rd8+ 2 Kxe6 Rxd4, but alas his
king would be mated by 3 Ra8+ etc.)
33 ... Nf3
Trying to get the knight activated by attacking h2 and putting pressure on
e5.
34 Ra2
Threatening 35 Rf2. Black finds the only good defence.
34 ... Rc8! 35 Kb5
 35 ... Ke8?

Question: Black had only one good defence and it’s difficult to find. But
have
a go anyway!

After 35 ... Ng5 36 Kb6 White can drive the black rook to e8 with 37
Kb7, then close in on it (and the e6-pawn) with 38 Kc7 and 39 Kd7.
However, there remained one fully adequate defence for Black:
Answer: 35 ... Rc7! and then:
a) 36 Rf2 Rf7 and if 37 Nd8? it’s trapper trapped after 37 ... Nd4+, while
37 Kb6? drops a pawn to 37 ... Nxe5!. White would have to run away with 37
Ra2, say, when 37 ... Rd7 is equal.
b) 36 Kb6 Rd7 (not 36 ... Rf7?? 37 Ra8 mate!) 37 Rf2 Rd3!. Perhaps
Wesley So missed this alternative way to defend the knight and keep the rook
active. White has no way to make progress.
36 Kb6
 The black rook is paralysed by the white king and knight.
36 ... g5
If 36 ... Kf7 then 37 Rf2 wins the knight, while after 36 ... Kd7 37 Ra7+
the rook can scoff the kingside pawns while the black rook remains trapped.
That leaves the exchange sac 36 ... Rxc6+, but it is hopeless: 37 Kxc6
Nxe5+ 38 Kd6 Nxg4 39 Kxe6 Kd8 (an unfortunate necessity to stop mate) 40
h3 Nf6 41 Ra8+ Kc7 42 Ra7+ Kd8 43 Rxg7 Ne4 44 Ke5 (not falling for 44
Rxh7?? Ng5+) 44 ... Nf2 45 Rxh7 and White wins.
37 h3

37 ... Nxe5
Desperation. By defending g4 with his previous move Karjakin has made
an exchange sacrifice fruitless after 37 ... Rxc6+ 38 Kxc6 Nxe5+ 39 Kd6. If
instead 37 ... h6, White can go hunting the kingside pawns with 38 Ra7 and
39 Rh7, or simply play 38 Ra1, when 38 ... Nh4 39 Kb7 is lethal as 39 ... Kd7
40 Rd1+ wins the rook.
38 Nxe5 Rc3 39 Rh2
You can afford to play humble moves when you are a piece up.
 39 ... Ke7 40 Kb5 Re3 1-0
Having reached the time control, Black resigned. White will exploit his
knight after 41 Nc4 etc.

Don’t Forget to use Your Pawns!

In the following example Black spurned the chance to call up a reserve piece.
He didn’t notice the beauty of Cinderella sitting modestly on h6 because he
was mesmerized by her flamboyant, attention-grabbing sisters: namely the
far-advanced passed pawns. Here’s a maxim which will be repeated several
times in this book: every plan needs the use of pawns at some point.

Game 7
J.Valmana Canto-P.Harikrishna
Spanish League 2006

It is White to move. He will win easily on the queenside if he manages to

give up his knight to eliminate or neutralize Black’s passed pawns.
 First of all let’s see how the game finished:
34 Ne2 f3 35 c4 a5
Black is already helpless. After 35 ... fxe2+ 36 Kxe2 h5 (or equally 36 ...
Kf4 37 c5 Ke5 38 Kxe3) 37 c5 Ke5 38 Kxe3 Kd5 30 b4, White wins the h5-
pawn at his leisure. Meanwhile, the h-pawn is one move too slow to save
Black: 35 ... h5 36 c5 h4 (or 36 ... Kf5 37 Nd4+ Ke4 38 c6 Kxd4 39 c7 and
wins) 37 c6 h3 38 c7 h2 39 c8Q+ with a fatal check.
36 a3 f2 37 c5 Kf5 38 Ng3+ Kf4
Or 38 ... Ke5 39 Ke2 Kf4 40 Nf1 and e3 drops.
39 Kg2 h5 40 Nxh5+ Ke4 41 Ng3+ 1-0

Question: Returning to the position after 34 Ne2, can you speed up

Black’s
kingside counterplay?

Answer: You might have noticed how the h-pawn is one move too slow
in the note to 35 ... a5. So, after 34 Ne2, let’s try 34 ... h5! at once. Then play
is more or less forced: 35 c4 h4 36 c5 h3 37 c6 h2 38 Kg2 (White is forced to
blink in the race between the pawns) 38 ... f3+ 39 Kxh2 fxe2 40 c7 e1Q 41
c8Q+ and again White queens with check, but after 41 ... Kf3 he is lost due to
Black’s excellently placed queen and king supporting their fearsome passed
pawn; e.g. 42 Qf5+ Ke2 43 Qc2+ Qd2 44 Qc4+ Kd1+ 45 Kh3 e2.
The move 34 ... f3? looks very natural, not least because it attacks the
white knight, but it turned out to be the fatal loss of a tempo. As we remarked
above, the pawn on h6 is like Cinderella, overlooked because of the loud and
glitzy pawns on e3 and f4. But she was the key to saving – and winning – the
game.
But we haven’t finished with this endgame.

Question: Returning to the diagram position before 34 Ne2, can you find
a
better plan for White?

Answer: White needed to block the advance of Black’s pawns with the
set-up: king on e2 and knight on f3. This can be done with 34 Ke2! (34 Nb5!
is also good enough) 34 ... h5 35 Nb5! h4 36 Nd4 h3 37 Nf3.
 It might seem a bit profligate to move the white king and then play three
moves with the knight when there are passed pawns racing down the board.
But White has achieve a blockade of the black pawns which holds out long
enough for the c-pawn to come to his rescue: 37 ... Kg3 38 c4 h2 39 Nxh2
Kxh2 40 c5 Kg2 41 c6 f3+ 42 Kxe3 f2 43 c7 f1Q 44 c8Q Qe1+. Both sides
have queened and it’s a draw after 45 Kf4 Qd2+ 46 Ke4 Qxa2.
An instructive endgame. In Think Like A Grandmaster Kotov relates how
he and some colleagues were stumped when trying to find the winning
continuation in an endgame they were studying. They therefore asked
Capablanca’s advice. After a quick glance at the position, the Cuban world
champion simply pushed the pieces to the squares where they were needed
for the implementation of the correct plan. He then walked away and left the
other players to find the individual moves.
Let’s imagine the ghost of Capablanca was revisiting San Sebastian,
where the game above was played. (He had his great debut success there and
even ghosts get nostalgic.) He might have been tempted to push the white
king to e2 and the knight to f3 and then left the (terrified) players to find the
moves.
The moves you need to make are not too difficult to find once you know
the correct set-up, as Kotov also remarked about the set-up Capa showed
him.
Finally, it should be mentioned that 34 Ne2 is the proverbial ‘second-to-
last mistake’ which wins the game. If White had played the correct 34 Ke2,
Harikrishna would have replied 34 ... h5 and drawn, not having had the
opportunity to lose (after 34 Ne2?) with 34 ... f3?.

Game 8
C.N.Ross-E.Spinu
IBCA Olympiad, Ohrid 2017

A position with only six pieces but full of complexity. It turns out that
White can’t win by capturing both black pawns with his rook, as after 55
Rh8? f2 56 Rxh4+?? (instead, 56 Rf8! still holds a draw) 56 ... Ke3 57 Rh3+
Ke2 the black f-pawn queens.
 Question: So can you think up a winning scheme for White?

Answer: The limited material on the board allows us to plan a long way
ahead. If White is to win, the process has to be:
1) Give up the rook for the f-pawn.
2) Capture the black h-pawn with his king.
3) Get his king off the h-file without being blocked in by the returning
black king.
4) Use his king to shut out the black king from stopping the white h-pawn
from queening.
As soon as the black king commits to supporting the advance of the f-
pawn a line of approach will open for the white king towards the h-file.
First of all, let’s see how the game ended:
55 Rf8
With the white king temporarily denied access to d5, it seems eminently
logical to begin by attacking the f-pawn, as the rook will have to go to f8
sooner or later to capture it. Now both kings go about their business. The
black king advances to support the f-pawn, while the white king tags along in
his wake and edges closer to the two h-pawns.
55 ... Ke3 56 Kd5 f2 57 h3
With his own king stymied, White moves his pawn and waits for the
black king to give way again.
57 ... Ke2 58 Ke4 f1Q 59 Rxf1 Kxf1
 Question: Black has regained the rook. Is that enough to save the game?

Answer: To clinch the draw Black needs either to get his king back in
time to stop the h-pawn from queening or shut in the white king on the h-file
in front of the pawn. Alas for him, the journey proves too far after:
60 Kf3!
The only way. He has to head for the h-file while blocking for a vital
move the black king’s approach.
You can verify that it’s a draw after 60 Kf4? Kf2 61 Kg4 Ke3 62 Kxh4
Kf4 – the white king has been successfully shut in on the h-file. The black
king can move towards f8, and the white king will remain shut in on the h-
file if he tries to keep his opposite number from going to g8 and h8 to block
the h-pawn.
60 ... Ke1
Around about here Black started banging the clock loudly, either out of
frustration or the hope it would make his king jump a bit faster down the
board.
61 Kg4 Kf2 62 Kxh4 Ke3 63 Kg5
Just in time, before 63 ... Kf4 imprisons him. Black loses because his king
hasn’t been able to prevent the white king leaving the h-file and clearing the
way for the h-pawn to advance.
63 ... Ke4 64 h4 Ke5 65 h5 1-0
The white king can prevent the black king getting to f8 after 65 ... Ke6 66
Kg6 etc.

From the above we can conclude that, if he is somehow going to draw,

Black needs to shorten the journey his king has to take to shut in the white
king on the h-file (as in the note to 63 Kf3).
Going back to the initial position imagine that, after 55 Rf8, Black had
played 55 ... h3.

Question: How does that affect the outcome of the game?

 Answer: We have a similar scenario to the game after 56 Kc5 (White has
to mark time for a move, so he can’t exploit Black’s delay in playing his king
to e3) 56 ... Ke3 57 Kd5 f2 58 Ke5 Ke2 59 Ke4 f1Q 60 Rxf1 Kxf1 61 Kf3.
This is the same as after 60 Kf3 in the game, but with the pawns on h2
and h3 rather than h3 and h4. This means that the black king has a one move
shorter journey to shut in the white king on the h-file: 61 ... Ke1 62 Kg3 Ke2
63 Kxh3 Kf3 and he arrives in time, with a draw as in the note to move 60.
So 55 ... h3! would have drawn.

Returning to the first diagram, instead of 55 Rf8, White could have

played 55 h3! at once. Then after 55 ... f2 56 Rf8 Ke3 57 Kd5 Ke2 58 Ke4
Ke1 (if 58 ... f1Q 59 Rxf1 Kxf1 60 Kf3, White wins as above) 59 Ke3!
(instead, 59 Rxf2? Kxf2 is only a draw as the black king is one square too
near and can shut in the white king in the style we have already seen) 59 ...
f1Q 60 Rxf1+ Kxf1, White wins as the black king ends up too far away on f1.
Finally, White would also have been winning after 55 ... Kf4 56 Kd5 f2
57 Rf8+ Kg3 58 Ke4 Kxh3 59 Rxf2.

I was Chris Ross’ Second at the Chess Olympiad for Visually Impaired
players in 2017. It’s fair to say I was watching this game with turbulent
emotions as, due to White’s omission of h2-h3! and Black’s omission of ...
h4-h3!, the assessment went from White’s point of view: 55 Rf8 (equal!) 55
... Ke3 (winning for White!) 56 Kd5 (equal!) 56 ... f2 (winning again!), when
Chris finally played 57 h3 and I could relax.

Getting Maximum Value from an Active Bishop

Game 9
Ma.Carlsen-Yu Yangyi
Stavanger 2019
 Question: It is White to play. Would you grab the pawn on c6?

Answer: Well, after 23 Bxc6 Rc8 you can support the bishop with 24 d5
(this is necessary as the c-pawn will drop if the bishop retreats). Then White
has connected passed pawns in the centre. Surely that is the way to go? Let’s
see how the game might continue:
24 ... Rb8
Black has an active rook, whose power can be further increased by 25 ...
Rb2, seizing the seventh rank. Meanwhile, White has been forced to advance
24 d5 which has blocked in his bishop on c6 and conceded the c5-square to
Black’s bishop. We could imagine Black generating counterplay with ... Rb2
and ... Bc5, attacking the pawn on f2.
The white pawns have been compromised. They have lost their
dynamism. Imagine if White’s pawn were still on d4. Then he would have the
plan of Bf3 followed by the pawn advance c3-c4-c5. His pawns would roll
through the centre. Instead, once White is committed to 24 d5, Black has a
dark square blockade. Nimzowitsch remarked many years ago that when a
pawn structure is prevented from advancing, the square to which it is
prevented from advancing automatically becomes a strong point for the
opponent. Here c5 is a great square for the black bishop.
An active black rook, a black bishop with a great blockade square, a
neutered white centre, a white bishop inert on c6 – well, 23 Bxc6 isn’t
looking so attractive now. Note that Black has to throw in the move 23 ...
Rc8! to force 24 d5 before playing 24 ... Rb8. If he played 23 ... Rb8 at once,
then 24 c4 gets the pawns rolling, and if 24 ... Rc8 then 25 Bd5! (under no
circumstances 25 d5) and White is ready for 26 c5.
Instead, Carlsen played 23 c4!

Rather than be turned into weaklings the white centre pawns maintain
their potency.
23 ... Rc8
What else? After 23 ... Rb8 24 c5 Be7 25 Bxc6 the white centre can’t be
restrained: 25 ... Rd8 is met by 26 Ba4!, guarding the rook on d1 to rule out
26 ... Bxc5. White can then build up by centralizing his king.
 24 c5 Be7
By avoiding the temptation to take on c6, Carlsen has reduced the black
rook to a defensive role and denied the black bishop a safe and strong square
on c5. It is now time to bring the king into the battle.
25 Kf1 f5
Yu Yangyi is a world-class player. He knows his only chance is to utilize
his kingside pawn majority to dislodge the white bishop from f3 and with
luck (a commodity normally in short supply when facing Carlsen) create a
passed pawn.
26 Ke2 g5 27 h3 Kg7 28 Kd3 Kg6
With the white king on its optimum square defending the d4-pawn it is
now time to activate the rook on the seventh rank.
29 Rb1 h5 30 Rb7 Bf6

Question: Now White can grab material with 31 Rxa7. Any thoughts on
why Carlsen declined the opportunity?
 31 Rd7!
Once again we see Carlsen spurn the chance of bagging a pawn in order
to prevent his opponent freeing his rook from the defence of c6.
Answer: If 31 Rxa7 then 31 ... g4 32 hxg4 hxg4 33 Be2 (alas for White
he can’t play the bishop manoeuvre we see in the game, as after 33 Bd1
Black has a skewer with 33 ... Bxd4!, when 34 Kxd4 Rd8+ 35 Ke3 Rxd1
regains the piece with good drawing chances due to his nimble rook) 33 ...
Rd8 34 Ra4 Rh8 and the rook will generate counterplay with 35 ... Rh2.
Things would remain difficult for Black, but Carlsen wants to give him
no hope whatsoever: Slow but inexorable is the motto of the world champion
in a winning position. By putting the rook on d7 he stops any skewer with ...
Bxd4 after his next move and threatens to win the key pawn on c6 after 32
Rd6. Black therefore has little choice but to drive the white bishop back.
31 ... g4 32 Bd1! Kg5
32 ... Rb8 33 Ba4 Rb2 34 Bxc6 is also hopeless for Black, as White’s
passed pawns are much too fast.
33 Ba4!
The bishop has ricocheted from f3 to d1 to a4 and is once again attacking
the c6-pawn.
33 ... f4
 Question: How should we put a stop Black’s kingside counterplay?

Answer: 34 f3!
Yu was aiming to create a passed pawn with 34 ... f3 35 gxf3 gxh3. Here
we see the inherent defect of a pawn majority that contains a doubled pawn:
even if Black got his pawn from f7 to f5 he still couldn’t create a passed
pawn just by advancing his pawn mass, assuming his opponent was vigilant.
34 ... Re8
A last attempt for activity.
35 fxg4 hxg4 36 hxg4 Re6
Or 36 ... Kxg4 37 Bxc6 and White’s bishop very conveniently defends the
g2-pawn against attack by 37 ... Kg3.
37 Bd1!
Now the bishop keeps the black king out by guarding g4.
37 ... Re3+ 38 Kc4 a5 39 Bf3
And here we are again, with the bishop back on f3 attacking c6. It also
destroys Black’s last chance of gaining counterplay by capturing the pawn on
g2.
I wonder what odds you could get on Magnus playing 39 d5??, allowing
39 ... Rc3 mate.
39 ... Ra3 40 Bxc6 Rxa2 41 Be4

The c-pawn will win the pawn race, as White’s king is at hand to deal
with the a-pawn.
41 ... a4 42 c6 Ra1 43 c7 a3 44 Kb3! 1-0
But not 44 c8Q?? Rc1+ 45 Kb3 Rxc8.

Using a Rook to Break into the Enemy Camp

Game 10
V.Kramnik-S.Shankland
Wijk aan Zee 2019
 Kramnik decided he would retire from classical chess at the end of the
2019 Tata Steel tournament, thus bringing down the curtain on a wonderful
career which included wresting the world championship from Kasparov in
2000. This game was played in the last round of the tournament. There is
something hard about saying goodbye to something which has defined your
life, as Kasparov found when he lost the last game of his classical career to
Topalov at Linares in 2005.
Kramnik fared no better. In the diagram position, instead of forcing an
immediate draw with 50 Bxa6 Rxa6 51 Rxb2, he “raged against the dying of
the light” with:
50 Ba2?
Despite White’s nominal material advantage, he has no chance to win due
to the power of the passed pawn on b2. His rook and bishop are both tied
down to stopping the pawn. Perhaps he hoped that his king would be able to
enter the fray and release his pieces but, as we shall see, that isn’t possible.
Furthermore, Black’s king can easily advance to d4, and the space advantage
conferred by the pawn on e4 becomes a vital factor.
 Question: How do you think Black built up his game?

Answer: Let’s see how Shankland exploited his advantage. He devised a

plan made up of a number of steps.
Step one: Advance the a-pawn to a3, where it defends the b2-pawn. This
will free the black rook from guard duty and create the constant possibility of
a breakthrough with ... a3-a2 should the white bishop be distracted or forced
from its blocking role.
50 ... a5 51 Kf1 a4 52 Ke2 a3 53 Kd2
The white king arrives one move too late to relieve his pieces of their
defensive duty by going to c2. This is because:
Step two: Cut off the white king from approaching the pawns.
53 ... Rc6! 54 h4

Step three: Bring the black king to d4.

54 ... Ke5 55 Re1
Or 55 Ke3 Rc3+ 56 Kd2 Kd4.
 55 ... Kd4 56 Bb1
Step four: Activate the rook on Black’s sixth rank and use it to drive the
white king away from the passed pawns. This involves checks and an attack
on f2.
Note that Kramnik has given his pieces a semblance of activity. If 56 Rb1
then 56 ... Rc3 intending 57 ... Rf3 would begin a procedure similar to that in
the game. The white king would be forced to e2 and then the black king
could infiltrate to c3 and then to c2.
56 ... Rc3
It’s vital that if White ever answers ... Rf3 with Rf1 while his king is on
d2, Black has the trick ... Rxf2+! and ... e4-e3+, regaining the rook by a fork
and then winning by breaking through with ... Kc3-b3, followed by ... a3-a2
(or with ... Kc3-c2 if the white bishop is on a2).
57 Rh1 Rd3+
A repetition while he devises the right idea.
58 Kc2
Naturally, after 58 Bxd3 exd3 the passed pawns roll through.
 Question: Can you find a way to stop White’s king aiding his resistance
on
the queenside?

Answer: 58 ... Rc3+ 59 Kd2 Rf3!

Exactly – the white king must be enticed to e2 to defend f2 so that it is cut
off from going to c2 when the black rook returns to d3.
60 Ke2
The king has to give way, as if 60 Rf1 (or 60 Rh2) then 60 ... Rxf2+! 61
Rxf2 e3+ wins for Black (see the comment to 56 ... Rc3).
60 ... Rd3!
Mission accomplished. The white king is barred from the queenside.
61 h5
Waiting with the rook is hopeless; e.g. 61 Rg1 Kc3 62 Rh1 Rd2+ 63 Ke3
Rc2!, threatening 64 ... Rc1 to break the blockade. As usual the black king
and passed pawns overwhelm the defence after 64 Bxc2 Kxc2.
This sequence would have formed step five for Black, but Kramnik
induces Shankland to amend his plan by sacrificing the h-pawn.
61 ... gxh5
Step five: Create another passed pawn in the centre or on the kingside to
end resistance.
62 Ke1 Rc3!
If you are winning and can stop any counterplay, then do it. The threat of
63 ... Rc1 ties down the white rook and king. It would be crazy to get
involved in 62 ... Kc3?! 63 Rxh5, when due to White’s active rook the plan of
63 ... Rd2? 64 Rxf5 Rc2? rebounds horribly after 65 Rc5+.
63 Kd2 f4!
 Making use of all his resources.
64 Ba2
Now White will be defeated by a centre pawn rather than a wing pawn.
After 64 gxf4 Rc1! 65 Rxc1 (or 65 Re1 Rxb1! 66 Rxb1 a2 etc) 65 ... bxc1Q+
66 Kxc1 h4, the h-pawn can’t be stopped: 67 f5 h3 68 f6 h2 69 f7 h1Q+ and
Black wins.
64 ... e3+ 65 fxe3+ fxe3+ 66 Ke2 Rc2+ 0-1
67 Kf1 Rc1+ 68 Kg2 Rxh1 69 Kxh1 e2 the e-pawn queens, while 67 Kf3
Rf2 is mate.
A highly instructive example of planning from Shankland.

An Example of Co-Ordination from the Middlegame

The technique of overwhelming the defence by bringing reserves into the
battle can be applied at any stage of the game. To end with here is an
example from the early middlegame

Game 11
A.Giri-F.Caruana
Wijk aan Zee 2013

It is White’s move. Caruana is reeling after being caught in a tricky

opening line.

Question: Can you suggest what White’s plan should be and how to
implement it?

Answer: It would be illogical for Giri to try to attack on the queenside.

Playing in the centre with 24 Rad1 makes a lot of sense. But why not attack
as far away as possible from the black queen? In chess there isn’t much of a
place for brotherly love: you should hit your opponent where he is weakest. If
White focuses his efforts on the kingside it’s pretty certain that no help for
the defence is going to come from the aforementioned black queen or the
rook on a8, while the black minor pieces are also some way off, especially
the bishop.
 In the game Giri piled his pieces in:
24 Bb2!
The bishop switches from staring at a brick wall on c5 to pointing at the
most vulnerable square on the black kingside.
24 ... a5
If Black prevents the next move with 24 ... b5 there are other ways to win,
the most direct being 25 Rg5 f6 26 Bd5+ Kh8 27 Rh5 h6 28 Qg6, threatening
mate in two with 29 Rxh6+. It’s never as good sign for the defence when a
computer program recommends a move like 28 ... Qxg3+ as the best way to
resist!
25 Ra4!

A splendid way to activate the rook. Instead, the computer program wants
to show off with the spectacular 25 Rh5 g6 26 Qf5!. The point is 26 ... gxf5
27 Rg5 mate; or 26 ... Ne5 27 Qf6, when the capture on e5 with 28 Bxe5
wins whether or not Black takes the rook; or finally 26 ... Qd6 27 Rxh7! Ne5
(to delay mate on h8) 28 Qf4 Kxh7 29 Bxe5 Qe7 30 Bf6, clearing the way for
the mating 31 Qh4+ unless Black gives up his queen.
 Nice variations to analyse. But if and when you have the chance to show
off against Caruana, or whoever is rated world number five at the time –
don’t do it! Play the moves that kill him dead in the quickest and simplest
manner. (Perhaps here I should point out I wish Caruana a long and
prosperous life; we are talking purely about chess matters.)
Don’t jeopardize a big win. On the other hand, if you are playing a
friendly club game or a tournament game of no great value, it could be worth
seeking adventure with 26 Qf5 to sharpen your tactical imagination.
25 ... Re8 26 Rg5 g6 27 Bd5

The attack on f7 is also an attack on g6, as the threat is 28 Rxg6+ hxg6 29

Qxg6+ Kf8 30 Qxf7 mate.
If you have a big advantage in firepower in a certain section of the board,
look for combinations. It’s no wonder the ‘brilliant’ 28 Rxg6+ appears when
it’s White’s queen, two rooks and two bishops versus Black’s shell of pawns
on f7, g6 and h7.
27 ... Kf8 28 Rf4 1-0
White’s intentions include 29 Rxf7+, or 28 ... Re7 29 Qc3 (threatening mate)
29 ... f6 (or 29 ... Ke8 30 Qh8+ Nf8 31 Bg7, winning the knight to start with)
30 Rxf6+ Nxf6 31 Qxf6+ Ke8 32 Bc6+ Kd8 33 Rd5+ and then 33 ... Kc7 34
Qd6 mate or 33 ... Bd7 34 Rxd7+ etc.
Chapter Two
Stopping the Opponent Playing Good
Moves
It is well known that positional chess isn’t just about playing good moves – it
is also about preventing the opponent playing good moves. This might mean
anticipating his plans in a general sense and taking measures to oppose or
defeat them.
But sometimes a simple step such as taking a square away from an enemy
piece can be of game-changing importance.

Zugzwang
The most extreme form of denying the opponent healthy options is to put him
in zugzwang. Then his pieces are completely dominated and any move he
makes will lead to disaster.

Game 12
V.Belov-B.Socko
Hastings 2004/05
 It’s Black to move. He isn’t material down or facing mate, but still Black
resigned.
The reason is zugzwang. His rook is shut in and his bishop and knight
have no safe moves. If he plays 44 ... Kf8 then 45 Nxd7+ Rxd7 46 Rb8 wins
the horse as 46 ... Rd8 47 d7+ is a decisive discovered check. This line only
works for White because the black king has moved from e8 – otherwise Nxd7
could be safely answered by ... Kxd7.
Black has a couple of pawn moves available, but White can quietly move
his king until they are exhausted.

Game 13
O.Tuka-P.Nikac
IBCA Championship, Cagliari 2019

It is Black to move. With two excellent bishops and a powerful concentration

of forces in the centre he surely has several ways to win. The method he
chose was surprising, elegant, and of a forcing nature.
 43 ... Rd3+!

Question: Can you see the idea behind this sacrifice?

44 Bxd3 exd3 45 Rc1 d2! 46 Kxd2

It seems that the IM from Montenegro has gone berserk, giving up first
the exchange and then his valuable passed pawn. But the point of his play
becomes clear after:
46 ... Kd4!
Answer: Suddenly White is in zugzwang and is doomed to drop a piece.
47 Rf1
Giving the knight up immediately. If instead 47 g4 Bxg4, White still can’t
escape the pin on c3; e.g. 48 Kc2 Bxc3 wins.
47 ... Bxc3+ 48 Kc1 Ke5
 My experience through the years is that two bishops almost always
triumph in the endgame against a rook and a pawn (other factors being more
or less equal), even if it doesn’t look too bad for the defender to start with.
And in this case things already look bad for White because of his inactive
king and fatally weak pawn on c5. Tuka’s sacrifice to activate his rook will
prove inadequate.
49 g4 Bxg4 50 Rf7 h5 51 Rxb7 Kd5 52 a4
I recall Capablanca observing that a passed pawn is either strong or weak,
and whichever it is becomes more pronounced as it advances up the board.
Here White has to use the a-pawn to generate counterplay, but it will become
like a lamb facing two wolves disguised as prelates.
52 ... Kxc5 53 Kc2 Be1 54 Kd3 Bxh4 55 a5 Be1 56 a6 Bc8 0-1
White will lose his last pawn with check.

Game 14
Hou Yifan-H.Toufighi
Subic Bay 2009
 Once again zugzwang will be our weapon of choice to incapacitate the
enemy pieces. Black is threatening mate in one, but Hou Yifan had prepared
a winning simplification.

Question: Try to discover her combination – it’s long but pretty much all
forced.

Answer: 32 Qe8+! Rxe8 33 Rxe8+ Qg8 34 Rxg8+ Kxg8 35 Bb3!

Now the black king must rush to d5 to defend the knight before White has
time to get her king to c3 and play Bxc4.
35 ... Kf7 36 Kc2 Ke6 37 Kc3 Kd5
It seems that Black has saved himself, but now comes a nasty surprise:
38 Ba2!
The bishop clears the way for the b-pawn to take the c5-square from the
black king.
38 ... Kc5 39 b4+ Kd5
 Now that Black’s king can’t oscillate between d5 and c5, he falls victim
to zugzwang.
40 a4
40 Bb3 is also good enough. But it’s not a good idea to touch the kingside
pawns (or at least not the g-pawn): 40 g3? needlessly creates an invasion
point for the black king on f3. After 40 ... Ke4 or 40 ... g4, even if he loses a
pawn on c4 Black has enough to draw through running after the f2-pawn with
his king.
40 ... a6 41 Bb3 g4 42 Ba2 1-0
It’s zugzwang. Black’s king must move away and he loses a pawn after
the exchange on c4. Note how White’s pawns on f2 and g2 form an
impenetrable barrier to the black king, warding off any attempt at
counterplay.

Game 15
Wang Yue-T.Radjabov
FIDE Grand Prix, Sochi 2008
 A doubled pawn is often a fatal liability in a simplified endgame. If we
add in White’s well-centralized king and advanced passed pawn there doesn’t
seem to be much hope for Black. Still, it isn’t obvious how White should
proceed. He’d like to infiltrate with his king on the queenside and capture the
b-pawns, which can’t be defended by the black king. But how? Black’s
bishop and b-pawns are combining well to shut out the white king who
cannot break through via c6, c5, c4 or c3 into the black queenside. The black
king meanwhile will go to e7 next move to guard the d6-square.
There is no obvious continuation, so we need to devise a plan.

Question: Before going further, can you work out how we can use
zugzwang to break through Black’s fortress?

Answer: Our scheme can be broken down as follows:

Stage one: Put the bishop on g4 to restrict the black king. This will
prevent him from opposing the white king with ... Kd7 or ... Ke6.
 Stage two: Advance the white king to d5. Black will have to keep his king
on e7, otherwise Kd6 will penetrate and win easily by shepherding the e-
pawn towards the queening square or else winning b6 with Kc7 etc
(remember that the bishop on g4 stops the black king going to d7).
Stage three: In the meantime Black’s bishop will have to stay on b5 to
stop Kc6 or Kc4, winning a queenside pawn. This means that a waiting move
such as Bf5! will leave Black in zugzwang once he has exhausted a couple of
pawn moves on the kingside. Either the bishop will have to move from b5,
when the white king gets to eat one or more of the queenside pawns, or the
black king will have to move from e7, allowing Kd6.
Let’s see how Wang Yue implemented his plan in the game:
43 Bf3! Ke7 44 Bg4
Completing stage one.
44 ... Bf1 45 Kd5 Bb5 46 Bf5
Stages two and three last just one move apiece. And yet without these
mini-plans White might have thrashed around for twenty moves trying to find
a way to break through before, in exasperation, giving up the game as a draw.
46 ... Be8
The black bishop is obliged to give way. After 46 ... Kd8 White could
almost close his eyes and play 47 Kd6, 48 e6 and 49 e7, finishing the game
off with 50 Bg6 mate if the black king insists on blocking the passed pawn.
Equally hopeless is 46 ... h5 47 h4, when 47 ... gxh4 48 gxh4 is zugzwang
again, while 47 ... g4 48 Bg6 wins the h5-pawn unless Black chooses a
completely resignable pawn endgame with 48 ... Be8 49 Bxe8 Kxe8 50 Ke6
etc.
47 Kc4
The king enters through the breach in the fortress and picks up a pawn.
47 ... b3 48 Kxb3 Bb5
 Nonetheless, resistance continues as the black pieces still form a barrier to
the entry of the white king on the queenside.

Question: How can Wang Yue add the pawn on b6 to his trophy
collection?

Answer: First of all His Majesty returns to his optimum square on d5.
49 Kc3 Be2 50 Kd4 Bb5 51 Kd5 Ba4
The black bishop has to stay on the a4-e8 diagonal to prevent Kc6.
Therefore White will build up with b2-b4, Bc8, e5-e6 and Bd7!. Then he
wins the king and pawn endgame with the help of zugzwang if Black
exchanges bishops, or else penetrates with the king to c6 if the black bishop
runs away.
52 b4 Bb5 53 Bg4 Ba4 54 Bc8 Kd8
Or 54 ... Bb5 55 e6 Ba4 56 Bd7! (an offer that is fatal whether accepted
or declined) 56 ... Bxd7 (or 56 ... Bb3+ 57 Kc6 as in the game) 57 exd7 Kxd7
58 g4! (depriving Black of pawn moves on the kingside to create zugzwang)
58 ... Kc7 59 b5 (or similarly 59 h3 Kd7 60 b5) 59 ... Kd7 60 h3 Kc7 (Black
has no choice but to give way) 61 Ke6 and White wins.
55 Ba6 Kd7
If 55 ... Ke7 then 56 b5! shuts out the black bishop and so threatens Kc6,
while 56 ... Kd7 57 e6+ Kc7 58 Ke5 and 59 Kf6 breaks in via the kingside.
56 e6+ Ke7 57 Bc8 Bb5 58 Bd7!

If Black now exchanges bishops it’s like in the note to 54 ... Kd8 above.
58 ... Be2 59 Kc6
Finally, the white king gets to demolish what was left of the queenside
fortress.
59 ... b5 60 Kb6 Bc4 61 Bxb5 Bxe6 62 Kc7 1-0
The e-pawn for the b-pawn was a good swap for White. Even if Black
gives up the bishop for the b-pawn he won’t be able to liquidate the kingside
pawns to achieve a draw.

Our final example on the subject of zugzwang is a remarkable feat of

planning.
Game 16
G.Meier-M.Vachier Lagrave
Grenke Classic, Karlsruhe/Baden Baden 2019

It’s White to move. Later in the book we shall see Caruana save himself
against Carlsen by exchanging off into a rook and knight versus rook book
draw. Here Georg Meier is tantalizingly close to pulling off the same trick,
but his rook is hanging.

Question: Should White move his rook to a1 or d1? A difficult but

crucial choice!

71 Ra1?
Allowing the white king to be driven to the back rank with fatal
consequences.
Answer: In contrast, after 71 Rd1! g3+ 72 Ke2! the white rook can block
the check 72 ... Rb2+ with 73 Rd2, when 73 ... Rxd2+ 74 Kxd2 Kg4 75 Ke2
draws. Black can instead continue to probe with say 72 ... Ke4, but White
holds on after 73 Rd7.
71 ... g3+ 72 Kg1
72 Ke2 Rb2+ wins the knight.
72 ... Nf3+ 73 Kh1
Or 73 Kf1 Rb2, when mate on f2 is a decisive threat.
73 ... Rb2!
White’s pieces are now under great stress. His king is deprived of any
move and mate will follow on h2 if his knight ventures from g2 without
giving a check. Meanwhile, his rook has to stay on the first rank to prevent ...
Rb1+, again unless it gives a check. Vachier-Lagrave’s task is to crowd out
the white rook by depriving it of all safe squares on the first rank. He
envisages a position where he has his own rook on f2 and his king on b2,
with the white rook on d1 and all the other pieces on the same squares as at
the moment. Then if it is White to move it will be zugzwang.

The super-grandmaster from France succeeds in achieving this set-up in

exactly ten moves. As we shall see, he begins by threatening to invade via f3
with his king in order to persuade White to play Kg1. Why this preparatory
manoeuvre is necessary is explained below.
74 Ra5+ Ke4 75 Ra1 Nd4 76 Kg1
If 76 Re1+ Ne2 77 Ra1 Kf3 78 Ne1+ (or 78 Rd1 Kf2 79 Ra1 Rb5, when
there’s no good way to stop 80 ... Rh5+ with mate next move) 78 ... Kg4 79
Rd1 (or 79 Ra8 Rb1 and wins) 79 ... Nc3 80 Rd4+ then 80 ... Kh3 81 Nf3
g2+ 82 Kg1 Kg3 (the most precise) and 83 ... Ne2 mate is a decisive threat.
Therefore Vachier-Lagrave has accomplished his first aim: make White
put his king on g1. The next phase is: put the rook on f2 and then head for b2
with the king.
76 ... Rf2 77 Ra4 Ke5 78 Ra5+ Kd6 79 Ra6+ Kc5 80 Ra5+ Kb4 81
Ra1 Kb3 82 Rd1

Imagine if the white king were on h1 here, rather than g1. White would
be able to answer 82 ... Nf3 (no check!) with 83 Rb1+, avoiding being
zugzwanged, while after 82 ... Kb2 83 Re1 (not 83 Rxd4?? Rf1 mate) 83 ...
Nf3 84 Rd1 we have the position reached in the game at move 83 below but
with Black, not White, to move – so again there is no zugzwang. This shows
how clever it was to persuade White to put his king on g1 before setting off
with the king to b2.
82 ... Nf3+ 83 Kh1 Kb2
The black king has finally reached its destination. The white pieces are
absolutely dominated as the rook dare not leave the first rank, due to mate on
f1, but has no other safe squares along it.
84 Rg1 Nxg1 85 Kxg1
A last try. There still remains some excitement: will Meier be able to
reach a book draw by eliminating the black pawn?
85 ... Kc3 86 Ne3 Kd3 87 Nf1 Ke2 88 Nxg3+ Kf3

89 Nf5
The knight will be slowly but surely cornered. Alas for White, it can’t
stay safely near the king; for example, 89 Nf1 Rg2+ 90 Kh1 Rc2 91 Nh2+ (or
91 Kg1 Rc1, picking up the knight) 91 ... Kf2 92 Ng4+ Kg3 and both the
king and horse are doomed.
89 ... Rd2 90 Nh4+ Kg3 91 Nf5+ Kg4 92 Ne3+ Kf3 93 Nf5 Rd5 94 Ne7
Rc5 0-1
A possible finish is 95 Kh2 Kf4 96 Kg2 Kg5 97 Kf3 Kf6 98 Ng8+ Kg7
99 Ne7 Kf7 and the knight drops.

Preventing the Opponent Activating a Piece

In the following games there is nothing as potent as a zugzwang, but shutting
a piece out of an important square still proves of tremendous value.

Game 17
A.Grischuk-B.Gelfand
FIDE Grand Prix, Khanty-Mansiysk 2015

It is Black to move. Given the chance White will recentralize his knight
with 18 Nc4 and put pressure on the d6-pawn. Therefore Gelfand has to take
some forceful measures.
18 ... a6! 18 Qb6
After 18 Qb3 b5 the c4-square is guarded and the white knight remains
passive on a3.
18 ... d5!
Not so much a freeing move as a preventive move. Once again it is
necessary to deny the white knight the c4-square.
19 exd5 Rxd5
This is the principled recapture, as reducing material by exchanging rooks
should ease Black’s game; though after 19 ... exd5 play was likely to
transpose in any case with 20 Rxd5 etc.
20 Rxd5 exd5 21 Rxd5
If White avoids snaffling the pawn he has the worst of it, as his offside
knight is a greater liability than Black’s isolated queen’s pawn (or IQP).
21 ... Nb4
The point of Black’s little combination. He regains his pawn with
equality.

22 Rd1 Nxa2+ 23 Kb1 Nb4 24 Qd6 Qe2

Counterattacking against d1.
 25 Rd2 Qe1+ 26 Rd1 Qe2
Not 26 ... Qxf2?? since the knight hangs. Now a draw by repetition
results.
27 Rd2 Qe1+ 28 Rd1 Qe2 ½-½
The most boring game in the whole book? Maybe. But it is a reminder
that you have to anticipate your opponent’s positional threats (that is, use
prophylaxis) and, if necessary, make a sacrifice to back up your plan.

Game 18
Vl.Fedoseev-S.Shankland
Wijk aan Zee 2019

Question: Here it is White to move. What is Black’s positional threat?

How should we stop it?

Answer: If he doesn’t do something fast Black will play 23 ... Nc6 and 24
... Nd4, putting his knight on a magnificent outpost square. The horse would
dominate the centre, leaving White struggling to draw. Therefore urgent
preventive measures, or prophylaxis, are needed to fight Black’s plan. Once
the knight gets to c6 it will be too late, so Fedoseev played:
23 Qe3!
A quiet queen move which will decide the course of the game. Black
can’t play 23 ... Nc6 as 24 Na4 wins the c5-pawn.
23 ... Qc7 24 Bf3
The space-gaining 24 h4 might have been played straight away.
24 ... Qe7
Still not 24 ... Nc6? due to 25 Qxc5.

Question: The positional threat of 25 ... Nc6 and 26 ... Nd4 has re-
emerged.
How should we pre-empt it this time?

Answer: 25 Na4! Na6

 An unfortunate knight: it dreamed of stardom on d4 and ended up in the
backwaters of a6. Having more or less permanently stopped Black’s threat,
Fedoseev can concentrate on seizing terrain in the centre and on the kingside.
26 h4 Qc7 27 Kh2 Qe5 28 Bg2
Admitting the imprecision of his 24th move as he needs to clear the way
for the f-pawn.
28 ... Bg6 29 f4 Qc7 30 Bf3 Qa7?
Black should have tried the time-honoured recipe of escaping trouble
through simplification. Here 30 ... Rxd1 31 Rxd1 Rd8 would ease his game,
as 32 Rxd8+ Qxd8 33 Nxc5? Qb6 would cost White a piece. Instead, 32
Ra1!?, keeping a pair of rooks on the board, would preserve a slight edge for
White, as having to defend c5 remains a nuisance for Black.
31 h5 Bh7 32 Rd3!
Now White builds up pressure on the d-file.
32 ... Rxd3
The exchange of rooks has come two moves too late for Black.
33 Qxd3 Nb8
Once again the dream of 34 ... Nc6 and 35 ... Nd4 has re-emerged.
34 Qd6!
And once again Fedoseev hits it on the head by attacking the c5-pawn.
34 ... Rc8 35 Rd1
 Fedoseev has built up a collection of advantages in a style which would
have impressed Steinitz, the first official world champion and the so-called
Father of Positional Chess. White has absolute control of the d-file; the black
queen and rook are tied down to the defence of a pawn; the black knight is on
the back rank; and the black bishop is shut out of the game by the white
kingside pawns. And most directly he is threatening mate in two with 36
Qd8+!.
35 ... Kh8 36 Qb6! Qe7
If 36 ... Qxb6 37 Nxb6 Rf8, White could go after the c5-pawn straight
away with 38 Na4 Na6 (or 38 ... Rc8 39 Nxc5! with a tactic similar to the
game) 39 Rd6 etc.
37 Nxc5! e5
Trying to mix things, as 37 ... Qxc5 (or 37 ... Rxc5 38 Qxb8+) 38 Rd8+
Bg8 39 Qxc5 Rxc5 40 Rxb8 is hopeless for Black.
38 Qd6 Qa7 39 Bg4!
White has managed to activate his bishop without having to release the
black bishop at the same time.
 39 ... f5
Not 39 ... Rxc5 40 Qf8+ Bg8 41 Rd8, when White is mating, nor 39 ...
Qxc5 40 Bxc8! (even better than 40 Qxc5 Rxc5 41 Rd8+, picking up the
knight) 40 ... Qxc8 41 Qd8+ and wins.
40 exf5 exf4 41 gxf4 Na6

42 f6!
42 Nd3 is also good enough. But not 42 Nxa6?, as Shankland had
prepared an amazing variation: 42 ... Qf2+ 43 Kh3 Re8! 44 Rd3 Bxf5!! (the
white bishop has to be deflected so that it doesn’t control e2) 45 Bxf5 Re3+
46 Rxe3 Qxe3+ 47 Kg2 Qe2+ and Black has escaped with perpetual check.
42 ... Rg8
After 42 ... Rxc5 there is the cute reply 43 Qe7!; e.g. 43 ... Qxe7 (or 43 ...
Rc7 44 fxg7+ and mate next move) 44 fxe7 Nc7 45 Rd7 Bg8 46 Rxc7 and
White wins.
With the black rook forced to a passive square, the rest is a carve-up:
43 Qxa6 Qxc5 44 fxg7+ Rxg7 45 Rd8+ Bg8 46 Qxh6+ Rh7 47 Qf6+
Rg7 48 Rd4!
 This was the last chance to fall for a perpetual check after 48 h6?? Qf2+.
48 ... Qe7 49 Qxe7 Rxe7 50 Rd8 Re3 51 h6 Re1 52 c5 1-0

Preventing the Advance of a Pawn Chain

Game 19
Vl.Fedoseev-A.Giri
Wijk aan Zee 2019

A double-edged position with Black to make his 20th move. We might

assume that Giri will try to gain counterplay on the queenside based on the
advance ... b4-b3 to break up the white pawn structure, or else attempt a
direct attack on the c2-pawn by getting his queen and a rook to the c-file after
moving the knight away from c7. But this seems terribly slow, as White will
advance on the kingside with 21 g5, already threatening to entomb Black’s
dark-squared bishop forever on h8 with 22 f6. A second wave of attack with
h3-h4-h5 would also be in the air.
 Question: So what is Black’s best way to avoid trouble on the kingside?

Answer: 20 ... Bf6!

A fine prophylactic move to stop the white attack. The bishop comes out
of its shell to prevent 21 g5. Giri isn’t so easy to outplay.
21 Kh2
Here 21 Bh6 is a critical alternative. Now 21 ... Re8? won’t do for Black
as 22 fxg6 fxg6 (or 22 ... hxg6 23 Qf3) 23 Qf3 gives White decisive pressure
on the f-file. For example: 23 ... Re6 24 g5! Bxg5? (but otherwise he loses a
piece) 25 Qf7+ and mate next move.
Black could try retreating the bishop with 21 ... Bg7. If then 22 Bxg7
Kxg7 White has the chance for a pawn advance of a type which tends either
to be very good or rather poor. After 23 f6+ Kh8 it might appear grim for
Black on the kingside. However, the f-file is blocked so there will be no
triumphal queen invasion on f7 of the kind we saw in the variation above.
Instead, White would be looking to mate on g7, but Black can always defend
with ... Rg8, and then how does White continue? In this specific case Black
has the extra option of 24 ... Ne6!, when the knight is well placed guarding g7
and ready to go to f4 at a favourable moment. Hence 23 f6+ is of doubtful
worth, at least immediately, as it closes lines and concedes the e6-square.
Going back, 22 Qd2 keeps some pressure, though Black is okay. White
also has the option of ‘calling it a draw’ with 22 Bd2 Bf6 23 Bh6 etc.
Although the lines after 21 ... Bg7 are perfectly okay for Black, I believe
that Giri had a different idea in mind, namely a positional exchange sacrifice
with 21 ... Bg5! 22 Bxf8 Qxf8, when the black bishop will dominate the dark
squares whether on g5 or e3. White’s kingside pieces have little scope: the
bishop on g2 is shut in by its pawns, the knight has no good squares, and the
rook on f1 has nothing to attack. The question would be: does Black have any
way to play actively? Yes. After 23 Ra1 (the open a-file is White’s only
source of counterplay) 23 ... Qc8! (to take the wind out of the rook invasion
on a7) 24 Ne2 Ne8!, Black plans 25 ... Nf6 followed by ... Kg7 and a well-
timed ... h7-h5 to undermine the white kingside structure. Then after ... h5xg4
and the recapture h3xg4 Black could assail the white king with an eventual ...
Qh8 etc.
Fedoseev clearly disliked this scenario and played a quiet king move.
21 ... b3
 Having prevented 21 g5 Giri switches his attention to the queenside.
22 Ra1
After 22 cxb3 Bg5 Black would aim to put pressure on b3 and d3 in the
style of the game. But 22 c3!? was an interesting alternative for White, when
after 22 ... dxc3 23 bxc3 Ba4 Black’s passed pawn gives him counterplay,
though the position remains double-edged.
22 ... Bb5!

The bishop targets the d3-pawn, which will soon become the base of
White’s pawn chain, and also controls a6, so that the knight can be brought
into the attack (see move 24).
23 Ra3
Fedoseev accepts his pawn chain being broken up. He hopes to elevate
the b-pawn from the status of ‘weak isolated pawn’ to ‘strong passed pawn’.
23 ... bxc2 24 Qxc2 Na6 25 b4 Bg5!
An indirect attack on the b4-pawn.
26 Rf3
White can’t avoid the exchange of dark-squared bishops as 26 Be1 Be3
leaves him too passive.
26 ... Rb6
Black doesn’t want to play 26 ... Bxd2 27 Qxd2, since he has conceded
the c1-h6 diagonal to the white queen. Giri’s move defends his knight and so
permits the bishop to retreat from b5. This would expose the b4-pawn to
frontal attack. White puts a stop to the idea with his next move.
27 Qa2!

The queen adds to the pressure down the a-file and also eyes the f7-
square, which dissuades Black from developing his rook from f8.
27 ... Bf4!?
It’s always difficult to judge the right moment to ‘go for it’. When should
you switch from quiet manoeuvring to direct aggression? Steinitz famously
said that when you have the advantage you must attack, or else the advantage
will disappear. It’s good advice not to shilly-shally when carrying out a direct
assault on the enemy king, as time is usually a key factor. You want to mate
or at least persuade your opponent to shed lots of material to avert it. If you
play with insufficient energy he might be able to bring up reinforcements, or
strengthen his defences, or even run away with his king.
The situation here is different as there is no mating attack or indeed any
objective advantage for Black. On the other hand, throughout the
middlegame once Giri found 20 ... Bf6!, the position has been harder to play
for White. He has had to hold together a large and unwieldy pawn structure.
Giri could have kept up simmering pressure with the quiet 27 ... Qe7, when
the queen lends a hand in the defence of f7. White doesn’t seem to have any
constructive plan to improve his position. For example, if after 28 Rf1 Nc7 he
tries to exploit the pin on c7 with 29 Ra7, Black has 29 ... Ra6! 30 Rxa6
Nxa6, when the exchange of rooks has reduced White’s dynamism and made
the b4-pawn more of a target. This would be a slow, positional approach by
Black, not forcing matters and trying to wear White down.
In contrast, 27 ... Bf4 leads to a sharp tactical struggle. It is a highly tricky
move and therefore a good practical choice against an opponent I’m
assuming (based on the high number of mistakes which follow) was in time
trouble. The pawn structure becomes broken and messy, and it is hard for
White to keep his scattered pieces safely defended. His king also becomes
vulnerable to tactical themes. It is no real surprise when he collapses, despite
the claims of computer programs that he was doing fine.
28 Bxf4 exf4 29 Ne2?
If anyone has the better of it after 29 Rxf4 Nxb4 30 Qd2 Qg5 31 Rf2!
Qxd2 32 Rxd2 Nc6, it is White.
29 ... Nxb4 30 Qd2 d5!
 Question: Giri deserves credit for the energy of his attack, but can you
see
how to hold it together as White?

31 Nxf4?
Answer: White missed the path to equality with 31 Qxb4 dxe4 32 Rxf4
Bxd3 33 Qxd4! (after 33 Qd2 Qb8! Black’s threats include 34 ... g5, 34 ...
Bxe2 and 34 ... Rb2) 33 ... Bxe2 34 Qxd8 Rxd8 35 Rxe4. Notice that in this
variation and the one at move 29 the correct path for White involved
exchanging off the queens. The old rules hold true!
31 ... Nxd3 32 Nxd5?
White’s position continues to slide downhill. There is an interesting
imbalance of material after 32 Raxd3 Bxd3 33 Qxd3 dxe4 34 Qxe4 Re8 35
Qd3 Qd6 36 h4 (stopping 36 ... g5) 36 ... Rc6 37 Qd2 etc.
32 ... Qd6+ 33 Rg3 Qxa3 34 Nxb6 Qd6 35 Nd5 Bc4
Black is clearly better due to the pin on g3, his passed pawn, base on e5
for the knight, and more active bishop.
36 f6?
Threatening 37 Qh6, but 36 Qa5 was the last chance to hold it together.
36 ... Bxd5 37 Qxd3
Perhaps Fedoseev was counting on 37 exd5 but missed the strength of 37
... Nf4!, and if 38 Qxd4 Black wins with 38 ... Ne2.
37 ... Be6 38 Kg1

Making things a bit easier for Black. However, in view of the awkward
position of the white pieces, including the terrible bishop on g2 (what did it
do in the game?) White won’t be able to prevent the advance of the passed
pawn. If 38 Qd2 then 38 ... Bc4 (threatening 39 ... Qxf6) or 38 ... Qe5 stops
any mating threats.
38 ... Bc4! 39 Qf3
Or 39 Qxc4 Qxg3.
39 ... d3 40 e5 Qxe5 41 Qe3 Re8! 0-1

When judging the value of a sacrifice, we should always consider what it

prevents the opponent’s pieces from doing as well as the extra power it
confers on our own pieces.

Game 20
Ma.Carlsen-A.Grischuk
Shamkir 2019

Grischuk has just played 28 ... c5. Given the chance his plan would be
something like as follows (for the sake of clarity we only give Black’s
moves):
Step one: 29 ... c4, toshut in White’s bishop on a2, open his own bishop’s
diagonal and create an outpost square on d3.
Step two: 30 ... exf4. Black would like to play 30 ... Nc5 at once, but this
would allow 31 fxe5, when 31 ... Rxe5 32 Bf4 costs him the exchange; hence
he begins by swapping on f4. We assume White recaptures with 31 gxf4, as
after 31 Bxf4 Black would stick his knight on the fine e5 blockade square.
Step three: 31 ... Nc5. Here White would defend his e-pawn from capture.
If he played it to e5 Black could retreat his rook to d7 and then carry on with
his plan.
Step four: 32 ... Nd3. Hey presto! Black’s knight reaches a dominant
outpost supported by the pawn on c4. His bishop also controls a long open
diagonal. White is in danger of falling into a positional stranglehold.
Returning to the diagram position, White could ward off the danger from
... c5-c4 with 29 c4! or 29 b3!, when his bishop pair gives him a promising
game in either case. But these moves would at least temporarily block in his
own light-squared bishop. Carlsen sees there is another way to stop 29 ... c4.
He will be able to activate both of his bishops and his rooks, and keep the
black knight and bishop passive – and all for the price of a pawn!
He did it with:
29 Be3! exf4
If Black refuses the offer, he is suffering for nothing.
30 gxf4!
There’s no going back. After 30 Bxf4? Ne5 the energy drains from the
white position. His rooks remain passive and the light-squared bishop is
denied an open b1-h7 diagonal. Meanwhile, Black’s knight has found an
ideal centre post.
30 ... Rxe4 31 Bb1 Re7 32 Rfe1
 Question: What has White gained from his sacrifice?

Answer: Rather than enjoying an open diagonal Black’s bishop is subject

to an annoying pin and has passively to defend c5.
Grischuk had dreamed of getting the horse to d3, whereas here it is
circumvented by its own pawn on c5 and the white pawn on f4.
In contrast, White’s light-squared bishop enjoys a lovely diagonal. If it is
allowed to go to g6 then Black’s king will be shut in and prey to ideas of a
back rank mate (see below).
White’s bishop on e3 shuts in its rival.
The white rooks are lined up on the open e-file in support of the bishops.
If Black does nothing active, White has the plan of 33 Bf2 and, after the
exchange of rooks with 33 ... Rxe2 34 Rxe2, then 35 Bg6 and 36 Re8 mate.
Black had only one saving move. He could play 32 ... Nb8!, intending 33
Bf2 Nc6, which defends e7 and avoids disaster down the d-file. Or if 33 Kf3
(with the threat of 34 Bxc5! and, since 34 ... Rxe2 no longer comes with
check, the killing zwischenzug 35 Bxd6+) then again 33 ... Nc6!, defending
e7 and holding everything together. White could probe with 34 Bg6 or 34
Bf5 (after either move 34 ... Nxa5? fails to the 35 Bxc5! trick), but Black can
defend soundly.
It is reasonable to assume that Grischuk was in his habitual time trouble
at this point. He might also have been dispirited by the turn of events and not
believed he still had a viable defence. In any case he missed the strength of
32 ... Nb8 and quickly crumbled.
32 ... f5? 33 Bxf5
Now White retains massive pressure without even being a pawn down.
33 ... Nf6 34 Kf3!

Once again Black faces the tactic 35 Bxc5!, when 35 ... Rxe2 (or 35 ...
Bxc5 36 Rxe7) 36 Bxd6+ and 37 Rxe2 wins for White.
34 ... Nd5
Stopping the threat but falling prey to a pin on the d-file.
35 Rd2 Rd8 36 Be4
Not 36 Red1?? Rxe3+.
 36 ... Red7 37 Red1 Nf6
Black avoids losing a piece, but the forced simplification leaves him
unable to defend his queenside pawns against the ferocious white bishops.
38 Rxd7 Nxd7
Or 38 ... Rxd7 39 Rxd7 Nxd7 40 Bb7 Nb8 41 b4 Ke7 42 bxc5 and wins.
39 Rd6 1-0

After 40 Rxa6 White will soon push the passed a-pawn through.
This was not just a magnificent positional display. Carlsen showed his
shrewdness in seizing the initiative with a pawn sacrifice. It was a good
practical chance and there was no real risk of losing. If his opponent had
found 32 ... Nb8 White might not have won, but that’s life!

Blocking in Multiple Pieces

It is a common technique to try to block a pawn or pawn structure, but in the
diagram below Ivanchuk sets out to blockade his opponent’s pieces. He
wants to deny them the use of important centre squares by ensconcing a
bishop deep in enemy territory. This is a mighty task, and who better than a
genius to show us how it can be done.

Game 21
I.Ivanisevic-V.Ivanchuk
European Championship, Batumi 2018

White is a pawn up and has stopped Black from castling. Surely it is

excellent for him? Not quite. As we shall see it becomes a race to support or
evict the bishop from c2.
19 ... Ke7!
Ivanchuk clears the way for his rook on h8 to enter the game. His king
will prove safe from attack on e7.
20 0-0 Qa5
And now the black queen vacates the back rank to allow the king’s rook
to go to c8 to support the bishop.
21 Rfc1
With an extra pawn on the kingside a general pawn advance there would
suggest itself as a plan for White. However, with his pieces ineffectually
placed it is unlikely to do more than leave holes in his centre. For example,
21 e3 Rhc8 22 f4 Qc5! leaves Black on top.
21 ... Rhc8

Just in time to defend the bishop. The success of a strategy – or here we

might say its birth – depends on getting the pieces to the right squares at the
right moment.
Here you might like to pause and reflect on the power of a blockade.
White has an extra pawn, but what is his plan? The co-ordination of his
pieces is ruined by the black bishop on c2.
His queen is shut in on b2, denied a path to d2 by the bishop. There are
two open files on the board. The white rooks can’t go to d1 as it is guarded
by the bishop. Nor can a white rook do anything on the c-file as the black
bishop, supported by the rooks, bodily blocks them off. And finally, if White
tries to open more lines with 22 a3? then 22 ... b3 leaves Black with a strong
passed pawn – you’ve guessed it, supported by the bishop!
In contrast the white bishop looks good on g2, but what does it do?
Normally, it can eye a weakness on a8, b7 or c6, but here there is nothing to
attack. We might conclude that White’s rooks and queen are dominated by
the enemy pieces. They are denied activity and the chance to take part in any
plan.
Nevertheless, we shouldn’t exaggerate the size of Black’s advantage. He
has the initiative and excellent co-ordination, but that is all. It is unpleasant
for White to have his pieces boxed in, but objectively speaking he is far from
having a lost position.
22 e3
Ivanisevic decides his only chance for activity is to bring his bishop to d1
to challenge its counterpart on c2.
22 ... Rc3
Putting the rook on a square where it is defended by the b-pawn. This not
only makes the piece more secure, it also means that if White ever dislodges
the bishop from c2 (or it retreats voluntarily as part of Black’s plan) the
exchange Rxc3, far from easing White’s game, would allow the recapture ...
b4xc3, creating a strong passed pawn.
The e5-pawn now hangs, though this is of no immediate consequence as
the black queen has to defend b4.
23 Bf3 Qc5
 Increasing the pressure with an immediate tactical threat.

Question: What trap does White have to avoid?

Answer: Black is threatening 24 ... Be4!, with the forcing sequence 25

Rxc3 (obviously 25 Bxe4 Rxc1+ wins) 25 ... bxc3 26 Qe2 Bxf3 27 Qxf3 c2
28 Rc1, after which 28 ... Qb5 followed by 29 ... Qb2 leads to the triumph of
the passed pawn.
24 Re1!
Wisely side-stepping the tactical threat. The question now is whether
Black can do anything himself besides maintain the pressure on the white
pieces. In that case he would have the better of an honourable draw, but no
win.
24 ... h6!
Ivanchuk has always impressed me with his ability to find constructive
waiting moves that give the opponent full rein to overreach himself or carry
out an incorrect plan. I regard this as the art of doing nothing. It sits well with
the advice of Tarrasch (as enthusiastically quoted by Fischer in My Sixty
Memorable Games): “when you don’t know what to do, wait for the
opponent to get an idea: it’s sure to be bad.”
25 Bd1?
White could wait himself with moves like 25 h4 (to deter any space
grabbing with ... g7-g5), 26 Bg2 and 27 Rec1, but for most players this is
torture. They don’t want to shuffle their pieces around and ask the opponent
“what can you do?” They want to be guided by a plan. It has been said that
any plan is better than no plan, which is demonstrably false, but it reflects the
way a (human) player’s spirit feels oppressed when he has nothing to do.
Therefore Ivanisevic completes his plan and his position deteriorates.
25 ... Be4!
The bishop takes control of the long diagonal which has been ceded to it
far too generously.
26 Ba4
Hoping to play 27 Rad1 and then 28 Rd7+ or 28 Rd4 to activate his
pieces. Naturally, Ivanchuk squashes the idea.
26 ... Ra8 27 Bb3
The retreat 27 Bd1 would be too miserable. Black could reply 27 ... Bc6,
when his pieces all dominate those of the opponent. In time, after securing b4
with ... Rb8, say, he could prepare ... Qxe5 or else ... Qd5 to force White to
weaken his centre with f2-f3. The white position would gradually be eroded.
27 ... Qxe5
Black finally regains his pawn as b4 is no longer hanging. White’s bishop
is performing a useful role in guarding the a2-pawn and the c2-square, but it
is still far from being an active piece. And, most importantly, it has deserted
the defence of its king. As we shall see, Black could have exploited this on
his next move.
28 Rac1
 28 ... Rac8?

Question: Can you see something more aggressive and stronger for
Black?

Answer: It’s very natural to double rooks, but the direct 28 ... Qf5!,
intending ... Qf3, would have been very strong. For example, 29 Qe2 Rd8
threatens 30 ... Rd2! 31 Qxd2 Qf3 32 Kf1 (or else mate on g2) 32 ... Qg2+ 33
Ke2 Bf3 mate. Then 30 f4 Qc5 leaves White with a weak pawn on e3 and an
exposed second rank. If 31 Rxc3 bxc3, the passed pawn plus 32 ... Rd2 would
be too much to bear; while 31 Rcd1 Rxd1 32 Bxd1 (other recaptures drop e3)
32 ... Ra3 means that a2 falls; finally, 31 Ra1 is truly miserable after 31 ...
Rdd3 etc.
29 Qd2 Qc5
Another solid and good positional move, but here 29 ... Qh5! 30 Bd1 Qh3
looks very strong. Perhaps Ivanchuk was in his habitual time trouble and so
underestimated his attacking chances against the white king.
 30 Rcd1
At last White has the chance for counterplay with a check on d7.
30 ... Bd3!
As so often in this game the black bishop slams the door in the face of the
white pieces.
31 e4 Bb5
Guarding the d7-square and preparing to go to c6 in some cases.
32 Qd4
White hopes the exchange of queens will ease his cramp and make b4 a
target. However, the black rooks are now able to infiltrate down the c-file.
32 ... Qxd4 33 Rxd4 Rc1 34 Rxc1
This leads to a king hunt, but the alternative 34 Rdd1 Rxd1 35 Rxd1 Be2
36 Re1 Bf3 is ugly for White as his king will stay passive and e4 is a target.
34 ... Rxc1+ 35 Kg2
Instead, 35 Rd1 Rxd1+ 36 Bxd1 Bc4 37 Bb3 Bxb3 38 axb3 Kd6 is a lost
pawn endgame for White as the black king invades via c5 (or e5) and d4.
35 ... Bf1+ 36 Kf3 Rc3+ 37 Kg4 Be2+ 38 Kh3 g5!
 The position has grown exciting. Black can’t defend b4 but has a
potential mating net around the white king.
39 Rxb4?
He had to make air for his king with 39 f4!, when 39 ... g4+? 40 Kh4 Rc1
41 Kh5! is suddenly winning for White. Instead, 39 ... gxf4 40 Kg2 (not 40
Rxb4 f3 and the pawn queens) 40 ... fxg3 (avoiding 40 ... f3+ 41 Kf2) 41
hxg3 Bb5 42 Rxb4 Bc6 keeps winning chances for Black, though White is
still strongly resisting.
39 ... h5!
The net tightens with the threat of mate on f1.
40 Kg2 Bf3+ 41 Kh3 Rc1 0-1
There is nothing to be done about 42 ... Rg1 followed by 43 ... Bg2 mate or
43 ... Bg4 mate.
Chapter Three
Full Grovel Mode
It happens to us all. We started out with the best of intentions: we were going
to play a fine positional game, organizing our pieces well and keeping our
pawn structure in trim. But here we are after 28 moves looking in disgust at
our backward pawn and the ugly holes on our dark squares. Then our gaze
switches to the opponent’s fine centre and his excellent bishop pair. Lucky
fellow. We can’t help shaking our heads in annoyance and asking ourselves:
where did it all go wrong?
Or at least that’s what Caruana might have been thinking when he
reached the position in the diagram below. But one thing is certain: a world-
class player isn’t going to cave in. He is going to devise a plan which offers
the maximum resistance. After all a half point saved can make all the
difference to the final tournament result – or maybe decide whether you
remain world champion or not! After he saved an unpleasant endgame
against Caruana in game two of his 2018 world championship match,
Magnus Carlsen described such last-ditch defence wryly as “full grovel
mode”.

Game 22
M.Vachier Lagrave-F.Caruana
Stavanger 2018

It’s Black to play. White has the advantages outlined above. Perhaps I
should take the chance to define what I mean by a backward pawn. It sits on
an open file, isn’t advanced beyond say the third rank and, crucially, can’t be
defended by another pawn. In the diagram position the pawn on b7 fits those
criteria, though a purist might argue it’s not a true backward pawn as it can
advance to b5 and be defended by the pawns adjacent to it. My reply would
be: the b7-pawn is never going to advance to b5.
 A backward pawn could well become a weakness as it lacks both a pawn
to defend it or the mobility to spring forwards when attacked. And it will be a
nuisance for any piece that gets consigned to protecting it. Furthermore, the
square in front of the backward pawn could also become a useful base for an
enemy piece, especially if it is in the centre. For more discussion of backward
pawns see Chapter Five and especially the game Nisipeanu-Radjabov.
We should return to our game. Another advantage White possesses is the
chance to engineer a breakthrough in the centre with d4-d5. If I remember
correctly Vachier-Lagrave went into the confessional booth around about
here (a soundproof room where players were allowed to give their opinions
of games as they were in progress). He said he thought he should win! But
Caruana found a great way to reorganize his pieces.
29 ... Ng8!
The plan is to go ... Be6 (to fight against the d4-d5 advance and vacate
the f5-square) and then ... Ne7-f5. On f5 the knight will attack d4 and also be
able to exchange itself on e3 if White’s dark-squared bishop has become
threatening.
 29 Rb4 Rd7 30 Bg2
Now 31 d5 is a huge positional threat: the bishop on g2 would exert its
power against the fragile black queenside pawns, and the bishop on e3 would
gain a strong square on d4.
30 ... Ne7
Just in time. Here MVL (the acronym by which Maxime Vachier-Lagrave
is apparently happy to be known) seems to have anticipated Caruana’s plan
and realized he no longer had winning chances. He therefore prepared a line
leading to a completely drawn opposite-coloured bishop endgame.
31 h4 Be6 32 Be4 Nf5
Completing his excellent defensive manoeuvre.
33 d5 cxd5 34 cxd5 Bxd5 35 Bxf5 gxf5 36 Rd4 Bc6 37 Rxd7 Bxd7 38
h5 Kh7 39 Kg3 Be6 40 Kf4 Bd7 ½-½
Black’s extra pawn is worthless.

Sometimes the only path to safety is to try to simplify to a so-called ‘book

draw’. If you can liquidate all the pawns, you might well be okay even a
piece down. As a rule it’s not too difficult to defend the endgame of rook and
knight versus rook. On the other hand, rook and bishop versus rook is
uncomfortable even though, objectively, it is also a draw – for example,
Caruana won it at the US Championship in 2019. Therefore I’d definitely
recommend you head for the rook and knight endgame if you are in trouble,
whereas practically-speaking you might choose another way to fight it out
rather than go into the rook and bishop scenario.

Game 23
F.Caruana-Ma.Carlsen
Grenke Classic, Karlsruhe/Baden Baden 2019
 It’s White to move. A pawn down against the world champion and with
his bishop hanging, Caruana decided it was time to call it a draw.

Question: Can you see how he escaped the danger?

Answer: 46 Bxg6! Nxg6 47 Rb6

White’s plan is straightforward: win the b-pawn by distracting the black
rook from its defence by advancing the d-pawn. He is greatly helped by the
pin on the knight which stops it lending a hand in holding back the white
pawn.
47 ... Rf4+ 48 Ke1 b4 49 d6 Rd4
Instead, 49 ... Ne5?? 50 d7+ would be very embarrassing for everyone.
50 d7
Planning 51 Rxb4 or simply 51 d8Q, so the b4-pawn will fall.
50 ... Rxd7 51 Rxb4 Kg5!?
If I hadn’t seen this game before and was asked to guess whether it was
Caruana or Carlsen playing Black, up until this point I wouldn’t have been
certain. But as soon Black tries to win this ‘book draw’ endgame I’d feel
pretty sure it was Carlsen with Black! Nonetheless, he had to concede a draw
after another 20 moves.

Game 24
K.Lagno-V.Ivanchuk
Cap d'Agde (rapid) 2008

Question: It is Black to move in this tense position. You might like to

analyse it yourself before reading on. Try to determine whether the
passed pawn will win the game for Black.

First of all, let’s see how the game finished:

58 ... d2
Threatening a decisive check on e1. The value of this move is discussed
below.
 59 Rxh6+? Kxh6 60 Re7
There is no time to take on d2 as the rook is hanging. Grandmaster Lagno
hoped that the pin on the e-file would do the job, but now comes a nasty
surprise.
60 ... Re1+! 0-1
Since 61 Kxd2 Nf3+ 62 gxf3 Rxe7 wins.

White was tantalizingly close to liquidating to a book draw. The problem

was that, in the combinational finale, the little black grouping of rook, knight
and pawn worked too well. Lagno needed to find a way to break their
streamlined co-ordination – or, in other words, get the black knight on the
wrong circuit.
Answer: She could have thrown a spanner in the works after 58 ... d2
with:
59 g4+! Nxg4
Not 59 ... Kh4? 60 Rxh6+, while 59 ... Rxg4 allows an immediate draw
with 60 Rxg4 Nxg4 61 Rxh6+ Kxh6 62 Kxd2.
60 Rxh6+ Kxh6 61 Rd7!
 Another way of framing White’s plan is: sacrifice a pawn to deflect the
black knight so that the white rook gains access to the d7-square. In any case,
White has achieved her aim as the pawn is indefensible. After, say, 61 ... Ne5
62 Rxd2 Kg5 we have the book draw of Caruana-Carlsen above.
It should be mentioned that Ivanchuk was in too much of a hurry at the
start of the endgame. Instead of jeopardizing the pawn with 58 ... d2?, he
could have won easily with 58 ... Ra4!, threatening instant mate on a1. Now
after 59 g4+ (or 59 Rxh6+ Kxh6) 59 ... Nxg4 60 Rxh6+ (White must free d2
for her king) 60 ... Kxh6 61 Rd7 Ne5 62 Rd5, the solid 62 ... Re4 is good
enough, but 62 ... Nc4! is even faster, exploiting the potential fork on b2 to
force the pawn through.

A knowledge of book draws will save you a lot of half points, whether you
are the defender trying to escape into a safe haven or the attacker trying to
steer well clear of the reefs.

Game 25
Ma.Carlsen-V.Anand
Grand Chess Tour, Zagreb 2019

After 65 Ng5!, it looks all over for Black as he is too far behind in the
race to queen.

Question: How well do you know your book draws? Can you see a way
for
Anand to slip out of trouble?

Answer: There is no time to take the knight. Before White can start
checking with his new queen Black has to get his pawn to a2 to set up the
draw:
65 ... a3! 66 Nxh7 Kb1 67 f8Q a2
A queen and a knight down against the world champion – it’s time to call
it a draw.
68 Qf5+ Kb2 69 Qe5+ Kb1 70 Qe4+ Kb2 71 Qe2+ Kb1
 72 Nf6
Of course Carlsen knows his book draws and so lets Black queen with no
more ado.
The basic drawing mechanism is seen if White arranges Qb3+ and Black
replies ... Ka1 to keep the a2-pawn guarded. This would seem the chance for
White to bring forward his king (or in this case king or knight) as the black
king is stuck in front of the pawn, stopping it from queening. And indeed that
would be the winning plan if Black had a b-pawn or a centre pawn rather than
the a-pawn. He would check the king in front of the pawn, move the king a
square closer, then begin a sequence of checks to force the king in front of
the pawn a second time, move the king a square closer, and so on. A
laborious process but one that works like clockwork.
But because it’s an a-pawn the king is stalemated when in front unless the
white queen gives him a breathing space. But that means White has no time
to bring forward his king (or in this case king or knight) to help mate the
black king. Indeed, if 72 Qb5+ here, Black could already reply 72 ... Ka1 and
White can make no progress: he must move his queen to free the black king
rather than bring up his reserve pieces to mate. If you remove the white
knight and put the black pawn on c2, it would also be a draw. White can
arrange Qb3+ and wins after ... Kc1 by bringing his king up a square as in the
process described above. But Black can again answer Qb3+ with ... Ka1!,
when the pawn threatens to queen, and Qxc2 is stalemate. In other words, the
king can’t be forced in front of the pawn.
Remember, in the battle of queen versus pawn on the seventh rank
supported by the king, it’s a draw with an a-pawn or c-pawn, and a win with
a b-pawn or centre pawn, as there are no stalemate tricks – and if the action is
on the kingside, it’s a draw with f-pawn or h-pawn, win with centre pawn or
g-pawn.
72 ... a1Q 73 Nd5
I remember in my youth watching a tournament on TV where Spassky
beat Karpov with queen and knight against queen, but that was because it was
a sudden death blitz finish and Karpov had less than a minute. IM Bill
Hartston recalls that before the tournament began they decided on a list of
positions from which it wouldn’t be permitted to win on time against your
opponent. Most of the players wanted to include queen and knight versus
queen, but Karpov said “no”! Evidently he thought he’d be the player with
the knight. Though Spassky did actually win the position on the board.
 The final moves were: 82 Qa8+ Kc7 83 Qa7+, when Black can hold by
putting the king on d6, but chose 83 ... Kd8?? and resigned after 84 Qb8+ as
84 ... Kd7 85 Nc5+ wins the queen.
Carlsen decides that, in these days of increment, Anand isn’t going to
make a one-move blunder and so sets up a jokey finish. (Should you as a rule
play on in this endgame in your games? I’d give it a go for some moves. Do
your opponents play better than Karpov does when he has a minute on his
clock?)
73 ... Qg7+ 74 Kd6 Qf8+ 75 Ke5 Qe8+
 Shock horror! World champion loses his queen to a skewer! (But all ends
well.)
76 Kd4 Qxe2 77 Nc3+ Ka1 78 Nxe2 ½-½

I’ll mention another book draw well worth knowing: say White has Kb1, Pc2,
Rd3 and Black has Kc4 and Qa4. It’s a draw if White swivels his rook from
d3 to b3 and back again, keeping it defended by the pawn. The black king has
no way to break through the barrier on the third rank to help the queen win
the c2-pawn or mate. White’s set-up is called a fortress.
I saw the Russian grandmaster Kupreichik trick his opponent into this
endgame in a simul he gave around about the time he won Hastings in 1982. I
was sitting next to the player who fell into it. He’d had a completely winning
position but, due to a lack of knowledge of book draws, allowed the fortress
and so lost the chance to beat one of the best players in the world. It was
rather pitiful watching him try to ‘reinvent’ endgame theory for about 30
moves before giving it up as a draw.
I probably don’t need to remind you that bishop and ‘wrong’ h-pawn or a-
pawn is only a draw with the defending king in front of the pawn. For
example, white Kg5, Bd3, Ph6 versus black Kh8 is only a draw as the black
king can never be ousted from h8, only stalemated. To win in this scenario
White needs to have a bishop that controls the queening square.
Slightly less well known is the endgame with white Kg4, Ng5, Ph6 versus
black Kh8. It is only a draw if White rushes with 1 h7?. Patience please! You
have to advance the king to g6, put the knight on f7 to check the king from
the corner, and then queen the pawn with h6-h7+ etc.

Besides exchanging off to a book draw, full grovel mode can also involve the
use of a fortress. It can be a laborious defence, but taking punishment has
always been part of the chess experience!

Game 26
V.Anand-H.Nakamura
London Chess Classic 2010
 It is Black to play. Nakamura looks in bad shape as he is a pawn down.
But like all great players he is strong in technical endgames and formulated a
plan to rescue himself:
Step one: Exchange rooks off.
Step two: Bring the king to f5 to block the advance of the f3-pawn and
keep the white king out – in other words, create a fortress.
Step three: Prevent White opening lines or creating a target on the
queenside to break the fortress.
Let’s see how he implemented his defence.
30 ... Rh1+! 31 Ke2 Rxd1 32 Kxd1
In principle it is a good idea to avoid exchanging off your last rook when
you are material down. Its long-range power and ability to switch its attack
between wings make it an ideal weapon to harass the opponent’s king and
pawns. Here, however, Nakamura wants to create an impenetrable barrier on
the kingside and doesn’t want the white rook interfering.
32 ... Kd7
Now the black king hurries towards f5. He needs to get there before
White’s king can reach e4, as then the fortress fails after the pawn advance
f4-f5, in conjunction with moves like Bf4 or even an invasion with Kd5 at
some point.
33 Bg5!
Anand sees his king will be outrun in the centre or on the kingside.
Therefore he focuses on trying to open a third path of entry on the queenside.
We’ll see at move 35 how the bishop can menace Black’s queenside if he
isn’t vigilant.
33 ... Ke6 34 a4 c6
Planning 35 ... a6 and 36 ... b5 to create an impenetrable barrier on the
queenside. Anand pre-empts this with his next move.
35 a5!
 Question: What is the best responsive to this aggressive stab at our
queenside?

Answer: 35 ... bxa5!

After 35 ... b5 36 a6! White’s advanced pawn is too strong. For example,
36 ... Kf5 (the king will be forced to go to this square once White finally
threatens to break the fortress with Ke4 or Kg4) 37 Bd8! has the terrible
threat of 38 Bb6, when the pawn queens after 38 ... axb6 39 a7, while if 38 ...
Bb8 then 38 Bxc5. If Black pre-empted this with 37 ... Bb8, then 38 Be7 wins
the c5-pawn and clears a route for the white king to enter.
Nakamura’s 35 ... bxa5! reminds me of the saying “beauty is as beauty
does”. It is a terribly ugly move, giving Black two sets of doubled and
isolated pawns. But “the East is not more beauteous than its service”. It
prevents White establishing a pawn on a6 and therefore avoids having
Black’s own pawn fixed on a7, where it is vulnerable to the idea of Bd8 and
Bb6, as we saw above.
 36 Kc2
But hasn’t a way been opened for the white king to penetrate via a4?
36 ... a4!
No, Black’s pawn commits hara-kiri to close the invasion square on a4
(this would also have been the reply to 36 Bd8).
37 bxa4 Kf5 38 Be3 a6!

A far safer square than a7 as it escapes the attention of the enemy bishop.
39 Kd3
Frustrated on the queenside, the white king shrugs his shoulders and
heads off on a long journey to the other wing. There is no need for White to
hurry in such positions; indeed, why not extend the opponent’s thankless
defence? A tough player like Nakamura is unlikely to crack under the strain,
but can you vouch for the fortitude of your own opponents?
39 ... Be7 40 Ke2 Bf8 41 Kf1 Be7 42 Kg2 Bd6 43 Kh3 Be7 44 Kg3 Bf6!
Here 44 ... Bd6+ 45 Kh4 Be7+ 46 Kh5 Bf8 looks a sound defence as the
white king is prevented from edging up the h-file. But I like the message that
giving up a second pawn sends to Anand. It reminds me of some advice that
David Bronstein, who drew a world championship match in 1951, gave me
when he came to England for a few years, later in his life: If you stand
slightly worse in a boring endgame, and your opponent is dragging the game
on, don’t just play the natural moves.
He showed me one of his games where he was a bit worse in a rook and
pawn endgame. His opponent captured a pawn, and rather than recapture with
a piece and keep an intact pawn structure, Bronstein retook with a pawn,
giving himself doubled pawns. Objectively, the pawn recapture was no better
or worse than the piece recapture, but the point is that he was showing his
opponent he wasn’t asleep. The opponent gave him a draw shortly
afterwards.
Here Nakamura is saying to Anand that he is totally confident he will
hold the position. He doesn’t even bother to hold onto the c5-pawn.
45 Bxc5 Bd8 46 Be3 Be7 47 Kg2 Bd8

The remainder of the game is pretty self-explanatory. White tries to break

in via c5 with 51 Kd4 but is stopped by 51 ... Bb6+. When he puts the bishop
on a7 aiming to play Kd4 without being bothered by ... Bb6+, Nakamura
replies 53 ... Be1 stopping the king advancing because of ... Bxf2+. These
two themes repeat themselves until Anand finally gives up the winning
attempt. The only other thing to mention is that 55 ... Kxf3? would miss the
point entirely, as after 56 Kd4 the white king threatens to enter via c5 or e5.
Then 56 ... Bb4 57 c5, intending 58 Ke5 etc, is decisive.
Here are the moves:
48 Kf1 Bc7 49 Ke2 Bd8 50 Kd3 Ba5 51 Kd4 Bb6+ 52 Kd3 Ba5 53 Ba7
Be1 54 Bb6 Kf4 55 Be3+ Ke5 56 Bc5 Kf4 57 Ke2 Ba5 58 Ba7 Kf5 59 Ke3
Be1 60 Bb6 Bc3 61 Bc7 Be1 62 Bd6 Bc3 63 f4 Be1 64 Be5 Ba5 65 Bd4 Bb4
66 Be5 Bc5+ 67 Bd4 Bb4 68 Ba7 Bc3 69 Kd3 Be1 70 Be3 Ba5 71 Kd4
Bb6+ 72 Kc3 Ba5+ 73 Kd3 Bc7 74 Kd4 ½-½
If you have ploughed through all this you’ll know by now that 74 ... Bb6+
would hold the draw.

Sometimes there is nothing to be done to save a position if the opponent finds

the correct plan, but that doesn’t mean we shouldn’t continue to defend
grittily. Our opponent might get overconfident or short of time or simply
careless. We have to be ready to take our chance.

Game 27
A.Ledger-A.Kosten
British League 2001
 36 Bxd7!
Exchanging minor pieces removes Black’s only source of counterplay
and clears the way to pin down his remaining pieces to the defence of the a6-
and d6-pawns.
36 ... Rxd7 37 Rea4
Now one black rook must stay on a8, while the other can only stagger
along a couple of squares on the second rank. Spoiler alert! White will resign
after another six moves because he can’t stop his king being mated by these
useless rooks.
37 ... Kd8
Black can only wait until ... well, he mates the white king.
38 Kf3!
Having paralysed the black pieces on the queenside, it is now time to
create a second front on the kingside. This will be done by threatening to
infiltrate with the king via f5 and g6.
38 ... Ke7 39 Ke4 Rb7 40 Kf5
Good enough, but the immediate 40 Ra1 followed by 41 Kf5 was more
thematic.
40 ... Rd7
Here 40 ... Rf8+ 41 Kg6 Rg8+ 42 Kxh6 wins for White as 42 ... Kf6,
threatening mate on h8, can be answered by 43 Rxd6+, driving the black king
away. Note that after the game move White no longer has a defence with
Rxd6+ to a threatened mate on h8, as d6 is defended by the black rook. This
fact becomes of decisive importance in a couple of moves.

41 Kg6

Question: What do you think of the immediate infiltration by the white

king?
What alternative plan might you suggest?

Answer: White is playing with fire to say the least. In the scenario of the
previous note, 43 Rxd6+ drove the black king away at a critical point to stop
mate and so win the game. As this method is no longer available, White
needed to prepare a second way to force the black king back. The way to do it
was with 41 Ra1!. For example: 41 ... Kf7 42 g4 (putting Black in zugzwang
so his king must give way) 42 ... Ke7 (42 ... Kg7 43 Ke6 wins the d6-pawn)
43 Kg6, when if 43 ... Rg8+ 44 Kxh6 Kf6, White stops the mate on h8 with
45 Rf1+! and wins easily after 45 ... Ke7 46 Rf5 Ke8 47 Rf6! etc.
41 ... Rg8+

42 Kh7??

Question: Can you work out why this move is graced with two question
marks?

Not 42 Kxh6?? Kf6 and the threat of mate on h8 wins, as there is no Rf1+
to save White. But it wasn’t too late to retrace his steps with 42 Kf5! and
then, after 42 ... Ra8, play 43 Ra1 followed by a second – and this time 100%
sound – infiltration with 44 Kf5 and 45 Kg6. But how difficult it is to change
your mind and admit you’ve done something wrong.
Answer: 42 ... Kf8+!
I guess White missed the strength of this king retreat. It defends one rook
and prepares a mating net by defending the other rook after his next move.
43 Kxh6 Rdg7! 0-1
Every other move for Black still loses, but this wins as mate follows on
h8.
A horrible swindle, but Tony Kosten didn’t give up hope and remained alert
for the one chance of saving himself. These sort of snap mates are not
unknown in double rook endgames.
Chapter Four
Punishing Faulty Freeing Moves
The aim of a lot of opening systems with Black is to free your position by
preparing a liberating pawn advance. In Chapter Ten we’ll use two games to
illustrate a freeing move in the Caro-Kann. Here we’ll focus on the downside
of such moves when incorrectly implemented.
In books and videos you’ll often see a player lauded for striking at the
white centre with a pawn move such as ... c5! or ... e5!. This can lead to the
false belief that any pawn move which doesn’t lose material and increases the
scope of your pieces must automatically be a good thing.
I recently heard an excellent piece of advice from my club-mate IM
Adam Taylor: When you advance a pawn, think about the squares and lines
that you are leaving behind.
Are we opening lines to the benefit of our pieces, or to the benefit of the
opponent’s pieces?
This chapter is a warning to be on your guard against making ‘false’
freeing moves. More positively, it shows you how to exploit this very
common strategic mistake. You’ll win a lot of games if you play in a steady
fashion, keeping your pieces in good order, and “wait for the opponent to get
an idea”.

Game 28
T.Banusz-L.Vajda
Szentgotthard 2010

It’s Black to play. Vajda decided that he had restrained the e4-pawn enough:
it was time to eliminate it and open lines for his pieces. However, this proved
to be very careless.
 27 ... f5?

Question: How would you punish Black?

Answer: 28 e5!
The pawn shoots forward with what Nimzowitsch called “an urge to
expand”. It turns out that the methods of restraint have been insufficient.
28 ... Rxe5
Black’s queenside structure crumbles after 28 ... dxe5 29 d6 cxd6 30
Nxd6 Rd8 31 Rxb6. The knight which looked invincible on c5 has lost all its
support. The black pieces are in such a tangle that 31 ... Nd7 32 Rc6 would
leave the bishop attacked with nowhere safe to go.
29 Rxe5 Rxe5 30 Nxc7 f4
My computer claims that Black is still okay after 30 ... Qe7! 31 Ne6 Re1+
32 Kh2 Bxe6 33 dxe6 Qxe6 34 Rxb6 Qe2!, though I think most humans
would be worried about the open lines around the black king and/or the
passed white a-pawn. Still, this line is much better than the game move which
gives up a pawn just to come out fast with the bishop.
31 Rxf4 Bf5 32 Qd2 Qh6 33 Ne6! Nd3?
Losing in abrupt style, though after 33 ... Bxe6 34 dxe6 Nxe6 35 Bd5
Black’s position is falling apart anyway.

Question: Find the move that immediately finishes the game.

Answer: 34 Rxf5! 1-0

Black’s queen hangs, and if 34 ... Qxd2 it’s mate on f8.

Returning to move 27, Black needed to add one more piece to the battle
for the e5-square with 27 ... Qg7!.
 Then he is ready to advance 28 ... f5, the vital difference compared with
27 ... f5? being that his queen is defending the c7-square ‘through’ the rook
on e7. This means that if White played in the style of the game with 29 e5
Rxe5 30 Rxe5 Rxe5, he wouldn’t have the option of 31 Nxc7.
It might seem strange to describe 27 ... Qg7 as fighting for control of the
e5-square when the main value of the move is to guard c7 in the key variation
above. But maybe this is consistent with Nimzowitsch’s observation that
strategically important points should be overprotected, as the pieces doing so
will “automatically” find themselves well placed for other tasks. That is, the
queen goes to g7 to overprotect the e5-square, and finds itself well placed to
defend the pawn on c7.
Of course after 27 ... Qg7 it would be White to move and he doesn’t have
to let Black get in 28 ... f5 ‘for free’. An interesting response would be 28
Qc3!?, offering the exchange of queens to remove the potential defender of
c7 and so once again deter 28 ... f5. Nevertheless, after 28 ... Ba6 (not rushing
to force matters) Black retains some edge as the white pieces have an
essentially defensive role in guarding the e4-pawn.
Sometimes it’s vital to keep lines blocked. You have to be cruel to be kind to
your pieces. Is it worth opening the cage door to let your pet hamster stretch
its legs, if a hawk is going to swoop down and carry it off?

Game 29
G.Serper-J.Becerra Rivero
Ledyard 2000

1 d4 g6 2 e4 Bg7 3 c4 d6 4 Nc3 Nf6 5 Nge2 0-0 6 Ng3 e5 7 d5

Anyone who has played the King’s Indian as Black knows that the
trademark freeing move is ... f7-f5. And so without more ado Black played:
7 ... Nfd7?

Clearing the way for the f-pawn.

Question: What are the drawbacks to this plan?

 8 Be2
Answer: White may have spent two moves getting his knight to g3 (rather
than playing the usual 5 Nf3), but this doesn’t justify moving the black knight
twice. As we shall see, g3 is a better square than f3 for taking advantage of
the ... f7-f5 advance. What’s more, the black knight has wandered away from
defending its king.
A far better approach was 7 ... a5!, followed by ... Na6. If 8 h4, Black can
keep lines closed with 8 ... h5.
8 ... a5
Evidently the plan is to play ... Nc5, but it’s all too extravagant.
9 h4! f5
The freeing move frees ... well, nothing but the h5-square for the white
knight. Still, White can also build up an attack after 9 ... h5 10 Bg5 Bf6 11
Qd2.
10 exf5 gxf5 11 Bg5 Nf6
Back again, but walking into a strong pin.
12 Nh5 Bh8
Saving the important bishop from exchange but losing more time.
 Question: How can White animate his kingside initiative?

Answer: 13 g4!
In contrast to Black’s pseudo-freeing move 9 ... f5 this pawn stab is the
real deal. If Black plays 13 ... f4 to block the g-file, then White’s knight on c3
gets the e4-square and his queen and light-squared bishop can take possession
of the b1-h7 diagonal.
13 ... Qd7
Getting out of the pin on f6, but this awkward move shows the complete
failure of Black’s plan.
14 Rg1!
Instead, 14 gxf5 was the obvious move. It would be very strong, but
Serper has seen that the rook on g1 plus the mobile g-pawn is a more potent
mix than just having the rook on an open file.
14 ... f4
We already know the drawbacks to this move, but Black can’t allow
g4xf5 to remain forever hanging over his head.
15 Bxf6!
Clearing the way for the g-pawn and winning the e4-square for the
knight.
15 ... Bxf6 16 g5 Bg7 17 Bg4!?
A very thematic move. Serper plans to eliminate the light-squared bishop
on c8 so that his queen and knight will be virtually unassailable once they
take up posts on the b1-h7 diagonal.
In fact, White’s attack was already so strong that there wasn’t the need for
positional niceties. Here is an interesting variation: 17 Bd3! Kh8 18 Bxh7!
(no check, no fuss: just stealing a vital pawn) 18 ... Kxh7 (not forced, but
with the loss of h7 it’s hopeless anyway) 19 Nf6+ Bxf6 20 Qh5+ Kg7 21
gxf6+ Kxf6 22 Ne4+ Ke7 23 Rg7+ Kd8 24 Qg5+ Ke8 25 Nf6+ Rxf6 26 Qxf6
and with mate on g8 and the black queen hanging this is where the proverbial
boxing match would be stopped.
17 ... Qd8 18 Ne4
The white knight has reached its dream square.
 Question: But can you see an even better post for it?

18 ... Kh8 19 Nxg7 Kxg7 20 Nf6!

Answer: A strong point near the opponent’s defensive line is even more
blissful than e4.
20 ... Na6
Here’s what might happen if Black challenged the monster on f6: 20 ...
Nd7 21 Be6! (clearing the h5-square for the white queen) 21 ... Nxf6 22
gxf6+ Kxf6 (or 22 ... Kh8 23 Qg4! Qxf6 24 Qg8+! Rxg8 25 Rxg8 mate) 23
Qh5 Bxe6 24 Qh6+ Kf5 25 Qxh7+ Kf6 26 Qg7+ and mate next move.
21 Bxc8 Qxc8 22 Qb1!
On c2 or d3 the queen might be attacked by ... Nb4 or ... Nc5
respectively, so Serper plays it safe. It’s always nice to see the queen boss
matters from the first rank.
22 ... Rh8 23 Nh5+
What’s this, the knight moves away from f6? Yes, as White needs to
utilize the g-pawn to break through.
23 ... Kf8 24 g6 Rg8
 Question: Can you see a convincing sacrifice for White?

25 g7+
More than good enough.
Answer: But ‘blundering’ a rook with 25 gxh7!! was the fastest and
prettiest way to conclude the game: 25 ... Rxg1+ 26 Kd2 Rxb1 (26 ... Kf7 27
Qxg1 quickly decides though is much less fun) 27 h8Q+ Ke7 28 Qg7+ Kd8
(or 28 ... Ke8 29 Nf6+ Kd8 30 Qf8 mate) 29 Qf8+ Kd7 30 Nf6 mate.
25 ... Ke7 26 Qxh7 Qh3 27 0-0-0
It’s surprising the game lasts until move 42 as White has a gigantic
passed pawn and a ready target in the black king.
27 ... Nc5 28 Qg6 a4 29 Qf6+ Ke8 30 Nxf4
 One more humiliation for Black, as 30 ... exf4 31 Rde1+ Kd7 32 Re7+
Kd8 33 Qf8+ would be brutal.
30 ... Qf3 31 Qg6+ Ke7 32 Qg5+ Kd7 33 Qf5+ Ke8 34 Qh5+ Qxh5 35
Nxh5 Kf7 36 Rg4 a3 37 b3 e4 38 Kc2 Rae8 39 Rf4+ Kg6 40 Nf6 Nd3 41
Rg1+ Kf7 42 Nxg8+ Kxg8 1-0
Black has finally had enough. The simplest way to win is 43 Rf6, 44 Rh6
and 45 Rh8+.

Game 30
L.Winants-N.Eliet
Belgian League 2014
 In the diagram position the pawns on e6 and f7 present a brick wall to the
bishop on b3. But Black became afraid of the other bishop and so decided to
block it with:
15 ... e5?
After this the brick wall becomes a paper sheet held together with sticky
plaster.

Question: Find the best way for Black to deal with the bishop on c3.

Answer: The correct way to neutralize the c3-bishop was by 15 ... Nf6!.
Some structural damage to the kingside can be tolerated if the centre is rock
solid; i.e. 16 Bxf6 gxf6 17 Qd4 (after, say, 17 Qg4+ Kh8 18 Qh4 Qe5 19 Bc2
f5, I’m starting to prefer Black) 17 ... Qe5! with equality.
16 Qh5 Be4 17 Rfd1
Black’s centre is creaking under the onslaught from the white bishops,
queen and rook. Note the contribution that the ‘quiet’ bishop on b3 is making
by pinning the pawn on f7 and so stopping the bolstering of e5 with ... f7-f6.
 17 ... Bg6?
Instead, 17 ... Rae8 guarding e5 was necessary, though White can start
another wave of his attack with 18 f4! and is going at least to win the pawn
on e5 in a few moves, as 18 ... exf4 19 Qg4 (attacking d7 and g7) 19 ... Ne5
20 Qxf4 Ba8 21 b5! leaves Black tied up. He has no good way to stop 22 Bb4
winning the exchange, as the rook on e8 has to stay guarding e5, and if 21 ...
Kh8 then 22 Bb4 Rg8 23 Bd6 Qb7 24 Bd5 (stopping mate on g2) would win
the knight.

Question: How to finish off the game?

Answer: 18 Rxd7! 1-0

The only way to play on is 18 ... Qxc3 (as after 18 ... Qxd7 19 Qxe5
there’s no way to stop mate on g7), but 19 bxc3 Bxh5 20 Raxa7 leaves Black
a pawn down in a terrible position.

Game 31
T.Radjabov-S.Vidit
Wijk aan Zee 2019

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Bb4 5 Qa4+

Question: Can you work out the point of this move?

Answer: At first sight this seems a useless check. In fact it is an

indication of the importance of the freeing move ... c7-c5 in Queen’s Gambit-
style centres (or more modestly ... c7-c6 to support the black centre). White is
willing to spend a whole tempo to make the black knight obstruct the pawn.
5 ... Nc6 6 e3 0-0 7 Qc2 Re8 8 Bd2 Bd6
Deprived of a ... c7-c5 break, it is natural for Black to align his pieces for
the alternative freeing advance ... e6-e5. There now follows a conversation
involving three little rook pawn moves:
9 h3
“I might be plotting g2-g4 to attack your king.”
 9 ... a6
“Well, in that case you’ll be castling long. I’ll get my queenside pawns
ready to attack your king with ... d5xc4 and, after you reply Bxc4, ... b7-b5.”
10 a3
“If that’s your plan then I’d better make a retreat square for my bishop on
a2. Not that I said I’m going to castle queenside ... ”
10 ... Bd7 11 Be2
Still keeping Black guessing about where the white king is going to end
up. Vidit presses on with the plan of ... e6-e5, keeping queenside action in
reserve.
11 ... dxc4 12 Bxc4

12 ... h6
Guarding the g5-square.

Question: Is it tactically safe for Black to play 12 ... e5 to free his game?

Answer: 12 ... e5 13 Ng5 (hitting f7) 13 ... Rf8 14 Nd5! (threatening to

take on f6 with mate to follow on h7) 14 ... g6 15 Nxf6+ Qxf6 16 Ne4 Qh4
17 d5 is uncomfortable for Black.
However, Radjabov now reveals his hand: he will attack on the kingside,
with the h6-pawn providing a ‘hook’ for his pawn lever g2-g4-g5.
13 g4
Earlier big name games had gone 13 0-0 and Black had replied 13 ... e5.
13 ... e5?
A classic example of a freeing move which fails. Vidit sticks to his plan.
Indeed, it is hard to resist, as Black’s pieces have been limbering up towards
it over the last few moves. But, as we shall see, White gets a strong attack
because he hasn’t committed his king to g1.
If Black had advanced 12 ... b5 on the previous move he would have
weakened his queenside pawns for no good reason: White could have
responded 13 Bd3 and 14 0-0, followed by the exploitation of Black’s
weaknesses down the c-file. But once he has played 13 g4 the situation has
changed: the white king’s residence on the kingside, if he chooses to live
there, will be ramshackle.
Therefore Black can now afford a structural weakness in his own camp
after 13 ... b5!, as White doesn’t have the luxury of whisking his king off to
safety and then chiselling away at it. If the white bishop retreats to a2 or b3
Black has 14 ... b4, creating counterplay. Meanwhile, 14 Bd3 takes the
pressure off the f7-square so Black can break with 14 ... e5; e.g. 15 dxe5
Nxe5 16 Nxe5 Bxe5 17 0-0-0 and both sides can attack the king.
14 g5!
It is surprising that White is able to carry out an attack with his own king
still sitting on e1.
14 ... b5 15 Ba2 exd4 16 gxh6!!
And this makes an astonishing impression. It’s more important that the
black queen is denied an active square on f6 than it is to capture a piece.
After 16 gxf6 Qxf6, both c3 and f3 are hanging, and things are excellent for
Black.
16 ... dxc3 17 Bxc3
White has two bishops lined up against the black king, a rampant pawn
empowering a breakthrough on g7 after Rg1, and the queen ready to support
Ng5 (which would attack f7 directly and menace an invasion on h7 after
Bxf6).
17 ... Be6 18 Bxe6 Rxe6 19 Rg1
 19 ... Ne8

Question: Can you see a breakthrough for White after 19 ... g6 - ?

Answer: We see the awesome strength of the advanced pawn after 19 ...
g6 20 Rxg6+! fxg6 21 Qxg6+ Kf8 (otherwise mate in one thanks to the
pawn) 22 Ng5 Qe7 (to guard f7 and e6) 23 h7! Rxe3+ 24 Kf1! (why even
think about taking the rook, although that wins as well) 24 ... Qg7 25 h8Q+!
(what a great career for the pawn!) 25 ... Qxh8 26 Qf7 mate.
20 Bxg7 Nxg7 21 Rxg7+ Kf8 22 Qh7 Qf6
The black queen finally gets to the f6-square. It is too late, but this
doesn’t stop Vidit putting up a stout resistance.
23 Ng5
Threatening 24 Rxf7+ as well as the black rook.
23 ... Rxe3+!? 24 Kf1!
After 24 fxe3? Bg3+ Black has enough for perpetual.
24 ... Nd8 25 Qg8+ Ke7 26 h7 Bg3
 Black’s counterattack looks like it might triumph after all, but now comes
a killer move.
27 Ne4!
Protecting f2 and gaining time to eliminate Black’s dangerous bishop by
attacking his queen.
27 ... Rxe4 28 Rxg3 Rh4 29 Rd1!
Cutting off the escape of the black king towards the queenside. As the h-
pawn has to be dealt with, there is no good way to prevent White picking up
the rook on a8 with his next move.
29 ... Rxh7 30 Re3+ Ne6 31 Qxa8
Threatening mate on d8.
31 ... Rh8 32 Qc6 Kf8
Allowing a neat winning combination to round off an excellent fighting
game by both players, but in any case 32 ... Rd8 to guard the d-file is
hopeless after 33 Qxc7+.
 Question: Can you see how to finish in style?

Answer: 33 Rxe6! Qxe6

If 33 ... fxe6 then 34 Rd7 threatening 35 Qa8+ wins.
34 Rd8+ Kg7 35 Qc3+ f6 36 Qxc7+ 1-0
Black loses his rook.

The following game shows the importance of the time factor in the battle
between a pawn structure’s urge to expand and the effort to restrain it. It
takes one move for the assessment to go from ‘fine for Black’ to ‘Black is
struggling’. And that move won’t be a tactical blunder or a spectacular
positional misconception. It will be a dull, unimaginative move which
squanders a tempo and gives White the chance to get an iron grip on the
centre.

Game 32
M.Vachier Lagrave-A.Tari
Gibraltar 2019

1 c4 c5 2 g3 g6 3 Bg2 Bg7 4 Nc3 Nc6 5 b3 d6 6 Bb2 e5 7 e3 Bf5 8 Nge2

Nge7
9 0-0 0-0

White has chosen a slow, careful opening. He hasn’t tried to stop his
opponent developing his pieces, castling his king into safety, or establishing
himself in the centre. One advantage of playing in this style is that you avoid
a heavyweight theoretical battle. The position isn’t going to burn out rapidly,
as can be the case when the two sides meet in a violent, but essentially pre-
planned opening struggle.
White’s opening might not be forceful, but it will preserve every piece
and pawn on the board until move 14. The position is therefore strategically
complex, if not tactically complex. Playing an outstanding but young
opponent, perhaps Vachier-Lagrave wanted to put the emphasis on quiet
positional manoeuvring ‘behind the lines’.
 We might sum up the position in a simple sentence: If Black manages to
advance ... d6-d5 in a safe way he will have solved all his problems. But as
we shall see, it is not so easy for him to work out when and how to achieve
this aim – or indeed for White to work out how best to restrain it.
10 Ne4?!

Question: Why might this move be dubious?

I’m here to praise MVL not bury him, as he handles the positional attack
superbly later on in the game. But it must be pointed out that at this stage the
omens aren’t good.
Answer: White puts the knight on a pleasant centre post, seeing that
Black can’t evict it with ... f7-f5 as his bishop is in the way. However, it
amounts to neglect of the key d5-square. It’s true that Black isn’t ready to
play ... d6-d5 yet as c5 would hang, but this can easily be remedied.
Instead, White should prefer 10 d3. For example: 10 ... Qd7 11 Nd5 Bh3
12 Bxh3 Qxh3 13 Nec3, consolidating his hold on the d5-square. If then 13 ...
f5 White has 14 f4 – the black queen isn’t going to mate White on its own. In
this sequence Black could try 11 ... Nxd5 12 cxd5 Ne7, when White no
longer has the d5-square for his knight, but he now has a superior wedge of
pawns in the centre. After 13 e4 Bh3 14 Qd2, White went onto win with a
well-timed f2-f4 in R.Keene-J.Penrose, England 1974.
10 ... Qd7 11 f4
At last a direct challenge to the black pawn structure. Nevertheless, it is
too much of a liberty when White hasn’t taken measures against ... d6-d5. He
should still prefer 11 d3 Rad8 12 N2c3 etc, though Black is more comfortable
after 12 ... Bh3 than in the scenario of the previous note.
11 ... Rad8 12 d3
Black would be well placed after 12 fxe5 dxe5 13 N2c3 (not 13 Nxc5?
Qxd2) 13 ... b6 etc.
12 ... b6
Now there is a massive positional threat of 13 ... d5!, overrunning the key
d5-square. It would ruin White’s build-up: after 14 cxd5 Nxd5 the e3-pawn
hangs, the d3-pawn is left backward on an open file, and the black knight
(which has hitherto been a poor piece on e7) springs to life.
Therefore restraint is necessary.
13 N2c3
 13 ... h6?

Question: How should White react to this slow move?

Black’s plan is to go ... Be6 and ... d6-d5 to achieve a good game without
being bothered by Ng5. Alas for him, this well-meaning but languid pawn
move gives White just the tempo he needs to impose his will on the position.
This is actually the critical moment in the game. Despite the slow
opening, time has become of massive value. Black needed to act at once with
13 ... Be6!, planning ... d6-d5. After 14 Ng5 (or 14 Qe2 d5, though it would
be better for White to submit to this) 14 ... d5 15 Nxe6 fxe6, a position arises
which at first glance might appear good for White due to his two bishops. In
fact his control of the dark squares d4 and e3 is decidedly shaky. Black’s
pieces are well centralized and can support moves like ... d5-d4 and ... Nf5 (a
superb post for the knight) to exploit this weakness.
The conclusion is that Black had no need for 12 ... h6. And it costs him
the game.
 Answer: 14 Nd5!
Stopping d5 once and for all and threatening a fork on f6. Black has fallen
into a bind, after which White’s 2780 technique takes over. The rest of the
game is a positional pummelling.
14 ... Nxd5
After 14 ... f6 15 Nec3 White has a firm grip on d5. If then 15 ... Nxd5 he
would reply 16 Nxd5 as, everything else being equal, he wants a piece rather
than a pawn on d5.
15 cxd5 Ne7

Black might have hoped that things aren’t too bad, as White has been
obliged to fill the d5-hole with a pawn, rather than occupy or control it with a
piece. If so, Tari underestimated the speed with which White is able to build a
dynamic centre after his next move.
16 Nf2!
MVL makes some sublime knight moves in this game to atone for the
‘sin’ of 10 Ne4. The knight retreat guards d5, stops ... Bh3, and threatens to
win a piece with 17 e4 or 17 g4. What more could you ask?
 16 ... Qc8
After 16 ... exf4 17 Bxg7 Kxg7 18 e4 the black bishop is trapped.
17 e4 Bd7
White has a space advantage and a solid centre. A couple of moves ago
he had less space and a loose structure. On the other side Black’s knight is a
poor piece with no squares, and he no longer has the freeing move ... d6-d5 or
the chance to ease his game by exchanging bishops with ... Bh3. In fact he is
deprived of any activity. White could already create a protected passed pawn
in the centre with 18 fxe5 dxe5, but MVL prefers to build up gradually.
18 Qd2 f5 19 Rae1 Qc7

Question: How should White reroute his knight?

Answer: 20 Nd1!
An awesome regrouping. The knight wasn’t doing much on f2 (it was
denied the d3, e4 and g4 squares) so it heads to e3 with ideas of Nc4 to add to
the pressure on the black centre.
 20 ... Rf7 21 Ne3 exf4
If instead 21 ... b5 then the c5-pawn will become a target to Rc1 after an
exchange of pawns on e5. White’s central pawn majority is working hard,
whereas Black’s 3-2 majority on the left of the board does nothing.
22 gxf4 Bxb2 23 Qxb2 fxe4 24 dxe4 Rdf8 25 e5!
MVL had to calculate that his 28th move would be decisive before
entering this line. The central breakthrough comes just in time before Black
can take the f4-pawn ‘for free’.
25 ... Rxf4
The way to fight on was 25 ... Bb5! 26 Nc4 Nf5, although this allows
White an advanced protected passed pawn with 27 e6!?.
26 Rxf4 Rxf4 27 exd6 Qxd6 28 Nc4
Winning the exchange in view of the loose black knight.
28 ... Rxc4 29 bxc4 Nf5 30 Bh3!

Forcing exchanges is the simplest way to clinch the point. 30 ... Nd4 31
Bxd7 Nf3+ (or 31 ... Qxd7 32 Qf2) 32 Kh1 Nxe1 would be hopeless for
Black, in view of 33 Be6+ Kh7 34 Qf6 Qc7 35 Bf7, threatening mate in two.
 30 ... Qf4 31 Qf6 Qd4+
Things have gone terribly wrong when you feel obliged to swap queens
while the exchange down.
32 Qxd4 cxd4 33 Kf2 Kf7 34 Ke2 Ke7 35 Kd3+ Kd6 36 Bxf5 1-0
An impressive middlegame by MVL after an inauspicious start.
Chapter Five
Exploiting a Hole
Sometimes our plan will consist of trying to gain control of a square near the
opponent’s defensive line or at some other point of strategic value.
The normal target is a square in the centre which can’t be defended by an
enemy pawn – a so-called hole. If we are successful, it can be used by one of
our pieces as an outpost. Having a piece in a commanding position will
facilitate an attack on the king or one of the operations against pieces or
pawns outlined in the previous chapters. Naturally, a strong opponent won’t
gladly submit to our domination of such a key square. He will try his best to
overrun it with a ‘freeing’ pawn advance or obstruct our plan in some other
way.

When the Hole is Part of an Opening Scheme

Game 33
L.D.Nisipeanu-T.Radjabov
Medias 2010

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5

The Sveshnikov Variation of the Sicilian Defence. Black gives himself a
hole on d5; that is, he no longer has a pawn which can defend this vital centre
square. The only way the weakness can be got rid of is by arranging the
freeing pawn advance ... d7-d5, which White will do his best to prevent.
In return for this positional concession Black gets to dislodge the white
knight from its strong central post and eventually drive it all the way to the
backwater of a3. He also gains an equal share of the centre – an unusual
achievement for Black in the Sicilian, as White’s space advantage tends to
last well into the middlegame and beyond.
 For these reasons Black is safer from a direct assault in the Sveshnikov
(and similar lines in the Najdorf where he plays ... e7-e5) than in Sicilian
systems where he settles for a small centre with pawns on d6 and e6 (for
example, the Scheveningen and most Najdorf lines) or leaves his pawn on e7
and combines ... d7-d6 with ... g7-g6 and ... Bg7 (as in the Dragon Variation).
White is deprived of his usual space advantage on which to build an
attack and, in any case, lacks the necessary firepower with a knight sitting on
a3. Therefore he adopts a positional approach rather than an overtly
aggressive one. If any player has attacking chances on the kingside it is
usually Black – though, as we shall see, his attempt to play sharply backfires
spectacularly in this particular game.
6 Ndb5 d6 7 Bg5 a6 8 Na3 b5 9 Nd5 Be7 10 Bxf6
Often White shudders at the idea of parting with a bishop for a knight, but
here he wants to remove a piece challenging his control of d5.
10 ... Bxf6 11 c4
Reinforcing one horse on d5 and clearing the way for the other to rejoin
the central battle via c2. Of course having a pawn on c4 cuts down the
options of White’s remaining bishop, but Nisipeanu will fianchetto it on g2.
The other approach is 11 c3, which likewise frees c2 for the knight and
helps guard the d4-square.
11 ... b4 12 Nc2 0-0

13 g3
Black is fine after 13 Ncxb4 Nxb4 14 Nxb4 Qb6 15 Nd5 (or 15 Nd3 Qd4,
when he regains his pawn) 15 ... Qxb2 etc.
13 ... Be6 14 Bg2 a5 15 0-0 Rc8 16 Qd3
White unable to bring his knight into contact with d5 by 16 Ne3 without
allowing 16 ... Nd4.
16 ... g6
Instead, 16 ... Bxd5 17 cxd5 (after 17 Qxd5 Nd4 18 Nxd4 exd4 the white
queen is awkwardly placed) and now 17 ... Nd4 18 Nxd4 exd4 is probably
okay for Black if you can overcome your instinctive horror of the doubled
pawns, but the plan of rerouting the knight to the excellent c5 post via d7
after 17 ... Nb8 is attractive.
17 Nxf6+!
 Nisipeanu says he thought a long time before giving up his ‘good’ knight
for the ‘bad’ bishop. Evidently he couldn’t see any advantage for White if he
allowed 17 ... Bg7.
17 ... Qxf6 18 b3 Qe7 19 Rad1 Rfd8
There is an impasse on the d-file, with the d6-pawn solidly defended.
Therefore Nisipeanu decides to fight for the key d5-square by playing a move
on the edge of the board:

20 h4
“Your eye on the wings, your mind on the centre, that is the deepest
meaning of positional play” – Nimzowitsch (quoted on page six of my Giants
of Strategy book).
The plan is to exchange White’s passive bishop on g2 for its counterpart
on e6 which is guarding the d5-square. This will be arranged by 21 Kh2 and
22 Bh3. If Black exchanges himself with 22 ... Bxh3 23 Kxh3 then White
gets control of d5; subsequently White can use the h-pawn to undermine the
black kingside with h4-h5. If Black leaves White to exchange bishops with
23 Bxe6 fxe6, then Black has a pawn guarding d5 but can no longer
counterattack against e4 with ... f7-f5; and White can again continue with
moves like Kg2, Rh1 and h4-h5 to attack on the kingside, which has been
weakened by the recapture ... f7xe6.
20 ... Kh8?!
Black could prepare a wing thrust of his own with 20 ... Ra8 and then 21
... a4. Chances would remain equal. Radjabov, who is a dynamic player at
heart despite his drawing propensities, wants instead to take the bull by the
horns on the kingside. He therefore tucks his king in the corner before lashing
out with his next move.
21 Kh2 f5?
Black won’t be able to adequately support the large centre he creates with
this move. He could still revert to the plan of 21 ... Ra8 and 22 ... a4, but it
would have been psychologically difficult. There’s a handy proverb, often
quoted by Russian chess authors: once you’ve said “A” you have to say “B”.
Having played 20 ... Kh8, it was just too tempting to follow up with 21 ... f5?.
22 exf5 gxf5
Notice that as soon as Black plays 21 ... f5 White ditches the plan of Bh3,
despite spending two moves preparing it. It is no longer relevant. Indeed,
after 23 Bh3? Black could break out from the bind with 23 ... d5! (23 ... f4 is
not bad either) 24 cxd5 (24 Bxf5? loses a piece to 24 ... e4) 24 ... Rxd5 with a
good game.
Many plans never see the light of day on the board, even if their
implementation and prevention have played a major part in the thoughts of
the players during a game.
23 Bxc6!
A great decision. It involves rising above years and years of stereotyped
thinking, which tells you that the fianchettoed bishop is a vital piece and
shouldn’t be exchanged for a knight. I believe the difficulty is in noticing
such a move is possible. It wouldn’t occur to most players, in the same way
that they wouldn’t examine a move that leaves a piece en prise, unless there
were some factors to prompt them, like an exposed king whose defences can
be removed with a knight sac. Once you’ve looked at 23 Bxc6 seriously for
even twenty seconds, you should begin to warm to it. The bulky black centre
loses a staunch defender and cannot be held together for long after White’s
next move.
If 23 f4 at once then after 23 ... e4 it is not only miserable for the white
bishop to have been shut in, it’s bad news for the whole of White’s strategy:
without its influence over d5 Black can break out after, say, 24 Qe3 d5.
23 ... Rxc6

Question: Can you see a vital preventive move for White?

Answer: 24 f4!
This cannot be delayed, as 24 Ne3 f4! 25 Nd5 Qg7 26 Qe4 Rg8 gives
Black exactly the type of attacking scheme he was hoping for when he played
20 ... Kh8.
Instead, the advance of the black f-pawn has now been blocked, so it
can’t be used as an attacking lever against White’s kingside. The c8-h3
diagonal remains closed, keeping the black bishop shut in. There is already
the threat of 25 fxe5, winning a pawn (we’re happy about giving up the
blockade on f4 if it destroys the black centre). And if Black replies 24 ... e4,
then 25 Qd4+ Kg8 (25 ... Qg7 loses the exchange after 26 Qxg7+ Kxg7 27
Nd4) 26 Ne3 leaves the black structure congealed.
 Question: How do you assess this analysis position? What is White’s best
plan?

Answer: Black’s centre is a lifeless mass, with the white knight on a

classic blockade square. The horse can’t be disturbed by the black pawns, and
it exerts pressure on key centre points and has the flexibility to join in
whatever plan White chooses.
For example, White can build up on the d-file with moves like 27 Rf2 and
28 Rfd2. But he should probably begin by threatening a breakthrough with
g3-g4 to undermine the e4-pawn, clear the way for f4-f5 (after the response ...
f5xg4), and exploit the vulnerable black king. This can be done by 27 Rf2
and 28 Rg2. Then 29 g4 (to answer 29 ... fxg4 with 30 f5) already looks a
threat, as after 29 ... Qxh4+ 30 Kg1 the white king is safe and Black is facing
collapse both in the centre and down the g-file.
So Black should stop the plan of g3-g4 with the excellent preventive
move 28 ... h5! (even in a bad position we are allowed to play excellent
moves – it’s when we really need them!). But the loose pawn on h5 becomes
one more liability. Having provoked it, White can then revert to the plan of
putting pressure on d6 with 29 Rgd2. Black faces a miserable defence having
to watch out for a breakthrough with c4-c5 or a sudden Nd5, hoping to
exploit the weakness of a5 and f5 after the exchange of minor pieces
(naturally, White would only allow this exchange if he had something
concrete in mind).
White could also indulge in acrobatics by rearranging his pieces with a
sequence such as 27 Rf2, 28 Rg2 (to provoke ... h7-h5), 29 Rgd2, and then 30
Qa1, 31 Rd4, 32 R1d2 and 33 Qd1, when the queen incidentally attacks h5.
This tripling of the major pieces on a file to attack a pawn (or sometimes a
pinned piece or other barrier), with the queen at the back, is sometimes called
Alekhine’s Gun. It is usually the best arrangement to have the relatively less
valuable rooks at the front to do the job of bludgeoning through the defence
(here it would be with the aid of the disruptive pawn move c4-c5) with the
queen controlling things from the back.
A computer program would perhaps hold the position as Black after 26
Ne3. But at elite level in modern chess a player is never content to defend a
compromised pawn structure with zero chances for activity while his
opponent has a clear plan to improve his game. Like Radjabov in the present
game, they will strive to keep some dynamism in their pawn structure.
24 ... Rcc8
Meeting the threat of 25 fxe5 by defending d8 a second time.
25 Qe3!
White has to be very careful to restrain the dynamism of the black centre.
If 25 Rf2? Black could break free with 25 ... d5!. Therefore the white queen
moves off the d-file and eyes the e5-pawn.
25 ... Qg7 26 Rf2
 26 ... Rd7?
Black had to try 26 ... d5 anyway, as after the game move he will be
positionally crushed. For example: 27 cxd5 Rc3 28 Qxe5 Rxd5 and now after
29 Rxd5?! Bxd5 30 Qxd5? Qxg3+ 31 Kh1 Qxf2 White has to thank his lucky
stars he can still draw by perpetual with 32 Qe5+ Kg8 33 Qe8+. Instead, 29
Qxg7+ Kxg7 leaves him a pawn up, but it won’t be easy to win in view of the
active black pieces. Alternatively, 27 Rfd2 Re8! 28 cxd5 Bxd5 29 Nxb4
(selling the knight for a pawn rather than allow 29 Rxd5 Rxc2+) 29 ... Be4 30
Nd5 Rg8 is again a pawn up for White, but Black is still fighting.
27 Nd4!
As 27 ... exd4 28 Qxe6 leaves Black’s pawns smashed up, the knight will
be able to join in the attack on the d6-pawn.
27 ... Qg4
Counterattacking against d1.
28 Rdd2 Re8 29 Nb5!
Playing it safe. After 29 fxe5 Black has the pretty response 29 ... Qh3+!?,
when 30 Kxh3? f4+ 31 Nxe6 fxe3 forks the white rooks and Black is at least
okay; but the cold-blooded 30 Kg1 Rg7 31 Ne2 would be good enough for
White. All the same, why allow the opponent to sacrifice his queen when you
can win a pawn without any fireworks?
29 ... d5
Black’s last bid for activity, but it is too late. Now 30 Qxe5+ Rg7 31 Rg2
is good enough, but Nisipeanu’s move is much more aesthetic.
30 Nd6!

A beautiful move. The knight pounces on the ghost of the pawn on d6. If
now 30 ... Rxd6, then 31 Qxe5+ and 32 Qxd6 wins.
30 ... Red8
After 30 ... Rg8 it’s not too late for White to lose with the catastrophic 31
fxe5??, when 31 ... f4 32 gxf4 Qg1 sees him mated, while after 32 Qxf4
Qh3+ 33 Kg1 Rxg3+ 34 Rg2 Rdg7 35 Qf8+ Bg8 Black has a decisive attack.
Instead, 31 Qxe5+ Rgg7 32 Rg2 would win easily for White.
Notice how Black has struggled throughout due to the inferiority of his
bishop. And that was all thanks to the blocking move 24 f4. If this barrier is
released even for a move, the bishop springs to life.
 31 Qxe5+ Qg7 32 c5!
Completing his domination of the central dark squares: c5, d6 and e5 are
all in his hands. Now White will be a pawn up in the endgame with a
dominating knight.
32 ... Qxe5 33 fxe5 Rc7 34 Rc2 d4 35 Rfd2 f4
Black finally frees his bishop, but it is a despairing combination which is
easily refuted.
36 gxf4 d3 37 Rxd3 Bf5

A brief moment in the sun for the bishop.

38 Nf7+
Simplest, as now 38 ... Rxf7 39 Rxd8+ wins. Things are so bad for Black
that even the ‘blunder’ 38 Nxf5 leaves White winning after 38 ... Rxd3 39 e6,
intending to queen the pawn after 40 e7 with the help of 41 Re2 and 42 Nd6
as needed.
38 ... Kg7 39 Rg2+ 1-0
In view of 39 ... Kxf7 40 Rxd8.
Game 34
L.Aronian-M.Vachier Lagrave
London Chess Classic 2018

1 e4 c5 2 Nf3 d6 3 Bb5+ Nd7 4 c3 Ngf6 5 Bd3 Ne5 6 Be2 Nxf3+ 7 Bxf3 e5

As in the previous game, Black accepts a hole on d5 in order to equalize
space in the centre.
8 d4 Be7 9 dxe5 dxe5 10 Qxd8+ Bxd8

Question: Have a look at the centre. What factors favour White, and how
should he manoeuvre his pieces to utilize the c4-square, assuming Black
plays ... Be6 on the next move?

Answer: White has a hole on d3 and Black on d6, but otherwise three
factors favour White in the central pawn structure:
1) White has the d4-square covered, whereas the black pawns can’t guard
the d5-square. Therefore d5 is an outpost for White (and another hole for
Black).
2) The c5-pawn slightly impedes the action of Black’s dark-squared
bishop.
3) The c5-pawn can also be targeted by Be3. If Black defends it with ...
b7-b6, the supporting pawn could be rammed in the future with a4-a5,
opening the a-file for the white rook and making the b6-pawn more shaky.
The first step in White’s plan is to manoeuvre his knight to c4, where it
attacks the pawn on e5, eyes the hole on d6, and can be redeployed, if
desired, to the outpost square on d5 via e3.
11 Nd2 Be6
Not giving up the c4-square without a fight.
12 Be2!
An important move. The bishop switches defensive duty on f3 for support
of the knight on c4.
12 ... Bc7
Guarding the e5-pawn in anticipation of White’s next move.
13 f3
Necessary, as e4 would hang if the knight moves from d2. Notice how it
is more economical for White to have a pawn guarding e4 than it is for Black
to have a bishop defending e5.
13 ... Ke7
The king stays in the centre to lend a hand guarding some squares. It
clears the way for the rook on h8 to go to the d-file and, as we shall see, the
knight to go to e8.
14 a4!
 An important move. White wants his knight to be safe from attack by ...
b7-b5 when it sits on c4. As stated above, White might further advance a4-a5,
if appropriate, once he has provoked ... b7-b6.
14 ... h5
A precaution against White gaining space with g2-g4 on the kingside,
when he might go g4-g5 to drive away the knight that is protecting the hole
on d5.
15 h4
Given the chance Black might have gained space himself on the kingside
with 15 ... h4, intending ... Nh5 and ... Nf4. Notice in this sequence that the
f4-square becomes in effect an outpost square for the black knight, as
evicting it with g2-g3 would be very problematic. It’s better for White to nip
the plan in the bud with the game move which fixes the kingside, when he
can focus on his own plan in the centre and on the queenside.
15 ... Rhd8
Black’s rook goes to the open file, but there isn’t anything aggressive for
it to do there. On the other hand, it further bolsters his defence of the d5-
square, which is useful as Black’s knight intends to relinquish its part in the
role – see his next move.
16 Nc4
Five moves after the knight aimed at c4 it’s finally time to go there.
16 ... Ne8
Preparing to challenge the white knight with ... Nd6. I hope you realize
that the white knight, although a nuisance for Black, is far more tolerable for
him than its removal with 16 ... Bxc4?, which would amount to positional
suicide. After 17 Bxc4 Black would be left horribly weak on the light
squares.
17 Be3
An attack we envisaged back on move ten.
17 ... b6

Necessary, to defend c5. We might say that White has completed the first
stage of his plan: the knight has reached c4, Black’s light squares on the
queenside have been loosened somewhat, and the b6-pawn is a target for a
future a4-a5. On the other hand, Black hasn’t been sleeping either. He has
brought all his pieces into the game and is planning to dislodge the white
knight with ... Nd6.

Question: Therefore White needs a new plan. Can you work out what it
might be? What pawn advances should you make?

Answer: Aronian’s scheme might be described as follows:

1) Prepare a breakthrough on the queenside with the pawn advance b2-b4.
This will open the c-file for the white rooks after the exchange ... c5xb4 and
recapture c3xb4. It will also increase the power of White’s dark-squared
bishop on the g1-a7 diagonal. A second pawn advance, a4-a5 (perhaps with
b4-b5 as a preliminary to fix Black’s queenside pawns as targets) would put
the b6-pawn and, beyond it, the a7-pawn under more pressure.
2) If Black challenges the white knight with ... Nd6 then it should be
rerouted to d5. There it would dominate the centre, so Black will exchange it
off for his light-squared bishop with ... Bxd5, and White will recapture
e4xd5. After that, in the style of #1, White’s bishop pair, especially the light-
squared bishop, will help him break open lines on the queenside to empower
the rooks and the passed pawn on d5. White will also have to take measures
against Black blocking lines on the queenside and seeking to gain
counterplay on the kingside.
18 Bf2!
It’s quiet but purposeful moves of this type which win games. The e3-
square is cleared for the knight.
18 ... f6
Guarding e5, as otherwise the intended ... Nd6 would leave it hanging.
19 0-0
At long last White castles because he wants to bring his king’s rook into
the game. Remember the saying that you should castle because you want to
or because you have to, not just because you can. White needed a lot of
moves to carry out his queenside build-up, and just one of them wasted on the
unnecessary 0-0 could have ruined everything. For example, Black might
have got the challenge ... Nd6 in before White was ready to play Ne3 in
response, or perhaps even have kept the white knight out of c4 altogether
with an earlier ... Nd6.
19 ... g6
Black is aware of White’s plan and prepares to support the advance ... f6-
f5 (see move 21).
20 Rfc1

Question: It might seem strange to move the rook to a blocked line.

What is White’s intention?

Answer: White put the rook on c1 as, according to part #1 of his plan, he
intends at a good moment to advance b2-b4, when the rook will be well
placed after the exchange of pawns on b4. Note that the plan of b2-b4 is all
the more effective because the black knight is nowhere near getting to the
hole on d4 which emerges after the exchange of pawns.
20 ... Nd6
Black decides not to wait for White’s attack.
21 Ne3
At long last the knight sees the ideal square on d5.
21 ... f5
And finally Black has some counterplay. Nonetheless, it feels like
White’s bishop pair are always going to give him some edge when the
position opens up. Play develops according to part #2 of the plan outlined for
White at move 17 above.
22 Nd5+!
Just in time, otherwise White would have had to concede the centre to
Black with 22 exf5 gxf5.
22 ... Bxd5 23 exd5

23 ... c4
Black’s knight is on a blockade square in front of the passed pawn –
compare this to the knight on e3 in the analysis to 24 f4 in Nisipeanu-
Radjabov above.
Instead, after 23 ... Kd7 24 Ba6!? White is ready to attack the e5-pawn
with moves like Re1 and Bg3, as well as keeping the option of advancing b2-
b4. He would be able to probe with his two bishops. Therefore MVL reduces
the scope of White’s light-squared bishop and also rules out b2-b4. The
drawback is that White can prepare b2-b3 as the beginning of a pawn
advance on the queenside: an amendment to his plan of b2-b4 outlined in part
#1.
24 Be3 Kd7 25 Rab1 Rf8 26 b3 f4
When a player makes a hasty bid for activity it is usually a sign either of
too much confidence or the exact opposite: too much fear. It’s unclear
whether Black overestimated his kingside chances or was anxious about
White’s potential queenside pressure. In any case, he should trust in his
kingside activity with 26 ... cxb3 27 Rxb3 e4, though White could play 28
Bd4, planning the pawn advance c4-c5.
27 Bf2 e4
Upon 27 ... cxb3 28 Rxb3 e4 (28 ... Nf5 29 Bd3 leaves Black with a hole
on e4 and a restrained centre) 29 fxe4 Nxe4 30 Bb5+ Kd6 31 Bc6 Nxf2 32
Kxf2 Rab8 33 Re1, we can add up White’s advantages: control of the e-file,
passed pawn, much superior bishop, and the awkward situation of the black
king.
28 fxe4 Rae8 29 bxc4 Nxe4
It looks as if Black is generating counterplay but he must have missed the
strength of White’s next move.
30 c5!
 White is able to break through on the key point on the queenside due to
the threat of winning the exchange with 31 Bb5+.
30 ... Nxc3
Now White acquires two passed pawns, but giving up the exchange
against a player like Aronian would have been equally hopeless.
31 Rxc3 Rxe2 32 d6 Bd8 33 c6+! Kc8
The neat point is 33 ... Kxd6 34 c7! Bxc7 35 Rd1+, winning the black
bishop.
34 a5!
Breaking through on the b-file is the simplest way to win.
34 ... Rxf2 35 Kxf2 Rf5 36 c7 Bxh4+ 37 Kf3 Rd5 38 axb6 1-0

Long-Range Manoeuvring to Exploit a Kingside Weakness

Game 35
F.Caruana-S.Karjakin
Stavanger 2018
Black’s doubled f-pawns, with the hole on f5 in front of them, make a ghastly
impression. Caruana already has his rooks lined up against the front pawn,
and his bishop is menacing the f7-pawn.
His dream would be to get his knight to f5, after which the black kingside
would surely collapse. The defence would be hampered not only by its
structural weakness but also by the vulnerability of the king to a mating
attack.

Nonetheless, Karjakin has his whole army, including the rooks, in action
and he is well known to be a doughty defender. And if the white pieces get
too carried away we could imagine the black queen announcing mate on g2
with the help of the ‘hidden’ bishop on b7.
Let’s see how White’s plan unfolded.
22 b4!
If you take home only one idea from reading this book, it should be what
I keep on pointing out to players I coach: Every plan needs the use of pawns
at some point. It is easy to focus on the machinations of the ‘big’ pieces,
especially when there are open lines and diagonals, and forget to use a pawn
either to improve your own position or wear down that of the opponent.
White will advance the pawn to b5 to drive the black knight away and
thus win the d4-square for his own knight.
22 ... axb4 23 axb4

23 ... Re7

Question: Try to work out the best response to 23 ... Nxb4. Think about a
sac with 24 Bxf7+ and the follow-up moves. Can you see an edge after
24 Qb3 instead?

Answer: With the game move Black adds a defender to c7 and f7. His
position collapses after 23 ... Nxb4 24 Bxf7+! Kxf7 25 Qxc7+ Re7 (it looks
as if everything is covered, but now comes the second wave of White’s
attack) 26 Ng5+! (deflecting the black queen from the defence of d6, as 26 ...
Kf8 27 Rxf6+ would be crushing) 26 ... Qxg5 27 Qxd6 Nd5 (again it appears
that Black is just about holding everything together, but now comes the third
wave of the attack in the shape of a double-attack on the black queen and
knight) 28 Rf5! Qg6 29 Rxd5 Bxd5 30 Qxd5+ Kg7 31 Qd6 Rb7 32 Qc6 (the
attacking 32 Rf3! is even stronger) 32 ... Rf7 33 Qxb6 and White is two
pawns up with an easy win.
I sympathize if you are thinking that you could never work out such a
long sequence of moves, especially as it includes some unexpected tactical
points (for example, 26 Ng5+! and 28 Rf5!). It is worth remembering that if
you want to be a strong positional player you still have to be able to calculate
ahead and see tactical nuances. Even if you succeed in winning a purely
strategic battle (and such games are rare) you will often have to finish off the
opponent in a tactical skirmish as he puts up a desperate last stand.
Nevertheless, you can often make things simpler for yourself. Here, for
example, if you are unsure about the piece sacrifice 24 Bxf7+!, you can settle
for 24 Qb3 (attacking b4 and f7) 24 ... Nd5 25 Nd4. You’ve got the knight to
d4 and are threatening a fork on f5. Black should try 25 ... Re5. Now both 26
Qa2 (intending 27 Qa7 in the style of the game) and 26 Qc2! (preparing 27
Nf5) are excellent for White – not as overwhelming as 24 Bxf7+, but good
enough for a clear edge.
In this way you can replace a long, difficult variation with a two- or three-
move calculation, followed by a positional assessment. But remember, there
is no excuse to reject calculation altogether. Notice after 25 ... Re5 in the line
above that 26 Qa2 or 26 Qc2 is recommended, rather than 26 Nf5 which
appears to win the exchange. It does, but after 26 ... Rxf5 27 Rxf5 Nxe3!
White is suddenly lost because of the mate threat on g2 (remember the bishop
on b7!). Whereas 26 Qc2 gives White the option (if allowed) of 27 Nf5 Rxf5
28 Qxf5, as the rook stays on f2 guarding g2.
Keep your eyes open for tactics and practice your calculation – your
chess can’t flourish otherwise.
24 b5 Ne5 25 Nd4
White has completed the first stage of his plan. There is a mega-threat of
26 Nf5, forking all of Black’s major pieces.
25 ... Bc8
The bishop is obliged to retreat and relinquish its pressure on g2, which
means that the white rook on f2 is freed from a defensive role.
 Question: What is Black’s threat?

26 Kh1
Answer: Not so fast! Before continuing with his plan White takes a
timeout to meet the threat of 26 ... Bxh3.
The impetuous 26 Nf5? would be a poor choice from a strategic point of
view as White doesn’t want to swap his great knight for the black bishop;
instead, he wants to exchange his own bishop for it (see the comment to the
next move). It would also be a tactical howler as White drops a pawn after 26
... Bxf5 27 Rxf5 Nxc4 28 Qxc4 Rxd2.
26 ... Kg7
Now Caruana has to decide on the next stage of his plan. His attention
focuses on the opposing bishop. He realizes that if he could get his own
bishop to f5 and challenge the black one, one of two scenarios would emerge:
1) The bishops are exchanged, leaving the white knight as absolute
master of the f5-square.
 2) The black bishop moves away to b7 to avoid the exchange. Then it will
be short of safe squares and will be a nuisance to defend if targeted by the
white queen via the open a-file.
But how is the white bishop going to get to f5 to carry out its mission?
The answer is: slowly and carefully.
27 Be2
The first step. Instead, 27 Bd3 Rxd4! 28 exd4 Qxd3 would be
embarrassing for White (always keep an eye on the tactics!). However, 27
Ba2 looks a good alternative, especially as 28 Bb1, getting the bishop to the
right diagonal, would come with a threat of 29 Qxh7+. I imagine Caruana
wanted to keep his bishop in contact with the kingside for as long as possible
to deter any possible attack by Black there.
27 ... Kh8 28 Qc3
Now the queen vacates the c2-square for the bishop.
28 ... Kg7 29 Bd1 Kg8

30 Bc2
The bishop finally makes contact with the f5-square.
 Question: Would 30 Rxf6 have been simpler?

Answer: 30 Rxf6?? would be quite ridiculous, allowing 30 ... Bxh3 31

gxh3 (or 31 R6f2 Rh6 with a lethal attack) 31 ... Qxh3+ 32 Kg1. It would be
bad enough if Black managed to escape with a perpetual check, but it is even
worse as White loses his rook on f6 after 32 ... Qg3+ 33 Kh1 Qh4+. This
variation suggests that Karjakin has serious attacking chances, but in reality
he can’t do much except wait and hope that White gets careless.
Sometimes a smashed structure is tolerable if it offers a semblance of
counterattacking potential. But here Black’s pawn chain on both sides of the
board is feeble and can do nothing to help his pieces.
30 ... Qh4 31 Rf4
Played so that the rook will support the bishop at move 33.
31 ... Qg3 32 Bf5
The challenge is finally issued to the black bishop.
32 ... Bb7 33 Be4 Bc8
Black’s bishop stubbornly refuses to yield to an exchange for its opposite
number. It is therefore time to implement the next step in White’s plan: attack
the bishop with the queen via the a-file.
34 Qa3! Kg7 35 Qa8 Bxh3
After 35 ... Bd7 36 Bf5 Be8 37 Bg4! White has won the battle of the
bishops. Black can no longer fight against the knight invasion on f5, as 37 ...
Bd7 38 Bxd7 (the trade Black has fought against tooth and nail) and 39 Nf5+
wins at once. Seeing no hope either in giving up the exchange with 36 ...
Rxd4 in this line, Karjakin makes one last effort to bamboozle his opponent
on the kingside.
36 gxh3 Qxh3+ 37 Kg1 Rxd4
 38 Bg2!
An important zwischenzug. After 38 exd4? Qg3+ 39 Bg2 Ng4 Black is
still fighting. Thus if 40 R1f3?? (instead, 40 Rxg4+ has to be played, when 40
... Qxg4 41 Qd8! would still require a bit of labour for White to win), there is
mate by 40 ... Re1+ 41 Rf1 Qh2.
By clearing the way to take on d4 with his rook next move, Caruana
keeps the e-file closed, so that 40 Rf3 will fend off the attack.
38 ... Qg3 39 Rxd4 Ng4 40 Rf3 Qe1+ 41 Bf1 1-0
There are no more good checks and Black is a rook down.

Exploiting a Hole on the Queenside

When the centre is blocked, and there are no attractive pawn advances
available, it is necessary to devise piece manoeuvres to exploit any open lines
or diagonals.

Game 36
Yu Yangyi-V.Artemiev
Gibraltar 2019

White has a space advantage on the kingside. The fractured black pawn
structure on that wing, in particular the pawn on h6, presents an object of
attack. Black, for his part, has to fortify his kingside and seek counterplay on
the queenside on the open c-file.

Question: Can you see a clever manoeuvre for Black to exploit the hole
on b4?

Answer: It is Black’s move and he began with:

14 ... Nb8!
The knight wasn’t doing much on d7, as the white pawns deny it the
central squares c5, e5 and f6. Therefore Artemiev intends to reroute it to b4,
where it sits safely in a hole in White’s queenside structure and has influence
over the d3- and c2-squares (the second of these is particularly important in
Black’s battle for control over the c-file). At the same time the queen is freed
to go to b6 as she no longer has to defend the knight, which in turn clears the
way for the rook on f8 to join the battle for the c-file.
15 Qd2
The queen announces White’s plan of attack by eyeing the h6-pawn.
Another attempt to exploit White’s superior kingside structure would involve
Nd2 followed by the pawn advances f2-f4, g2-g4 and f4-f5. It would be great
if White were able to carry that out safely, but already after 15 Nd2? Black
could attack d4 with 15 ... Qb6, when 16 Nb3 leaves White’s knight on the
wrong side of the board, while 16 Nef3 just means his last move was a waste
of time.
15 ... Kg7
Over the next few moves both players build up on their stronger side of
the board.
16 Nd3 Na6 17 Nf4 Qb6 18 h4!
The h-pawn vacates the h2-square to make possible the manoeuvre Nh2-
g4 to increase the pressure on h6. By going to h4 (rather than h3) the pawn
guards the g5-square, so that after Nh2 Black can’t reply ... Bg5, pinning the
knight on f4 (the h4-pawn will have to be defended by g2-g3 before Nh2, so
that Black can’t play ... Bxh4). White also has the option of h4-h5, dislodging
the bishop from g6, though he’d think carefully before playing it as it takes
the h5-square away from his knight and gives g5 back to the other black
bishop.
18 ... Rfc8
The health of Black’s position depends on activity down the c-file.
19 Rfc1
 Question: Is it time for 19 ... Nb4 or should we play a preliminary move?

19 ... Rxc1+!
In view of the previous comment this might seem a strange decision. Why
does Black hand over the c-file? The simple answer is he doesn’t: he leaves it
temporarily under the control of the white rook before seizing it back. As we
shall see, this exchange helps Black’s plan as it allows his knight to utilize
the a2-square.
Answer: If 19 ... Nb4 at once, 20 Rc3! is annoying. Then 20 ... Rxc3? is a
huge positional mistake, strengthening White’s centre after 21 bxc3, while
otherwise Black’s counterplay has been slowed down (i.e. there’s no ... Na2!
move).
20 Rxc1 Nb4 21 g3
Guarding h4 in preparation for his 23rd move.
21 ... Na2! 22 Rd1
Instead, 22 Ra1 Nb4 23 Rc1 would offer a draw by repetition. This was
the financially crucial last round of the Gibraltar Masters and both players
were going for broke.
22 ... Rc8 23 Nh2 Qc7!
The queen adds her power to the attack down the open file. She prepares
to answer 24 Ng4 with 24 ... Qc2!, when the exchange of queens is
unwelcome to White.
Note that 23 ... Rc2? would be a mistake, falling for 24 Nh5+! Bxh5
(after 24 ... Kh8 25 Qxh6 White’s attack is lethal) 25 Qxc2 Bxd1 26 Qxd1
and the forced exchange of Black’s light-squared bishop has left his king
very exposed.
24 Bd3 h5!
The doubled pawn proves its value by stopping the knight going to g4.
25 g4
A forceful reply, which Artemiev meets with a sacrifice.
25 ... Nb4!? 26 Bxg6 hxg6 27 gxh5 Qc2 28 hxg6
Black is left with a commanding rook on the seventh rank after 28 Qxc2
Rxc2 29 hxg6 Bxh4 30 gxf7 Bxf2+ 31 Kh1 Kxf7.
28 ... Qf5!
 The point of Black’s plan. He is temporarily two pawns down but his
queen is well placed to nullify White’s attacking chances on the kingside. In
fact, due to his broken pawn structure, it is Yu Yangyi’s king who is in most
danger. Black is sure to get one pawn back, when his dynamic pieces give
him full compensation for the material.
29 Qe3
White hurries his own queen over to g3 to stabilize matters. He didn’t like
29 gxf7 Kxf7, when Black is poised to attack down the open g-file with 30 ...
Rg8+.
29 ... fxg6 30 Qg3 Rf8 31 Ng2 Qc2 32 Ne3?
The first real slip by either player. The knight was doing an important job
on g2: guarding h4, shielding the king, while ready to go to f4 to attack e6
and g6 in some lines. White should settle for 32 Rf1, when 32 ... Nc6 33 Qg4,
protecting d4 and attacking e6, is unclear.
32 ... Qe4
Preparing to target the d4-pawn.
33 Neg4?
 Yu Yangyi has had no luck with this knight. The last chance was to
retract his last move with 33 Ng2, when 33 ... Nc6 (33 ... Bd8!?, planning to
attack the d4-pawn with ... Bb6, also looks awkward for White) 34 Qc3
grimly defends the d4-pawn (and if 34 ... Bxh4? Black runs into trouble due
to 35 f3!). White is passive but he can hold on.
33 ... Nc6 34 Qe3 Rf4!
A very strong entrance, after which White’s contortions to avoid losing
the key d4-pawn will leave him with a lost endgame.
35 Nf6 Bxf6 36 exf6+ Kxf6 37 Qxe4 Rxe4 38 Nf3 Ne7!

The knight heads for f5, when White won’t be able to defend the pawns
on d4 and h4 and also keep the black rook from wreaking havoc on his
second rank.
39 Kg2 Nf5 40 Rd3
Or 40 Kh3 Re2 41 Rd2 Rxd2 42 Nxd2 Nxd4 and Black wins.
40 ... Nxh4+ 41 Nxh4 Rxh4
Artemiev has won his pawn while keeping the white rook tied down to
defending d4. He smoothly wraps up the endgame.
 42 Kg3 g5 43 Kg2 Re4 44 Kg3 Kf5 45 Kg2 g4 46 Kf1 b6 47 Kg2 Re1
48 b3 Ke4 49 Rd2 Rb1 50 Kg3 Kf5!
If you are winning, don’t give your opponent the slightest counterplay.
After 50 ... Rxb3+ 51 Kxg4 you might still win easily enough, but why let
White activate his king?
51 Rd3 Rh1 52 Kg2 Rh3! 0-1
This way there can be no doubt. Either White gives up the b3-pawn or has
an entirely lost pawn endgame after 53 Rxh3 gxh3+ 54 Kxh3 Ke4 55 Kg4
Kxd4.

Provoking the Opponent into Creating a Hole

It’s important to have good defensive skills and remain calm when your king
is under attack. Even if you play steady, solid openings, you are bound to
make mistakes in some games which expose you to a fierce assault. Or you
might do nothing wrong but still be obliged to hand over the initiative if your
opponent makes a speculative or ‘crazy’ sacrifice.

Game 37
J.Van Foreest-D.Lewtak
European Championship, Skopje 2019

1 e4 e5 2 Nf3 Nc6 3 Bb5 Nf6 4 0-0 Nxe4 5 Re1 Nd6 6 Nxe5 Nxe5 7 Bf1
Be7 8 Rxe5 0-0 9 d4 Bf6 10 Re1 Nf5 11 d5
White grabs space before Black is able to play 11 ... d5 with an equal
share of the centre. Given the chance he will harass the bishop on f6 with 12
Nd2 and 13 Ne4, when 14 Nxf6+ would gain him the bishop pair. Black has
to take measures against this.
11 ... g6?
Clearing the way for the bishop to retreat to g7, but it allows Jorden Van
Foreest to create an outpost square on e6. The best way to meet the positional
threat outlined above was 11 ... d6 12 Nd2 Bg5!. After 13 Ne4 Bxc1 14 Rxc1
Black has avoided weakening his kingside and made a not unwelcome
exchange of bishops. He hasn’t quite equalized but is solid after 14 ... Bd7.
12 Nd2 d6 13 Ne4 Bg7
 Black has completed his plan. His bishop looks good on g7 where it
controls a long diagonal. But alas, this happy state of affairs only lasts one
move.

Question: Can you see how White can create a central outpost for his
knight?

Answer: 14 Bg5! f6
Black is forced to shut in his bishop and concede a hole on e6 as 14 ...
Qd7 15 Nf6+ Bxf6 16 Bxf6 is intolerable.
15 Bd2 Nh6
Lewtak intends to reactivate his bishop with the advance ... f6-f5, but this
will create a route to e6 for the white knight.
16 c4 Bd7 17 c5
The attack on d6 cajoles Black into carrying out his plan. As we shall see,
White is offering a strong positional pawn sacrifice.
17 ... f5 18 Ng5
 18 ... Ng4
Positionally busted, Black decides to go for broke on the kingside. After
18 ... dxc5 White has no need to occupy e6 immediately with his knight. He
could build up with moves like 19 Bf4, 20 Qb3 and 21 Rad1. Then his pieces
dominate the centre. He can stab his knight into e6 to create a passed pawn
(after the more or less forced exchange ... Bxe6 and recapture d5xe6).
Instead, taking the b2-pawn with 18 ... Bxb2 would allow 19 Ne6 Bxe6
20 dxe6, when White has three threats: 21 Bxh6, pocketing a knight; 21 e7,
forking the queen and rook; and 21 e7 and 22 Qb3+, winning the bishop by
double-attack.
In these types of position it is seldom a good idea to go stealing the so-
called poisoned pawn. Even if there was no retribution of the kind above,
White could just answer 18 ... Bxb2 with 19 Rb1 Be5 20 Rxb7, regaining the
pawn and keeping all his positional assets.
19 h3!
A fearless move. Not everyone is willing to put their king on e4 in the
middlegame.
 19 ... Nxf2
There’s no going back as 19 ... Nf6 20 c6! mangles him; e.g. 20 ... bxc6
21 dxc6 Bc8 (or 21 ... Bxc6 22 Ne6, winning the exchange), and with Black’s
bishop buried on c8 White can attack down the open e-file after 22 Qb3+ d5
23 Bb4 etc.
20 Kxf2 Bd4+ 21 Kf3
Also possible was 21 Ke2!? to answer 21 ... Re8+ with 22 Kf3, when
having been enticed to e8 the black rook doesn’t support the f-pawn as in the
game.
Definitely not 21 Be3? Bxe3+ 22 Rxe3 Qxg5 and Black regains his piece
with an extra pawn and attack.
21 ... f4

Attacking the white knight and clearing a line for the bishop on d7.
22 Ne6!
The only move. It is essential that the horse shelters the king on the e-file.
22 ... Qh4
Throwing another piece onto the fire to keep the attack burning. There
was no joy in 22 ... Bxe6 23 dxe6 Qh4 24 e7 Qf2+ 25 Ke4 d5+ 26 Kd3 and
Black’s initiative has fizzled out.
23 Nxd4 Qg3+
One of the secrets of defending against a sacrificial onslaught on your
king is to return some of the material to break the back of the attack. Even if
you only emerge with an extra pawn in the endgame, that can be good
enough. After 23 ... dxc5 White can surrender one of his two extra pieces
with 24 Qb3! cxd4 25 Ke2, when his king will slip back to the safety of d1.
24 Ke4
A pretty picture. The white king leads the battle from the front.
24 ... Rae8+ 25 Ne6
The knight once again takes up its post on the outpost square.
25 ... dxc5

Damian Lewtak deserves credit for his vigorous play after his inferior
opening. His attack has reached its high point. All his remaining pieces are
involved in the action. Alas for him, there aren’t quite enough of them.
Still, it looks terribly dangerous now for the white king. The threat is 26
... Rxe6+! 27 dxe6 Bc6+ 28 Ke5 and mate with 28 ... Rf5 or 28 ... Qg5. There
is in fact only one good defensive move.

Question: Only one move will do for White. Any idea what it might be?

Answer: 26 Qf3!
The lazy white queen has sat peacefully on d1 for the whole game while
her king has been blown around in a storm. But now she rouses herself just in
time to save him from mate by creating a retreat square on d3. At the same
time the power of the black queen is seriously reduced: her view along
White’s third rank has been blocked and she is being bullied by an
unacceptable offer to exchange herself for the white queen.
26 ... Rxe6+ 27 Kd3!
The remainder is pretty straightforward. Remember what we said at move
23 about giving back some material. White sacrifices his noble knight as 27
dxe6 Bc6+ 28 Kd3 Bxf3 drops the queen, though even here White has good
winning chances after 29 gxf3 due to his continuing material advantage
coupled with a strong passed pawn.
27 ... Rd6
27 ... Rxe1 28 Rxe1 Qg5 29 Kc2 is hopeless for Black. Not only is he a
piece down, his king will be wide open to an attack by White’s dark-squared
bishop.
28 Qxg3 fxg3
After 28 ... Rxd5+ the white king can venture even further up the board
with 29 Kc4 fxg3 30 Kxd5.
29 Kc3 b5
Of course if 29 ... Rxd5 then 30 Bc4 wins.
30 Re5 a6 31 Rae1 c6 32 Be3 b4+ 33 Kb3 1-0
 Black loses the exchange after 33 ... Rxd5 34 Bc4 or 33 ... cxd5 34 Bxc5.

When the centre is completely open and both sides are developed, unless
there is a clear object of attack you’d imagine that a draw is very likely.
However, appearances can be deceptive, as for example in the next diagram
position below.

Game 38
A.Giri-P.Harikrishna
Shenzhen 2019
 Question: How do you assess the position? What features give one player
the advantage, and how big is it?

Answer: At first sight the position looks equal, even dead equal. But
White has a series of minute advantages based on the superior mobility of his
forces and the restraint they impose on the enemy pieces. This becomes
apparent if we consider who has ascendancy over the centre squares.
White’s bishop is attacking c7 and tying down a rook to its defence. His
pawns impede the black minor pieces by denying the knights the c5- and d4-
squares and restricting the bishop (if it could be posted on f6 without any
harm being done it would still have nothing to attack there). The white minor
pieces are further up the board than Black’s and control the d6- and e5-
squares. Both the white rooks are placed on open files.
Nonetheless, there still remains the problem of how to turn this pressure
into something more tangible. Giri came up with a brilliant solution. White’s
pieces and queenside pawns are working hard, but he has one positional asset
that is doing nothing: his kingside pawns.
19 g4!
I would say that White has a substantial advantage. Most players would
be ground down by Giri here, but being a great defender he could probably
defend it against himself!
19 ... Bf8
After 19 ... Bf6 20 Ndc5 the pressure on Black’s queenside would soon
win a pawn.
20 Kg2
White has no need to hurry as there is no way for Black to shake off the
pressure.
20 ... Re8 21 Bg3
Clearing the f4-square for the advance of the f-pawn. If Black replies 21
... Re7 he also has to reckon with 22 Nf4, threatening 23 Nd5.
21 ... f6
Creating a hole on e6. However, if Black waits, Giri could push his
kingside pawns as far as f5, g5 and h5 with a stranglehold on both wings. A
well-engineered breakthrough with a move like h5-h6 would in time smash
up the black kingside.
White’s next three moves are all aimed at getting a knight to e6.
22 f4 Re7 23 f5 Rce8
It seems like Black has achieved some activity, but White can ignore the
threat to e4.
24 Nf4! Nd8
A desperate attempt to stop 25 Ne6. After 24 ... Rxe4 25 Rxe4 Rxe4 26
Rxd7 Re7 27 Rd2 Ne5 28 Ne6 White has a crushing advantage. In view of
the threat of 29 Rd8, Black has nothing better than 29 ... Rd7 30 Rxd7 Nxd7
31 Bxc7 with a lost endgame.
 Question: How did Giri turn his pressure into something more concrete?

Answer: 25 Rxd7!
A combination to win a pawn.
25 ... Rxd7 26 Nxf6+ gxf6 27 Rxe8 Kf7 28 Re3 Rd2+ 29 Re2 Rd1 30
Ne6!
The e6-square still hounds Black. If 30 ... c6 then 31 Nxf8 Kxf8 32 Bf2
Rd7 33 Bd4 leaves Black a pawn down and his pieces dominated. As played
he loses a second pawn.
30 ... Nxe6 31 fxe6+ Ke8 32 Bxc7 Rd3 33 Bf4 Rd5 34 Kf3 Ke7 35 Re4
Bg7 36 Be3 f5 37 Bg5+ Bf6 38 Bxf6+ Kxf6 39 g5+ Kxg5 40 Re3 1-0
 After 40 ... Rd8 41 e7 Re8 42 Re6, with the rook tied down Black soon
falls into zugzwang, while if 42 ... f4 then 43 Re5+ Kf6 44 Kxf4 wins.
Despite an objective analysis revealing White’s advantage before he
played 19 g4 I still greatly admire how Giri won ‘out of nothing’.
A Knight isn’t Always Good on an Outpost Square

Game 39
F.Caruana-Ma.Carlsen
Tromsø Olympiad 2014
 Question: Magnus Carlsen is on the back foot in the diagram position.
Can you see the positional threats he is facing? How should he stop them?

Answer: The world champion has to deal with disruptive advances such
as:
a) 23 h6, breaking up Black’s kingside pawns (or leaving him with a
weak pawn on the f-file if he replies 23 ... g6).
b) 23 g5, dislodging the knight from f6 and building up on the kingside.
c) 23 e5, again pushing back the knight and this time intending to invade
with 24 Ne4 and 25 Nd6+.
But it is Carlsen’s move and he took preventive measures with:
22 ... h6!
Directly stopping 23 h6 or 23 g5 and, as we shall see, reducing the power
of White’s third option.
23 e5 Nh7
The knight has been driven to the edge, but it is a stepping stone to a
square even better than f6.
24 Ne4
At first sight putting the knight on d6 appears a mighty manoeuvre, but it
doesn’t turn out as strong as Caruana anticipated.
24 ... Rf8
Defending f7 and preparing to dissolve the potential weakness with ... f7-
f6.
25 Nd6+ Kc7 26 Bg2 Ng5

The knight re-emerges on a square where it blocks the advance of

White’s kingside pawns, helps to guard the structure f7 and e6 (which could
be useful once Black plays ... f7-f6), and is sitting pretty as it cannot be
evicted by a pawn. As Black’s position appeared to be tottering in the initial
diagram, it’s easy to see the good work Carlsen has done in holding it
together.

Question: But what do you think about the white knight on d6?
Have a think before reading on.
 Answer: The knight might be marvellously centralized on d6 but, as
Capablanca and other great players have observed, the co-ordination of the
pieces is the most important principle in the middlegame. How is White
supposed to build on the power of the knight to execute a plan involving his
other pieces? If he plays 27 c5 the knight is strengthened even further, but it
would take away any chance to arrange an expansion with d4-d5 and cede the
d5-square to a black knight. The suspicion arises: would White’s knight
actually be better on e3 than on d6, as it would help support an advance with
d4-d5, when the bishop and rooks would come to life?
27 Rhf1 f6
Not giving White time to build up pressure against f7, when having the
knight on d6 might be justified.
28 Kc2 fxe5 29 dxe5 Nc8!
An awkward moment for White. If he exchanges on c8 not only is he
deprived of his star piece, he is left with the worse minor piece: a bishop with
no scope versus a knight which has the chance of attacking e5 from f7 at
some point in the future.
30 c5
White therefore reinforces the knight, but this means handing over d5 as a
base for a black knight.
30 ... Ne7 31 b4?
Caruana is striving to play actively when he should be settling for a draw.
31 Ra1 Nd5 32 Kd2 would hold, as Black has no obvious way to improve his
position.
31 ... Nd5 32 Bxd5
With b4 hanging and a fork looming on e3 White has little choice other
than to strengthen the black pawn centre. If not actually losing at this point
Caruana’s position is certainly deteriorating fast. The chances of holding
everything together against Carlsen are pretty slim.
32 ... cxd5
 Already the threat of 33 ... Nf3, winning the e5-pawn, is hard to meet.
Caruana sacrifices it in an attempt to drum up queenside counterplay.
33 b5 axb5 34 Nxb5+ Kc6
Carlsen is aided by having a king who can threaten the overstretched
white pawns, while their own king is a passive observer.
35 Nd6 Nf3 36 b4 Ra8 37 Ra1 Rxa1 38 Rxa1 Nxe5 39 Ra7 Rb8
Black is happy to defend for the moment as his knight can gobble the
pawn on g4 as well if the white rook stays on a7.
40 Ra3 b6!
Better than 40 ... Nxg4 41 Rg3 Nf6 42 Rxg7. The pillar on e5 supporting
the white knight has already vanished and the other one on c5 is about to
crumble away.
41 Ra7 bxc5 42 Ra6+ Kc7 43 bxc5 Nd7 44 Ra7+ Kc6 45 g5 Nxc5
 The knight has carried out a demolition job on the white centre (38 ...
Nxe5 and 45 ... Nxc5). The remainder is easy for a world champion.
46 Nf7 d4 47 Ne5+ Kd5 48 Nd7 d3+ 49 Kc1 Nxd7 50 Rxd7+ Ke4 0-1
There’s no hope in 51 Rxg7 Ke3 52 Rd7 hxg5 etc.

Avoiding a Hole Being Created in your Pawn Structure

Game 40
F.Caruana-S.Karjakin
Sinquefield Cup, St. Louis 2018
 Let’s try to work out what Karjakin and Caruana might have been
thinking when they discussed the position in the diagram above.
Karjakin: My plan is to drain the dynamism from White’s pawns. I will
begin with the thrust 21 ... e5. White will have to reply 22 d5 to keep his
centre intact, upon which I will retreat my knight to b8. Then there is a
newly-created hole in White’s structure on c5.
I will manoeuvre my knight there via d7, perhaps throwing in ... Ra5 to
support the knight once it reaches c5. I’ll have the horse on a great square and
the white pawns will be blocked. Having reduced them to inertia I’ll have a
safe game. After that I can try and work out a plan to undermine and destroy
them. Therefore:
21 ... e5
Caruana: I can’t take on e5 as my knight will hang, but if I’ve calculated
right I will get a good game by advancing my pawn to d5.
22 d5 Nb8

Question: Who has assessed the position correctly and what move will
prove it?

Answer: Caruana: This feels like the critical moment. If I let him get his
knight to c5 with ... Nd7 and ... Nc5, he’ll have a good game. My pawns will
be stopped in their tracks. So there’s no time to lose. I have to strike straight
away.
23 c5!
Karjakin: Well, that’s a surprise. If it works he’s breaking up my
queenside before I have time to get the blockade in place. But can’t I take the
pawn on d5?

Question: What happens now after 23 ... Qxd5 - ?

Answer: Karjakin: Of course 23 ... Rxd5 24 Qe8 is mate (that’s another

disadvantage of having the knight on b8), but 23 ... Qxd5 allows me to
simplify after 24 cxb6 Qxb5 25 Rxb5 cxb6. It looks a shade uncomfortable
for me. But wait, after 24 ... Qxb5 he doesn’t have to recapture the queen, he
can play 25 bxa7! and I’m lost. If I retreat my queen to e8, he queens on a8,
while if she goes somewhere to guard a8 then he takes on b8 and queens.
That’s quite a blow. Let’s drive the white queen back before taking on d5.
23 ... Ra5
Caruana: I’ll retreat my queen to b2. If he plays 24 ... Rxc5 25 Rxc5
bxc5 26 Qxe5 he’s left with a mouldy pawn structure. After that he still can’t
take the pawn on d5 due to his weak back rank (26 ... Qxd5 27 Qxd5 Rxd5 28
Rxb8+).
24 Qb2 Qxd5 25 cxb6 cxb6 26 Nc4
Karjakin: Of course he is in no hurry to take on b6. Things have
happened too fast for my pieces. I thought the position was going to stay
quiet long enough to justify the knight manoeuvre ... Nb8 and ... Nd7.
Instead, I’m left with a knight stuck on the back rank in an open position. All
his pieces are excellently placed, mine are disorganized. I should never have
played 21 ... e5, I was in too much of a hurry to equalize. Well, let’s at least
centralize the rook and attack c4.
26 ... Rc5
Caruana: I need to strike while the iron’s hot or he might find a way to
develop his knight. Let’s capture with the queen on b6 and attack his rook on
d8, so that he can’t take twice on c4.
27 Qxb6
 Question: Can you see how White will reply to 27 ... Rdc8 and 27 ... Rf8
-?

Answer: Karjakin: That’s very strong. If I play 27 ... Rdc8 then 28 Rd1!,
and if I move my queen to safety, 29 Rd8+ wins. I could play 27 ... Rf8 28
Qxb8! Rxc4 29 Rxc4 Qxc4 30 Qxe5, but then I’m a pawn down and a4 is a
target. Let’s try to hold it together by defending e5 and making a hole for my
king to end these pesky back rank mates.
27 ... f6
Caruana: Now 28 e4 looks winning after 28 ... Qd4 29 Qe6+ Kh8 30
Nd6, planning a fatal fork on f7 or a smothered mate combo. But I’ll play
more simply.
28 Rd1 Qxd1+ 29 Rxd1 Rxd1+ 30 Kg2 1-0
Karjakin: I had better resign. 30 ... Rc8 31 Qe6+ wins a rook, and 30 ...
Rxc4 31 Qe6+ Kf8 32 Qxc4 wins easily for White once the a4-pawn drops.
Or if 30 ... Nd7 31 Qe6+ Kf8 32 Nd6, the threat of mate forces 32 ... Rxd6 33
Qxd6+ Ke8, when White has at the least 34 Qa6, winning the a4-pawn.
We might add that 21 ... e5 was a poor decision. And after 22 d5 Black
should try 22 ... Na5! rather than retreat the knight. If then 23 Qxa4? Nb7,
Black would regain the pawn on a3 while getting his knight to c5. So White
should prefer 23 Qb4 when 23 ... Nb7 (23 ... f6 is more solid) 24 Ne4! keeps
the knight from c5 and maintains the energy in his set-up.
After 22 ... Nb8?, thanks to Caruana’s vigorous response, things
happened too fast for the black knight: stranded on b8 it was not only a target
but also harmed the king by interfering with the queen’s defence of the back
rank.

Game 41
V.Anand-A.Giri
Shamkir 2019

Question: It’s Black to move in a position which is difficult to assess, but

have
a go at evaluating the various positional features. Would you rather have
White or Black?

Answer: Both players have their pieces centralized and a king safe from
attack. There are no gaping weaknesses on either side. Black has the two
bishops which in a general way constitutes an advantage.
To balance this White has the superior structure: his extra pawn in the
centre can be contrasted with the doubled black b-pawns. If he can mobilize
his kingside pawn majority in a favourable way then he will get the
advantage, assuming Black doesn’t manage to utilize his queenside majority
for counterplay. But the proviso “in a favourable way” is of great importance.
The advance e4-e5 turns the d5-square into a hole in White’s centre which
can be occupied by ... Bd5, leaving him potentially weak on the light squares.
Therefore White will only advance e4-e5 if the effect of the hole left on d5
has been neutralized in some way, or at least lessened by other positional
factors. For this reason White has to watch out that Black doesn’t manage to
play the pawn stab ... f6-f5 in a favourable manner to win control of the d5-
square.

Question: Strategically speaking, the best move for Black is 20 ... f5, but
can
you see any tactical drawbacks?

20 ... Qh5
Instead, 20 ... f5 21 e5 leaves White with the aforementioned hole on d5
which can be exploited in careful style after 21 ... h6 etc (not 21 ... f4?,
attempting to win a piece, as 22 Ng5, hitting the black queen and h7, would
cost Black a pawn); and 21 exf5 Bxf5 leaves the white centre broken up,
while 21 Ne5 Qh5 also favours Black.
Answer: But tactics come before strategy: the finest move or manoeuvre
has to be rejected if it is unsound. And here 20 ... f5? fails to 21 Ng5 Qg6 22
Nxe6 Qxe6 23 exf5, when White has won a pawn.
The game move, while not doing much actual harm, is a little too
generous as it lets White reroute his bishop to a superior post on the
queenside. Giri might have thought about getting his own bishop over there
with 20 ... Bb4!. For example, 21 Ne1 Bc8 22 Nd3 Ba5 (not 22 ... Rxd4?? 23
Nxb4 Rxb4 24 a3 and the rook is trapped) 23 b4 (after 23 Nc5 b6 24 Nd3
Rxd4 White has just lost a pawn) 23 ... Bb6 24 Nc5 a5 and Black has
sufficient counterplay.
21 Bc7!
After 21 Bh2 or 21 Ne1 Black could play 21 ... f5!.

21 ... Rd7
Here 21 ... Rc8 looks more accurate. After 22 Bb6 f5 23 Rde1 (23 e5?
looks horrible because of the hole on d5 and even drops material after 23 ...
f4 24 Re2 Bxh3 or 24 Rc3 Bb4 25 Rcd3 Bf5, winning the exchange) 23 ...
fxe4 24 Qxe4, then 24 ... Bxh3 leads to a sharp tactical exchange and
perpetual check: 25 Qxe8 Rxe8 26 Rxe8 Bxg2! (but not 26 ... Qg4? 27 Nh4!
Qxh4 28 Bc5, when the threat of mate on f8 wins for White) 27 Kxg2 Qg4+
28 Kf1 Qxf3 29 Bc5 Qd3+ 30 Kg1 Qg6+ and Black can keep on checking.
Instead, 24 ... Bd6 keeps the game going. After 25 Ne5 Bd5 26 Qd3 Black’s
bishop on d5 is a beauty, but White’s well-entrenched knight and doubled
rooks are also imposing. The position remains unclear.
22 Bb6
 The bishop now supports the white centre by defending the d4-pawn
again, which frees the knight and rook on d1 for other duties (jumping ahead,
we’ll see the knight go to d3 and the rook to e1). It also restrains the black
queenside pawns from advancing and gives White extra control over c5,
which helps support his coming knight manoeuvre to that square.
22 ... Bb4?
This time 22 ... f5? is well answered by 23 Ne5, attacking the black rook.
Although Black’s two previous moves didn’t endanger his position, they
were perhaps a sign that he was ‘drifting’. It’s no wonder that he now makes
a definite mistake.
A better option is 22 ... Qg6. White can continue to probe with 23 Qc1!?,
say, intending to build up with 24 Rde1 etc, and if 23 ... Bxh3? 24 Nh4 White
wins a piece, or 23 ... Bd6? (tempted by the chance to threaten 24 ... Bf4 but
allowing a central breakthrough) 24 d5! cxd5 25 exd5 Bf7 26 Rxe8+ Bxe8 27
Re1 Bb4 28 Re6! and White is better. Instead, Black can play more carefully
with 23 ... Be7 or seek counterplay with 23 ... a5!?, when 24 Bxa5 Ra8 25
Qd2 Bd6 leaves White facing the awkward 26 ... Bf4. So 24 Rde1 Bb4 25
Nd2, intending 26 a3, looks the way to keep the game going.

23 Ne1!
Now the knight will get to the d3-square with gain of time, as 23 ... Bxe1
24 Rdxe1 leaves Black with no bishop pair to compensate for his inferior
pawn structure. It also clears the way for a future f2-f4 advance.
23 ... Bf7 24 Nd3 Bd6 25 Rde1 Bb8
Notice how White has achieved a superb co-ordination of his pieces over
the last couple of moves. The rooks and minor pieces are all ready to support
the central push.
26 f4!
Planning to push again with 27 e5.
26 ... f5
 Question: How should White respond to this pawn advance?

Black’s counter-strike against e4 has been a fly in the ointment for the
whole middlegame as far as Anand has been concerned. In particular, he has
been reluctant to advance f2-f4 as long as Black can use the ... f6-f5 pawn
thrust to liquidate the e4-pawn, or force it to advance to e5, when d5 becomes
a massive hole in his centre. But now the Indian former world champion is
satisfied he can concede the d5-square to the enemy bishop without it
draining the energy from his advancing kingside pawns.
Answer: 27 Ne5! Bxe5 28 dxe5 fxe4 29 Qxe4 Rd2
After 29 ... Bd5 30 Qc2 Black has a great-looking bishop, but he doesn’t
have the thing White feared: a blockade on the light squares holding back the
centre pawns. Instead, they are free to advance with 31 e6 or 31 f5 etc, while
an attempted blockade with 30 ... g6 looks fragile and leaves the black
kingside full of dark square holes.
30 R3e2 Bd5 31 Qe3 Rxe2 32 Rxe2
 Bringing an end to the black rook’s counterattack.
32 ... Qf5

Question: Give yourself a pat on the back if you can find White’s next
move!

Answer: 33 g4!!
An extraordinarily important lesson in positional chess. If White plays a
slow move such as 33 Kh2, then Black has time for 33 ... h5!, when he
succeeds in constructing a blockade on the light squares after all. Deprived of
the advance g2-g4, the dynamism has vanished from White’s kingside pawn
majority. Black could even start dreaming of playing for the advantage as his
bishop is superior to the white one blocked in by its pawn on e5.
After the game move Black’s queen gets to hassle the white king for a
while, but the key thing is that the white pawns maintain their vitality.
33 ... Qb1+ 34 Kf2 h5 35 f5 Qh1 36 Kg3 Re7
The black rook can’t help its queen, so her attack is easily rebutted by the
white pieces.
37 Bc5 Re8 38 e6
Now Black has to reckon with 39 Bd4 and 40 Qg5, aiming to mate on g7.
38 ... Kh7 39 Qg5! 1-0
A possible finish is 39 ... Qf3+ 40 Kh4 Qxe2 41 Qg6+ Kg8 42 Qxe8+ Kh7 43
Qg6+ Kg8 44 Bd4 Qe1+ 45 Kxh5 and the mate threat on g7 decides.
Chapter Six
Manoeuvring Against Pawns
If we are unable to impose direct pressure on the opponent’s pieces or attack
his king, we can still disturb the harmony of his set-up by targeting his
pawns.
The aim is to oblige one or more of the enemy pieces to defend a pawn,
so that they are hampered in their movement and work less well with the
other pieces. Our plan might also involve restriction – that is, preventing a
desired advance of a pawn or a group of pawns. Then the enemy pieces will
be denied the chance to take part in an active plan supporting the forward
moving pawn or pawns. Our manoeuvring might lead to the win of a pawn,
but by reducing the energy of the opponent’s pieces it could equally be the
groundwork for scenarios like those in Chapters One or Two.

The Backward Pawn

You might like to reread the brief discussion of backward pawns in the game
Vachier Lagrave-Caruana at the start of Chapter Three.
In our first game below Black misses the chance to gain counterplay by
striking with a pawn at the white centre. As a consequence the pawn itself
festers in a backward state and interferes with the co-ordination of the black
pieces.

Game 42
A.Korobov-V.Anand
German Bundesliga 2019

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Be7 5 Bf4 0-0 6 a3

In my game with Nabaty which follows, White played 6 e3 and we
quickly ended up with a different type of pawn structure.
6 ... b6 7 e3 Bb7 8 cxd5 Nxd5 9 Nxd5 exd5 10 b4
 A new move at the time. Development with 10 Bd3 is usual, when Black
responds with 10 ... c5. Instead, Korobov immediately fights for the c5-
square.
10 ... Nd7?
A demonstration of the power of a theoretical novelty. There are very few
players who react well to being surprised in the opening, and that includes
world champions. If Black chooses to play 10 ... c5, the natural reply to free
his game, he has to carry out a difficult piece of analysis when his brain
hasn’t yet warmed up. According to my computer program Black would be
okay after, say, 11 bxc5 bxc5 and now:
a) 12 Rb1 Ba6 13 Bxa6 Qa5+! 14 Qd2 Qxa6 15 dxc5 Nd7!, planning 16
... Nxc5. Black has enough play due to the awkward position of the white
king who is prevented from castling for the time being.
b) 12 dxc5 Nd7 (preparing to retake on c5 with the knight, when it is
actively placed) 13 c6 Bxc6 14 Nd4 (White has won time to get his knight to
the blockade square) 14 ... Qa5+ 15 Qd2 Qxd2+ 16 Kxd2 Ba4 and Black is
ready to play 17 ... Nc5 with counterplay.
 Of course Anand would have looked at 10 ... c5 during the game, but
(unlike his opponent, who had analysed 10 b4 before the game) he couldn’t
consult a computer program to confirm that it is tactically watertight. He
would have been wondering: “has Korobov prepared a clever move which
I’ve missed?” And so he settled for the safe-looking game move, which
ducks the messy tactics after 10 ... c5. This was exactly what Korobov was
hoping for.
11 Bd3

11 ... Nf6

Question: Can you see a tactical reason why 11 ... c5 is no longer viable
for Black?

Answer: 11 ... c5 now runs into 12 bxc5 bxc5 13 Bxh7+! Kxh7 14 Qb1+
Kg8 15 Qxb7 and White has won a pawn. Black can’t claim sufficient
compensation, as after 15 ... Qa5+ 16 Ke2 any bid to exploit the awkward
position of the white queen or king is hampered by the fact that ... Rb8 isn’t
possible as the white bishop controls b8 and, besides, the knight and d5-pawn
are hanging.
Tricks of the 13 Bxh7+ kind, setting up a double attack by the white
queen on the black king and another piece, are common in the Queen’s
Gambit when Black has castled kingside and doesn’t have a knight on f6.
Perhaps the most common pattern is: Black has an undefended pawn on c7,
no knight on f6, White has a bishop on d3. After Qc2 there is a double threat
of Qxc7 and Bxh7+, so White wins a pawn.
I doubt there are many players who haven’t fallen for this type of trap at
some point. The board can seem a large place, especially when you are
starting out in chess, and to see that the fate of a bishop on b7 is connected
with events on h7 isn’t easy. Once a player gains experience with the
Queen’s Gambit, he’ll be forever wary about falling for a tactic if his
opponent has a bishop pointing at h7 and he only has his king defending it.
12 0-0 Bd6
Because Black has moved his knight away from the queenside, his
position there is too understrength to justify the opening of lines with 12 ...
c5?. Thus after 13 bxc5 bxc5 14 Qb1!, the bishop on b7 again runs into
trouble. If 14 ... Bc6 15 dxc5, Black can’t recapture as 15 ... Bxc5 16 Rc1
will lose material, while 14 ... Qb6 15 dxc5 Bxc5 16 Qc2, intending 17 Rab1,
is very awkward for Black.
13 Qc2!
Correctly deciding that the exchange of bishops on f4, not on d6, will
benefit him the most. Black has failed to engineer the ... c7-c5 advance,
which would have left him with either hanging pawns on c5 and d5 or an
isolated queen’s pawn on d5 (we’ll examine these alternative structures in the
next two games). Instead, he is left with a backward pawn on the open file c-
file: a worse weakness than the other two pawn formations.
13 ... Bxf4
Here 13 ... Qe7 would have been more resistant, not allowing the opening
of lines.
14 exf4
 Question: We’ve criticized Black’s backward pawn on c7, but how
would
you evaluate White’s isolated d-pawn and doubled f-pawns?

Answer: Well, a weakness is only a weakness if it can be attacked. In

contrast to the pawn on c7, there is no open file on which Black can attack
the d4-pawn. (A little thought experiment: remove the black pawn from d5.
Suddenly Black can smash up the white pawn structure with ... Bxf3 and,
after g2xf3, then ... Qxd4. Even if we remove the pawn from d5 and give
White the move, it still looks healthier for Black than the game position.)
Meanwhile, the f4-pawn can be defended simply by g2-g3.
So much for the non-existent static vulnerability of White’s pawns. The
dynamic value of letting them be split up becomes apparent over the next
couple of moves (see, for example, the comment to 16 Ne5).
14 ... Ne4
The knight won’t be able to stay permanently on e4, but Anand wants it
out of the way of his f-pawn so that he can challenge a white knight on e5
with ... f7-f6 at some future point.
15 Rac1
Forcing Black to choose whether he wants his backward pawn to be weak
on c7 or c6.
15 ... Rc8
One of the principles of positional play is to make your opponent use his
big pieces to defend pawns. As long as White makes sure that Black can
never play the ‘explosive’ freeing move ... c7-c5, the rook won’t be a happy
piece on c8.
After 15 ... c6 it would be at least premature to snatch a pawn with 16
Bxe4 dxe4 17 Qxe4, as after 17 ... Re8 18 Qd3 Qd6 Black gains some active
play (compare our thought experiment in the notes to 14 exf4 above).
Therefore White would do better to continue in the style of the game with 16
Ne5, after which Black is left with a pawn on c6 blocking his bishop’s
influence over the centre squares.
Here’s a spectacular continuation (after 15 ... c6 16 Ne5): 16 ... Rc8
(defending c6) 17 Rfe1 f6 (it looks as if Black can evict the knight, but ... ) 18
Rxe4! dxe4 19 Qb3+ Kh8 (19 ... Qd5 20 Bc4 wins the queen) 20 Ng6+! hxg6
21 Bc4!, threatening mate on h3. There is no good way to stop it; e.g. 21 ...
Qd7 22 Rc3 renews the threat. Black would have to try 21 ... Qd5 22 Bxd5
cxd5 23 Rxc8 Rxc8, but the endgame after 24 f3 is lost for him as not only is
he material down, he also has a bad bishop.
As we can see, whether Black chooses to defend or push the backward
pawn it mars the harmony of his piece co-ordination.
16 Ne5
 The d4- and f4-pawns are two pillars supporting the knight on e5. Besides
being on a commanding centre square, it is in touch with c6, a hole in front of
the backward pawn which is an outpost square.
16 ... g6
Preparing to retreat the knight to d6 without dropping the h7-pawn. Upon
16 ... f6 17 Nc6 Bxc6 18 Qxc6 it might appear that Black has benefited from
exchanging his ‘bad’ bishop for the active white knight. In reality the bishop
was defending some key squares, even if it was of no attacking value. Thus
after its exchange Black is faced with threats such as 19 Ba6, ousting the rook
from c8 to win the c7-pawn; or 19 f3, when retreating the knight to d6 costs
the d5-pawn; or finally the simple 19 Bxe4.
Remember when you devise a plan that the ‘bad’ bishop can be a useful
piece in a supporting role. Not every piece can or should be the star of the
show.
17 Rfe1 Re8
Once the knight voluntarily retreats from e4, or is pushed back by f2-f3,
the white rook will gain the most benefit from the open file. How do we
know this, when the black rook is also sitting on the open file? It’s easy: the
player with the superior overall piece deployment comes out on top in the
battle for an open file.
18 g3
Korobov refuses to be rushed. He has the time to make little
improvements – or safety first moves – as his bind on the queenside won’t go
away. Thus he guards f4 and gives himself the chance to expand with h2-h4
in some scenarios. And, as we shall see, there is another advantage to the
pawn move which will be revealed at a key moment in the game.
18 ... Nd6 19 b5

Finally fixing the pawn on c7.

19 ... Re7
Black should try 19 ... Qf6, though 20 h4 intending 21 h5 looks a good
response (here, incidentally, we see the value of 18 g3). With the black
bishop unable to affect matters on the kingside, White would have an
advantage in firepower if he launched a direct attack on the black king.
 Question: As a tactical exercise have a look at 19 ... f6 20 Bxg6!. It’s
good for
White, but would you rather play 20 Nc6 to keep a smaller, more
positional
edge? What choice would you make in a game?

Answer: After 19 ... f6 such an assault could be implemented

immediately with the help of the sacrifice 20 Bxg6! and now:
a) 20 ... fxe5 21 Bxh7+ (the black king is denuded of cover) 21 ... Kh8 22
dxe5 (White isn’t even material down as he has three strong pawns for the
piece) 22 ... Re7 23 Bg6 Ne4 24 f3 Nc5 25 f5 and White rolls his pawns
forwards with a crushing game.
b) 20 ... hxg6 21 Qxg6+ Kh8 (White builds up relentlessly after 21 ... Kf8
22 Qh6+ Kg8 23 Ng4 Rf8 24 Re6! – attacking f6 again – 24 ... Ne8 25 Qg6+
Kh8 26 Rce1 and 27 Re7 will win next move) 22 Ng4 Rxe1+ 23 Rxe1 Ne4
(or 23 ... Ne8 24 Qf7, when 25 Re7 is one threat) 24 Rxe4! dxe4 25 Nxf6 and
Black has to give up his queen to stop mate on h7, as 25 ... Qe7 doesn’t help
after 26 Qh6+.
These are only sample variations, but I hope you can see the strength of
the attack. If White demurs to sacrifice he could indeed try 20 Nc6!?.
Objectively this is inferior, but it nevertheless gives White excellent pressure
and avoids any tactical complexities.
20 f5!
There is no need to arrange h4-h5 as the f-pawn has suddenly grown
fangs.
20 ... Qf8 21 f6 Re6
 Question: You’re doing really well if you find White’s next move.
Any suggestions?

It seems like White might be overstretched, but now comes a beautiful

move:
Answer: 22 Bf1!
A simple retreat which exemplifies what we said at move 17: the player
with the superior overall piece deployment comes out on top in the battle for
an open file. Black has a lousy bishop imprisoned on b7, whereas its white
counterpart is ready to go to h3 with decisive effect (here we see another
benefit in playing 18 g3 as it clears the bishop’s path to h3).
22 ... Rxf6
Black must give up the exchange as 22 ... Ree8 23 Bh3 Ra8 24 Bd7 Red8
25 Qxc7 is abysmal for him.
23 Nd7 Qd8 24 Nxf6+ Qxf6 25 Bh3 Rd8 26 Qxc7
The loss of the backward pawn means that Black’s position falls apart.
 26 ... Ba8 27 a4 1-0

Hanging Pawns
Here is another development from the Queen’s Gambit line which featured in
the previous game.

Game 43
T.Nabaty-N.McDonald
London Chess Classic 2017

1 d4 d5 2 c4 e6 3 Nc3 Be7 4 Nf3 Nf6 5 Bf4 0-0 6 e3 b6 7 Bd3 Bb7 8 cxd5

exd5 9 0-0 c5 10 Qe2 Nc6 11 dxc5 bxc5 12 Ba6 Qb6 13 Bxb7 Qxb7 14
Qb5 Qxb5 15 Nxb5 a6 16 Nc3 Rfd8 17 Rfd1

Black has hanging pawns on c5 and d5. That is, they are on adjacent open
files, with no pawn on either side of them to offer support. They are
vulnerable to attack: indeed, you might describe them as two isolated pawns
joined together. On the other hand, if either pawn needs defending it can
advance a square and be guarded by its brother-in-arms. Also, in contrast to
an isolated pawn structure, there is no safe square for an enemy piece in front
of either pawn as long as they remain on the same rank.
Speaking in general, hanging pawns have dynamic potential and if left
unrestrained one of them, usually the centre pawn, can sometimes advance in
‘explosive’ fashion. In the diagram position Nabaty has very sensibly
exchanged queens to reduce the aggressive value of the hanging pawns.
Nonetheless, they still have the capacity to expand. Black has to make a
decision: should he advance in the centre with 17 ... c4 or 17 ... d4 in order to
obtain maximum benefit from the hanging pawns?
Sitting at the board with my clock ticking, I recalled the games of two
chess heroes: Capablanca and Fischer. They both advanced ... c5-c4 in
similar positions. Let’s have a look at the games:

This is O.Bernstein-J.R.Capablanca, Moscow 1914. Here Capa played 15

... c4 and gave his move an exclamation mark in his book My Chess Career.
He says he wasn’t worried about conceding the d4-square to the white knight:
he wanted to activate his bishop and fix the pawn on b2 as a target. The game
continued 16 Rfd1 Rfd8 17 Nd4 Bb4 18 b3 Rac8 19 bxc4 dxc4 20 Rc2 Bxc3
21 Rxc3 Nd5 22 Rc2 c3 23 Rdc1 Rc5 24 Nb3 Rc6 25 Nd4 Rc7 26 Nb5 Rc5
and now, rather than 27 Nd4 with a likely draw, there came a famous
blunder: 27 Nxc3?? Nxc3 28 Rxc3 Rxc3 29 Rxc3 Qb2! and White resigned
as his weak back rank destroys him however he plays; e.g. 30 Rc8 Qa1+ or
30 Rd3 Qb1+. Note Capa’s excellent judgment in allowing White’s knight to
d4 where it unwittingly helped shield the pawn on d5 from attack.

This is M.Bertok-R.J.Fischer, Stockholm Interzonal 1962, which

continued 17 ... c4 18 Nf4 Rfb8 and White was already under pressure down
the b-file. There followed a swift collapse: 19 Rab1? (the best chance to
confuse things was 19 b3 cxb3 20 axb3 Qxb3 21 Qe7, though after 21 ...
Qb4! 22 Nxd5! Bxd5 23 Qxd7 Be6 Black remains with a strong passed
pawn) 19 ... Bf5 20 Rbd1 Nf6 21 Rd2 g5! 22 Nxd5 (desperation, as he’s
being crushed after 22 Nh5 Ne4 23 Rc2 Qb4! etc) 22 ... Nxd5 23 Bxc4 Be6
and Black soon exploited his extra material to win at move 31. I had seen this
game with notes by Fischer in My 60 Memorable Games.
Returning to my own game, I tried to calculate 17 ... d4 but couldn’t quite
work out what was going on. Did I have enough dynamism or was I just
going to be left with a weak pawn on either c5 or d4? Another notion was
urging me to choose 17 ... c4 – if it was good enough for Capablanca and
Fischer, then surely it was good enough for me?
Unable to make up my mind, I settled on a solid move:
17 ... h6?

More than a hundred years ago the second world champion, Emanuel
Lasker, was warning us against making careless or pointless moves with
pawns. But still we don’t listen. I only have to recall this game to know that
I’m not above wasting time in positions where tempi are as precious as gold
dust.
Maybe have a look through your own games and see if you make similar
time-wasting moves. Were they a way to put off having to make the effort to
think or make a decision?
18 Rac1 c4
Only now did I take the plunge. We should examine the rest of the game
and then return to see how Black could have done better at move 17 and 18.
 19 Ne5!
A good active response. After something slow like 19 h3 Black has 19 ...
Nb4, planning to invade on d3, when 20 Ne5 is too late after 20 ... Bd6,
intending to exchange on e5 and then go ... Nd3 anyway.
19 ... Na5!
After 19 ... Nxe5 20 Bxe5 the black centre pawns are near to death due to
the threat of 21 Bxf6 Bxf6 22 Nxd5 Bxb2 23 Rxc4. Note that Black is once
again hampered in his counterplay by the fact that White’s bishop controls
the b8-square and so prevents ... Rab8, attacking b2.

During the game I didn’t think things were too bad for Black here. With
luck there would be time for ... Ra7 and ... Rb7 to attack b2, or even ‘Capa’s’
move ... Bb4 to activate the bishop. But I had missed something.

Question: Can you see Nabaty’s next move which begins a fine plan to
undermine the d5-pawn?

Answer: 20 h4!!
 The plan is simple and strong: White will advance g2-g4-g5 to drive the
black knight away from f6 and then win the d5-pawn. In the Capa and
Fischer games the advance of the white kingside pawns was never a ‘thing’.
20 ... Bb4
To counter the plan of g4-g5 Black has to take the pressure off d5 by
giving himself the resource ... Bxc3.
21 g4 Rac8
I didn’t like the look of 21 ... Bxc3 22 Rxc3 Rac8 23 Ra3 Nc6 24 Rxa6
Nxe5 25 Bxe5 Nxg4 26 Bc3, when White has a passed pawn, the superior
minor piece, and continuing pressure on d5.
22 Na4!
Now the knight invasion on b6 is very difficult to meet.
22 ... Bd6
Losing quickly. Black should have played 22 ... Rb8 and hoped for the
best, though 23 g5 Nh5 24 Bh2 looks pretty grim.
23 Nb6 Rc5
Or 23 ... Rb8 24 Nxd5, when 24 ... Bxe5 25 Bxe5 wins as b8 is hanging,
while 24 ... Rxb2 25 Nxf6+ gxf6 26 Nxc4 is a carve-up.
24 Ned7!
 Winning material however Black plays.
24 ... Rb5
Or 24 ... Nxd7 25 Bxd6 Rc6 26 Be7 Re8 27 Nxd7 Rxe7 28 Rxd5 and
wins. Black could resign already. I waited until he disentangled his pieces.
25 Bxd6 Nxd7 26 Bc7 Re8 27 Nxd7 Nc6 28 Nb6 d4
A little late!
29 Nxc4 1-0
An excellent game by Nabaty, who showed complete mastery of the
tactical and strategic requirements of the specific situation after 18 ... c4.

However, we haven’t finished our analysis of the key moments.

We’ve seen that 18 ... c4 gave Black a far from easy game, but would 18
... d4 have been okay for him?
 Let’s see: 18 ... d4 19 Na4 dxe3 20 Bxe3 Nb4! with ideas of 21 ... Nxa2
or 21 ... Nd3 gives Black enough play. So 19 exd4 Nxd4 20 Nxd4 cxd4 21
Na4 Nd5 22 Be5 is critical, and here Black has to find 22 ... d3!, when 23
Rxd3 Nb4 24 Rxd8+ Rxd8 25 g3 Nxa2 holds, but 23 Kf1! Re8 (threatening
24 ... Bg5, attacking e5 and c1) 24 Bd4! is still uncomfortable for him.
So we have to conclude that Black wouldn’t equalize easily with 18 ... d4
either. He has to find hard computer moves just to stay alive. But was there a
chance to play the pawn advance at a more opportune moment?
Yes: Black had played the slow and unnecessary 17 ... h6 on the previous
move. If Black instead plays 17 ... d4! immediately, he gets to initiate a hard
fight in the centre before White has had time to play Rac1. This makes a huge
difference.
 Now after 18 exd4 Nxd4 19 Nxd4 cxd4 20 Na4 Black has 20 ... Rac8!,
getting to the c-file first with at least an okay position. Or 18 Na4 Nd5! (if
White’s rook were already on c1, this would just lose a pawn to Nxc5) 19
Rac1 dxe3 20 Bxe3 Nxe3 21 fxe3 Rxd1+ 21 Rxd1 Ra7 with equality.
It is worth observing that 17 ... c4 also has more merit when it is played
without the ‘extra’ Rac1 for White, though 18 Ne5 still gives White the edge.
Thus the culprit for Black’s problems was 17 ... h6?. He avoided 17 ... d4
or 17 ... c4 for a turn in favour of a quiet move he could ill afford in such a
double-edged position. In effect he lost a tempo.

Losing a game is painful, especially when we feel we haven’t done ourselves

justice (which for most of us feels like every game we lose). Nevertheless, a
lot can be learned through studying our defeats in detail and trying to get to
the root cause of our loss. It is too easy to make excuses and come up with
false reasons which protect our ego but do nothing to stop us losing the same
game again.
In the game above I was indecisive at the key moment, partly because I
failed to analyse 17 ... d4 properly. This suggests I would benefit through
practising calculation more. Also, I haven’t had much experience with
hanging pawns, so I should look at some modern games featuring them, not
just rely on the memory of two games, one of which was played more than a
100 years ago.
It should also be mentioned that book learning and a knowledge of
famous games is often a great help when it comes to choosing a plan.
However, there is the danger that you might ignore the specific features of a
position in front of you and rely on your memory of how similar (but not
identical) set-ups were handled in the past. Sometimes a move or idea has
made such a strong impression on you that you think it must always be the
correct plan. Thus in the game above I couldn’t escape the impression that
Capablanca and Fischer had made on me.

The Isolated Queen’s Pawn (IQP)

In the following game White lets himself be pushed around, as he hopes in
the long term to quell Black’s activity and then exploit the static weakness of
the IQP.

Game 44
J.K.Duda-D.Navara
Prague 2019

1 Nf3 d5 2 d4 Nf6 3 c4 c6 4 e3 e6 5 Nbd2

 In one of his books Nimzowitsch calls putting the knight on d2, rather
than c3, in this type of set-up a decentralizing move.

Question: Any idea why he might have said this?

Answer: At first it sounds absurd, as the knight has most definitely

moved to the centre. But there is a kernel of truth in the statement once you
think about it (and Nimzowitsch’s greatness as a chess teacher was that he
got players thinking about accepted chess values).
The move is certainly against the spirit of centralization, as by spurning
the chance to go to c3 the knight loses influence over the d5-square and cuts
off the queen’s view of the action on d4. White is in a sense less well
centralized than he was before he put the knight on d2.
Black’s response reflects this: with White’s grip on the centre at least for
the moment lessened, he feels justified in pushing the c-pawn a second time
to challenge d4.
5 ... c5 6 dxc5 Bxc5 7 a3 a5
 It’s wise to stop White gaining ground with 8 b4.
8 Qc2 0-0 9 b3
Black has nothing to fear after 9 cxd5 Qxd5 10 Bc4 Qh5, so Duda settles
for a quiet development of his bishop on b2.
9 ... b6 10 Bb2 Bb7 11 Bd3 Nbd7
Instead, Black could liquidate in the centre with 11 ... dxc4, when after 12
Nxc4 Nbd7 13 0-0 Rc8 14 Qe2 (or 14 a4 h6) 14 ... Qe7 15 Rfd1 Rfd8 he has
full development and no structural weaknesses.
12 cxd5 exd5
Black could still avoid an IQP with 12 ... Bxd5, though he has to satisfy
himself that threats to his h-pawn are innocuous after 13 e4 Bb7 14 e5 Ng4
15 0-0 Rc8! (threatening 16 ... Bxf2+), when 16 Bxh7+ Kh8 17 Qd3 Nxf2 18
Rxf2 Bxf2+ 19 Kxf2 g6, intending 20 ... Kxh7, favours Black.
13 Bd4

It is somewhat unusual for a bishop, rather than a knight, to take up

residence on the blockade square in front of the IQP. However, Duda wants
to clear the way for his queen to go to b2 if indirectly threatened by 13 ...
Rc8. (Otherwise she’d have to retreat to b1 or d1, where she is less active and
interferes with the co-ordination of the rooks.) Besides, after 13 Nd4 Black
can activate his knight with the immediate 13 ... Ne5.
13 ... Bxd4
Black might consider 13 ... Bd6!?, leaving White with a bishop rather
than the ‘ideal’ set-up of a knight on d4, and planning to activate his own
knight with 14 ... Nc5 (rather than on e5 as in the game). Also possible was
13 ... Qe7, keeping the tension between c5 and d4.
14 Nxd4 Ne5
The reason for exchanging on d4. Black’s knight gets to an attacking
square with gain of time as 15 0-0? Nxd3 16 Qxd3 Ba6 would be awkward
for White.
15 Be2 Ne4
A second knight jumps forward. Black is getting full value out of the IQP
by utilizing the outpost square it supports on e4.
16 0-0 Qc8 17 Qa2
The queen retreats into the corner, not liking the look of 17 Qb2 Qc3 18
Qxc3 Nxc3, though this would give White the chance to expel the invader
with a move in the laid-back style of the twelfth world champion, Anatoly
Karpov: 19 Nb1!, and if 19 ... Nxe2+ (Black should keep more tension with
19 ... Ne4) 20 Nxe2, White can continue with Nbc3 and Rfd1, looking to
exploit the IQP.
17 ... Nc3 18 Qb2 Nxe2+ 19 Nxe2 Nd3 20 Qb1 Ba6
White’s queen has been pushed around and the black knight has taken up
residence on d3. That’s a lot of indignity Duda has had to endure. On the
other hand, Black hasn’t been able to land a telling blow. The basic shape of
the position hasn’t changed, with the white pawn structure solid and Black
having an IQP. White can oust the black knight from d3, after which Black’s
strategy is revealed to have been a lot of huffing and puffing, as what exactly
has it achieved?
21 Rd1 Re8 22 Nf3
White finally gets to push the knight back.
22 ... Ne5 23 Nxe5 Bxe2
 In general, every exchange increases the ‘endgame features’ of an IQP
position, which makes the pawn’s structural weakness more pronounced and
reduces potential dynamism.
24 Rc1

Question: Why is this preferred to either 24 Re1 or 24 Rxd5, capturing

the pawn?

Answer: After 24 Re1 Rxe5 25 Rxe2 Black can be rid of the IQP with 25
... d4!. If instead 24 Rxd5 Qe6 25 Rd2 Qxe5 26 Rxe2 Red8, the passive white
pieces and Black’s control of the open d-file (which he can increase with
moves like ... Rd6 and ... Rad8) would give him compensation for the pawn.
Duda would rather build up slowly against the IQP and tie one or more of the
black pieces down to its defence. But to begin with he must rescue his knight
from its rather precarious situation. The first step is to attack the black queen.
24 ... Qb7
After 24 ... Qe6 25 Nc6 the knight will reach d4 and with gain of time by
hitting the queen.
25 Qb2
Not only defending the knight but forcing the black bishop to choose
which diagonal to retreat along. Whichever way it goes, a retreat square will
open for the horse.
25 ... Bh5
Or 25 ... Ba6 26 Nf3 and the knight is finally ready to land on d4.
26 Nd3 f6 27 Nf4 Bf7
Some exemplary positional play by Duda has forced the black pieces
back. There isn’t much left of Navara’s initiative: his queen and remaining
minor piece (the worst one of the whole set as it is impeded by the IQP) have
taken up defensive roles guarding the d5-pawn.
It seems that Duda relaxed now. After all the hard work of rebuffing his
opponent’s initiative and getting his knight home to f3, he takes a break and
plays a ‘luxury move’. But as we shall see Black’s dynamic chances haven’t
yet been exhausted.
28 h3?
 Question: Can you see Navara’s powerful response to this move?

Instead, 28 Ne2! was called for. White could then stick the knight on d4
and play to double rooks on the c-file with Rc3 etc. If Black opposed this
with ... Rac8, the resulting exchange of rooks would accentuate the weakness
of the IQP and the inferiority of the bishop to the knight. Black would still be
far from lost, but the endgame would certainly be unpleasant for him.
After the game move we should recall Nimzowitsch’s dictums that: In the
last resort position play is nothing other than a fight between mobility (of the
pawn-mass) on the one side and efforts to restrain it on the other. And: The
passed pawn is a criminal who should be kept under lock and key. Mild
measures, such as Police surveillance, are insufficient.
When you are playing a Goldfinger of chess like David Navara – I mean
a player rated well over 2700 – you should be especially careful with
preventive measures. The lack of restraint with 28 Ne2 is immediately
punished.
Answer: 28 ... d4!! 29 Qxd4!
To be fair to Duda he recovers well. After 29 exd4, in return for the pawn
Black has an open file for his rook on e8, clear diagonals for his queen and
bishop, and weaklings on d4 and b3 to attack. He can regain his pawn with a
good game; e.g. 29 ... Qe4 30 Qd2 g5 (30 ... Rad8 is good too) 31 Nd3 Qxd4.
29 ... Bxb3
Now Black intends to mobilize his queenside pawns with 30 ... b5, so
White plays a restraining move.
30 a4! b5?
Anyway. Navara still seems to be feeling euphoric at freeing his game
with 28 ... d4!! and gets carried away. He thinks that the passed pawn he
creates on the a-file, supported by the bishop, will be a strong force.
Unfortunately, the white pieces are too well placed on the queenside to make
this a viable plan.
Chances would be balanced after 30 ... Rad8 31 Qb2 Bf7 32 Rab1 Rd6.
31 axb5 a4
After 31 ... Qxb5 32 Rc5 Qb6 (or 32 ... Qb4 33 Qxb4 axb4 34 Rb1 Ra3
35 Rb5, winning the b-pawn) 33 Nd5! (perhaps Black missed the strength of
this move, which attacks the queen and also has ideas of a fork on c7) 33 ...
Bxd5 34 Qxd5+ the a-pawn drops.
 32 Rc5
Keeping the extra pawn. Black is running out of activity as 32 ... a3 drops
the passed pawn after 33 Qb4.
32 ... Rec8 33 Ne2!
Intending to reorganize his pieces with 34 Qc3 and 35 Nd4, when the
knight finally reaches the d4-square where it defends b5 and attacks b3.
Looking for counterplay Black blunders his passed pawn.
33 ... Qe7

Question: Try to see a very neat tactic for Duda and work out the
variations.

Answer: 34 Rxa4! Rxc5

Black also loses after 34 ... Qxc5 35 Rxa8, or 34 ... Rxa4 35 Rxc8+, or 34
... Bxa4 35 Qc4+ and then 36 Rxc8+.
35 Rxa8+ Kf7 36 b6 Rd5 37 Ra7
Simplification is the surest way to win.
 37 ... Rxd4 38 Nxd4 Bd5 39 Rxe7+ Kxe7 40 Nf5+ 1-0
If Black defends the g7-pawn, then 41 Nd6 and 42 b7 wins the bishop,
while 40 ... Kd7 41 Nxg7 wins easily for White.
In many ways this was a typical tournament game. Blunders caused by
over-confidence and carelessness were mixed with moments of positional
mastery and tactical acuteness.

Preparing an Attack on a Fixed Pawn Centre

Every pawn move has to be carefully considered. There is no going back
once you’ve committed it to a square.

Game 45
O.S.Phillips-C.N.Ross
Solihull 2017

I have had the good fortune to coach Chris Ross, the leading British blind
player, at various international events through the years. I have to admit that
his understanding of the opening lines he plays is superior to my own
knowledge of them, and I wouldn’t presume to outwit him in a tactical battle.
Fortunately, there are still ways I can help him. The diagram position was
reached in one of the games I studied with him as a warm-up to a major
event.
 Question: It is Black to move. How would you assess the position?
What is Black’s best plan and what is the best starting move?

Here Chris played 27 ... e5. He gains space, restrains the white centre and
puts a pawn on the opposite square to his bishop. Surely a good idea? But as
soon as I saw this move I instantly disliked it. It might be a good move in
itself, but it isn’t part of a good plan. In fact, it hinders Black’s plan.
Answer: Black should instead be trying to break up the white pawn centre
and increase the scope of the bishop by advancing ... f7-f5 at a good moment.
If the white rook remains on f1, this will have to be done with ... g7-g6 and
then ... f7-f5. If the rook goes away ... f7-f5 might be possible ‘in one go’.
Let’s imagine that, instead of 27 ... e5, Black played 27 ... Kf8, a sensible
move aiming to centralize the king as a prelude to a future ... f7-f5 break.
 Here are some sample lines:
a) 28 Kf2 Ke7 29 Ke2 Rc1 and White is tied up, as 30 Nf3 Rc2+ loses a
pawn, while after 30 Kd2 Rb1 31 Kc2 Ra1 the pin on e1 is awkward (32 ...
f5! is already on the cards).
b) 28 Rf2 Ke7 29 Rc2 Rxc2 30 Nxc2 Kd6 31 Kf2 f5 32 exf5 (32 Ke3
fxe4 33 dxe4 Ke5 wins the e-pawn) 32 ... exf5 and Black has achieved his
aim: great bishop, broken white pawns.
c) 28 Nf3 Ke7 29 Nd4 g6 30 g4 (or 30 Kf2 f5) 30 ... e5 (only now, as part
of a plan to enforce ... f7-f5) 31 Nb3 Rc2 32 Rf2 Rxf2 33 Kxf2 Bc8 34 Ke3
Kd6 and Black is ready to break up the white kingside with ... h7-h5!
followed by ... f7-f5 (note that White can’t play 35 g5 as the h3-pawn would
hang).
In the game 27 ... e5 was answered by 28 Kf2 and Black, with a slight
edge, went on to win a complex struggle after 28 ... Rc1 29 Ke3 f6. Instead,
28 Nf3! looks a better approach.
 For example: 28 ... f6 (after 28 ... Rc2 29 Nxe5 f6 30 Nd7 White has
counterplay, as 30 ... Rxb2? would be a serious mistake due to 31 Rc1,
threatening 32 Rc7, winning a piece) 29 Rd1 Rc2 (29 ... Kf7 30 Kf2 a5 31
Ke3 is similar) 30 Rd2 Rc1+ 31 Kf2 Kf7 32 Ke3 and White’s pieces are in a
compact group. After 32 ... Ke7 he can activate his game with 33 d4!; e.g. 33
... exd4+ 34 Nxd4, when 34 ... Re1+?! 35 Kf2 achieves nothing for Black as
35 ... Rxe4? 36 Nf5+ Ke6 37 Nd6 is a deadly fork. Against other moves
White can also consider 33 d4.
As we can see, 27 ... e5 slows down Black’s build-up, makes the pawn on
e5 a target for both Nf3 and a future d3-d4 pawn stab, and takes away the
pillar on e6 that, along with ... g7-g6, would support Black’s own pawn lever
with ... f7-f5.
When I first saw 27 ... e5 I couldn’t articulate these objections to the
move. I just knew I didn’t like it at a gut level.

In the next example Carlsen is happy to improve his opponent’s pawn

structure so that it becomes vulnerable to a flanking attack.
Game 46
Ma.Carlsen-S.Karjakin
Sinquefield Cup, St. Louis 2017

Question: It is White to move. What should he attack as the first stage of

his plan?

Answer: Karjakin’s knight on c3 is securely defended but it is adrift from

the other black pieces. On the other hand, if he gets the chance to challenge
for control of the c-file with ... Rc8, the horse might prove well placed for
active operations on the queenside. Black might even gain the initiative.
Therefore Carlsen acts quickly. He starts a sharp struggle in the centre and on
the kingside to highlight the fact that, with the knight offside, Black’s
defence is essentially a man down. (Technically speaking White’s knight is
even more offside being on the edge of the board. But unlike the black horse
it is co-ordinated with the other white pieces and doing a sterling job in
defending the rook.)
28 Bf4!
The first step is to threaten the d6-pawn in order to provoke the exchange
of bishops.
28 ... Be5
There’s not much choice as 28 ... d5 29 Rc7 Qd8 30 Nc6 sees the white
pieces invade (note what we said above about the effectiveness of the knight),
when after 30 ... Qa8 31 e5 the d4-pawn is dropping.
29 Bxe5
Carlsen isn’t squeamish about straightening out the black pawn structure
as it is a prelude to dismantling it.
29 ... dxe5 30 f4!
An assault not only on the e5-pawn but also on the d4-pawn which
depends upon it for vital support. The wearing away of Black’s centre will
also undermine the knight’s base on c3.
Besides making possible the attack on e5, the exchange of bishops has
weakened the dark squares on Black’s kingside and emphasized the liability
of the runaway knight by depleting the number of pieces that can cover for its
absence.
30 ... Qe7
Looking for counterplay against the b4-pawn. Black can’t hold things
together with 30 ... f6 as his kingside crumbles after 31 f5! gxf5 32 exf5 Bxf5
33 Rxf6.
 Question: How does White keep control?

Answer: 31 Rc5!
Note that 31 fxe5? would be massively wrong, as 31 ... Qxb4 then gives
Black a passed pawn and disturbs the white knight. Even worse, after 32 Nb7
Black frees his knight with the trick 32 ... Nd5!, when it is inviolable because
the white queen hangs. This would mean the total defeat of White’s plan.
Instead, with his simple but elegant rook retreat, Carlsen blocks the attack
on b4, clears the way for Nc6, and attacks e5 again.
31 ... Rc8?
This leads to a collapse on the kingside. Karjakin must have
underestimated White’s 35th move. 31 ... exf4 was necessary. After 32 Qxf4
(32 gxf4? allows 32 ... Qxh4+), White is ready to play 33 Nc6 etc, when d4 is
a serious weakness. Nonetheless, Black could have resisted with 32 ... Nd1!,
intending 33 ... Ne3, when the knight is finally useful (in some cases it can
even go to g4 to rejoin the other pieces). The game would then be far from
over.
32 Rxc8+ Bxc8 33 Nc6 Qd6 34 Nxe5 Qxb4
It looks as if Black has gained good counterplay in the form of the passed
pawn, but Carlsen reminds him his kingside is full of dark square holes.
35 f5! Qd6
The trick 35 ... Nd5 no longer works. Among other winning replies White
has 36 Qa2!, when the knight can’t save itself as f7 would fall.
36 Nf3!

Clearing the way for the e-pawn and preparing to go to g5 in some key
lines.
36 ... gxf5
There are some fun variations after 36 ... b4 37 Qg5 b3 38 e5 Qf8 (if 38 ...
Qb6 39 Qh6, the threat of 40 Ng5 followed by 41 Qh7+ and mate on f7 is
decisive) 39 e6! b2 40 Ne5! (the tactics might seem complex, but remember,
the essential point is that White is attacking a flimsy pawn structure with
overwhelming force; everything just flows for him) 40 ... b1Q (or 40 ... fxe6
41 f6 b1Q 42 Qxg6+ Kh8 43 Nf7+ Qxf7 44 Qxf7 and mate follows on g7) 41
exf7+ Qxf7 (upon 41 ... Kh7 42 Qxg6+ Kh8 43 Qf6+ Kh7 44 Nf3 Qc1 45
Ng5+ Qxg5 46 hxg5 the advancing pawns will squash the black king and
queen) 42 Nxf7 Kxf7 43 Qxg6+ Ke7 44 f6+ Kd8 45 f7 Qb4 46 Qg8+ Kc7 47
f8Q and after all the mayhem White wins ‘on points’.
37 Qg5+ Kh7
After 37 ... Qg6 38 Qd8+ the bishop is lost.
38 e5!
Clearing the queen’s path to d8.
38 ... Qg6
The queen is forced away from the d-file as 38 ... Qd7 39 Qxh5+ Kg8 40
Ng5 wins for White.
39 Qd8
By attacking the bishop White gains time to get his knight to g5 and
capture the d4-pawn.
39 ... Be6 40 Ng5+ Kg7 41 Qxd4 Na4 42 Nh3! 1-0

Retreating to victory. Black has survived the initial wave of the attack,
but the passed pawn offers him no counterplay and his knight is still cut off
from the kingside. White’s next move will be 43 Nf4, when he could win the
h5-pawn with Bf3 and Nxh5+, or prepare simply to advance his centre pawns
(most likely Carlsen would do both). Rather than watch his position be
eviscerated, Karjakin resigned.

The Perils of Pawnless Play

Our next game is yet another reminder about the need to utilize your pawns
in any plan you undertake. White’s queen and rooks are an impressive sight,
tripled on an open file and assailing a backward pawn on f7. But there is no
pawn push available to break through the defence. White’s attack therefore
goes to seed, while Black is able to carry out a flanking attack on the other
wing because his initiative there has the support of his pawns.

Game 47
E.Sutovsky-P.Eljanov
Poikovsky 2014

1 e4 c6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5 5 Ng3 Bg6 6 Nf3 e6 7 h4 h6 8 Ne5

Bh7 9 Bd3 Bxd3 10 Qxd3 Nd7 11 f4 Bb4+ 12 c3 Be7 13 Bd2 Ngf6 14 0-0-0
c5
In Chapter Ten we’ll examine this opening line, including the over-
aggressive 15 d5. Here White builds up slowly in the centre.
15 Be3 0-0 16 Ne4 cxd4 17 Bxd4
It looks a better idea to throw in 17 Nxd7 Qxd7 18 Bxd4 to try for
pressure in the centre. That way White’s bishop enjoys open diagonals
pointing left and right.
17 ... Nxe5!
A good moment to exchange, as we see from the next note.
18 fxe5
White can no longer keep the bishop active with 18 Bxe5, as he would
lose a piece after 18 ... Qxd3 19 Nxf6+ (or 19 Rxd3 Nxe4) 19 ... gxf6! 20
Rxd3 fxe5. Therefore White has to recapture with the pawn.
18 ... Nxe4 19 Qxe4 Qc7 20 g4 Rad8
 Question: What do you think of White’s attacking chances?

Answer: An initial evaluation of the position might be: “White stands

better. He has a space advantage in the centre and attacking chances on the
kingside.”
But if we look more deeply we might start to have doubts. The exchange
of all but one of the minor pieces means that White’s kingside attack is
somewhat lacking in supporters. And White’s remaining minor piece isn’t
any great attacking weapon as it is boxed in by its own pawn on e5 (so much
for White’s central space advantage: he’d much rather have the pawn on f4 to
add oomph to the kingside pawn storm and leave the bishop with a clear view
of g7). If the bishop has any value it is as a barrier on the d-file. It prevents
Black offering the exchange of rooks, when White’s attacking chances would
dwindle to nothing.
When we consider White’s kingside pawn advance things aren’t as rosy
as they might appear. If he pushes h4-h5 he will concede the g5-square to the
black bishop, when a breakthrough with g4-g5 becomes infeasible – he
wouldn’t be able to arrange Be3 and one of the rooks to g1 to support g4-g5
without Black exchanging rooks on the d-file (if it’s Rhg1) or (after Rdg1)
counterattacking against e5 and on the d-file in general with ... Rd5 and ...
Rfd8 etc.
It is well worth remembering the adage that a wing attack can’t succeed if
you stand badly in the centre.
In the game Sutovsky made the natural pawn advance.
21 g5 h5!
Locking the pawns together to keep the kingside closed. We should
repeat that Black’s defence is greatly facilitated by the exchange of three
minor pieces. There are no white knights or light-squared bishop to attack h5
or sacrifice themselves on that square to open lines. The feeble bishop on d4
takes no part in the proceedings.
22 Rhf1
In this type of set-up 22 g6, to ram the black kingside and open lines,
would be the natural move. But once again White would be hampered by the
lack of pieces supporting the (so-called) attack. Black could reply 22 ... Qc6!
to offer an unwelcome queen exchange. If White then plays 23 gxf7+ Rxf7
he has achieved less than nothing with 22 g6, as Black’s rook has been given
the f-file and his bishop can attack the h4-pawn.
22 ... g6
Completing the blockade of White’s kingside pawns. This is bad news for
him because a plan seldom succeeds without the use of pawns. As we
remarked about the h5-pawn, the g6-pawn is also a solid barrier as, due to his
limited attacking resources, White can’t sacrifice anything to smash through
it (assuming Black shows a modicum of care, as we shall see). It’s tough
trying to break through to mate when the cheapest cannon fodder available to
you isn’t a pawn, or even a knight, but a rook.
White decides to triple his pieces on the f-file and ... hope something
turns up.
23 Rf4 Rd7
Defensively, the rook prepares to bolster f7; aggressively, it is ready to be
doubled and crash through the d-file once the obstacle on d4 is removed.
24 Rdf1
 Question: Now Black has to be a bit careful. What is White’s threat?

Answer: The most splendid positional build-up can be spoilt by one

moment of tactical madness. The threat is 25 Rxf7! Rxf7 26 Qxg6+ Rg7 27
Qxe6+, when White has a draw if no more than that: 27 ... Kh8 28 Qh6+ Kg8
(he has to avoid 28 ... Rh7 as 29 Rf8+! mates next move) 29 Qe6+ etc with
repetition. Black has no need to let White force an honourable draw. On the
contrary, he can start thinking about getting the advantage himself. He
therefore adds his king as a defender to g6.
24 ... Kg7! 25 Qe3 b6!
Not only defending a7 against capture, but also preparing his next move
to challenge the bishop on d4.
26 Qf2
The white pieces have reached maximum power on the f-file and are
threatening to crash through on f7. But Black is prepared.
26 ... Bc5! 27 Kb1
 After 27 Bxc5 bxc5 the e5-pawn is weak and Black can prepare
counterplay on the b-file with 28 ... Rb8 etc.
27 ... Bxd4 28 cxd4

White has let the exchange take place on d4 so that he can recapture with
the pawn and no longer has to worry about defending e5.

Question: Try to list the features in the position which favour Black.

Answer: There are two backward pawns in the position: on f7 and d4.
Black’s king can help defend the pawn on f7 without much trouble, whereas
while there are queens on the board the white king would never venture to d3
to help his pawn. Therefore the d4-point is the greater weakness.
White’s pawns on the kingside are blocked and can’t add to his pressure
on the f-file. In contrast, Black’s pawns on the queenside are free to advance
and can co-operate with the pieces in any plan of attack on that wing.
Finally, Black’s pieces have the open c-file. His queen can aim to break
through via the holes on c4 and d3.
 As long as White is unable to land a bloody blow on the f-file, we have to
conclude that Black has the advantage.
28 ... Qc4
After many moves of methodical defence Eljanov finally goes on the
attack. The d4-pawn is now attacked twice, tying down the white queen and
f4-rook to its defence.
29 Rc1
The rook takes the open file, though he won’t get to enjoy it for long.
29 ... Qd3+ 30 Ka1 Rc8!
After waiting on f8 since move 15 the rook has a stirring entrance. It is
immune due to a back rank mate.
31 Rf1 Rcc7!

A little lesson in improving the co-ordination of the pieces. The rook

resumes its duty guarding f7, but this time from a square where it is also
active.
White is in effect in zugzwang. He can’t move his king. If he moves the
rook on f4 to f3, the d4-pawn drops. If he lessens the pressure on f7 by
moving the rook on f1, then ... Rc2 invades his second rank as White doesn’t
have Rxf7+ in response. If Qg1, keeping d4 guarded, then ... Rc2 again
follows. White is therefore reduced to moving his queenside pawns, but as
we shall see this isn’t without a drawback either.
32 a3 b5!
With the black queen and rooks having tied down the white pieces, it is
now time for the queenside pawns to show their worth. The pawn on a3 will
provide a ‘hook’ for their advance.
33 Rf6 b4
Whittling down the white king’s defences. After 33 ... Qxd4? 34 Rxf7+!
(34 Rxg6+? Kxg6! wins for Black) 34 ... Rxf7 35 Qxd4 Rxf1+ 36 Ka2 White
has better drawing chances than in the game.
34 axb4
If 34 Rxe6 then 34 ... bxa3 is crushing for Black; but not 34 ... fxe6?? 35
Qf8+ Kh7 36 Qh6+ and mate next move.
34 ... Qb3

Not only to regain the pawn but also defending e6 with the queen so that
35 Rxe6 isn’t an option.
35 Qf3
White’s attack is tantalizingly close to getting him a draw or even a win
with 35 Rxg6+ Kxg6 (35 ... fxg6?? 36 Qf8+ will mate as before) 36 Qf6+
Kh7 37 Qh6+ Kg8 38 g6, but the attack is beaten off by 38 ... Qa4+! (the only
way) 39 Kb1 Qc2+ 40 Ka1 Qxg6 and Black wins.
35 ... Qxb4 36 Rf4 Qa4+ 37 Kb1 Qc2+ 38 Ka1 Kg8!
Ensuring that a rook capture on f7 doesn’t come with check. This means
that 39 ... Rd5 (with the threat of 40 ... Ra5+) and other active rook moves are
on the cards.
39 Qe4 Kg7
Evidently Black was short of time and so decided to repeat moves. With
longer to think he would have seen that 39 ... Qc4, planning 40 ... Rd5, is
decisive. Presumably he was afraid of 40 Rxf7 Rxf7 41 Qxg6+, but the king
evades the checks after 41 ... Kf8 etc. Sutovsky decides not to repeat moves
as, with the time control at move 40, Eljanov would soon be able to work out
an attacking win against his king. Instead, he exchanges queens and tries to
survive the rook and pawn endgame.
40 Qxc2 Rxc2 41 d5
 The only chance for active play. Otherwise the black pieces will gang up
on the d4-pawn with ... Rd2 or the b2-pawn with ... Rb7 etc.
41 ... exd5 42 Re1 d4 43 Kb1
There’s no time for 43 e6 fxe6 44 Rxe6 d3 45 Rff6 to attack g6 as the d-
pawn will roll through after 45 ... d2.
43 ... d3 44 Rd1 Re2 45 Rf3 Rb7
Black exchanges the passed pawn for two rooks on the seventh rank – a
good deal!
46 Rdxd3 Rbxb2+ 47 Kc1 Ra2 48 Kb1 Reb2+ 49 Kc1 Rh2 50 Kb1
Rae2 51 Rf1 Rxe5 52 Rd7 Rf5!
White has spent the whole game trying to break through on the f7-square
and here, after 52 ... a5??, it’s mate in three with 53 Rdxf7+. Of course
Eljanov was never going to let that happen.
53 Rxf5 gxf5 54 Rxa7 f4 55 Ra3 Kg6!
White is only one pawn down but the disparity in activity between the
two kings proves fatal.
56 Ra6+ Kf5 57 Rf6+ Ke4 58 Rxf7
 He finally gets to take on f7, but it is too late to change anything.
58 ... Rxh4 59 Kc2 Rg4 60 Rg7 h4 61 Kd2 Kf3!

Putting the king on g2 is the cleanest way to win. There’s no need to get
involved in 61 ... f3 62 Ke1, though that also wins.
62 g6 Kg2 63 Rf7 h3 0-1
It remains to be asked where things went wrong for White. We can trace
the petering out of his attacking chances right the way back to 17 Bxd4, when
his bishop became an ineffective piece. In general White’s initiative was
diminished by the exchange of three minor pieces. As far as being in danger
of losing, we could point to White’s insistence on trying to exploit the f-file.
He could instead have kept his rooks centralized, say on the d-file, and
fended off Black’s initiative. But it is psychologically hard to assume a
defensive pose with the white pieces when the siren of attack is calling you.
The f-file was a blind alley that sucked all the dynamism out of White’s
position.

Slipping into Quicksand

If a centre pawn loses its power to advance, can’t be defended by another
pawn, and is vulnerable to frontal attack then it can become as serious a
liability as a backward pawn. When I played through the following game I
was reminded of a comment made by Ulf Andersson in Learn from the
Grandmasters about a game between Reshevsky and Bronstein at Zurich
1953: “The way Black manoeuvred his forces, especially the knights, to
increase his advantage as the game proceeded, reminded me of a duel
between two equal giants, but one standing on rock and the other in
quicksand.”

Game 48
A.Naiditsch-P.Svidler
Grenke Classic, Karlsruhe/Baden Baden 2019

1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Bxc6 dxc6 5 0-0 Qf6!?

An interesting retort to the Exchange Variation. Black intends to build up
pressure with 6 ... Bg4, which would also facilitate queenside castling, and so
White forces the issue in the centre.
6 d4 exd4 7 Bg5 Qd6 8 Nxd4
 White has made some aggressive-looking moves, but now I’m reminded
of Tarrasch’s lament (quoted from azquotes.com): “As Rousseau couldn’t
compose without his cat beside him, so I cannot play chess without my light-
squared bishop. In its absence the game to me is lifeless and void. The
vitalizing factor is missing, and I can devise no plan of attack.”
Having exchanged off his light-squared bishop, Naiditsch’s set-up lacks
oomph.
8 ... Be7 9 Bxe7 Nxe7 10 Nc3 Bd7 11 Nde2
I don’t like this retreat of the knight. 11 Nb3 is more active, when 11 ...
Qxd1 (after 11 ... 0-0-0 12 Qh5 Ng6 13 Rad1 White has some dynamism) 12
Raxd1 b6 (guarding the c5-square) 13 Rd2 0-0-0 14 Rfd1 looks fairly equal.
11 ... 0-0-0 12 Qc1 Ng6 13 Qe3 Kb8 14 Rad1 Qe7 15 f4
White could avoid loosening his kingside pawns; e.g. 15 Rd2 Ne5 16 b3
Bc8 17 Rfd1 Rxd2 18 Rxd2 though after 18 ... f6 19 f3 h5 20 Rd1 g5 Black
had chances for the initiative in T.Radjabov-Mi.Adams, FIDE Grand Prix,
Baku 2008.
15 ... Bc8 16 Nc1 Rxd1 17 Rxd1 b6 18 Nd3 a5 19 Re1 Ba6
 Question: You’ll learn a lot about positional chess if you take your time
to
decide whether White should play 20 Nf2, 20 e5 or 20 f5.

Answer: At first glance there doesn’t seem to be much to choose between

the two sides. Black has a bishop versus a knight, which in a general sense is
an advantage. On the other hand, White’s pristine kingside structure looks
more impressive than Black’s doubled queenside pawns.
20 Nf2?

Question: It is startling that Peter Svidler thought this was, strategically

speaking, the decisive mistake for White. With this new information,
maybe you would like to pause again and try to understand his reasoning?

Answer: If instead 20 e5 then 20 ... Qe6 leaves White with no good way
to improve his position, whereas Black has a host of possibilities; for
example, ... c6-c5 and ... Bb7 to put the bishop on a great diagonal, or ... Rd8
seizing the d-file, or rerouting the knight with ... Ne7 and then ... Nf5 (and
maybe ... Nd4 if defended by ... c6-c5). The white kingside pawns have
stalled leaving him with no good plan while the black pieces have all the
dynamic chances.
For this reason the eight-time champion of Russia told chess24 com that
White “absolutely had to play 20 f5 here with unclear consequences”. After
20 ... Ne5 21 Nxe5 Qxe5 22 Rd1 White has possession of the d-file. In
addition to 23 Rd7, his ideas include 23 Qd4! to exchange off queens. This
would enable his king to walk to the centre in support of the e4-pawn and
also allow a general advance on the kingside with g2-g4, h2-h4 and g4-g5
etc.
White’s pieces wouldn’t be unduly troubled by the need to defend e4, and
he would have a clear plan. In this sequence Black could decline the
exchange of queens but then he has lost his grip on the e5-square, and g7
might be left en prise.
After the game move, White’s pawn on e4 is defended four times and
attacked once. Nimzowitsch espoused a theory of overprotection, by which
he meant that key centre squares or other posts of strategic importance should
be strongly guarded, the more times the merrier. Pieces defending such
squares are likely to find themselves well placed ‘by accident’ to carry out
active plans or manoeuvres necessary for the health of the position. In a sense
Nimzo’s idea is a refinement (cynics might say an exaggeration) of the
principle of centralization: if you put your pieces on centre squares (or posts
where they influence the centre) they are more likely to find themselves co-
ordinated and ready to take part in whatever plan you formulate.
This is good advice, of a similar kind to if you don’t know what to do,
improve your worst-placed piece. You are unlikely to make any terrible
mistake if you rely on centralization (or overprotection) to guide you. Indeed,
there are many positions in the opening where it is too early to devise a
strategy: you simply have to develop and centralize and wait for the features
of the position to congeal before you can embark on a definite plan of action.
But most of the time centralization and overprotection or improving your
worst piece are feeble rules to live by. It’s far better to have a carefully
worked out strategy, or even just a little plan lasting a couple of moves. (Can
a plan last just one move!?)
I should mention that having all his pieces defending e4 doesn’t mean
that White is employing overprotection in the way that Nimzowitsch
recommended it. Guarding the e4-point isn’t adding any energy to the white
pieces: it isn’t providing a jumping off point for other operations. The
problem is that White’s natural plan is to advance his kingside pawn
majority, but he has been deterred from playing e4-e5. Instead, as discussed
above, he needed to play 20 f5!, accepting a backward pawn on e4 but
exchanging off pieces and not becoming unduly tied down to the e4-pawn. In
the likely event of a queen exchange, his kingside pawns are filled with the
energy to expand.
It’s interesting that the computer assessment hardly changes after 20 Nf2
compared with 20 f5. And yet Svidler, as quoted above, was of the opinion
that 20 f5 was indispensable. Indeed, playing through the rest of the game
you get the impression it is downhill all the way for White. Though it must be
admitted that your opponents are unlikely to manoeuvre as superbly as
Svidler: you are by no means doomed if you play the equivalent of 20 Nf2 in
your games.
20 ... Re8
Black’s plan will proceed one step at a time. First of all he centralizes his
rook and adds pressure on e4.
21 a3
Not liking 21 e5 Qb4, when both b2 and f4 are hanging.
21 ... f6
 A more direct method of restraint. Svidler does his utmost to make sure
the advance e4-e5 is unattractive for White. Deprived of its power to expand,
the e4-pawn becomes in effect a backward pawn on an open file. It is already
attacked from the front by the black queen and rook. But that isn’t enough for
Svidler: he wants to attack it with both of his minor pieces as well.

Question: So what is the procedure to get the bishop attacking e4?

22 Ne2 Bb7 23 Ng3 c5

Answer: The bishop is now lined up against the selected target. Black has
also increased his influence over the d4-square. Meanwhile, White is
threshing around with his knight.
24 Nf5
Completing a rather useless three-move knight manoeuvre.
24 ... Qf7 25 Rd1
 Question: Can you see a fantastic five-move manoeuvre by the black
knight
(it will require a move of the bishop as well and a retreat by the white
knight from f5) to get it into position to attack e4?

Now comes the most elegant part of this game, and perhaps of any
example in this book.
Answer: 25 ... Nf8!!
The start of a very long journey. It isn’t normally advisable to carry out a
luxurious manoeuvre with the knight on the back rank. An alert opponent will
disrupt it by doing something active, when the absence of the knight from the
immediate struggle will be keenly felt. But here White’s position is too
lacking in vitality to allow any riposte.
26 c3 a4
Fixing the pawns on the queenside to prevent 27 b4.
27 h3 Ne6 28 Ng3
 The horse retreats before it is pushed back by 28 ... g6.
28 ... Bc6!
Defending the a4-pawn and the d7-square, but most importantly clearing
the b7-square for the knight.
29 Rd2 Nd8!
Beginning a three-move gallop to d6.
30 Qd3 Nb7 31 Kh2 Nd6!
An impressive manoeuvre from g6. The knight attacks e4, blocks the
white pieces from using the d-file, and is ready in some cases to go to c4 to
attack the b2-pawn.
32 Rd1

Question: How do we continue the process of undermining the e4-pawn?

Answer: 32 ... h5!

So far Black has been concerned with bringing fresh forces into the attack
against the beleaguered pawn. But now Svidler is concerned with undoing the
work of the white pieces (an expression I’ve borrowed from Lasker) by
driving the knight from its post on g3 where it defends the pawn.
33 Re1
After 33 h4 Black can undo the work of the white kingside pawns with 33
... g5!, breaking them up and leaving their king in peril.
33 ... h4 34 Nf1

34 ... Qb3!
Overloading the defence, as e4 and b2 can’t both be adequately defended.
It was also possible to target the e4-pawn directly with 34 ... Qg6!, when
35 f5 loses to 35 ... Nxf5!, winning a pawn as 36 exf5 allows mate on g2. Or
if 35 Nd2 then 35 ... f5! and e4 drops, as mate again follows on g2 if White
moves the pawn from e4.
35 e5
Or 35 Qb1 Nc4, when 36 Re2 or 36 Nd1 both allow 36 ... Nxa3, winning
a pawn to start with.
35 ... fxe5 36 fxe5 Qxb2 37 Re2
Or 37 exd6 Rxe1 38 d7 Bxd7 39 Qxd7 Qxf2 and Black wins.
 37 ... Qb5!

38 Qc2
If 38 exd6 Rxe2 39 c4 Qb2 40 d7, then 40 ... Bxd7 41 Qxd7 Rxf2 is a
steady win, while letting White queen with 40 ... Rxf2 41 d8Q+ Kb7 is the
fastest and most elegant method – disaster follows on g2 for White.
38 ... Nc4 39 e6 Nxa3 0-1
After 40 Qd1 Nb1 41 Re3 a3 Black can think about queening one of his extra
pawns.
Chapter Seven
Promoting a Pawn
A common plan is to mobilize a pawn majority in order to create a passed
pawn. We can expect to meet enormous resistance, as any worthy opponent
knows that to let us queen will almost certainly mean his defeat. Therefore he
will use techniques of the kind discussed in Chapter Three to thwart the
advance of our pawns, or might try to deflect us from our plan by launching
an assault on our king.
We’ll begin by looking at examples of an overtly tactical nature where a
pawn cannot be stopped from queening, so the opponent must seek
immediate tactical salvation. Then we’ll examine games of a more strategic
nature which feature a pawn majority versus a kingside attack. Finally, we’ll
consider the use of a blockade to avert the advance of the pawns.

Game 49
A.Demchenko-D.Gukesh
Ho Chi Minh City 2019
 42 e6!
With this move White was planning not only to create a passed pawn but
also to break open lines against the black king. Already the exchange down
and in a desperate situation, Gukesh has to maximize his chances for a
swindle. King safety? Stopping the opponent queening? Nah, forget it. What
matters is counterplay! And so:
42 ... Rd8! 43 Qxd8 Qxb3 44 e7
Demchenko has to let his opponent queen first. But he is fine with that as
he judges that his own king will be able to evade checks, whereupon the
threat of mate to the black king on g8 or h8 once he himself queens will be
decisive. A correct assessment, but it puts enormous pressure on him to find
the only refutation to Black’s next move.
44 ... Qb7+!
 Question: A tricky check. Can you work out how White should respond?

45 Kh3?
An entirely natural move – which loses the game.
Answer: The way to win was 45 Rf3! b1Q 46 e8Q, when Black has the
next move, which is often decisive with four queens on the board. However,
the mate threat to the black king trumps everything. For example:
a) 46 ... Qxf5 47 Qg8+ Kg6 48 Qb6+ Qxb6 49 axb6 Qc2+ 50 Kh3 wins.
Vitally, the f7-pawn is hanging to the white queen.
b) 46 ... Qc2+ 47 Kg1! (again the only move but it leaves Black with no
good checks) 47 ... Qa7+ 48 Kh1 or 47 ... Be3+ 48 Qxe3.
45 ... b1Q 46 e8Q
 Question: Can you see the killer blow White had missed?

It looks as if Black is going to have to resign in view of unstoppable

mate, but:
Answer: 46 ... Qxf5+!
A horrible surprise for White. If he takes the queen it is mate on h1.
47 Kh2 Qc2+ 0-1
It will be mate on g2.
It feels as if White was somewhat unlucky in that the logical course of his
plan required him to find the ‘only’ move 45 Rf3, without which he was lost.
When the opponent queens first, the stakes on the accuracy of your moves
become very high. Meanwhile, Black had to find the tricky 44 ... Qb7+! and
hope White would overlook the deadly idea behind it. Gukesh was a 12-year-
old grandmaster at the time of this game, and not likely to miss such a tactical
chance!
Game 50
T.L.Petrosian-H.Martirosyan
Aeroflot Open, Moscow 2019

Here Black played:

41 ... Rd1!
The intention is 42 ... Qe1 which, in view of the threat of mate on g1 or
h1, would force the exchange of queens. Then it would be a fairly long but
essentially easy process to force the c-pawn forward (the black king could
come up the board to help). Petrosian therefore stakes everything on an
initiative against the black king.
42 Qg5 c3
Definitely not 42 ... Qe1? 43 Qxg6+ and White’s attack gets in first.
Instead, 42 ... Qf5 looks like a good idea as it attacks the rook and offers the
exchange of queens, but White can fight on with 43 Qe7!, as after 43 ...
Qxc2?? 44 Be5! the threat of mate on g7 dooms Black.
43 Qh6
 The queen inveigles her way further into the black kingside, as if 43 Be5
then 43 ... Rd5 wins a piece.
43 ... Qf5
Forcing the further advance of the c-pawn. Nonetheless, everything had
to be carefully calculated as the white rook will be able to join in White’s
kingside assault.
44 Re2 c2 45 Be5

White had staked everything on his attack and it indeed seems menacing.
There is the threat of mate on g7 as well as a scary check on h8. If Black
plays 45 ... Qxe5+ 46 Rxe5 c1Q then 47 Qxg6+ leads to a quick mate.

Question: How should Black respond?

Answer: 45 ... Rh1+!

Another way to win was 45 ... g3+!, when 46 Kxg3 Qg4+ is mate next
move, while otherwise the white king is denied the g3-square and so can’t
avoid the fatal exchange of queens once Black promotes: 46 fxg3 Rh1+! 47
Kxh1 c1Q+ or 46 Bxg3 c1Q (no need to queen with check as there is no
longer mate looming on g7).
46 Kg3
Evading the queen exchange that follows 46 Kxh1 c1Q+. But now it
looks like White is winning after all. In fact there is only one way for Black
to stop the threats and that also wins for him:
46 ... Rh3+! 0-1
After 47 gxh3 Qf3+ 48 Kh2 Qxh3+ 49 Kg1 queening the pawn will force
mate in two.
“A nice combination, but Tigran Petrosian must have been very old when
this game was played,” said one person when I showed him this game. In fact
the player of White was Tigran Levonovich Petrosian (born 1984), a very
strong modern grandmaster. He was born about a month after the death of the
ninth world champion, Tigran Vartanovich Petrosian (1929-1984)

Game 51
R.Rapport-A.Giri
Wijk aan Zee 2019
 Rapport has completed his build-up on the queenside. All his pieces are
poised for action against the queenside pawns. The problem is that there is no
good way to continue the attack. Well, it’s not a problem if you are a
computer or a staid old grandmaster. You just settle for a way to repeat
moves: for example, 21 a3 Be7 22 Rd2 (but not 22 Nc3? Bxd3) 22 ... Kh7 23
Nc3 Ba6 24 Na4 Bb5 etc with a draw by repetition. However, Richard
Rapport is a young and ambitious player with a creative style. If there is a
brick wall on the queenside he will go straight through it:
21 Nxb6? Qxb6 22 a4
This is the idea. White will regain his piece, after which his rooks will
bust through on the c-file. However, Giri finds a simple positional refutation
of the plan:
22 ... Ra8! 23 g4
Rapport sees what’s coming on the queenside and so prepares a counter
demonstration on the kingside.
23 ... Qd8!
 Question: What should Black play after White recaptures on b5?

Answer: 24 axb5 a4
This is the point: the white queen can’t stay guarding b5, so Black gets a
powerful queenside pawn majority.
25 Qc3 cxb5 26 Qc6
White’s plan to dominate the c-file turns out to be an idle fantasy as there
is nothing much to attack along it.
26 ... b4 27 Bc7 Qg5 28 f4 Qg6 29 e4
Defending the d3-pawn. Rapport presses on with his bid for kingside
activity since he can’t just wait for Black’s queenside pawns to roll forwards.
29 ... h5!
The only real drawback to Black’s set-up is that his queen has become
separated from her rooks and has no influence on the queenside. Giri hopes to
use her actively on the kingside – or at least facilitate her return to the other
black pieces – by punching a hole in White’s structure.
 30 g5 h4 31 Kf2 a3 32 b3 Rec8 33 Qb7 dxe4 34 dxe4 f6!
Now White’s position is tottering. The need to watch over the a3-pawn
interferes with the co-ordination of his pieces; the g5-pawn is hanging; and
the black queen is poised to slip back to f7, rejoining the rest of her army and
pinning the white bishop. Rapport tries one last throw of the dice but it is
easily defeated.
35 Be5 Rxc2+ 36 Rxc2

Question: How would you swindle Black after 36 ... Qe8 37 gxf6 a2 - ?
And what happens if 37 ... gxf6 instead?

36 ... Re8!
Answer: Rapport had set up a swindle upon 36 ... Qe8? 37 gxf6, when:
a) 37 ... a2? 38 f7+! Qxf7 39 Qxa8 wins for White.
b) 37 ... gxf6! 38 Bxf6 a2 39 Rc7! and now 39 ... a1Q 40 Rg7+ mates,
though Black can hold a draw with 39 ... Qg6! 40 Qxa8 Qg3+ as he can give
perpetual check.
 37 gxf6 gxf6 38 Kf1
Hoping for 39 Rg2, winning the queen, but Giri cleverly makes sure it is
Black, not White, who seizes the g-file:
38 ... Re7! 39 Qc6 Rg7 40 Qxe6+ Kh7 0-1
Black is threatening 41 ... Qg1+ 42 Ke2 Rg2+ 43 Kd3 Qf1+ 44 Ke3 Rg3+
45 Kd2 Rd3 mate, to say nothing of 41 ... fxe5 or 41 ... Qxe4.

The two basic plans in a game of chess – trying to mate the enemy king and
attempting to queen a passed pawn – often come into conflict. While one
player is throwing his pieces forwards trying to crash through a king’s
defences, the other player is trying to hold on long enough to queen a pawn
and then win due to his superior material forces.

Game 52
S.Mamedyarov-Ding Liren
FIDE Candidates, Berlin 2018
 27 ... a5!
Ding Liren makes his intentions clear: advance the queenside pawn
majority, break through the white defences, and queen a pawn. Of course this
isn’t as easy as it sounds, especially when you are playing an opponent at the
peak of his game and rated 2809.
28 g4
It’s worth mentioning that White had another possibility: 28 Nxe6!? fxe6
29 Bxe6+ Kf8 30 d5 Qxe3+ 31 Kxe3 with a double-edged endgame.

White has only two pawns for the piece, but they are passed pawns, and
his king and rook quickly become very active; e.g. 31 ... Bd7 32 Bxd7 Nxd7
33 Rc7 Ke8 34 Kd4. However, Black can hold the balance by giving a pawn
back to activate his own rook: 34 ... b4! 35 Ra7 Rc8 36 Rxa5 Rc2 with equal
chances.
Alternatively, Black could play 31 ... Bxd5, when 32 exd5 Nxd5+ 33
Bxd5 Rxd5 has burnt out to a draw, but 32 Rd1!? Bxe6 33 Rxd8+ Ne8 34
Ra8 a4 35 a3 is another hard to judge endgame. White can push his centre
pawns but if he loses control (for example, if the knight gets to c4) he could
suddenly be in trouble. The best assessment is probably that Carlsen in top
form would win it with either colour.
Let’s see what happened in the game:
28 ... a4 29 Bc2
Perhaps the best line of play is 29 Bd1! Nd7 30 Nd3 (hoping to get the
chance for 31 d5!, when Black is in trouble as the exchange of queens will
leave his bishop hanging) 30 ... Ba8 (avoiding the trap) 31 Rb1 Rc8 (or 31 ...
Qc7 32 Kg2, guarding the h2-square, when White is safe) 32 Bxa4!? Qc7 33
Bd1 Qh2+ 34 Ke1 Qxa2 35 Rxb5 Qh2 36 Qf2 Qh1+ (36 ... Qxh3?? 37 Rh5!
unexpectedly wins the black queen) 37 Qf1 Qh2 38 Qf2 with a draw by
repetition.
29 ... Nd7!

30 Bd3?
It’s a serious mistake to allow the exchange of knights as, hereafter,
White has one fewer piece to oppose the advancing queenside pawns or
complicate matters on the kingside. Therefore 30 Nd3 was called for. Then
the impetuous 30 ... b4 meets with a trap we have already seen in the
previous note: 31 d5!, attacking c6 and offering the exchange of queens.
Black would lose his precious b-pawn. Instead, 30 ... Ba8 keeps a marginal
edge.
30 ... Nxc5 31 Rxc5 b4 32 Bc4 Bd7
As we shall see, White’s problem isn’t a black pawn storming to b1, since
he has the b3-square covered three times and his bishop can always oppose
the pawn’s queening. What troubles him is a pawn sac with ... b4-b3, clearing
the way (after the recapture a2xb3) for the advance ... a4-a3. His defensive
resources on the a-file are much worse than on the b-file.
33 g5
White was feeling either optimistic or desperate at this point. He goes
active as it would be a kind of slow death to wait for Black to arrange a
breakthrough on the queenside.
33 ... hxg5 34 Qxg5 Be8
He could already play 34 ... b3!.
35 Qe7 b3! 36 axb3 a3!

The pawn slips through the white defence.

 37 b4 Ra8 38 d5!
At last White utilizes his pawn majority in the centre. Mamedyarov’s aim
isn’t to queen a pawn but rather to mate the black king. Thus we see a clash
between the two main plans mentioned above: promoting a pawn (and mating
later) versus mating by direct attack. When two great players use their
ingenuity to try to impose their contrasting ideas on the struggle, it becomes
an even more enthralling spectacle.
In any case White has to do or die, as 38 Ba2 Qxb4 is hopeless: both 39 ...
Qxd4+ and 39 ... Qb2+ are threatened.
38 ... a2 39 dxe6
Mamedyarov has to stake everything on this central breakthrough. His d-
pawn bludgeons its way to f7, but is it enough?
39 ... a1Q
Black has achieved his aim.

Question: But can White can complete his side of the deal by mating the
black king, or at least forcing an honourable draw by perpetual check?
 Answer: Without wishing to spoil the denouement for you, I think the
distribution of exclamation marks and question marks over the preceding
moves should give you a big clue to the answer! But have a go at working out
the tactics.
40 exf7+ Bxf7 41 Bxf7+ Kh7 42 Qh4+ Qh6 43 Rh5!
It looks as if White might draw the game by nabbing one of the queens,
but:
43 ... Qa7+! 0-1
Mamedyarov was hoping for 43 ... Qd4+ 44 Kg2 Ra1 (note that the white
bishop guards a2) which looks great as mate on g1 is threatened, but 45 Qf2!
holds on (whereas inserting 45 Rxh6+? gxh6 loses as 46 Qf2 is met by 46 ...
Qg7+, winning the bishop).
Rather surprisingly Ding Liren’s move in the game is the only one that
wins. White resigned as 44 Kg2 Qxf7 45 Rxh6+ gxh6 leaves him a rook
down.

In our next example the attack on the king and the advance of the passed
pawns happen on the same wing. You can bet that leads to one tough fight!

Game 53
H.Melkumyan-S.Shankland
Batumi Olympiad 2018

The position below arose from the super-sharp Botvinnik System in the
Semi-Slav. When they devise a plan players are trying to tell their own story
about a position. Their gallant pieces armed with an immaculate strategy are
going to defeat the ill-led enemy rabble. For instance, in the diagram: are the
black pawns on c3 and b4 heroes or villains? It depends on whether you are
playing White or Black.
 Material is even. Black has a protected passed pawn only two squares
from queening. The white pawn on b3 would be very weak in an endgame: its
loss would mean a complete collapse on the queenside, allowing Black to
acquire connected passed pawns. White has his own protected passed pawn
on the kingside, but the idle fellows on h4 and g3 are no match for the
dynamic duo on c3 and b4.
So there we have it: the pawns on c3 and b4 are heroes.
Or at least that is what Black would claim. White might disagree: we are
still in the middlegame; who cares about the endgame if I can mate the black
king? The two pawns on g3 and h4, rather than being lazy, are noble fellows
who have sacrificed their chance of glory by staying back to shield their king.
Meanwhile, the reckless pawns on c3 and b4 have abandoned the defence of
their own king in the pursuit of fame. What scoundrels!
Who is right? Let’s see how the game unfolded:
27 Qe2!?
It looks as if White is right: the advance of the pawns to b4 and c3 has left
the f1-a6 diagonal open, allowing the white queen to slip into b5 or threaten
what looks a terrible check on a6.

Question: Can you see a strong entrance by the black queen?

Answer: 27 ... Qd4!

The power of centralization can turn defeat into victory. The black pieces
have such a solid grip on the d-file that they can escort their king away from
the danger on the queenside.
28 Qb5
A check on a6 meets with the same reply.
28 ... Kc7! 29 Na6+ Kd6
The white pieces are decentralized or disorganized. Meanwhile, the black
pieces are in a single unit guarding their king.
30 Bxd5 exd5?!
Stronger was 30 ... Qxd5. For example, 31 Qe2 Ke7, when there is no
more attack and the white knight is hopelessly entangled on a6.
31 a5?!
If immediately 31 Qb7 (with the dire threat of 32 Qc7+) then 31 ... Qb6
forces a fatal exchange of queens. But Shankland is able to guard the c7-
square and so defeat the attack with his next two moves. So now was the time
for the h-pawn to create a diversion with 31 h5, intending 32 h6 with
counterplay.
31 ... Re8! 32 Qb7 Re7 33 Qc8 Qd2
 The passed c-pawn has the last laugh after 34 Rf1 c2, so White makes a
desperate bid for perpetual check.
34 Nb8 Qxc1+ 35 Kg2 Nxb8 36 Qxb8+ Kc5 37 Qc8+ Kd4 0-1
The king runs forward to d3 etc.
We might say that White lost because he failed to appreciate that the
black pawns on b4 and c3 were not only heroes for the endgame, but were
also heroes in protecting their king. It is true they left the f1-a6 diagonal
gaping open. But, more importantly, they formed an invincible barrier to the
white rook on the c-file, so that White had no way to follow up his attack
with the queen and knight.
Just before the initial diagram position arose, White had in fact made a
serious positional mistake: 1 d4 d5 2 Nf3 Nf6 3 c4 e6 4 Nc3 c6 5 Bg5 dxc4 6
e4 b5 7 e5 h6 8 Bh4 g5 9 Nxg5 hxg5 10 Bxg5 Be7 11 exf6 Bxf6 12 Bxf6
Qxf6 13 g3 Bb7 14 Bg2 Na6 15 Ne4 Qe7 16 0-0 0-0-0 17 h4 c5 18 a4 Rxd4
19 Qe2 b4 20 Rac1 Bd5 21 Rfd1 Rxd1+ 22 Rxd1 Rd8 23 Rc1 Nb8 24 Qe3
Nc6 25 Nxc5 Qf6 26 b3? c3. He had misjudged the defensive power of
Black’s position with the c-file blocked and the d-file inaccessible to the
white pieces. The attack on the f1-a6 diagonal proved a will-o’-the-wisp.
Instead, either 26 Bf1 or 26 a5 would have led to obscure complications.

Game 54
C.Sandipan-V.Ivanchuk
Gibraltar 2018

1 d4 Nf6 2 c4 e6 3 Nc3 Bb4 4 e3 c5 5 Nge2 cxd4 6 exd4 d5 7 c5 Ne4 8 Bd2

Nxd2 9 Qxd2

Remember a saying of Nimzowitch quoted earlier: “In the last resort

positional play is nothing other than a fight between mobility (of the pawn
mass) on the one side and efforts to restrain it on the other.” In the present
game we see this taken to extremes.
White has a 3-2 pawn majority on the queenside which is free to advance.
Ivanchuk endeavours to restrain it with his a-pawn:
9 ... a5 10 a3 Bxc3 11 Nxc3 a4
It seems that Black has achieved his aim. The plucky a-pawn prevents 12
b4, and if 12 Bb5+ then 12 ... Bd7 defends it safely. Sandipan decides upon a
drastic course to escape from the restraint imposed by the pawn on a4.
12 Nxa4!?
This position has been reached over 100 times in international play, and
Sandipan is the first to offer this sacrifice. (In truth, if White wants to play on
the queenside, he has little choice but to sacrifice once he has allowed 11 ...
a4. After anything else he would remain paralysed there.) Objectively
speaking, it is probably dubious, but it sets Black some serious problems.
And it has psychological value as Ivanchuk is taken by surprise.
Instead, a typical line is 12 Bd3 b6 (undermining the pawn on c5) 13 cxb6
Qxb6 14 Bc2 Bd7 15 0-0 0-0, when White is left with a backward pawn on
b2, but Black’s pawn on a4 needs constant defence. It looks about equal.
12 ... Rxa4 13 Bb5+ Bd7 14 Bxa4 Bxa4

Question: Try to evaluate the position. What features favour White and
Black?
What is the best move?
 Answer: In a middlegame position the exchange of two minor pieces for
an enemy rook and pawn is normally a bad idea. Indeed, it wouldn’t feel
strange to describe it in terms of sacrificing two pieces for a rook and pawn,
despite the fact that in numerical terms it is an equal exchange.
Two minor pieces usually do more work than a rook and a pawn in a
middlegame. They can combine their power with the other pieces to guard
centre squares or launch an attack on the enemy king. A rook is often no
more useful as a defender of its king than a bishop or knight, so the defence
would be overstrained.
A rook can’t usually show its full strength until the endgame, when more
of the pawns have been exchanged off, granting it open lines. And even then,
for advantage the rook usually needs to be doing something special, such as
sitting on the seventh rank or supporting a far-advanced passed pawn.
Normally, the exchange of queens is a big help to the player with the rook
and pawn – there is no more danger of a mating attack and the long-range
rook can try to overstretch the two minor pieces.
But Sandipan reasons something as follows: the black minor pieces are a
long way from creating threats to White’s king (we assume he has castled
kingside). The knight on b8 is still asleep, and the bishop on a4 can’t aim
itself against g2 any time soon. Meanwhile, the white queenside pawns are
released from the bind and will be able to advance rapidly.
White has another advantage: the next move. If he plays slowly with 15
0-0 then Black can organize his pieces with ... Nc6. Recalling the advice of
Pillsbury that “you should castle because you have to or want to, not just
because you can”, he played:
15 Qb4!
The queen attacks both a4 and b7. If Black responds 15 ... Bc6 then his
knight is deprived of its natural square. Furthermore, on c6 the bishop is a
target for a future b4-b5 advance.
15 ... Qd7
The natural reply. We could describe 15 Qb4 as a preventive or
prophylactic move, since by side-tracking the black queen it stops her
attacking on the kingside with ... Qg5. Instead, a big positional battle is about
to begin on the queenside.
16 0-0
 Have a look through your games and see if you tend to castle too early.
Did you lose a vital tempo which affected the struggle in the centre or
elsewhere? Were you a tempo short at the critical moment because you had
castled, when your king could have been still sitting happily on e1? Or did
you by castling kingside let your opponent know where your king lived too
early, so he could commit himself to an attack on it? Did you miss the chance
to castle queenside and start an attack on the opponent’s king because you
went kingside too early? Did you castle and then the queens quickly came
off, meaning your king would have been better left in the centre for the
endgame? (N.B. If you don’t castle and get mated in the centre in 12 moves,
then all I can say is sorry.)
In this position Black should in fact now castle with a double-edged game
in prospect. White can edge his pawns forward on the queenside beginning
with 17 b3. He would play a pawn to a4 and only then advance his pawn to
b4 in order to prevent the enemy bishop using the light squares a4 and b5.
After that, he would be ready for the advance b4-b5. However, this takes a lot
of time and Black could aim for a set-up such as: bishop on a6, knight on c6,
and then play ... e6-e5! to create a passed pawn on the d-file and enhance the
power of the knight. If he could widen the struggle in this manner, his two
minor pieces might eventually come out on top.
Ivanchuk chooses a different plan: immediate action against the
queenside pawns.
16 ... Qb5?
He couldn’t play 16 ... Nc6 at once as a4 would hang. But if now 17
Qxb5+ Bxb5, Black is ready to play 18 ... Nc6, when one of the white rooks
will be tied down to the defence of d4. It seems that White’s initiative is
running out of steam, but Sandipan had prepared another sacrifice.
17 b3!
Giving Black no choice but to allow the opening of the a-file.
17 ... Qxb4 18 axb4 Bb5
If 18 ... Bxb3 then 19 Ra7!, followed by 20 Rxb7, creates monstrous
passed pawns.

But after the game move it seems like Black is going to triumph
according to Nimzowitsch’s motto for dealing with an enemy passed pawn
(which equally applies to dealing with an advancing enemy pawn complex):
“first restrain, second blockade, third destroy!” White’s pawns are blocked,
and if he saves his rook with 19 Rfc1 then 19 ... Kd7 prepares 20 ... Nc6 to
eat b4 or d4. After 20 Ra8 Rc8 or 20 Ra7 Kc7 Black is again ready for 21 ...
Nc6. Nor does 19 Ra8 help White after 19 ... Kd7, followed by ... Rc8 and ...
Nc6. Black’s king and minor pieces have reduced the pawns to inertia and his
horse then gets to feed on them.
It seems like another great positional victory for Ivanchuk. But it turns
out to be a mirage after Sandipan’s next move.
19 Ra7!
Black’s king could meet 19 Ra8 with 19 ... Kd7, but is one move too far
away to meet the immediate attack on b7. If 19 ... Ba6 or 19 ... Bc6 then 20
b5! follows and b7 or the bishop will drop. Therefore Black takes the rook.
19 ... Bxf1 20 Rxb7
A vital zwischenzug to establish connected passed pawns. If Black saves
his bishop he loses both the knight and the rook.
20 ... Nc6 21 Kxf1
 21 ... 0-0
“Castle because you have to or want to ... ” According to my computer
Black could have held on with 21 ... Rf8, keeping the king close enough to
fight the advance of the queenside pawns. For example: 22 b5 Nxd4 23 b6
Kd8 24 Ra7 (a key line is 24 b4 Kc8 25 Ra7 Kb8, holding up the pawns) 24
... Nc6 25 Rc7 Ne5 26 b4 Nd7 27 Ke2 e5 and White can cash the pawns in
for the knight whenever he wants, but he can’t win.
Ivanchuk is a fabulous player with a quirky style, but it is beyond even
him to find such a sequence of moves.
22 b5
The blockade has failed. The passed pawns, abetted by the rook on the
seventh rank, prove too powerful for the black rook and knight.
22 ... Na5
Or 22 ... Nxd4 23 c6 etc.
23 Rc7 1-0

Ivanchuk resigned as he could see no good way to stop the pawn

advances b5-b6-b7 and c5-c6 followed by Rc8 to force a new queen.
He could have fought on a bit with 23 ... Nxb3, but the pawns get through
eventually: 24 b6 Na5 25 c6 g5 26 b7 Kg7 27 Ke2 (27 Rc8 allows 27 ...
Nxc6, though 28 Rxc6 Rb8 29 Rc7 should still win) 27 ... h5 28 Kd3 (the
king approaches to harass the knight) 28 ... h4 29 Kc3 Rb8 30 Rxf7+! Kxf7
31 c7 Rxb7 32 c8Q and the queen soon wraps up the endgame since Black’s
pieces are disorganized.

Once Sandipan had committed himself with 12 Nxa4, his subsequent

sacrifices were all necessary. If he had hesitated for even one move, his
queenside pawns would have been blockaded. Deprived of their ability to
expand, they would have lost their dynamism and been picked off by the
black knight.
You have to learn to make sacrifices, even if you are by nature a careful,
cautious player. They aren’t a luxury but a vital part of your chess armoury.
Chapter Eight
Using a Pawn as a Battering Ram
It is time to consider an exciting but challenging aspect of chess: the direct
attack on the king. A common way to empower this plan is to prepare and
carry out a disruptive pawn advance. If successful, it will clear lines for our
pieces, drive back the defenders, and shatter the protective wall around the
enemy king. The stakes are high: a resourceful opponent will force us to co-
ordinate our pieces perfectly and play with flair and imagination if we want
to break through and win.

Game 55
J.K.Duda-R.Rapport
Wijk aan Zee 2019
 A tense situation in which two of the most promising young players in the
world are battling it out. Enticed by the dark square holes around the black
king, Jan-Krzysztof Duda ‘went for it’ with:
31 Nb3!? Bxb3 32 Qd8?
As a general rule, if you think you see a good move, you should play it.
It’s better to lose the occasional game than not to trust what your chess
instinct or judgment or calculation recommends. On the other hand, if
experience has taught you that you tend to blunder in time pressure, then it
might be a wise decision not to provoke a tactical crisis when short of time.
Or if a certain opponent has the habit of outplaying you in double-edged
battles, then you might decline to play the ‘best’ but messy move in favour of
a solid, but not so objectively advantageous, alternative.
As you get better at chess you’ll learn what tends to work and what
doesn’t. If, for example, you calculate a line which seems good for you, but
your intuition is telling you “no, no, no!” should you trust your analysis or
your gut feeling? (Your intuition isn’t always right – it can make you afraid
of ghosts. But an uneasy feeling of “there’s something not quite right here”
can equally indicate your calculation has been too optimistic and you’ve
missed something good for the opponent.)
With hindsight it was better for White to call it a draw with 32 Qb2
(threatening mate as in the game and, vitally, keeping the black rook out of
c1), when after 32 ... Rc8 33 Rxc8 Qxc8 34 Qxb3 Qc1+ 35 Kf2 Qd2+ 36 Kf1
Kg7 37 Qa3 it’s equal.

Question: Well, can you work out Richard Rapport’s counter-

combination?
He has to do or die with a counterattack on the white king, as otherwise
there is no good way to stop mate on h8.

Answer: 32 ... Rc1+ 33 Kf2 Qf1+ 34 Kg3

Duda had probably envisioned a finish such as 34 ... Qe1+ 35 Kh3 Qc3
(to stop mate) 36 Qg8+ Kh6 37 Rxb3 Qxb3?? 38 Qh8+ Kg5 39 f4 mate,
though even here 37 ... Qg7 holds the draw. What he hadn’t reckoned with
was Black’s powerful pawn stab:
 34 ... f4+!!
When carrying out your plans, tactical or strategic, you always have to be
on the lookout for powerful pawn moves. By opening the c8-h3 diagonal, the
pawn strike transforms the light-squared bishop into a strong attacking piece
and takes away the white king’s refuge on h3.
35 Kxf4
Instead, 35 exf4 Qe1+ 36 Kh3 Be6+ is similar to the game, while after 35
Kh3 Be6+ 36 Kh4 Qf2+ 37 Kg5 Rc5+ the black attack gets in first.
35 ... Rc4+ 36 Kg3 Qe1+ 37 Kh3
Once again the white king appears safe, but now comes another surprise
on the c8-h3 diagonal.
37 ... Rc8!!

Better than a draw with 37 ... Rh4+ 38 Qxh4 Be6+ 39 g4 Qf1+ 40 Kg3
Qe1+ 41 Kh3 Qf1+ etc.
38 Rxc8
Of course 38 Qxc8 Be6+ wins the white queen.
38 ... Be6+ 39 g4 hxg4+ 40 fxg4
 After 40 Kg2 gxf3+ 41 Kxf3 Qf1+ 42 Ke4 (or 42 Kg3 Qh3+) 42 ... Qf5+,
White loses in prosaic style as the rook drops.
40 ... Qxe3+ 41 Kh4
Or 41 Kg2 Qe2+, when going to h1 or h3 with the king allows mate,
while otherwise 42 ... Qxg4+ picks up the rook.
41 ... Qf2+ 0-1

For 42 Kh3 Qf3+ mates on g4, while 42 Kg5 Bxc8 leaves White unable
to recapture on c8 without allowing another mate in one.

One thing I’ve noticed is that most inexperienced players hate making
positional sacrifices. If they give up a pawn they want a mate in return or,
failing that, at least to be shown how they will get the pawn back. To be told
“you have a strong knight and long-term pressure” isn’t enough. I can
sympathize, as pressure is something intangible: what if it suddenly vanishes
and you are still a pawn down?
I hope the following game will satisfy everyone. There is a splendid
double pawn sacrifice which transitions smoothly into a mating attack.
 A basic aim of dynamic chess is to add power to your pieces while
diminishing, or at least not improving, the energy level of the opponent’s
pieces. In pursuing this strategy a sacrifice can be highly effective if it opens
lines for your own forces while keeping the enemy’s locked in.
In the game below, the opening of lines completely wrong-foots the white
pieces. It becomes one of the most severe punishments you will ever see for
decentralizing a piece.

Game 56
S.Karjakin-A.Esipenko
World Rapid Championship, Riyadh 2017

Karjakin’s knight is off limits on h5, but he probably thought it didn’t

matter too much as the centre is blocked. If so, he got a rude awakening after:
15 ... c4!
A pawn sacrifice which energizes the black pieces. Black’s bishop on f8
didn’t look any better than its counterpart on f1, but now it will gain an
attacking role, while the white bishop remains passive. At the same time the
c5-square becomes available to the black knight on d7, and the way is cleared
for a second sacrifice to free d4 for the other knight.
Notice that White has the bishop pair, but it is impeded by his own
pawns.
16 dxc4
White eliminates the pawn before it can do damage with ... c4-c3.
16 ... Ba3
Not only bringing the black bishop to life, just as importantly forcing
White’s bishop into a defensive role: it can’t guard b2 and also fight for the
d4- or c5-squares.
17 Bc1
Or 17 b3 Nc5 18 Qe1 d3! with a decisive attack for Black.
17 ... Nc5
The black knight jumps to a menacing square and gains time by hitting
the white queen.
18 Qf3 d3!!
 The aforementioned second sacrifice to clear d4 for the other black
knight. It also cuts off the white queen’s attack on a3 and her general
defensive power along the third rank. Thus 19 Bxd3 (the move White would
like to play to get developed) 19 ... Bxb2! 20 Bxb2 Na4 leaves White with no
way to defend b2.
19 cxd3 Na4
Not giving White the option to defend b2 immediately with his queen
after 19 ... Nd4 20 Qf2.
20 Rd2 Nd4 21 Qf2

Question: Take the chance to contemplate Black’s sacrifice. How would

you explain his compensation for the pawns, and what should be the
next stage in his plan?

Answer: Why has Black’s takeover of the centre gone so swellingly? The
positional basis is the terrible white knight which is doing nothing to contest
the key squares c5 and d4. White’s bishop on f1 is unable to contribute
anything to the fight against the black minor pieces. Therefore Black has
been able to establish a blockade on the central dark squares.
I hope you can see how Black’s pieces have been empowered by the
double pawn sacrifice and, equally, the white pieces diminished – the bishop
on c1 deprived of the ability to fight for d4 and the white queen blocked from
using her power along the third rank.
The task for Black now is to drive home a direct attack on the white king
before the knight on h5 is able to return to the fray. We’ll soon see how this
is done.
21 ... Nc3+ 22 Ka1 Qb3!
Threatening mate in one, while 23 axb3 Nxb3 mate would be a fitting
tribute to the black knight. All your ingenuity will fail to discover winning
combinations in equal positions. But when you have central domination,
fantastic (and sound) moves appear of their own accord.
23 bxc3 Qxc3+ 24 Bb2
24 Kb1 Qxc1 is mate, while 24 Rb2 Qxc1+ is the same as the game but
with Black’s bishop still on the board.
24 ... Bxb2+ 25 Rxb2 Qc1+ 26 Rb1 Nc2+ 27 Qxc2 Qxc2 28 g3 b5!
 Energetic to the end. One of the signs of a very strong player is that they
don’t ease up once they have a winning position. Esipenko batters his
opponent with the most precise move, not ‘coasting’ despite having a large
material advantage. If now 29 Rxb5 Qc1+ 30 Rb1 Qc3+ 31 Rb2 then 31 ...
Kd7! clears the way for 32 ... Rb8 and White will soon be pulverized down
the open file.
29 cxb5 Rd4 0-1
There is no good way to meet the threat of 30 ... Qc3+ 31 Rb2 Rb4 32
Rhh2 Qc1+ and mate next move.

Game 57
M.Matlakov-L.Aronian
FIDE World Cup, Tbilisi 2017

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 Nxd5

A massive decision. Black could recapture 4 ... exd5, giving him an equal
share of the centre, but Aronian wants an imbalanced structure with all its
risks and opportunities.
5 e4 Nxc3 6 bxc3
White has established a mobile centre, which Black immediately seeks to
undermine.
6 ... c5 7 Rb1
A sign of aggressive intentions. Matlakov prevents Black from
exchanging another piece, as would be the case with 7 Nf3 cxd4 8 cxd4 Bb4+
9 Bd2 Bxd2+ 10 Qxd2, though after 10 ... 0-0 11 Bc4 White’s centre would
still give him slightly the better chances.
7 ... Be7 8 Nf3 0-0

The position is very similar to the Grünfeld Defence but with Black’s
bishop on e7 rather than g7. For example, compare the line 1 d4 Nf6 2 c4 g6
3 Nc3 d5 4 cxd5 Nxd5 5 e4 Nxc3 6 bxc3 Bg7 7 Nf3 c5 8 Rb1 0-0. In the
Queen’s Gambit version having a pawn on e6 and no structural loosening
with ... g7-g6 would suggest that Black is more secure against an attack. On
the other hand, he doesn’t have Grünfeld counterplay based on the bishop on
g7 attacking d4, nor the option of ... Bg4 to pin the f3-knight (the pawn on e6
is in the way). We might conclude that with 4 ... Nxd5 Aronian mixed things
up, but he still hasn’t achieved as much activity as he would have with a
bishop on g7.
9 Bc4 Qc7 10 Qe2
Guarding the bishop against discovered attack by 10 ... cxd4.
10 ... a6 11 a4
Another preventive measure, this time against the positional threat of 11
... b5, gaining space on the queenside.
11 ... cxd4 12 cxd4 Bd7
After the natural 13 Bb3 to defend his a-pawn, Black can activate his
queenside pawns with 13 ... b5!. This is awkward for White, who definitely
doesn’t want to be deprived of the right to castle after 14 axb5? Bxb5, but
otherwise Black will get a passed pawn with 14 ... bxa4. So Matlakov ignored
the threat with:
13 0-0 Rc8
After 13 ... Bxa4 White gets a similar initiative to that in the game. By
attacking the bishop on c4 Aronian gives his opponent another chance to
‘chicken out’ with 14 Bb3, when 14 ... b5 once again activates the black
queenside pawns. But Matlakov insists on sacrificing the pawn.
14 Bd3 Bxa4
If Black refuses the offer with 14 ... Nc6, he has no compensation for the
fact that White has a centre with a lot of expansive potential, while Black’s
queenside pawns remain passive. In effect, he would have lost the battle of
pawn structures. So Aronian grabs the pawn to cheer himself up. If White’s
initiative dies out then Black’s queenside pawns (now connected passed
pawns!) will have the last word.
15 d5!
 White has to stake everything on the burst of energy given to his pieces
by his advancing centre. Kasparov won a lot of games in this style when he
was a young man. He would break open the centre and then slice through the
enemy king’s defences before his opponent had time to put up effective
barriers or mobilize all his pieces. Indeed, in an early book I argued with
some seriousness that the secret of Kasparov’s success was getting his rooks
into key positions while those of the opponent were still sleeping.
Of course no style of play is without drawbacks. If the opponent succeeds
in weathering the storm, the attacker will have spent all his active pawn
moves. There will be no chance to revert to a slow manoeuvring game with
the intention of grinding down the opponent in the endgame. Even worse, the
attacker could be left material down.
Korchnoi described the young Kasparov’s play as “one big punch”. This
wasn’t meant as a compliment: Korchnoi was hoping before their 1983
Candidates semi-final match that he could avoid being overwhelmed by
Kasparov’s opening prep and then outplay him in the later stages of the
game. As it turned out Korchnoi started off well in the match, before losing a
key game to a sucker punch ... in a rook and pawn endgame.
Nonetheless, as we see in the present game, this style of play can be very
effective, especially if backed up with a knowledge of theory.
15 ... Nd7
Not wanting to open a diagonal for the bishop on d3 and the e-file for the
queen after 15 ... exd5 16 exd5. Instead, 15 ... e5 at least temporarily blocks
lines in the centre, but it leaves White with a permanently strong pawn on d5.
Therefore he no longer has to rush to do-or-die: he can build up his game
with 16 Be3 Nd7 17 Rfc1.
16 e5!
Another pawn stab in the centre. It threatens to win a bishop by both 17
d6 and 17 Qe4!, attacking a4 and menacing mate in two on the kingside. So
Aronian takes a second pawn and safeguards both bishops.
16 ... exd5 17 e6
The third pawn spears the soft underbelly of the black kingside – the f7-
pawn – which has become exposed by the two earlier pawn offers.
17 ... Nf8 18 exf7+ Kxf7
The king decides to brave the open air rather than hide in the corner, as
after 18 ... Kh8 19 Ne5 etc the pawn on f7 would be an eternal torment for
the defence, even if it came to an endgame.
19 Nd4?
I like the nonchalance of this move. It calmly brings the knight in contact
with the e6- and f5-squares and clears the way for the queen to move to f3, g4
or h5. It works a treat by provoking Black’s next move.
Nevertheless, the ‘boring’ 19 Re1!, putting the rook on the open file,
looks more precise as it ties the black king down to the defence of e7 and so
stops it retreating to g8, while 19 ... Bf6 is met by the same move as in the
game.
19 ... Bf6?
It is very natural to attack the knight, but he had to try 19 ... Kg8!. White
keeps up his initiative after 20 Nf5 Bf6 21 Qg4, say, but there is no killer
blow.
 Question: Can you see a clever move to stop the black king escaping
from
the centre? Maybe work out a couple of variations?

Answer: 20 Bxh7!
The knight proves overworked on f8 as it has to defend both the e6-
square and h7-pawn. The loss of h7 not only severely weakens the black
kingside, it also prevents the king slipping back to g8.
20 ... Qe5
Aronian covers the e-file, stops a check on h5, and offers the exchange of
queens – something anathema to anyone trying to mate the enemy king.
White now has a knight and a bishop hanging, so his attack needs to break
through or he’ll be in trouble. It will turn out that the rook’s entry on b7 will
add the necessary oomph to his attack.
We should also consider what would have happened if Black had taken
the bait on either h7 or d4.
 After 20 ... Nxh7 21 Qe6+ Kf8 22 Ba3+ Be7 23 Rbc1! Black’s queen
can’t defend both c8 and e7 in key lines. For example: 23 ... Qd7 24 Bxe7+
Qxe7 25 Rxc8+ wins, and 23 ... Bc6 24 Nxc6 bxc6 (or 24 ... Bxa3 25 Ne5
Bxc1 26 Ng6 mate) 25 Rxc6! Qxc6?! 26 Bxe7+ Ke8 27 Bd6+ Kd8 28 Qe7 is
mate.
Alternatively, 20 ... Bxd4 leaves the black king hopelessly stuck in the
centre: 21 Qf3+ Ke7 (or 21 ... Bf6 22 Qxd5+ Ke8 23 Re1+ Be7 24 Rxb7 is
crushing, while 22 ... Ne6 23 Rxb7 wins the queen) 22 Re1+ Ne6 23 Qg4
(attacking both d4 and e6) 23 ... Be5 24 Ba3+ Kd7 25 Qxa4+ Kd8 26 Qg4
and Black is being pulverized with only a measly pawn for comfort.
21 Rxb7+ Bd7
21 ... Nd7 22 Qg4! is similar.
22 Qg4!

22 ... Qxd4
After 22 ... Nxh7 23 Rxd7+ Kg8 24 Nf5! White has a decisive attack.
There is already the threat of 25 Nh6+ Kh8 (or 25 ... Kf8 26 Rf7+ and it’s
mate next move) 26 Nf7+, winning the queen.
 If instead 22 ... Rd8, then 23 Nf3! Qe6 24 Qh5+ g6 25 Bxg6+! Nxg6 26
Qh7+ Kf8 (26 ... Bg7 27 Ng5+ wins the queen again) 27 Bh6+ Ke8 28
Qxg6+ is disastrous for Black.
23 Rxd7+ Nxd7 24 Qxd7+ Be7 25 Re1 Qe5
Attempting to save himself by utilizing a back rank mate after 26 Rxe5??
Rxc1+. But Matlakov isn’t ruffled. 25 ... Qf6 26 Ba3 Re8 27 Qxd5+ was also
hopeless.
26 Bd2 Rd8
Black is also crushed after 26 ... Qd6 27 Qf5+ Ke8 28 Bg5 Rc7 29 Bg6+
Kd8 30 Qf8+ Kd7 31 Qxa8 etc.
27 Qg4 1-0
If the black queen retreats, White can sacrifice a bishop to get control of
the e6-square with a quick mate: 27 ... Qf6 28 Bg5! Qxg5 29 Qe6+ Ke8 (or
29 ... Kf8 30 Qg8 mate) 30 Bg6+ Qxg6 31 Qxe7 mate; or 27 ... Qd6 28 Bb4!
Qxb4 29 Qe6+ Ke8 30 Bg6+ Kf8 31 Qf7 mate.

Game 58
L.Aronian-R.Rapport
Wijk aan Zee 2017

1 d4 Nf6 2 c4 e6 3 Nf3 b6 4 g3 Bb7 5 Bg2 Be7 6 0-0 0-0 7 Nc3 Ne4 8 Bd2
Bf6
A typical Queen’s Indian set-up with Black delaying the advance of his
centre pawns in favour of controlling the key e4-square with his bishop on
b7.
9 Be1
 The quiet nature of this move might have made Rapport careless. After
all, putting a bishop on e1 isn’t normally the prelude to a crushing attack.
9 ... Re8?
It’s hard to fathom the purpose of this very mysterious rook move.
Instead, 9 ... Nxc3 10 Bxc3 Be4 (not wanting to be shut in after the next
move) and then 11 ... d5 fights for the centre.
10 Qc2 d5?
A serious mistake, as the resulting opening of the d-file proves very
unfavourable for Black. He had to make do with 10 ... Nxc3 11 Bxc3, though
11 ... d6 12 Rad1 and 13 e4 leaves White with a pleasant space advantage,
while 11 ... d5 12 cxd5 exd5 13 b4 gives White pressure down the c-file.
11 Nxe4 dxe4 12 Nd2
Now Black has no choice but to take on d4 or he will be a pawn down
once e4 falls.
12 ... Bxd4 13 Rd1
The e4-pawn can’t run away as b7 would drop. After the immediate 13
Nxe4 Black could try to wriggle out with 13 ... Nc6 14 Rd1 Qe7. With the
game move Aronian keeps the option of answering 13 ... Nc6 with 14 Bxe4,
with ideas of Bxh7+ as well as Nf3 and Bxc6 to embarrass the bishop on d4.
In any case the rook move has the desired effect as the black queen goes to c8
(rather than e7) and cedes the g5-square to the white knight.
13 ... Qc8 14 Nxe4 Bc5

Question: You could inflict damage on Black’s queenside pawns with

15 Nxc5, but can you see a better plan?

Answer: 15 Ng5!
Aiming at h7. We are told if you see a good move, look for a better one.
We might add:
If you see a good plan, look for a better one!
After 15 Nxc5 Bxg2 16 Kxg2 bxc5 17 Bc3 White can add the superior
minor piece and control of the d-file to his structural advantage. If you are a
steady positional player who dislikes making sacrifices then I can understand
why you’d prefer to play like this. The doubled pawns on c7 and c5 aren’t
going to vanish in a hurry, whereas an attack on the king can disappear after
one or two suboptimum moves.
But in my opinion you shouldn’t shun a big opportunity just because it
isn’t your type of chess. I’m not saying that you should play sharp openings
and seek tactics if you are happier with a quiet strategic approach. But when a
chance to exert powerful dynamic pressure appears in the course of your
solid openings, you should seize it.
Here, for example, Aronian has played a slow, positional opening but is
quite ready to switch to tactics. A move like 9 Be1 is as slow and unobtrusive
as it gets, but leads to violence after Rapport’s rather careless response.
Therefore the knight heads for g5 as the prelude to an attack on the black
king.
15 ... f5 16 Bxb7 Qxb7 17 Bc3

First of all, provoking 15 ... f5 with the knight has weakened the e6-pawn.
Then, by exchanging on b7, White has driven the black queen away from the
defence of d7 and e6. Finally, the bishop has aimed itself at g7.
All these measures focus on obstructing the development of the black
knight. Thus, with the queen gone from c8, 17 ... Nd7 is ruled out. Or if 17 ...
Nc6 then 18 Rd7 attacks g7 (thanks to the bishop) and 18 ... Re7 in response
drops a pawn (thanks to the knight) due to 19 Rxe7 Nxe7 20 Nxe6.
Not only is Black unable to bring his knight out to a decent square, his
rook on a8 has to stay in the corner as well. The upshot is that Black is
seriously outgunned in the balance of forces in the centre and on the kingside.
But such advantages tend to evaporate over time, so Aronian is obliged to
attack as quickly and violently as possible.
17 ... Bf8
Defending g7 and, I expect, preparing a sequence such as 18 ... c5, 19 ...
Nc6 and 20 ... Rad8 to complete his development (he can answer 21 Rxd8
with 21 ... Nxd8 to keep e6 guarded). Aronian gives him no time for any of
that.
18 e4!
A lead in development needs open lines. The f5-pawn provides a hook
onto which White can latch his attack.
18 ... h6 19 exf5! hxg5 20 f6
 Question: How would you sum up White’s compensation for the piece?

Answer: In return for the piece White has a powerful pawn supported by
the bishop. His queen has an open diagonal and is ready to go to g6,
threatening the rook on e8 as well as mate with f6-f7+ followed by a queen
check on the h-file. Or White might choose to play 21 fxg7 and only then
Qg6.
20 ... c5
Allowing the queen to defend along the second rank. For instance, 21
Qg6 would now be rebuffed by 21 ... Qf7!.
21 f4!
Here we see another point of the piece sacrifice. White is able to bring his
reserve pieces into the attack faster than Black can reinforce the kingside’s
defences. Once White has wrenched open the f-file, he has two rooks, a
queen and a bishop besieging the black king, to say nothing of the rampant
pawn on f6. Meanwhile, Black still has a knight and a rook dozing on the
queenside. So, despite being nominally a piece down, White actually has
more pieces where it matters.
21 ... g4
A despairing attempt to keep things blocked up.
22 f5 gxf6 23 fxe6 Qh7
To meet the threat of 24 Qg6+ which would have won a rook. Note that
23 ... Nc6 would have lost at once to 24 Rd7.
24 Qg2!
The threat to slay the sleeping rook on a8 obliges the black knight to
abandon the defence of the d7-square.
24 ... Na6 25 Rd7 Qh5 26 Rxf6
Now both rooks crash through,
26 ... Rad8
Too late! Too late!
 Question: Can you see the tactical finish to the attack?

Answer: 27 Rxf8+! Kxf8

After 27 ... Rxf8 28 Rg7+ Kh8 29 Rg5+ Kh7, Black is actually
threatening mate beginning with a rook check on d1, but White can easily
sidestep it with 30 Rxh5+ Kg6 31 Qe2 or 30 Qb7+, when he has an entirely
crushing game in either case.
28 Qf1+ 1-0
Since if 28 ... Kg8 29 Rg7+ Kh8 30 Rg5+ Kh7 31 Qf5+ Kh6, there is a
double nightmare for Black who loses his king and queen on the same move
with 32 Rxh5 mate. (32 Bg7 mate is just as effective, if less brutal.)

Game 59
B.Lalic-M.Vicas
Dublin 2019
1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 Be7 6 e3 c6 7 Bd3 0-0 8 Qc2
h6 9 Bf4 Bd6 10 Nge2

A decision that determines the nature of the middlegame struggle. The

knight development to e2 indicates that Grandmaster Bogdan Lalic is aiming
to castle queenside. The knight usually goes to f3 if White plans to castle
kingside, one reason being its greater defensive value in guarding h2. Thus
10 Bxd6 Qxd6 11 Nf3 Re8 12 0-0 would lead to a positional battle in which
White would try for pressure on the queenside with a plan such as 13 Rab1
and 14 b4, aiming to besiege c6 with 15 b5.
When two plans of equal worth are available, one aggressive and the
other positional, the temperament and competitive aims of the player should
decide which one is chosen.
Would you say you prefer to attack the king or are you happiest hacking
away at a weakness in the opponent’s pawn structure? You might choose to
look through your games and see what sort of playing style works out best for
you.
10 ... Re8 11 Bxd6 Qxd6 12 h3
 This would be a defensive move if White had castled kingside, but here it
is the start of a kingside pawn advance. It also keeps Black guessing for one
more turn about White’s real intentions with his king.
12 ... Nbd7
Imagine if Black had anticipated White castling queenside and lashed out
with the immediate 12 ... b5?. Then Lalic could change his mind with 13 0-0!
and put pressure on the backward pawn on c6 with moves like Ng3, Nce2 and
Rac1. Black would have no compensation for the weakness of his queenside
structure.
Pawns can’t move backwards, so it is essential you advance the right
pawns at the right time.
13 g4
Only now is White more or less committed to castling queenside.
13 ... b5!
The horrible positional mistake 12 ... b5? has become the dynamic 13 ...
b5!. And it all comes down to the position of the white king.
14 0-0-0
With his queen and king both sitting on the c-file, a plan to win the c6-
pawn would be a hot potato. It’s no surprise that Lalic continues with his
kingside attack.
White could still have avoided castling queenside with, say, 14 Rc1 Nf8
15 Ng3, when his king could sit on f1 or g2 in the future. Nevertheless, in
that case his pressure on c6 would be no positional free lunch for him, so to
speak. Black can point to the somewhat ragged state of the white kingside
pawns.
14 ... a5!
 There is more to attacking an enemy king than flinging the pawn
forwards. Here Marijus Vicas has a specific idea in mind. He clears the way
to challenge the white bishop with ... b5-b4 and ... Ba6. It would be good to
swap a passive piece sitting idly on c8 for an active one that can potentially
threaten his king. The removal of the bishop from c8 will also vacate the
square for a rook – a good post in view of the chance to break through with ...
c6-c5 and embarrass the white queen and king.

Question: Lalic now decides to bring a second rook over to the kingside,
leaving the other on the h-file, so that they are both ready to support an
advance with h3-h4 and g4-g5. Do you see any drawbacks to this plan?

15 Rdg1?
Answer: It is very hard to checkmate in the middlegame, especially if you
are playing an opponent who develops his pieces quickly, doesn’t neglect the
centre, and takes no great risks with his king. You need as much fire-power
as possible to help you break down resistance. With this in mind Lalic
shuffles his rook from d1 to g1.
However, there is a well-known maxim that for a wing attack to succeed
you need superiority in the centre. A proviso needs to be added that if the
central pawn structure is blocked, or at least fixed with the opponent having
no good way to disturb the equilibrium, then you don’t necessarily need
superiority there. It’s good enough to have a solid standing in the centre and
be left undisturbed there while you carry out your attack.
As we shall see, White has neither superiority nor a secure centre
position. For this reason the queen’s rook needed to remain on d1 to help
fight Black’s counterattack.
White’s decision can be classed under neglect of the centre (an unjustified
decentralization) or – if we accept that 15 Rhg1 was the correct move – the
wrong rook. When a player has connected rooks (usually on the first rank) he
has to decide where he should deploy them for maximum value. Even great
players have made mistakes, such as (after castling kingside) moving the
rook from f1 to centralize it on e1, when the rook should have remained on f1
to support kingside action, and so Rae1 should have been preferred. The term
“wrong rook” is often a useful shorthand for “White (or Black) is beginning
an attack on the wrong side of the board”.
After having seen how the position plays out over the next few moves,
we’ll return to see how things would have been different after 15 Rhg1.
15 ... b4 16 Na4 Ne4
Preparing a pawn sacrifice to fight for the initiative. It wasn’t necessary,
as the immediate 16 ... Ba6 looks okay for Black; e.g. 17 Bxa6 Rxa6 18 g5
hxg5 19 Rxg5 Ne4 20 Rg2 c5! 21 Rhg1 g6, when after 22 f3? Rc6! the threat
of 23 ... cxd4 gives Black a good game.
17 Ng3 Ba6!? 18 Bxe4
If 18 Nxe4 dxe4 19 Bxe4, Black should try 19 ... Bb5!?, threatening to
win a piece with 20 ... Bxa4, as after 19 ... Rac8 White can change his mind
with 20 Rd1!.
18 ... dxe4 19 Nxe4 Qd5
The point of Black’s combination is that a2 and e4 are both hanging.
20 Nec5 Nxc5 21 Nxc5
 Question: Should Black play 21 ... Qxa2 to infiltrate with his queen and
get
a pawn back?

21 ... a4!
An excellent decision. Vicas is prepared to make further pawn offers to
keep up the momentum of his attack.
Answer: 21 ... Qxa2? would not be in the spirit of the position. After 22
b3! Qxc2+ (instead, 22 ... Qa3+ 23 Kd2 leaves Black with no good answer to
the threat of 24 Ra1, trapping the queen) 23 Kxc2 Black has been forced into
a bad endgame – his pawns on a5 and c6 make a terrible impression
compared with White’s sleek mass of centre and kingside pawns and,
besides, the knight is very strong on c5.
It should also be mentioned that White is a veteran grandmaster while
Black was a young and upcoming player rated 2194 Elo at the time. Lalic
would have loved the chance to grind his opponent down in an endgame even
half as favourable as that after 23 Kxc2, rather than be embroiled in a
middlegame mêlée. Of course this begs the question of why White played so
aggressively in the opening rather than steering the game along positional
channels. Perhaps he underestimated his opponent and thought he would cave
in quickly when faced by an attack. He couldn’t have guessed that he was
going to find such great moves.
22 Nxa4
If he doesn’t take the pawn then 22 ... b3 will ram his queenside. It looks
like the threat of a knight fork on b6 might give Black pause for thought, but:
22 ... c5!
A third pawn sacrifice to complete his plan of breaking open lines on the
queenside. Bogdan Lalic showed me this game before the start of a match in
which we were playing for Wood Green (a team that, due to the efforts of our
captain Brian Smith, has won the London League Championship 17 years in
a row). He told me had hoped for 22 ... Bd3 23 Qxd3 Rxa4, when he regains
the initiative from Black with 24 Qb3 Qa5 25 g5! (it’s White’s turn to make
some line-clearing sacrifices) 25 ... hxg5 26 h4! gxh4 27 Rxh4 Rxa2 28 Kd2!
Ra1 29 Rg2! and White’s attack is the stronger – he threatens to double rooks
on the h-file or g-file (note that after Rhg4 Black can’t defend with ... g7-g6
as Rxg6+ follows, due to the pin on f7).
 23 Nxc5
If 23 Nb6 then 23 ... Qxa2 24 Nxa8 cxd4 25 Nc7 d3 26 Qb1 Qc4+ 27
Kd2 Qxc7 and Black has a decisive initiative. One plan would be to
manoeuvre his queen to f6 via c6 to attack f2 and threaten to infiltrate via f3
into e2. This can be combined with ... Rc8, intending ... Rc2+ once the white
rooks have been tied down by the black queen.
23 ... Rec8
Threatening 24 ... Rxc5! 25 dxc5 Qxa2, intending the deadly 26 ... b3 27
Qb1 Qa4 followed by a killer check on c4.
24 Rd1
The rook is recalled to the defence of the centre, having achieved nothing
on the g-file. If 24 Kb1 then 24 ... Bb5! (threatening a2) 25 b3 Rxc5! 26
Qxc5 (after 26 dxc5 Bd3 the queen is lost) 26 ... Bd3+ 27 Kc1 Qf3 28 Rd1
Bg6, intending 29 ... Rxa2, gives Black a decisive attack (note that 29 Rd2
drops the rook on h1), while 27 Kb2 allows an unexpected mate: 27 ...
Rxa2+! 28 Kxa2 (or 28 Kc1 Qxb3 mates) 28 ... Qa8+ 29 Kb2 Qa3 mate.
It stands to reason that in the complications discussed above and on
previous moves keeping the rook on d1 (rather than 15 Rdg1) would have
aided White considerably: the fact that the rook is now compelled to return to
d1 shows that White in effect lost two tempi by decentralizing it on g1. If the
rook had stayed on d1, and 15 Rhg1 been played instead, we would have had
to re-evaluate the worth of Black’s plan with 16 ... Ne4. For example,
imagine one white rook is on d1 and the other on g1, and the line above with
24 Kb1 Bb5 25 b3 Rxc5 had occurred. White can then reply 26 dxc5 because
with the rook on d1 he doesn’t lose his queen to 26 ... Bd3. This line
therefore ends not in a clever mate by Black but in a trivial win for White.
Black would have to try another sequence of moves after 15 Rhg1 to get
his attack going, but it would be by no means as easy.
24 ... Qxa2
After 24 ... Rxc5 25 dxc5 Qxa2 26 Qe4!, attacking a8, White comes out
on top.
25 b3!

The only way to stay alive. Black has nothing better than to force a draw
by repetition with a final sacrifice.
 Question: Can you work out his combination?

Answer: 25 ... Rxc5! 26 dxc5 Qa1+ 27 Kd2

Or 27 Qb1 Qc3+ 28 Qc2 Qa1+ etc.
25 ... Rd8+ 28 Ke1 Rxd1+ 29 Qxd1 Qc3+ ½-½
Black’s queen gives perpetual on c3 and a1.
An eventful game. Black deserves a lot of credit for his string of
sacrifices, and White defended well after his initial inaccuracy with 15 Rdg1.
The next time you are playing a grandmaster, remember this game and come
out fighting!

Game 60
I.Nepomniachtchi-S.Karjakin
Grand Chess Tour, Zagreb 2019

1 e4 e5 2 Nf3 Nc6 3 Bc4 Bc5 4 c3 Nf6 5 d3 d6 6 0-0 a6 7 Bb3 Ba7

The usual moves here are 8 Nbd2 or 8 Re1 or sometimes 8 Be3. But
Nepomniachtchi saw nothing wrong with:
8 h3
 A typical move in this set-up and favoured by some strong players. Along
with 8 Nbd2 or 8 Re1 it seems just another way to get a standard middlegame
position after Black plays 8 ... 0-0. White begins by preventing ... Bg4 and
the game carries on with a struggle in the centre.
However, it’s the habit of strong players to think concretely about a
position. Often choosing a certain move order makes no difference, but here
it is of game-changing significance, as Karjakin shows with his next two
moves.
8 ... h6!
Rather than 8 ... 0-0. It looks like Black is ‘returning the compliment’ by
stopping Bg5, but in fact there is a whole attack contained in his little pawn
move.
9 Re1 g5!
This is the point. Black is going to gain maximum value from the hook
for an attack with ... g5-g4 provided by 8 h3.
10 Nh2 Rg8
Notice the role in the attack played by the bishop on a7. White can’t stop
Black’s next move with 11 f3 as the pawn is pinned. So Nepo tries to reduce
the impact of Black’s attack by exchanging off bishops – the best try in the
circumstances.
11 Be3 g4 12 hxg4
The attempt to keep lines blocked rebounds after 12 h4 Bxe3 13 Rxe3
Nh5 14 g3 Nf4!, planning 15 ... Nh3, when the knight is a thorn in White’s
kingside. Note that 15 gxf4? exf4 16 Re1 Qxh4 gives Black a crushing attack
with moves like 17 ... g3 or 17 ... Ne5 in mind.
12 ... Nxg4 13 Qf3
The queen takes up a useful defensive post. Black mustn’t get carried
away and miss the mate on f7!
13 ... Rg7 14 Bxa7 Nxh2 15 Kxh2 Qh4+ 16 Kg1 Bg4

17 Qe3
When facing an attack less experienced players often get scared and want
to exchange off queens no matter what. In contrast, a strong player would
rather take his or her chance against the attack than defend a miserable
endgame of the kind reached after 17 Qg3 Qxg3 18 fxg3 Nxa7, when Black
would continue 19 ... Ke7 and 20 ... Rag8 to beleaguer the pawn on g3.
There’s more likelihood of an experienced opponent messing up in a complex
middlegame than in a simplified endgame. Besides, there’s always the
cheering thought that you might even win if he goes wrong, whereas in the
endgame you’re just holding on for a draw.
17 ... Nxa7 18 Nd2 Nc6 19 g3 Qh3 20 Bd1 Bxd1
Despite what we said in the last note Nepo still seeks safety in exchanges,
as long as he doesn’t compromise his position too much in arranging them.
21 Raxd1 0-0-0 22 Qf3 Rh8 23 Nf1
Things might not took too bad for White, but the second wave of Black’s
kingside pawn storm is about to break.
23 ... h5 24 Ne3 h4

Question: Can you give a reason why 24 ... Ne7 would be a superior
decision?
As a positional and tactical exercise you might like to consider how Black
should then meet the moves 25 Nf5 and 25 Qf6. We shall discuss 24 ...
Ne7
below after looking at the rest of the game.

Nonetheless, the game move looks like a powerful lever to open the h-file
and put pressure on g3. It seems that the white position is about to collapse;
for example, 25 Nf5 Rgh7 leaves him facing a strong attack which can be
strengthened by the manoeuvre ... Nd8-e6-g5, driving the white queen from
her vital defensive post on f3. Then 26 Nxh4 fails to 26 ... Rxh4, while if 26
Qg2 hxg3 the recapture 27 Nxg3 allows 27 ... Qg4!, declining the exchange
of queens to keep up an overwhelming attack; otherwise, Black is prepared to
exchange queens to win a pawn in the event of 27 Qxh3 gxf2+ 28 Kxf2 Rxh3
or 27 fxg3 Qxg2+ 28 Kxg2 Rh2+ 29 Kf3 Rxb2.
Notice how Black is more than happy to give up his attack to win a pawn
in these lines. Sometimes an attack is an adventure that can only be justified
if you bring back the head of the opponent’s king. You’ve sacrificed material
or made a major concession which means there is no way to return to normal
positional play. You have to do or die. But there are a lot of occasions when
you can happily give up your attack for a humbler, but still decisive prize.
Here the pawn on b2 was minding its own business but turned out to be the
crowning point of Black’s strategy after 29 ... Rxb2 in the line above.
An initiative can evaporate for no clear reason, whereas a material
advantage is forever (or at least seems like forever to the poor guy defending
a long endgame, which is another psychological reason for going for
material). Be flexible in your thinking. Keep an open mind and don’t go nuts
in chasing the mirage of a checkmate when there is a good, healthy, real
pawn to feast on.
None of the above happened in the game. Nepo held it together very
easily by getting the queens off.
25 Qf5+! Qxf5 26 Nxf5
Besides defending g3, the knight attacks g7 which gains another move for
the defence.
26 ... Rg5 27 Re3!
 A good swap: White prepares to exchange off the active black rook on g5
for a rook that was doing nothing on e1.
27 ... Kd7
Not 27 ... Rgh5? 28 Nxh4, while 27 ... h3 28 Kh2 leaves the advanced
pawn as a liability, perhaps doomed to death.
28 Kg2
The final consolidating move. Black has to liquidate or else 29 Rh1 will
capture the h-pawn. The rest was uneventful:
28 ... hxg3 29 Rxg3 Rxg3+ 30 Kxg3 Ne7 31 Nxe7 Kxe7 32 Kf3 Rh3+
33 Kg2 Rh8 34 Kf3 Rh3+ 35 Kg2 Rh8 ½-½

Answer: We should return to the critical moment in the game and see if
Black could have done better with 24 ... Ne7.
 This rules out the immediate queen exchange, while after 25 Nf5 Black
can again renounce the attack in favour of eventually winning a pawn: 25 ...
Nxf5 26 Qxf5+ Qxf5 27 exf5 h4 (loosening up White’s kingside before going
after the real target on f5) 28 Kg2 hxg3 29 fxg3 Rg5 30 Rf1 Rhh5. The f5-
pawn is going to fall. If it advances to f6 Black could put his king on e6 to
help the rooks round it up with, say, ... Rh6 and ... Rgg6.
In the game Nepo got to exchange queens at no cost, whereas in this line
White secured his king from attack but is left with a pawn on f5 which can’t
be held.
The alternative 25 Qf6 looks awkward for Black as both his knight and
g7-rook are hanging. But the direct attack on the king proves strong after 25
... Rhg8! 26 Ng2 (taking the knight allows a forced mate: 26 Qxe7? Rxg3+
27 fxg3 Qxg3+ 28 Kf1 Qf3 mate or 28 Kh1 Qh3 mate) 26 ... Rg5! (to play
his next move without allowing White to take on h4 with the queen, though
26 ... h4 27 Qxh4 Qe6 also looks menacing for the white king) 27 Qxe7 (or
27 d4 h4, keeping up the attack) 27 ... h4.
 It might not seem obvious at first glance but there is no way for White to
hold his kingside together. For example:
a) 28 Re3 hxg3 29 Rxg3 Rxg3 30 fxg3 Qxg3 31 Rd2 Qe1+ wins the rook
with a quick mate.
b) 28 Rd2 hxg3 29 fxg3 Rxg3 30 Ree2 Rh8 and there is no good way to
stop 31 ... Qh1+ 32 Kf2 Rxg2+ 33 Ke3 Rh3 mate.
c) 28 Qxf7 hxg3 and White can’t keep the g-file closed, as 29 f3 Qh2+ 30
Kf1 Qh1+ 31 Ke2 Qxg2+ 32 Ke3 Qf2 is mate.
Chapter Nine
Sacrificing to Gain the Initiative
Here we are concerned with investing material in order to force the opponent
onto the back foot: he is compelled to respond to our manoeuvres and threats,
rather than indulge in plans of his own. The focus is on sacrifices which
create attacking chances against the enemy king, so there will be some
overlap with the procedures outlined in Chapter Eight.

All-out Attack
In his manual Chess Fundamentals Capablanca makes the point that if you
start an all-out attack on the king with sacrifices it has to be carried through
to victory – you can’t break it off halfway through or you will lose. Of course
the need to mate isn’t quite as mandatory as Capa suggests: we could imagine
discontinuing an attack if the opportunity arose to recoup any material
invested and exchange into a good endgame. In such a scenario we might
even say that reaching a good endgame completed the attack.
But Capa is certainly right to warn us against attacking in a half-hearted
or haphazard manner. If we launch a committal attack and it doesn’t hit
home, we are likely to be left with disorganized pieces which can’t hold on to
key points in the face of the opponent’s counterattack. Even if we haven’t
sacrificed any material we might find ourselves positionally lost.

Game 61
D.Paravyan-S.Golubov
Korchnoi Memorial, St. Petersburg 2018

1 e4 e5 2 Nf3 Nf6 3 Nxe5 d6 4 Nf3 Nxe4 5 d4 d5 6 Bd3 Bd6 7 0-0 0-0 8 c4

c6 9 Qb3!?
Objectively speaking, this move isn’t particularly good, but it has the
virtue of being a reasonable choice and different from the main line. Making
the opponent think for himself is a sure way to provoke blunders. After the
usual sequence 9 Nc3 Nxc3 10 bxc3 dxc4 11 Bxc4 Bg4 12 Qd3 Nd7 White
has more space but Black is fully developed.
9 ... dxc4
 This is a valid response, but it falls in with White’s plans by giving him
pressure on f7. If Black had expected 9 Qb3 he would probably have learnt
an active alternative with 9 ... Na6!?, when White can’t take twice on d5 as
he loses the queen to a ... Bxh2+ trick.
10 Bxc4 Nd7 11 Re1 Ndf6 12 Nbd2 Nxd2 13 Bxd2 Qb6
Also possible was 13 ... b5; e.g. 14 Bd3 Bg4 15 Ne5 Be6 16 Qc2 Rc8
with unclear play.
14 Qd3 Qxb2?!

Black still looks okay after 14 ... Nd5. A pawn on its starting square
whose capture leads to trouble is often referred to as a poisoned pawn. The
great Bobby Fischer lost two games by grabbing a poisoned pawn in his 1972
match with Spassky: with the infamous 29 ... Bxh2? in game one, and then in
game eleven in the Poisoned Pawn Variation of the Sicilian Najdorf: 1 e4 c5
2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Bg5 e6 7 f4 Qb6 8 Qd2 Qxb2 (he
didn’t respond well to 9 Nb3!? rather than the usual 9 Rb1) Apart from these
poisoned pawn and Poisoned Pawn mishaps, and a game lost by default,
everything in the match went splendidly for Fischer.
 Why then do players still insist on take these venomous pawns? Perhaps
it is because they know that – in games that don’t make it into books – greed
often triumphs.
15 Rab1 Qa3 16 Qc2 Nd5?
Allowing White to begin a glorious combination. Instead, 16 ... b5 17
Bb3 Qa6 stops the black queen becoming a target (though there’s nothing
pretty about having your most powerful piece stuck on a6). If then 18 Ne5
Be6, White has the initiative but Black can still hope to defend soundly.
After the game move White is able to target both the black queen and
king. If he could only attack the king, or only harass the queen, Black would
be able to defend against the threats and wriggle out. But as we shall see, he
isn’t able to deal with the double whammy of threats to his two tallest pieces.
17 Rb3
The third rank is the route by which the rook can travel to the kingside.
17 ... Qa4 18 Bxd5! cxd5
Everything else being equal, the exchange of bishop for knight looks a
poor deal for White. Perhaps this is why Golubov underestimated the danger.
In this specific case Paravyan is removing a key piece opposing his kingside
assault.
19 Ng5
The attack begins in earnest with a good old-fashioned threat of mate.
19 ... g6
 Black has an extra pawn and the bishop pair. He is also poised to develop
with 20 ... Bf5, putting the bishop on an excellent square with gain of time.
Materially and positionally White is busted unless his attack breaks through
and wins. The dynamic power which makes the attack possible is going to
evaporate quickly if he doesn’t do something fast: Black needs only to
complete his development with the aforementioned 20 ... Bf5 and then
centralize his rooks.

Question: Can you see the first of White’s sacrifices? Remember, we are
trying to combine threats to the black king and queen.

Answer: 20 Nxh7!
As you will see, Paravyan’s attack is by no means easy to carry out. That
was why Golubov was seduced into thinking he would be able to brush it off.
We should mention that White would also have had a decisive attack after
19 ... f5 20 Nxh7! Rf7 21 Qd1! etc.
20 ... Bf5
 After 20 ... Kxh7 21 Rh3+ Bxh3 22 Qxa4 Black doesn’t have enough for
the queen.
21 Nf6+ Kg7

With his queen and knight both hanging it seems as if White has overshot
his bolt, but now another piece joins in the attack.

Question: How do we continue to drive the black king out into the open?

Answer: 22 Bh6+! Kxf6

Again 22 ... Kxh6 23 Rh3+ costs Black his queen. Golubov’s king has
been driven out of his defensive shell, but it’s too soon to despair as he is a
piece up and still attacking the white queen.

Question: If you see the next move then you have really taken the chapter
on using a pawn as a battering ram to heart!

Answer: 23 g4!!
 In fact 23 Qd2 was also strong, when Black would have to answer 23 ...
Be4. However, the splendid game move is not only objectively best but also
gives me an opportunity to repeat advice close to my heart: don’t forget to
use your pawns as part of a plan.
23 ... Bf4
If 23 ... Bxc2, White loses (e.g. 24 g5+? Kf5) unless he plays the one
move that wins: 24 Rf3+, when 24 ... Bf5 25 g5 is mate.
Or 23 ... Be4 24 Rxe4! (winning the queen with 24 Rf3+ Ke7 25 Qxa4
isn’t bad either, and 24 ... Bxf3 25 g5 is mate) 24 ... dxe4 25 Qxe4 leaves
Black defenceless against the twin threats of 26 g5 mate and 26 Rf3+. For
example: 25 ... g5 26 Qe5+ Kg6 27 Qg7 mate.
After the game move it seems like Black is going to escape. If 24 Bxf4
Black can at last safely play 24 ... Bxc2 and win. But Paravyan has yet one
more surprise up his sleeve.

Question: Now think about how you can overload the bishop on f4 which
has to guard against mate on g5.
 Answer: 24 Qc7!!
Threatening mate on e7 and ready to answer 24 ... Bxc7 with the familiar
25 g5 mate. I wonder if White enjoyed making this move during the game or
if he was too caught up in calculating and the general stress of play to
appreciate its beauty.
24 ... Bxh6 25 Qe5+
Only here. 25 Qe7+? would be a tragic mistake as the black king escapes
back to g7.
25 ... Kg5 26 h4+
Let’s forgive Paravyan that 26 Rg3! was a faster mate.
26 ... Kxh4
Or 26 ... Kxg4 27 Rg3+ Kh5 28 Qe2+ Kxh4 29 Kg2 (apparently, 29 Qf3
mates one move faster, but clearing the path for a rook check on h1 is the
‘human’ way) 29 ... Qd7 (or 29 ... Be4+ 30 Qxe4+ dxe4 31 Rh1 mate) 30
Rh1+ Bh3+ 31 Rhxh3+ Qxh3+ 32 Rxh3+ Kg5 33 Rg3+ Kf4 34 Qg4 mate.
27 Rh3+! Kg5
Or 27 ... Kxh3 28 Qg3 mate.
 Question: Now what is the quickest win?

Answer: 28 Qe7+! 1-0

Checking on e7 forces 28 ... f6 (as 28 ... Kxg4 29 Qh4 is mate and 28 ...
Kf4 29 Qe3+ is mate next move), when the black king’s retreat to f6 is cut
off. Then 29 Qe3+ Kxg4 30 Qg3 is mate.
White’s combination was so unexpected and beautiful that it is tempting
to give up on logical analysis. Instead, we could shrug our shoulders and
describe it as a black swan, the product of a lucky alignment of the stars that
day, or as an act of genius on Paravyan’s part. (What did he have for
breakfast?)
Well, we are unlikely ever to play a move as startling as 24 Qc7!. But if
the enemy king has been driven out of its defences, and we are in effect
attacking with two extra rooks, then magic does happen. A big advantage in
firepower leads to fireworks!
How to Make Carlsen Play Stupid Moves
If you trap one of your opponent’s pieces he will probably have to play some
convoluted manoeuvres to free it. If it is his queen who needs to be rescued
then the ‘ransom’ to be paid can be moves which in other circumstances
would seem absurd.

Game 62
L.Aronian-Ma.Carlsen
Stavanger 2017

1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 Nc3 e6 5 e3

The Meran Variation of the Semi-Slav. White defends the c4-pawn and in
most cases builds up towards an e3-e4 centre break. It seems he can’t do
anything very aggressive once he has shut in his bishop on c1. However, as
Aronian will demonstrate, sometimes you can breathe fire even into the most
tranquil-looking positions.
5 ... a6 6 b3!
 Restraining the advance of Black’s queenside pawns as would occur after
6 Bd3 dxc4 7 Bxc4 b5 8 Bd3 c5, when 9 0-0 Bb7 gives Black an active game.
Whereas now 6 ... dxc4? 7 bxc4 merely strengthens White’s pawn centre.
6 ... Bb4
Putting the bishop on b4 is a measure against White advancing e3-e4 in
the future.
7 Bd2 Nbd7 8 Bd3 0-0 9 0-0 Qe7 10 Bc2 Rd8
Carlsen is a long way towards equalizing. He has a compact centre, a safe
king, and an almost complete development. The fly in the ointment, as in
many Black openings after 1 d4 d5, is the bishop on c8, which is blocked in
behind a wall of black pawns on light squares. We might expect a tense
manoeuvring game, with the world champion attempting to advance ... e6-e5
to free his bishop, and Aronian trying to gain a structural or other type of
advantage when he does so. Instead, after ten moves of quiet play, the game
abruptly takes a tactical twist.
11 a3!? Bxa3
Carlsen was never going to turn down the pawn offer unless he saw a
definite drawback. The alternative 11 ... Bd6 would allow White to advance
12 e4 with impunity, as Black no longer has ... Bxc3 followed by ... Nxe4 to
deter it.
12 Rxa3! Qxa3 13 c5
 Question: How do you assess the sacrifice? What is White’s
compensation?

Answer: White has cut off the retreat of the black queen. It’s time to take
stock. Assuming she can’t actually be captured by force (and she is indeed
safe with best play), what has White gained by sacrificing the exchange and a
pawn? Black’s kingside, queenside and centre remain rock solid. The only
way they will fall into danger is if Carlsen plays some outrageously anti-
positional moves, the sort of thing he might have done when he was seven
years old. And no player rated over 2800 will ever do that, even in a bullet
(one minute) game. Right?
Ah, but in this position Carlsen’s queen is being held to ransom. The
price for her release is a couple of terrible moves which open lines to the
benefit of the white pieces on both the queenside and kingside. In other
words Carlsen is obliged to play stupidly.
13 ... b6
 If 13 ... Qa5? 14 Nxd5, Black has lost an important centre pawn without
achieving any freedom for his queen.
14 b4 Ne4
Here is the stupid move: Black gives up a pawn and exposes both his king
and c6 to danger. But the binds around the queen had to be loosened
somehow.
15 Nxe4 dxe4 16 Bxe4 Rb8

Question: Is there a way to disturb the black king? Can you calculate
the variations?

With 16 ... Nf6 Carlsen could put a stop to the coming attack against his
king. Of course he is more than familiar with the Greek Gift sacrifice. But
after 17 Bxc6 Rb8 18 Ne5 the black queen is still surrounded. And what’s
more, White would have a tremendous pawn on c5. Carlsen decides it is
better to put his king in peril than collapse on the queenside.
You can see the power of alternating threats to two points. Carlsen would
have effortlessly swatted off the danger to his king or the c6-point. But he
can’t do both at the same time.
With 16 ... Rb8, Black now has counterplay after 17 Bxc6 a5. Instead,
there came:
Answer: 17 Bxh7+! Kxh7 18 Ng5+ Kg8
After 18 ... Kg6 19 Qg4 f5 20 Qg3 the black king is for the chop.
19 Qh5 Nf6
This is an unusual form of the Greek Gift, as Black has a knight available
to guard the h7-square. Therefore his king is safer than usual. However,
Aronian has seen that his queen’s entrance on f7 will pick up material while
keeping up the initiative.
20 Qxf7+ Kh8

Question: Can you see the clever move Aronian had prepared when he
made the sacrifice on h7?

Answer: 21 Qc7!
 Another form of alternation. Black can defend both rooks from the white
queen, but not at the same time deal with the knight swooping into f7.
21 ... Bd7 22 Nf7+ Kh7 23 Nxd8
It looks like it’s all over, but Carlsen is able to keep the fight alive by
winning the knight.
23 ... Rc8 24 Qxb6 Nd5
Crowding out the white queen from the defence of d8.
25 Qa7 Rxd8
It looks at first glance that Black has survived, as the material balance is
not bad: three white pawns for the black knight. But now begins the next
wave of White’s attack.
26 e4! Qd3
26 ... Nf6 27 Bg5, threatening both 28 e5 and 28 Bxf6 gxf6 29 Qc7, looks
gruesome for Black.
27 exd5 Qxd2 28 Qc7 Qg5 29 dxc6 Bc8
The material balance has changed again: now White has four pawns in
return for a bishop which has no safe moves.
30 h3
Stopping any back rank mate and so freeing his rook to become active on
the third rank.
30 ... Qd5
Of course 30 ... Rxd4 31 Qxc8 wins.
31 Rd1 e5?
 We’ve been lauding Aronian for the energy and invention of his attacking
play. But the world champion also deserves a lot of credit. Considering he
was taken by surprise in the opening and had to find his way through a mass
of complications, Carlsen has actually defended superbly up until this point.
Only now, when it seems that Black might finally be co-ordinating his pieces
and getting some counterplay, does he falter. Instead, 31 ... Rf8 32 f3 Qb3
gives Black enough activity to equalize chances.
32 Rd3!
Aronian has calculated that the entrance of his rook will be decisive.
32 ... exd4 33 Qe7!
Precisely does it. After 33 Rg3 Black can defend with 33 ... Rg8, when 34
Qe7? Qxc6 35 Qh4+ Qh6 defends. White should still win with the superior
34 Qd6! Qxd6 35 cxd6, but it would be more laborious.
33 ... Bf5
Now 33 ... Rg8 34 Qh4+ Kg6 35 Rxd4 wins. Black can’t even play 35 ...
Qxc6 because of 36 Rd6+.
34 Rg3 Bg6
 It’s mate in one after 34 ... Rg8.
35 Qh4+ 1-0
After 35 ... Kg8 36 Rxg6 Black can’t even fight on with 36 ... d3 as 37
Qf6 is immediately decisive.

Catching the Opponent’s Pieces on the Hop with Dynamic Play

Sometimes when the opponent is in the middle of an over-elaborate build-up
it is possible to throw the chess equivalent of a spanner in the works. This
might mean opening lines before the enemy pieces have co-ordinated
themselves properly: they are gradually gelling together with a view to
exerting long-term pressure and aren’t ready for an immediate tactical
slugfest.

Game 63
A.Grischuk-And.Volokitin
Baku Olympiad 2016

1 Nf3 Nf6 2 g3 g6 3 Bg2 Bg7 4 c4 c6 5 b3 Ne4 6 d4 d5 7 0-0 0-0 8 Bb2 Bf5

9 Nbd2 Qa5 10 e3 Nd7 11 Qe2 Rfe8 12 Nxe4 Bxe4
 After a slow opening White played:
13 Bh3
His bishop evacuates the g2-square and introduces the idea of 14 Nd2,
embarrassing the black bishop on e4. As Black’s knight is also hanging
Grischuk hopes his opponent will acquiesce to 13 ... Bxf3 14 Qxf3 e6, when
White can grind away with his bishop pair after 15 Rfd1. (Admittedly, it’s
difficult to see any real advantage for him.)
Instead, Volokitin has a more enterprising take on the position. He seeks
to prove his bishop is an asset rather than a liability by combining its power
with that of his queen.
13 ... dxc4!?
It’s rather remarkable to offer a piece sacrifice when you only have a
queen and bishop ready to attack!

Question: What is the point behind this sacrifice?

14 Bxd7 Qh5!
 Answer: This is Black’s idea. Besides the direct threat to f3, two other
tactical features give Black a stronger initiative than is apparent. Firstly, the
plight of the white bishop on d7. As things stand, when it is attacked by a
black rook, it will have no safe retreat squares. Furthermore, as we shall see
in the note to the next move, in a critical line attacking the bishop will be a
way to ferry a rook into the attack on f3 with gain of time.
The second tactical resource for Black is the pawn on c4. It didn’t just
clear the way for the queen to h5, it is also ready to support the move ... Bd3
should the bishop need a safe square or Black wish to regain the exchange
with a subsequent ... Bxf1.
15 g4!
A clumsy-looking move, but he must break the pin on f3.

Question: How should Black bring up reinforcements after 15 Kg2 - ?

Answer: At first it looks like 15 Kg2 would solve White’s problems by

defending f3 and supporting the retreat Bh3. But as indicated above, after 15
... Red8 16 Bh3 (White has to try 16 g4, though 16 ... Qd5 would regain the
piece with an excellent game) 16 ... Rd6! there is no good way to stop 17 ...
Rf6 with a fatal strengthening of the pin on f3. If 17 g4 then 17 ... Qh4 and 18
... Rf6 follows.
15 ... Qh3

Question: What is the only way for White to defend now?

Answer: 16 Ne1!
Of course 16 Ng5 allows mate in one, while 16 Bxe8 Qxg4+ is also lethal
to the white king. Grischuk finds the only move, saving his knight, guarding
g2, and taking some of the sting out of Black’s next move.
16 ... Bd3!
Black has to carry on dynamically. 16 ... Red8? 17 bxc4 Rxd7 18 f3
leaves him with a lost position.
17 Nxd3 cxd3 18 Qxd3 Rad8! 19 Bxc6!
Selling the bishop’s life as dearly as possible. After 19 Bxe8 Black can
(and must) force perpetual with 19 ... Qxg4+ 20 Kh1 Qf3+, leaving honours
equal between attack and defence.
19 ... bxc6
Of course not 19 ... Qxg4+?? 20 Bg2.
20 Qe4
The queen defends g4 and gets out of the way of a pin on the d-file.
Grischuk still has an extra pawn, but it isn’t easy to hold everything together.
20 ... c5
Over the next few moves Black’s initiative is aimed at wearing down
White’s pawn structure in the centre and on the kingside. As we shall see, the
looseness of the white king’s defences aids this plan.
21 Rad1 h5 22 gxh5 Qxh5 23 f4?
This and his next move cost Grischuk the game. Instead, 23 f3! would
keep the black queen out, though Black has ample compensation for the pawn
after 23 ... Rd6 etc.
23 ... Qe2!

There is a Russian saying that “one man in a field isn’t an army”. Clearly
they never met a lady like Black’s queen on e2. She not only hits the bishop
but also creates a pin on the e3-pawn. There is no good way to evict her as 24
Rf2? drops the rook on d1 and 24 Qg2 loses the e3-pawn.
24 Ba1?
My computer suggests rather desperate measures to evict the black queen:
24 Bc1! cxd4 25 Rd2, when 25 ... Qxe3+ 26 Qxe3 dxe3 27 Rxd8 Rxd8 28.
Bxe3 leads to a fairly equal endgame, while 25 ... Qa6 26 exd4 e6 gives
Black good play for the pawn but no more than that.
24 ... cxd4 25 Rde1
Grischuk may have overlooked that after 25 Bxd4 Bxd4 26 Rxd4 Rxd4
27 Qxd4 Black has 27 ... Rc8!, threatening to win at once with 28 ... Rc2,
mating the white king. After 28 Qd1 (what else?) 28 ... Qxe3+ White faces a
horrendous defence with broken pawns and an exposed king.
25 ... Qg4+ 26 Qg2
After 26 Kh1 d3 Black suddenly has a strong passed pawn. The exchange
of queens finally removes any danger to White’s king, but it leaves him in a
pawn down in the endgame.
26 ... Qxg2+ 27 Kxg2 dxe3 28 Re2
Or 28 Rxe3 Rd2+, winning the a2-pawn, as 29 Rf2? would drop the
bishop after 29 ... Rxf2+.
28 ... Bxa1 29 Rxa1 e5!
Ensuring White can’t regain his pawn.
30 fxe5 Rxe5 31 Rae1 Kg7 32 Kg3 Rde8
 33 h3
After 33 Kf3 Black can arrange to swap his e3-pawn for the h2-pawn: 33
... Rh8 34 Kg3 Reh5! (planning 35 ... Rh3+ to win the h2-pawn) 35 Rxe3
Rxh2 and the kingside passed pawns will win the day for Black.
33 ... f5 34 Kf3 Rh8 35 Kg3 g5!
Now 36 Rxe3 would lose to a fork on f4.
36 Rh2 f4+ 37 Kf3 Kf6 38 Rc1 Kf5 0-1
One threat is 39 ... Rxh3+! 40 Rxh3 g4+ 41 Kg2 gxh3+ 42 Kxh3, when
the way is clear for Black to advance the passed pawns.

Warding off Danger With a Positional Sacrifice

Sometimes you can with complete justification turn down the chance to
sacrifice. Maybe you are uncomfortable giving up material, or your style is
better suited to exploiting small advantages than committing yourself to a
dynamic fight. Especially in time pressure it can be best to avoid a situation
where every move you make might instantly win or lose the game.
But sometimes there is no quiet and promising path to follow: the
position demands that you make a sacrifice. Vigorous and bold play is no
longer a luxury but a necessity to get you out of an awkward spot.

Game 64
W.So-D.Sadzikowski
Gibraltar 2019

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 e5 6 Ndb5 d6 7 Nd5 Nxd5

Black has no choice in view of the threatened check on c7.
8 exd5 Nb8 9 a4
Stopping Black expanding with 9 ... a6 and 10 ... b5. This method of play
with 7 Nd5 against the Sveshnikov came into prominence when Caruana tried
it in his 2018 world championship match with Carlsen. It differs from the
standard main line in that White makes no attempt to gain control of d5, a
hole in front of the backward d6-pawn, and occupy it with a piece. Instead, he
is happy to have the square filled by a white pawn. This means Black no
longer has to worry about the structural weakness. On the other hand,
White’s 4-3 majority on the queenside could become dangerous if he
manages to engineer a c2-c4-c5 advance.
9 ... Be7 10 Be2 0-0 11 0-0 Nd7 12 Bd2 f5
White has his queenside majority, but that means Black has an extra pawn
on the kingside. It is unfettered by a pawn on e4 so can advance immediately.
13 a5
 The reasoning behind this second advance of the a-pawn is that Black
will want to push the knight back from its commanding position on b5. A
typical line is 13 ... a6 14 Na3 e4 15 Nc4 Ne5 (if Black keeps his knight on
d7 he can’t use it in kingside operations, nor indeed develop his queenside)
16 Nb6 (the white knight reaches an excellent outpost deep in enemy
territory) 16 ... Rb8 17 f4! (stopping the black pawns expanding further with
17 ... f4, when they start to look very menacing as 18 ... f3 is looming) 17 ...
exf3 18 Bxf3 Bf6. We are following P.Svidler-F.Caruana, Grenke Classic,
Karlsruhe/Baden Baden 2019. (Having studied this variation in massive
depth and played it as White in his world championship match, Caruana
decided to give it a go as Black!) After 19 c3 g5!? the position was
imbalanced. White has a bind on the queenside, but Black’s dynamic chances
on the kingside shouldn’t be underestimated. Indeed, I suspect most club
players would rather have Black here.
13 ... f4
Sadzikowski decides he can get by without 13 ... a6.
14 Re1 Nf6 15 Bd3
 Deterring 15 ... e4. It leaves d5 en prise but, as we shall see, it is too risky
for Black to take the pawn.
15 ... f3?
At first 15 ... Nxd5? 16 Bc4 Be6 17 Qf3! (17 Bc3 is also good) 17 ... Nc7
18 Nxc7 Bxc4 19 Nxa8 Qxa8 doesn’t appear too bad for Black with his big
centre, the bishop pair and a pawn for the exchange, but 20 a6! is
unexpectedly strong: if 20 ... Bxa6? then 21 Qd5+ Kh8 22 b4 Qe8 23 c4 and
the threat of 24 b5 wins a piece, so he has to try 20 ... bxa6 or 20 ... b6, but
then 21 Qxa8 Rxa8 22 Bxf4 exploits the pin on e5 to cut the black centre
down to size.
Instead, 15 ... Ng4!? gives Black counterplay; e.g. 16 f3 Nh6 17 c4 Bh4
18 Rf1 Nf5 and so on.

Question: We have seen pawns rammed with great effect into the
enemy structure in the previous chapter. Can you see how Wesley So
used the chance of a positional sacrifice to neutralize the danger?
 Answer: 16 Qxf3!
Only by sacrificing his queen will So be able to justify his play. After 16
g3 a6 17 Nc3 Qe8 and 18 ... Qh5 White’s king would be in great danger. 16
gxf3 Nh5 is also unsatisfactory for him, as Black can assail his weakened
defences with moves like ... Bg5 and ... Nf4.
16 ... Ng4 17 Qe4
The only way to curtail Black’s pressure against f2 is by threatening h7.
17 ... Bf5

Question: What must we do now?

Answer: 18 Qxf5! Rxf5 19 Bxf5

The queen sacrifice has deprived Black of three assets of his attack: the
rook on f8, the light-squared bishop, and the pawn on f3 which was its
spearhead. Meanwhile, White hasn’t had to weaken his king’s defences.
Years ago I listed three conditions your position should fulfil before you
contemplate a positional queen sacrifice: firstly, your king has to be safe
from attack; secondly, your pawn structure has to be compact and free of
weaknesses; and thirdly, your opponent mustn’t have a passed pawn. Here
White is untroubled by any of these shortcomings.
19 ... a6
This leads to an opening of lines which proves unfavourable for Black.
But otherwise his pieces remain passive; for example, 19 ... Nf6 20 Be3 Qb8
(after 20 ... Kh8 21 c4 g6 22 Bh3 or 20 ... a6 21 Bb6 Qf8 22 Nc7 Rb8 23 Ra3
White maintains a bind) 21 c4 a6 22 Nc3 and Black can only wait for White
to build up an attack, probably with an eventual c4-c5 breakthrough.
Incidentally, these variations show how much Black misses his light-squared
bishop: there is no way to create counterplay in its absence.
 Question: What should White do before capturing on g4?

Answer: 20 Nxd6!
A desperado sacrifice to destroy the base of Black’s centre before taking
the knight.
20 ... Qxd6 21 Bxg4 Qxd5
Sadzikowski must have hoped that freeing his queen from confinement
behind its own pawn structure and getting rid of the strong white pawn on d5
would benefit him. However, lines have also been opened for the white
pieces and two targets have become accessible to them: the pawn on e5 and –
even more seriously – the pawn on b7. So’s next two moves tie down black
pieces to their defence.
22 Bc3 Bf6 23 Bf3 Qb5
 Question: What is wrong with the centralizing move 24 Rad1 here?

24 Re4!
A key moment. Putting the rook on e4 is vital prophylaxis (or preventive
play) against the advance of the black e-pawn.
Answer: It would be very easy to play the routine 24 Rad1?, which looks
great as it seizes the d-file. But then Black would have 24 ... e4!, which
allows him to exchange off his passive bishop for the white one guarding the
a5- and b2-pawns; i.e. 25 Bxe4 (25 Bxf6 exf3 26 Bc3 Re8 sees the black
pawn reincarnated on f3 to intimidate the white king) 25 ... Bxc3 26 bxc3
Qxa5 and White’s queenside is broken up.
In the game Black is unable to disrupt White’s rock-solid structure.
24 ... Rd8 25 g3
Making a hole for his king on g2 before pressing on with his winning
plan, which will focus on Black’s weakness on b7.
25 ... Qc5 26 Raa4!
 The rook will infiltrate via the c4-square.
26 ... Rd7
After 26 ... Rc8 there is the trick 27 Rec4! (rather than 27 Rac4, when 27
... Qf8 resists) 27 ... Qxc4 (or 27 ... Qf8 28 Bxb7) 28 Rxc4 Rxc4 29 Bd5+
(the point) 29 ... Kf8 30 Bxc4, which is hopeless for Black.
27 Rac4 Qe7 28 Rc8+ Rd8
Black is pulverized after 28 ... Kf7 29 Bb4 Qe6 30 Bg4 Qd5 31 Re3, as
he loses the exchange to start with. But now the b7-pawn will drop sooner or
later.
29 Rec4 e4
This advance is too late to make any difference.
30 Bxe4 Bxc3 31 Rxc3

31 ... g6
If 31 ... Qxe4 then 32 Rxd8+ Kf7 33 Rc7+ carves Black up, while 31 ...
Rxc8 32 Bd5+! (a deadly zwischenzug which wins with a lot less effort than
32 Rxc8+ Kf7) 32 ... Kf8 33 Rxc8+ wins the black queen.
32 Bxb7!
 The key pawn drops without Black even being able to liquidate the white
pawn on a5. The rest is simple for Wesley So.
32 ... Qe1+ 33 Kg2 Rxc8 34 Rxc8+ Kg7 35 c3 Qe2 36 b4 h5 37 Rc7+
Kh6 38 h4 g5 39 Bf3 Qb5 40 Rc5 1-0
To add to his woes on the queenside Black loses both his kingside pawns
after 41 Rxg5 and 42 Rxh5+.

Sacrifices for Dynamic Equality

Here is one grandmaster’s opinion of the Petroff Defence: “This is a boring,
turgid opening line used at the highest level to neutralize the advantage of the
first move. Unfortunately, you will also occasionally meet it in your games,
as some ordinary players like to copy the play of the great masters, even at
the cost of enjoying their chess.”
Who wrote these intemperate words? Was it Steinitz or Tarrasch on a bad
day? I have to confess it was yours truly in a book I wrote some years ago.
Since I have played 1 e4 d5 2 exd5 Qxd5 3 Nc3 Qd8 as Black, and 1 e4 e6 2
d4 d5 3 Nd2 dxe4 4 Nxe4 Bd7 5 Nf3 Bc6 has been one of my favourite
openings, it seems a bit curmudgeonly (to say the least) to accuse Petroff
players of not deserving any pleasure from their chess. I enjoy my chess (or I
think I do) while devoting my efforts at the start of the game to neutralizing
White’s attacking chances. (My only excuse is that I wrote the criticism in a 1
e4 repertoire book and I went too far in cheerleading White’s chances. We
live and learn.)
As Caruana has shown in recent years, the Petroff can be a valid winning
attempt for Black. The purpose of such unobtrusive openings is to get a
playable middlegame while taking the emphasis away from theoretical
knowledge. In the present game Caruana shows that it can also be used for
counterattack if White chooses an aggressive set-up against it.

Game 65
R.Robson-F.Caruana
US Championship, St. Louis 2018

1 e4 e5 2 Nf3 Nf6 3 Nxe5 d6 4 Nf3 Nxe4 5 Nc3

For a long time 5 d4 d5 6 Bd3 Nc6 (or 6 ... Be7, while 6 ... Bd6 was
played in Paravyan-Golubov above) 7 0-0 Be7 8 c4 was the standard way to
take on the Petroff. Then White discovered that he didn’t have to castle
kingside and play a symmetrical or IQP centre: he could castle queenside and
attack.
5 ... Nxc3 6 dxc3
This is the idea. Taking away from the centre clears a path for the bishop
on c1 to go to e3 or f4, which facilitates the plan of queenside castling.
6 ... Be7 7 Be3 0-0

Sometimes Black arranges to castle queenside out of respect for White’s

potential kingside attack; e.g. 7 ... Nc6 8 Qd2 Be6 9 0-0-0 Qd7, when after 10
Kb1 a6 (stopping 11 Bb5) Black is ready for 11 ... 0-0-0. Note that 9 ... Bxa2
would be no free lunch as 10 b3 shuts in the bishop, but nor is it a trivial loss
for Black. I remember spending one sunny afternoon in Montenegro looking
at the consequences of 10 ... a5 11 Kb2 a4 12 Ra1 axb3 13 cxb3 Bxb3 14
Rxa8 Qxa8 15 Kxb3 Qa1 16 Qb2!, when White should win. If you study an
opening line you shouldn’t just look at the moves played by masters.
Otherwise, using the line above as an example, you might reach the position
after 8 ... Be6 in a real game and think: “Hmmm, I don’t remember Black
being able to take my pawn on a2 if I castle queenside! Have I forgotten the
theory? Maybe after 8 ... Be6 I’m not meant to castle queenside?!”
It’s a good idea, as you work through your opening lines, to ask yourself
if there are any moves not mentioned in your book (or not among the top
choices in your database) which would surprise or upset you if you faced
them for the first time in a game.
8 Qd2 Nd7 9 0-0-0 c6 10 Kb1
White has more aggressive ideas in 10 h4!? or 10 Bd3!?.
10 ... d5

Black’s basic plan is to use centralization to negate White’s attacking

chances. Thus, after 11 h4 Re8 12 Bd3, he can challenge the bishop on e3
with 12 ... Bc5. Or if 11 Bd3, he could play to exchange off the other bishop
with 11 ... Nc5.
As Caruana is always brilliantly prepared in the opening, Robson is
understandably reluctant to try to mate him right from the start. He therefore
chooses to apply mild pressure to the black centre.
11 c4 Nb6!
Already Black is thinking about a pawn sacrifice, as White would have a
pleasant and safe edge after 11 ... dxc4 12 Bxc4.
12 cxd5
After 12 Bxb6 Qxb6 13 cxd5 Rd8 14 Bc4 Bf6 Black’s dark-squared
bishop is king of the road (or at least of the diagonal heading to the vital b2-
square). White’s pawn on d5 will soon drop; for example, 15 Qc1 (if 15 Bb3
then 15 ... Rxd5! exploits the weakness of the b2-square, and 15 c3 Bf5+ 16
Ka1 cxd5 isn’t pleasant for White either seeing that 17 Bxd5? allows the
similar theme 17 ... Be6 18 c4 Bxd5 19 cxd5 Rxd5!) 15 ... Qc5 16 Bb3 cxd5
and White has no compensation for conceding the two bishops.
12 ... Nxd5 13 Bc4 Bf5!?

Caruana gambits the d5-pawn to reach a position of dynamic equality. He

proves the naysayers wrong when they claim the Petroff is a boring opening,
whereas they would have a case after 13 ... Be6 14 Rhe1, when White is fully
centralized and can probe for an edge in what is a statically equal (and dull)
position.
14 Bxd5!
Capablanca gave some great advice when he said that every move, no
matter how obvious, should be checked before playing it. Here the natural 14
Rhe1 (to complete development) runs into 14 ... Bb4! and White suddenly
loses the exchange. Therefore Robson accepts the pawn offer.
14 ... cxd5 15 Qxd5 Qc8
Of course exchanging queens isn’t part of Caruana’s plan. Instead, he hits
the vulnerable c2-square.
16 Nd4 Bg6 17 Ka1
A later game continued 17 h4 Rd8 18 Qb3 (more active is 18 Qf3! – and
safer for the queen) 18 ... h6 19 h5 Bh7 20 Rd2 a5 21 a4 Bb4 22 Re2
(Ga.Papp-I.Sipos, Hungarian League 2018) and now 22 ... Be4! (threatening
23 ... Bd5 24 Qd3 Bc4, winning the exchange) would have given Black a
strong initiative.
17 ... Re8!
Much more creative than the obvious 17 ... Rd8, when White could reply
18 Qb5. Caruana wants to attack the white queen with the rook on e5 so that
she is forced to retreat back to b3.
18 Rhe1 Bf6 19 c3 Re5 20 Qb3 a6 21 Bf4 Rxe1 22 Rxe1 Qd7
White is still a pawn up and there is no attack on his king in sight. But if
you think he is better you just have to ask yourself: how is he going to queen
his extra pawn? Black has two fine bishops and there are no structural
weaknesses in his camp. If you try to promote the c-pawn (the most likely
candidate) the knight on d4 will lose its support, much to the pleasure of the
bishop on f6.
 As a first step Robson decides to neutralize the pressure on d4, but he
runs into an awkward pin.
23 Be5 Re8
Not 23 ... Bxe5? 24 Rxe5, when White suddenly has a great position.
Black needs to keep the bishop pair to maintain the dynamic balance.
24 f4 Bd8!
The situation on the e-file is now awkward for White as ... f7-f6 would
win a piece if Black breaks the pin on f7.
25 a4
Preventing the space-gaining 25 ... b5. Nonetheless, the loosening of the
white king’s defences gives Black another potential target to add to his
pressure down the e-file.
25 ... h6 26 Rd1 Qg4 27 Rd2 b5!?
 Here 27 ... Be4 was safe, but offering a second pawn was an excellent
practical decision.
28 axb5 axb5 29 Qd1?
A computer is happy to defend the white position after 29 Qxb5! f6 (or 29
... Re7 30 b3!, stopping mate on the a-file) 30 Bd6, when 30 ... Re1+ (or 30 ...
Bf7 31 b3!) 31 Ka2 Bb1+ 32 Kb3 holds on. It will only take one double
attack by the queen, or one quiet but deadly move by a bishop, for Black’s
attack to crash through ... but in fact there is no such move!
It’s no surprise Robson avoided this, as the white pieces are scattered
(with the rook and bishop undefended) and the king is sitting in front of his
defences rather than behind them. Perhaps he was already in time pressure.
Caruana took a risk with 27 ... b5, but it was a well-judged one. He
played the percentages and was rewarded. No doubt his reputation also
played a role: would you want to expose your king against a world
championship challenger if he pushed 27 ... b5 with a confident flick of his
hand?
29 ... Qd7!
 Once again Black declines the exchange of queens, this time with a threat
of mate in two on the a-file.
30 f5?
It’s tempting to shut out the bishop, but 30 b4!, allowing the rook to
defend along the second rank, was the way to stay in the game.

Question: Can you work out the best reply to White’s move?

Answer: 30 ... Bg5!

It’s only equal after 30 ... Bxf5? 31 Nxf5 Qxf5 32 Rxd8 Qxe5, while 30
... Rxe5? 31 fxg6 is bad for Black as 31 ... fxg6? allows 32 Nf3! Qa7+ 33
Kb1, when Black has two pieces hanging.
31 Rd3
He loses the exchange after 31 fxg6 Bxd2 32 Qxd2 Rxe5.
31 ... Bxf5!
31 ... Rxe5 32 fxg6 is still good for White.
32 Nxf5 Qxf5 33 Bg3
 The upshot of the exchanges in the centre is that the black queen and rook
can force the white king into a fatal pin.
33 ... Ra8+ 34 Kb1 Rd8 35 Kc2

Question: You’ve got him tied up, but how do you continue to attack
White’s position?

Answer: 35 ... b4!

As so often a pawn is needed to cap off an attack by the pieces. The threat
is 36 ... b3+, winning the rook.
36 cxb4

Question: Try to calculate the winning sequence.

Answer: 36 ... Rc8+ 37 Kb3 Qe6+ 38 Rd5

If 38 Ka3 then 38 ... Qa6+ 39 Kb3 (or 39 Qa4 Qxd3+) 39 ... Qc4+ 40 Ka3
Ra8+ wins.
 38 ... Rd8
The pin on the rook resumes, two ranks up the board.
39 Kc4 Qc6+ 0-1
White loses the rook, since 40 Rc5 Qe4+ would win his queen.

The course of the game might suggest that Caruana was already planning an
onslaught against the white king when he played 13 ... Bf5. In fact his aims
were more modest: his bishop pair would give him enough dynamism to
counter any attempt by White to utilize his extra pawn. It was only when
White pressed forward that he decided to gamble on an attack on the king
with 27 ... b5.
Chapter Ten
Deciding the Character of the Game in the
Opening
It is now time to consider how the opening phase will determine the typical
plans we have seen in the middlegame.
At the start of a game White has a small but definite advantage by virtue
of having the right to move first. In most cases he seeks to seize territory in
the centre by advancing pawns there (beginning 1 d4 or 1 e4) and supporting
them with an efficient development of his pieces.
In general Black has two possible responses to this land grab. They are
very different in style and spirit and will shape the nature of the ensuing
struggle in the early middlegame, indeed perhaps for the whole game. It
might be said that in the opening phase solidity and activity are at the
opposite ends of a sliding scale.

Plan A: Equalizing Space in the Centre

Firstly – let’s call it Plan A – Black can mirror White by advancing one of his
own pawns to the fourth rank and then guarding it with his other pawns and
pieces. Thus the game might begin 1 e4 e5 or 1 d4 d5. Black will achieve a
solid and safe standing in the centre but, unless White goes wrong, will never
quite equalize. The disadvantage of moving second will linger as a long-term
irritation, even if it never causes Black serious trouble. The game is likely to
be of a manoeuvring nature, though that doesn’t preclude an outbreak of
violence if the opportunity arises.
Here to illustrate Plan A are two examples, one from the Italian Game and
the other from the Ruy Lopez. In the first, White wins the manoeuvring
battle, while Black returns the compliment in the second game.

Game 66
M.Vachier Lagrave-D.Navara
Biel 2018

1 e4 e5 2 Nf3
 Instead, 2 f4 is the good old King’s Gambit, which looks for an
immediate and violent struggle. It might be said of all gambits after 1 e4 e5
that White is expecting too much too soon. He is in danger of handing over
the initiative to Black, or at least allowing the position to burn out too quickly
with advantage to neither player.
Such at least is an objective assessment. If you love gambit lines and
know them well, then feel free to ignore any doubters. Only reconsider your
approach if your results with your gambit lines start to dip or you feel
dissatisfied with the positions you are getting with them. In that case it might
be time to move on. But until then keep on winning your games against
unprepared opponents because you know the traps and middlegame schemes
better than they do.
2 ... Nc6 3 Bc4 Bc5 4 0-0 Nf6 5 d3 0-0 6 Re1 d6

Question: What positional factors allow White to claim a small edge

here?
 You might be wondering why this position isn’t dead equal. Black has
established himself in the centre, developed his pieces to sound squares,
castled his king into safety, and avoided any weaknesses in his pawn
structure. You can’t ask for more than that after six moves!
Going back to move two, White attacked e5 with 2 Nf3 and Black replied
2 ... Nc6. Of all the ways to defend the e5-pawn this was the most natural as
it developed the knight to a good centre square. But it meant that the pawn on
c7 was blocked from contributing, for at least a time, from any strategic plan
Black devised. In contrast, White hasn’t been obliged to defend his e4-pawn
with Nc3. Therefore he can play:
7 c3!
Answer: While Black’s pawn on c7 is inert, White can use his c-pawn to
control the d4-square. It supports a bid to advance in the centre with a well-
timed d3-d4, or maybe gain space on the queenside with b2-b4 (though White
would have to be sure that pushing the queenside pawns was worthwhile).
Black’s bishop on c5 is a useful target.
Therefore we can see that 2 ... Nc6 is probably the best move in the
circumstances but has the long-term disadvantage of making Black’s build-up
less flexible or pawn friendly than White’s. (Remember Black has done
nothing wrong: it’s all a consequence of White getting to move first.)
7 ... h6
Taking the g5-square away from White’s bishop.
8 Nbd2!
Putting the pawn on c3 takes that square away from the knight on b1.
Therefore it wouldn’t amount to much as a plan if there wasn’t some other
way to get the horse to a good square. In steps a manoeuvre celebrated in the
Ruy Lopez/Italian structure: Nbd2, Nf1 and Ng3 to get the knight to a strong
post on the kingside.
8 ... a5
Ruling out b2-b4 and providing the bishop with a haven on a7 out of the
way of the d3-d4 advance. With the same idea it is more usual for Black to
play the more modest 8 ... a6 with a typical sequence 9 Bb3 Ba7 10 Nf1 Ne7
11 Ng3 Ng6 12 h3 Re8 13 d4 c6 (or 13 ... Be6 to challenge the white bishop)
14 Bc2 Qc7, when Black is solidly placed. But David Navara has an
interesting style and is always looking to gain the maximum counterplay in a
position.
9 Nf1 Ne7
 Black responds in kind: he sends his own knight to the equally good g6-
square. At the same time he frees the c7-pawn for future action in the centre.
10 Bb3!
A discreet retreat which removes the bishop from the range of a ... d6-d5
break-out. It also deters Black for the moment from advancing further on the
queenside with ... a5-a4.
10 ... Ng6 11 d4
At last White presses forward in the centre.
11 ... Ba7 12 h3!
Preventing Black from playing 12 ... Bg4 which would put indirect
pressure on the d4-pawn.
You can see that both players are indulging in a mixture of manoeuvres
and preventive measures. They are trying to maximize the co-ordination of
their own pieces while keeping a close eye on the opponent’s attempt to do
the same thing.
12 ... Bd7
Here 12 ... c6 or 12 ... Re8 are steady moves in the style of 8 ... a6 lines
mentioned in the notes to 8 ... a5 above.
13 Ng3 a4
Driving the bishop to a more passive diagonal (though, jumping ahead, it
won’t stay passive for long!).
14 Bc2
Instead, 14 Bxa4 Bxa4 13 Qxa4 Bxd4 would amount to a good swap for
Black of a wing for a centre pawn.

14 ... Nh7?
Navara plays in the enterprising style which has brought him a rating well
over 2700. Here, however, the plan of putting the knight on g5 leads to
disaster. It might have worked against a lesser grandmaster, but it is playing
with fire against a superbrain like Vachier-Lagrave who loves to play
dynamically and make sacrifices. The solid move was 14 ... Re8, while 14 ...
b5 would be consistent with Black’s plan of advancing on the queenside.
15 Nf5
The knight takes up residence near the opponent’s defensive line. It
wasn’t too late for Black to play 15 ... Re8 with a defensible position, but as
we have already discussed in the notes to Nisipeanu-Radjabov in Chapter
Five: once you’ve said “A” you have to say “B”.
15 ... Ng5?
Black’s reasoning is as follows: “after the exchange on g5, I’ll be able to
play ... Nf4, putting my remaining knight on an excellent square defended by
both the e5- and g5-pawns. Then I can evict the white knight with ... g7-g6
and begin my own attack on White’s king.”
Optimism is a great quality for a chess player, but here Black has
overstepped the mark. Nevertheless, it will take some imaginative play by
MVL to show the fallacy of Black’s concept.
16 Nxg5 hxg5 17 Qh5!
White had to assess the impending knight offer before committing
himself to this queen foray.
17 ... Nf4 18 Bxf4 exf4
After 18 ... gxf4 19 dxe5 dxe5 20 Rad1 the pin on the d-file is fatal; e.g.
20 ... Qe8 21 Rxd7! Qxd7 22 Qg4 and the mate threat on g7 costs Black his
queen after 22 ... g6 23 Nh6+.
 Question: Now what move adds oomph to the white attack?

Answer: 19 h4!
The following piece sacrifice energizes White’s kingside pawns and
enfeebles the black pawns by disrupting their defensive role.
19 ... g6
Now Black’s king will face a huge attack, but the endgame after 19 ...
gxh4 20 Nxh4 Qf6 21 e5 Qh6 22 Qxh6 gxh6 23 exd6 cxd6 24 Re7 Rfd8 25
Bf5 was very ugly for him.
20 Qh6 gxf5 21 exf5
Threatening to win at once with 22 f6, followed by mate on g7 or h7.
21 ... f6 22 Qg6+ Kh8 23 hxg5 Bc6
Clearing the d7-square to allow the queen to defend along the second
rank. Instead, after 23 ... Rg8 24 Qh6 we have an epaulette mate.
24 Rad1!
Calling up the reinforcements with a mighty rook manoeuvre via d3 to
h3. Note that White had the luxury of calling it a draw by perpetual check,
either here or on the previous move, if there was something he didn’t like
about the position. This means that, as regards any risk of losing, he didn’t
need to calculate beyond the 22nd move when he made the sacrifice. As a
rule, if your instinct tells you a sacrifice is good and you see there is at least a
forced draw after a few moves, and maybe more, don’t bother looking any
further down that variation. Save time and energy by crossing bridges when
you come to them.
24 ... Qd7 25 Rd3 Qg7 26 Rh3+ Kg8 27 Re7!!
 A brilliant strengthening of the attack. Normally, the exchange of queens
weakens an attack, but here the black king will be at the mercy of White’s
rampaging rooks, strong pawn on g6, and bishop supporting things from a
distance.
Remember our previous comment: MVL didn’t have to see this
unexpected entrance by the rook way back on move 17. He only needed to
spot it before renouncing a draw by perpetual in favour of 24 Rad1.
27 ... Qxg6 28 fxg6 f5
Navara shuts the white bishop out as 28 ... Rae8 allows the pretty
combination 29 Rh8+! Kxh8 30 g7+ Kg8 31 Bh7+! Kxh7 32 gxf8Q+ and
White wins.
29 Rhh7 Rfe8 30 Reg7+ Kf8 31 Rxc7
Not so much to grab a pawn but to get the rook out of the range of offers
to exchange it, and with gain of time by threatening mate on h8.
31 ... Kg8 32 Bxf5
 The bishop rejoins the action. White now has four pawns for the piece as
well as an irresistible attack.
32 ... Bb6 33 Rcg7+ Kf8 34 Rf7+ Kg8 35 d5!
Another vigorous pawn stab which wins the d7-square for his rook.
35 ... Bxd5
If 35 ... Bb5 then 36 Rfg7+ Kf8 37 Rxb7 is crushing for White.
36 Rfg7+ Kf8 37 Rd7 Kg8 38 g7
Threatening mate on h8.
38 ... Bf7 39 g6 1-0
Vachier-Lagrave made epic use of his pawns.

Game 67
V.Topalov-R.Wojtaszek
Shamkir 2018

1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 0-0 Bc5 6 c3 0-0 7 d4 Ba7 8 Re1?!
A natural move, but it seems that Topalov should play more aggressively
if he wishes to preserve his opening advantage. This could, for example, be
done with 8 Bg5 – the pin is somewhat awkward for Black with his bishop on
a7 rather than its more usual e7-square in the Ruy Lopez.
8 ... b5 9 Bb3 d6 10 h3 Bb7 11 a4
It wasn’t too late to try the pin with 11 Bg5, which would rule out Black’s
reply in the game as 12 Bxf6 gxf6 would leave him with a broken kingside.
11 ... Ne7! 12 Bc2 Ng6 13 Na3 c6 14 Bd3 Re8 15 Nc2 h6

So far the players have been concerned with developing their pieces,
building a solid centre, getting their kings to safety and avoiding structural
weaknesses. Notice how Black played 11 ... Ne7 and 12 ... Ng6 to redeploy
the knight to a good post on the kingside. At the same time he cleared the c6-
square so that, after White attacked the b5-pawn with 11 a4 and 13 Na3, his
queenside structure could stand its ground with 13 ... c6 rather than be broken
up.
At the moment the manoeuvre of the knight to g6 is a minor positional
achievement for Black, but it will become of decisive importance later in the
game.
 16 dxe5?

Question: Why is this a poor decision, and how instead can White firm
up his centre?

Answer: White didn’t try to take advantage of Black’s bishop being on

a7, rather than e7, with 8 Bg5 or 11 Bg5. (Perhaps Topalov wasn’t very well
acquainted with the theory of this line and so didn’t want to play these sharp
moves?) Now after this insipid exchange we must regard Black’s opening as
a complete success as the bishop on a7 is much stronger than it would have
been on e7 or f8.
It was better to play 16 Be3. For example, 16 ... Bb6 (Black moves the
bishop to a safer square as he doesn’t want to risk it becoming the butt of a
combination – after the exchange a4xb5 and recapture a6xb5 – of the kind
Rxa7! ... Rxa7, d4xe5, when both a7 and f6 are hanging) 17 axb5 axb5 18
Rxa8 Bxa8 19 Qb1 and White can probe in this level position.
16 ... dxe5 17 Be3
White doesn’t want to leave the black bishop enjoying a powerful
diagonal, but after this exchange the black knight can land on f4, from which
it will be practically immovable.
17 ... Bxe3 18 Nxe3 Nf4
Here we should start examining the stages by which Wojtaszek exploits
his positional advantage (and some significant inaccuracies by White) to
eventually win the game. First of all, he puts the knight on a fine square and
gains time by attacking the white bishop.
19 Bc2 Qc7
 Next Black prepares to seize the open d-file and drive the white queen to
a passive square.
20 Nf5?
This leads to more trouble for White. He could try to hold things together
with 20 Nh2 Rad8 21 Qf3 c5 22 Nhg4. The queen is often a staunch defender
on f3 in the Ruy Lopez. In any case this was a better option than having her
stuck on the back rank.
20 ... Rad8 21 Qc1 Bc8
An excellent decision. Black plans to reroute the bishop to e6 where it is
more active than on b7. There is also the latent threat of ... Bxh3 in some
lines, giving up the bishop for the sake of a kingside attack (such a plan
would aim to follow up, after White’s capture g2xh3, with ... Qc8 and ...
Qxh3). Black doesn’t expect the white knight to remain on f5, but he is
pleasantly surprised.
22 Nh2?
For better or worse he had to avoid the exchange on f5 by retreating the
knight. After 22 Ng3 he is suffering unpleasant pressure due to Black’s
strong knight on f4, control of the d-file and aforementioned idea of ... Bxh3
in some lines. But allowing Black’s e-pawn to join the show is strategically
catastrophic.
Up until this point Wojtaszek’s thoughts have been about centralizing and
improving the co-ordination of his pieces in a general way: the knight going
to f4, the rook going from a8 to d8. But after seeing this move he must have
started planning more concretely.

Question: How can Black now get a grip on the d3-square?

Answer: 22 ... Bxf5 23 exf5 e4!

Black creates a base on d3 for his knight. It will be so dominant there that
White will have little choice but to remove it with Bxd3, when ... e4xd3 will
give Black a passed pawn in the centre.
24 axb5 axb5 25 Nf1 c5
The plan is to be able to support the future pawn on d3 with ... c5-c4,
when it becomes a protected passed pawn.
26 b3 Nd3
Forcing matters at once. 26 ... c4 was solid and good, but Wojtaszek sees
that White’s attempt to get counterplay with 29 c4 will rebound on him ... so
he doesn’t stop him playing it!
27 Bxd3 exd3
You can see how the different stages of Black’s build-up (solid
centralizing moves: ... Nf4, ... Qc7, ... Rad8, ... Bc8, followed by moves with
a specific aim: ... Bxf5, ... e5-e4, ... c6-c5 and ... Nd3) have been crowned
with the creation of a passed pawn.
28 Rxe8+ Rxe8 29 c4

White not only stops 29 ... c4 but also attempts to generate queenside
counterplay.

Question: Should you meet the attack on the b5-pawn by taking on c4,
blocking things up with 29 ... b4, or carrying on with your attack in the
centre?
 Answer: A lot of good players would be satisfied to block the queenside
with 29 ... b4 (29 ... bxc4 30 Qxc4 gives White some undeserved freedom),
so that they can focus on exploiting their passed pawn. But a top player is
always looking for the most incisive moves. He won’t give his opponent the
slightest breathing space if he can help it. And so Wojtaszek ignores the
attack on b5 and, without wasting a tempo, seeks to exploit his initiative.
29 ... Re2! 30 cxb5 Qe5!
First of all, the rook goes to the seventh rank, and then the queen comes
up to support it.
31 Ra8+ Kh7 32 Qc4 Qxf5!
It is vital to combine mating threats to the white king with support of the
passed pawn. After 32 ... d2 33 Rd8! White suddenly has everything covered.
33 f3 Rc2!
Not just ignoring White’s threat to f7, actually forcing him to carry it out.
34 Qxf7 Qg5!

White’s initiative is at a standstill and his king is defenceless.

 35 g4 Qh4 0-1
There’s no good way to stop a check on f2 followed by mate. After all the
fuss about the passed pawn the game was decided by a direct attack on the
white king. But remember it was the advance of the passed pawn that broke
the co-ordination of the white pieces: a vital prerequisite to a mating finale.

Plan B: Conceding Space for Counterplay

A second approach, Plan B, involves Black holding back his centre pawns
and declining, for at least some time, the chance to build a pawn structure
with its advanced point on d5 or e5. Instead, he will try to establish a durable
set-up, which allows him to co-ordinate his pieces despite the lack of space,
and seek to counterattack against the white centre. In this scenario it is usual
for Black to fianchetto a bishop, either on b7 (e.g. in the Queen’s Indian) or
g7 (as in the King’s Indian and related openings), as the bishop’s way out via
centre squares is blocked by pawns.
If Black plays in this style he must ensure that his opponent isn’t given a
complete free hand in the centre. He has to fight for key squares even if he
isn’t occupying them with pawns. In the first of the two Nimzo-
Indian/Queen’s Indian games below Black fails in this task and ends up in a
stranglehold. In the second game he strongly challenges White’s centre
control with blows from the queen’s wing.

Game 68
Ma.Carlsen-E.Hossain
Baku Olympiad 2016

1 e3!?
Carlsen likes to try every opening at least once. Apparently, this is called
Van’t Kruijs’ Opening after a Dutch player who tried it against Anderssen in
1851. As we shall see, it soon transposes to more commonplace opening
territory.
1 ... Nf6 2 Nf3 e6 3 c4 b6 4 Nc3 Bb7 5 d4 Bb4 6 Bd3
Now we have the familiar shape of a Nimzo-Indian structure. Black is
willing to sacrifice the bishop pair in order to give his opponent doubled
pawns and increase his own influence over the key e4-square. We might say
that the Nimzo-Indian is a positionally complex opening and so a good
choice if you are looking to win with Black, even if it lacks some of the
tactical fire of the Indian defences (the King’s Indian, Benoni, etc) which are
based on a fianchetto on g7. Here Black isn’t usually looking for a
counterattack against the white king: his thoughts are on restraining and then
wearing down the white centre.
6 ... Ne4 7 0-0
Offering a gambit. After, say, 7 ... Nxc3 8 bxc3 Bxc3 9 Rb1 d6 10 Qc2
Ba5 11 e4 we have exactly what Black is trying to prevent in the Nimzo-
Indian: a mobile white centre which has overrun the e4-square and is difficult
to restrain.
7 ... Bxc3 8 bxc3 0-0 9 Ne1!

Preparing to dislodge the black knight from e4 without any more ado.
9 ... c5?
A serious mistake. The whole point of Black’s opening is to fight for the
e4-square. As we saw in the note to the 7th move, White is even willing to
sacrifice a pawn to get control of it, and yet here Black hands it over for free!

Question: What vital preventive move has Black failed to play?

 Answer: 9 ... f5! is the only logical move, preventing the white centre
expanding with impunity. After 10 f3 Ng5 11 Qe2 Black can then consider
11 ... c5 as 12 e4? is premature due to 12 ... fxe4 13 fxe4 Rxf1+ 14 Kxf1 (if
14 Qxf1, the e4-pawn drops) 14 ... cxd4 15 cxd4 Qf6+ with active play for
Black. So White would have to build up with 12 Nc2; e.g. 12 ... Nc6 13 e4
fxe4 14 fxe4 Rxf1+ 15 Qxf1 d6, when Black could aim to block the white
centre with ... e6-e5.
10 f3 Nf6 11 e4 Ne8
Stopping White from pinning the knight, but 11 ... d6 12 Bg5 Nbd7 might
have been a better approach. In any case Black would remain with a passive
position.
12 d5!
Shutting in Black’s bishop, preventing the active development of the
queen’s knight to c6, and ruling out any future strike against the white centre
with ... d7-d5.
12 ... d6 13 Nc2 Nd7 14 Ne3
 The knight is excellently placed on e3 to support both defensive and
attacking measures. For example, it protects c4 a second time in case Black
tries ... Ba6 and ... Ne5.
14 ... Nef6 15 a4!
The first purpose of Carlsen’s move is to deter any attempt at counterplay
with ... b6-b5 and, in general, to increase his grip on the queenside. The
second is to give himself the option of a4-a5 at an opportune moment. And
thirdly, it clears the a2-square so that the queen’s rook can slide into battle
along the second rank.
15 ... a5
Putting a stop to any idea of a4-a5, though I believe Carlsen was always
planning his main attack on the kingside. In fact he was probably pleased to
see the queenside blocked up as now Black won’t have the slightest hope for
counterplay there.
16 Ra2 Qc7 17 f4
The world champion is satisfied with his precautions on the queenside
and his positional build-up (Ne3 and Ra2), and so switches to direct
aggression.
17 ... exd5
 18 exd5!
A highly instructive moment.

Question: Why does he recapture this way?

Answer: 18 cxd5? would keep White’s centre broad and mobile, but it
would release the stranglehold on the black pieces. With 18 ... c4! Hossain
would clear the c5-square for his knight. Then 19 Nxc4 Nxe4! breaks up the
white centre, while 19 Bb1 Rae8, intending ... Nc5, would besiege the e4-
pawn. White is no longer better after 18 cxd5?, whereas he is (strategically-
speaking) close to winning after 18 exd5!. If you want to master the art of
planning you have to handle your pawns well and avoid stereotyped
decisions.
In the game, Black has no freeing move on the queenside and there is no
white pawn on e4 to target.
18 ... Rfe8
 Question: After White’s previous move you might be thinking “we no
longer
have a pawn on e4. So where is our pawn storm going to come from?”
Have a further think about it!

Answer: 19 g4!
Carlsen immediately provides an answer. If you have a commanding
initiative you can afford to push all the pawns in front of your king. Hossain
has no way to exploit the open spaces. His pieces are going to be pushed
back, leaving his own king in mortal danger. (Even if Black could start a fire,
Carlsen’s rook on a2 is a fire engine ready to trundle over to quell the
flames.)
19 ... Nf8 20 g5 N6d7
You can see how the black knights have been denied centre squares by
the white pawns on d5, f4 and g5. Meanwhile, Black’s bishop is staring at a
brick wall. It’s no wonder White’s initiative grows stronger and stronger.
 21 h4 Rad8 22 h5 Bc8 23 Ng4 Re7 24 Rg2!
A breakthrough on the g-file begins to loom.
24 ... Kh8 25 Qf3 Rde8 26 Qg3 Rd8 27 Bd2 Rde8 28 f5!

From an aesthetic point of view it is annoying to give the black pieces

even the slightest freedom. But giving up the e5-square is a small price to pay
for the destruction of the pawn cover around the black king.
28 ... Ne5 29 Nxe5 Rxe5 30 Bf4 Nd7 31 f6!
There’s no reason to settle for 31 Bxe5 Nxe5.
31 ... g6 32 hxg6 fxg6 33 Bxg6! 1-0
After 33 ... hxg6 34 Qh3+ Kg8 35 Rh2, mate can’t be averted for long;
e.g. 35 ... Nxf6 (or 35 ... Nf8 36 Qh8+ Kf7 37 Qg7 mate) 36 Qh8+ Kf7 37
Qxf6+ and mate next move.

Game 69
N.Radovanovic-Iv.Saric
European Championship, Skopje 2019
1 d4 Nf6 2 c4 e6 3 Nc3 Bb4 4 g3 0-0 5 Bg2
By avoiding Nf3 and making a quick fianchetto on the kingside White
obstructs the immediate ... b7-b6 and ... Bb7, which is the standard
development for Black in the Nimzo-Indian/Queen’s Indian complex. Players
who know everything about everything in the openings (e.g. Karjakin, So,
Anand, etc) are happy to play 5 ... d5 here, when 6 Nf3 dxc4 transposes into a
line of the Catalan where White has played an early Nc3, which is regarded
as slightly less promising than the main line where he can readily regain the
pawn on c4 with his queen. Thus the typical Catalan sequence is 1 d4 d5 2 c4
e6 3 Nf3 Nf6 4 g3 Be7 5 Bg2 0-0 6 0-0 dxc4 7 Qc2 a6 8 Qxc4 b5 9 Qc2 Bb7
10 Bd2 with an edge to White in an endlessly debated position.
It’s all well and good to play 5 ... d5 if you know something about the
Catalan, but what if you always play the Nimzo-Indian and don’t know
anything about 1 d4 d5 lines?
This encapsulates a frequent dilemma in the openings: should you play
the ‘best’ move even if it takes you away from your tried and tested opening
structure, or should you play something decent but suboptimal which keeps
you in more familiar territory. (Here’s another example: after 1 f4 I’ve a
feeling 1 ... e5 is the best response, but would you want to risk 2 e4 with the
King’s Gambit if you never answer 1 e4 with 1 ... e5 - ?)
During your opening preparation you need to take particular notice of
move orders which aim to trick you out of familiar lines. If you can find a
way to stay in familiar territory without compromising too much, then stick
to your guns. Otherwise you’ll have to learn a non-repertoire line specifically
for the rare occasions when you face a move order trick.
As we shall see, Saric has a line ready to keep it as a Nimzo-Indian.
5 ... Bxc3+
There is no point in delaying this exchange as White has no intention of
ever squandering a tempo provoking it with a2-a3.
6 bxc3 d6 7 Nf3 Nc6 8 Qd3
After 8 0-0 e5 Black gains space in the centre. The point of the rather
unusual game move is to answer 8 ... e5 with 9 Nd2, when White’s queen
remains guarding d4 and he has control over the e4-square.
8 ... Rb8!?
Saric avoids action in the centre in favour of a mysterious rook move on
the queenside.
9 Nd2 Na5 10 Nb3
 Question: Can you see a creative way of answering this move?

It looks as if Black’s queenside manoeuvres have misfired, as 10 ... Nxb3

11 axb3 strengthens the white centre, 10 ... c6 would leave the d6-pawn weak
after 11 Bf4, and 10 ... Nc6 just proves it’s all been a waste of time. But now
we see the point of Black’s play:
Answer: 10 ... b6! 11 Nxa5 bxa5

Question: What has Black gained through splitting up his pawns?

Answer: Black has accepted doubled pawns in order to activate his rook
on b8 and give his bishop access to the a6- and b7-squares. The a5-pawn
might appear weak but it is difficult to strike at, whereas the c4-pawn is easy
to target with ... Ba6 and ... d6-d5. The contrasting health of the pawns on a5
and c4 illustrates the adage that a weakness is only a weakness if it can be
attacked.
 We should also mention that Nimzowitsch would have applauded the idea
of ... Rb8 and ... Na5. Nowadays, the term “mysterious rook move” (which he
coined) tends to be used of any unusual or surprising move which puts a rook
on a blocked line. But Nimzowitsch meant more specifically placing the rook
on a blocked file to anticipate it being opened by the opponent, most usually
with an exchange of pawns. Here it was with the exchange of knights on a5.
12 c5!
Very sensibly getting rid of the weakling on c4 before 12 ... Ba6 can be
played. We’ve been applauding Black’s imaginative ideas, but the position is
no more than dynamically balanced, with White still retaining some of his
opening advantage. He has the bishop pair and an extra pawn in the centre (a
downside of Black allowing his b6-pawn to be deflected to a5).
12 ... dxc5 13 Ba3
The point of his previous move. White’s play hereabouts is excellent.
13 ... Bb7

14 e4!
Sacrificing a pawn to retain the bishop pair and keep up the pressure.
 If 14 Bxb7 Rxb7 15 Bxc5, then 15 ... Qd5! is an important zwischenzug
attacking h1. There is a lovely variation if White tries to throw a spanner in
the works with 16 e4?, namely 16 ... Nxe4 17 Bxf8 Nxf2! 18 Kxf2 Rb2+ 19
Ke3 (to stop the mating attack White has to give up the queen with 19 Qe2,
but of course that is also losing) 19 ... Qg5+ 20 Kf3 Qf6+ 21 Kg4 h5+ 22
Kxh5 (or 22 Kh3 g5! and there’s no good way to stop 23 ... g4 mate) 22 ...
Qh6+ 23 Kg4 Rf2! 24 h4 f5+ 25 Kh3 g5!! 26 Bxh6 g4 mate! I should confess
that this is all courtesy of the genius of the Stockfish computer program.
After 15 ... Qd5 White should settle for 16 f3, though Black then has time
to double rooks on the b-file with 16 ... Rfb8, when his pieces are well
organized.
14 ... Nd7 15 0-0 Qc8
Clearing d8 for the rook and planning ... Ba6. Over the next two moves
Radovanovic prudently evacuates his queen and rook on f1 from the potential
firing line of the black bishop.
16 Rfe1 Rd8 17 Qc2 cxd4 18 cxd4 Nb6
 Question: Would you choose safety with 19 Bc5 or adventure with 19
Bb2
if you had to decide?

19 Bb2
Answer: Here 19 Bc5, followed by 20 Rac1, looks a safe way to keep up
the pressure. Black has an extra pawn but his pieces, especially the queen,
remain boxed in. On the other hand, White is giving value to the black pieces
by launching an attack in their vicinity.
Instead, Radovanovic initiates the general scheme of Rac1, Ba1 (to get
the bishop out of the way of an attack by ... Na4 after his next move), Qc3
and d4-d5 to create threats against g7. This makes a lot of sense as he has a
powerful dark-squared bishop and Black’s kingside is lacking defenders – all
his pieces are bunched together on the other wing.
White’s plan is therefore strong but it has a practical drawback: it forces
him to play with great precision and vigour. Sometimes a higher-rated player
will create a tense situation, even if it is objectively unfavourable for him, in
the hope that the opponent will collapse under the pressure. Here Saric is
higher rated by 277 Elo points and evidently doesn’t want to draw, even with
Black. He has therefore tempted his opponent to avoid the safe and solid 19
Bc5 in favour of the more double-edged 19 Bb2. An elite player such as
Caruana would probably break through Saric’s defences and win in efficient
style, but it isn’t easy for lesser mortals.
If you reached the position after 18 ... Nb6, would you choose to
consolidate with 19 Bc5, giving you a risk-free edge, or play more boldly
with 19 Bb2, when you have more pressure but also more chance to be
outplayed? At such moments a number of factors – your style and
temperament, who you are playing, your tournament situation and how much
time you have on the clock – will form your decision.
19 ... Ba6
Black looks to gain counterplay through controlling the c4-square.
20 Rac1 Bc4 21 Ba1 a4
Instead, 21 ... c5!? is a non-human concept as it opens up lines for the
white pieces, but my computer says that Black is okay.
 Question: White’s dynamism depends on advancing d4-d5 in a
favourable
manner. Can you see a quiet move which undermines the pawn barrier
Black has erected against it?

22 a3?
Just the type of slow, inconsequential move that Saric was hoping for.
White wants to play his next move without dropping the a2-pawn, but there
was no need to be so methodical.
Answer: Instead, 22 Bh3! would involve the hitherto passive bishop on
g2 in White’s build-up. It pins the e6-pawn and introduces the threat of a
breakthrough with 23 d5. After 22 ... c6 23 Qb2! (aiming for 24 d5), here are
some variations:
a) 23 ... f6 24 Qb4 Bxa2 25 d5!, when Black has no good way to take on
d5; e.g. 25 ... Nxd5 26 Qa3 Nb4 27 Bc3 and White will win the knight or
bishop.
 b) 23 ... Nd5 24 Qd2 Nb6 (White has got his queen to a superior square
with gain of time, after which the central breakthrough is again killing) 25
d5! cxd5 26 Qd4 f6 27 exd5 Rxd5 (again Black has no adequate way to
recapture on d5) 28 Qf4 (eyeing the rook on b8) 28 ... Rc5 29 Bxe6+ Bxe6 30
Rxc5 Qxc5 31 Qxb8+ and White wins.
c) 23 ... Kh8 24 d5 Rg8 25 d6 is also pretty unpleasant for Black, but he
can continue to resist.
It’s not easy to notice the quiet move 22 Bh3 when all the attention has
been on the queenside, while 23 Qb2, putting the queen on a square where
she is exposed to a discovered attack, isn’t an obvious decision either. Still,
having asked yourself “how can I add power to the d4-d5 breakthrough?”, it
should be possible to find both moves. Of course the subsequent
complications aren’t simple to wade through, but White’s initiative flows
move by move once you have found the key moves to build it up.
22 ... Qd7 23 Qc3
Tartakower once described chess as “the tragedy of a single tempo”. After
23 Bh3 Qe7 the black queen has escaped from the pin, meaning that the e6-
pawn is a firm defender of the key d5-square again.
23 ... f6 24 Qa5 c6 25 Qc5 Bb5 26 e5?
 Question: Why is this a shocking decision, positionally speaking, and
what is Black’s best reply?

Answer: 26 e5 is a horrible move. White gives up on the idea of d4-d5

and leaves the d5-square as a gaping hole in his pawn structure. Instead, after
26 f4 or 26 Bh3 Kh8 27 f4, White is still dynamic and chances would remain
balanced.
26 ... f5
Just one move ago White’s dark-squared bishop had the potential to
become the best minor piece on the board. But now, in a terrible reversal of
fortune, it is shut in behind an immovable barrier on d4. Radovanovic should
hasten to give it a semblance of activity with 27 Bc3! Nd5 28 Ba5. Then it
would be difficult for Black to make any use of his extra pawn. Instead
comes another positional howler.
27 Bxc6?
 Question: Why does this move also deserve harsh criticism?

White would be just about okay after 27 Bc3, pulling himself back from
the brink. But if you have spoilt your pawn structure with the sole intention
of capturing a pawn, how hard it is to then change your mind about grabbing
it!
27 ... Bxc6 28 Qxc6 Nd5
Answer: The exchange of bishops on c6 has left White grievously weak
on the light squares. In particular, it is disastrous for him that the black knight
can’t be moved from its blockade square on d5. This point is the natural
expansion square for the white pawn structure – see, for example, the
variations with an explosive d4-d5! advance given at move 22 above.
Nimzowitsch has observed that: It is an enigmatic fact that the blockade
square tends to become a strong point for yourself and a weak one for the
opponent. (Note that Nimzo is optimistically assuming we’d be Black in the
present scenario!)
In other words, once you’ve restrained the opponent’s pawn from making
its natural advance, the square it would have advanced to usually becomes a
useful post for your pieces. A wise comment, though I’m not sure why this
should be particularly mysterious.
In the game White has a useless bishop and a weak pawn on a3, in
addition to – or perhaps we should say because of – his light square woes.
The rest runs smoothly for Saric.
29 Qa6 Rb6 30 Qa5 Rb3 31 Rc4
After 31 Qc5 Rdb8 Black can gradually prepare a move like ... R8b5 to
dislodge the white queen from the defence of a3.
31 ... Rdb8
Or 31 ... Rxa3 at once.
32 Rec1 Rxa3 33 Bc3 Rab3 34 Bd2 a3
 The white pieces are distracted by the need to hold back the advanced
passed pawn, which allows Saric to strike a deadly blow on the kingside.
35 Ra4 R3b5 36 Qa6 Rb2 37 Be1 Rb1 38 Rxb1 Rxb1 39 Qe2
After 39 Kf1 Nb6 40 Rxa3 Qc6! the threat of 41 ... Qh1+ is decisive.
39 ... Nc3 0-1
And 39 ... Qxa4 is not bad either. White’s last move suggests time
pressure. Indeed, perhaps he was already afflicted with it when he played 26
e5? as his play deteriorated sharply after a good start.

The Fianchetto on g7
Plan B appeals to many players – they want to make it harder for White to
keep the initiative he was granted by moving first, even if they are,
objectively speaking, allowing him to make more use of it with best play.
When there is a dynamic fight with tactical and strategic pitfalls it is much
easier for White to lose his way and hand over the initiative to Black. And
such a fight is more likely to occur when the position is imbalanced from the
first moves.
 A kingside or ‘Indian’ fianchetto is a very popular response to 1 d4. It
includes the King’s Indian (the main line being 1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4
e4 d6 5 Nf3 0-0), the Grünfeld (with one typical sequence 1 d4 Nf6 2 c4 g6 3
Nc3 d5 4 cxd5 Nxd5 5 e4 Nxc3 6 bxc3 Bg7 7 Nf3 c5) and the Benko Gambit
(which often goes 1 d4 Nf6 2 c4 c5 3 d5 b5 4 cxb5 a6 5 bxa6 g6 6 Nc3
Bxa6). But perhaps the most double-edged is the Benoni.
How much risk (tactical or strategic) are you prepared to take in the
opening? In my early days as a chess player I had problems playing White
against counterattacking openings such as the Benoni. I just couldn’t work
out how to deal with them with a sufficient mixture of restraint and
aggression. My position would crumble in the face of a queenside advance of
the ... b5-b4 kind. I remember sitting helplessly at the board, knowing that
something disastrous was going to happen but with no idea of how to prevent
it. That is often the case when players learn theory but don’t understand the
plans and motivations behind it.
I was sensible enough to switch to playing 1 Nf3 and 2 g3 to avoid the
whole issue. If you don’t put pawns on c4 and d5 they can’t be undermined
by the ... b7-b5 advance. Of course Karpov would put his pawns on c4 and d5
and win, but what was that to me? I lost.
After a period playing 1 Nf3 (incidentally, a much longer period than was
good for my chess development) I switched to 1 e4. It seemed a better move
than 1 d4 as I couldn’t get crushed by the Benko or King’s Indian etc. Of
course this was a false impression as you can be destroyed by the Najdorf, or
a counterattacking line in the Ruy Lopez, just as easily as by any opening
after 1 d4 if you don’t know what you’re doing. But by the time I started
playing 1 e4 I was more experienced and understood the ideas behind the
openings much better. Nonetheless, there is something to be said for a fresh
start with a new first move as White. It doesn’t have any of the residue of
failure lingering around it from when you were a less strong player. And,
speaking more technically, it could be that there are bad habits you have
picked up in the way you handle certain openings which won’t be transferred
to new openings.
When I saw the following game I was reminded of my own plight against
the Benoni. Buhr is a strong player and for a time makes good preventive
moves, but then some positionally unjustified exchanges. The result is that,
like me, he can only watch as his position collapses on the queenside.
Game 70
C.Buhr-V.Malakhov
European Cup, Kallithea 2008

1 d4 Nf6 2 c4 e6 3 Nc3 c5 4 d5 exd5 5 cxd5 d6

Getting to the Benoni via a Nimzo-Indian move order is a common
stratagem. The standard move order is 2 ... c5 3 d5 e6 4 Nc3 exd5 5 cxd5 d6
etc.
6 Nf3 g6
In the Benoni Black quickly challenges the white pawn on d4. When the
pawn (quite sensibly) advances to d5 he puts the bishop on g7 where it aims
right into White’s queenside. He hopes to increase the pressure on the
diagonal by advancing his queenside pawns: the 3-2 majority could
eventually become a passed pawn.
This all sounds great for Black, but the Benoni is often described as
strategically risky – and with good reason. He has conceded White a majority
of pawns in the centre and the chance to build up towards a central break with
e4-e5. Black must therefore ensure he keeps sufficient control over the e5-
square. But even if he succeeds in doing so, he’ll still be labouring under a
space disadvantage in the centre: White’s most advanced pawn is on d5, his
own most advanced pawn is on c5. Furthermore, the pawn on d6 is
performing a vital role, but it is also backward: both the c-pawn and e-pawn
have abandoned its defence in search of adventure.
If White is astute he will do his best to restrain Black’s queenside pawns.
More positively, he will either probe the weakness of d6 or build up and
maintain his pawn centre, or even carry out both plans.
7 Bf4
Immediately attacking the structurally suspect base of Black’s centre. The
purpose of this book is to discuss plans rather than theory, but I should
mention that a sharp and much debated position arises after 7 e4 Bg7 8 Be2
0-0 9 0-0 Re8 10 Nd2 Nbd7 11 a4 Ne5. If you are attracted by Black’s active
pieces then the Benoni could be the opening for you.
7 ... a6
 After 7 ... Bg7, 8 Qa4+!? can be a nuisance as the natural reply 8 ... Nbd7
drops the d6-pawn, while 8 ... Bd7 9 Qb3 leaves the d7-bishop sitting on the
square his knight wants. The game move allows Black to answer 8 Qa4+?!
with 8 ... b5, when 9 Nxb5 Bd7 pins the knight. In the long term, putting the
pawn on a6 prevents a raid by White on the d6-pawn with Nb5 and also
prepares to expand with ... b7-b5. So Black hasn’t really lost anything by
playing 7 ... a6 first, while he has gained by preventing the annoying queen
check on a4. That is the value of playing moves in the best possible order.
When you learn a new opening pay attention to so-called nuances in move
order. Don’t just automatically play a move like 7 ... Bg7 because you know
that’s where the bishop always goes.

Question: How should White respond to Black’s plan of grabbing space

on
the queenside?

Answer: 8 a4!
 A good preventive move. There is another sharp line after 8 e4!? b5 9
Qe2, which sets a nice trap: after the natural 9 ... Bg7? White has 10 Bxd6!,
the point being 10 ... Qxd6 11 e5 Qe7 (otherwise the discovered check on f6
wins) 12 d6 Qe6 13 Ng5 and Black is lost after 13 ... Qc4 14 Qf3, hitting a8
as well as f6, or 13 ... Qg4 14 f3! Qxg5 15 exf6+ Be6 16 fxg7 etc.
So Black probably does best to ‘change his mind’ and play 9 ... Be7!. He
doesn’t have his bishop where he wants it, but on the plus side his queenside
attack is going swimmingly with 10 ... b4, to push back the white knight, and
then ... Re8 in the air to target the e4-pawn (behind which the white queen
and king crouch awkwardly).
The importance of 9 ... Be7 reminds us once again that we can’t afford to
play according to a plan conceived before the game. You have to adapt your
strategy according to the specific circumstances which arise on the board. A
knowledge of theory will also help considerably by showing us how to avoid
tricks like 9 ... Bg7 10 Bxd6!.
8 ... Bg7 9 h3
 Question: What are the benefits of this little move?

If White wants to seize more space then 9 e4 was the time for it.
However, while a big centre has its pluses, it would also be subject to
pressure (after 9 ... 0-0) by moves such as 10 ... Bg4 and 10 ... Re8.
Answer: Instead, White chooses a restrained approach. He makes a hole
on h2 for the bishop (should it be attacked by ... Nh5) and deprives the black
pieces of the use of the g4-square. Thus a future manoeuvre with ... Ng4 and
... Ne5 is prevented, and Black can’t ease the congestion in his camp by
exchanging off a piece with ... Bg4 and ... Bxf3. Note that the white knight on
f3 is involved in the fight for the e5-square and, as we shall see, can be
manoeuvred to c4 to attack d6. Therefore Black is by no means reluctant to
exchange it for a bishop with limited scope. In fact White would most likely
answer ... Bg4 with Nd2, keeping the knight and allowing the exchange of
bishops instead after ... Bxe2.
9 ... 0-0 10 e3 Qe7
Readying himself to develop the queen’s knight to d7 without dropping
the d6-pawn.
11 Be2 Nbd7 12 Nd2
The knight heads for the strategically important c4-square.
12 ... Ne5
 13 Bxe5?
The positional jockeying would continue after 13 0-0 Rb8 14 Bh2 Ne8
etc. In the long term White would be aiming to advance in the centre while
doing his best to deter ... b7-b5 (if it became unstoppable he’d probably
throw in a4-a5 so that Black is left with a weak pawn on a6 in the event of ...
b7-b5 and a5xb6 en passant). For his part, besides ... b7-b5, Black might try
... f7-f5 to gain some control over the e4-square. He’ll be hoping that if the
white centre expands it will become overextended and implode.

Question: I hope that, having reached the last chapter of a book on

planning,
you can tell me why taking the knight is an obscene decision.

13 ... Qxe5 14 Nc4 Qe7

Answer: It’s nice to get the horse to c4, but giving up the dark-squared
bishop was too big a price to pay for it. You only have to compare White’s
passive bishop on e2 to its counterpart on g7 to realize the damage done by
his faulty 13th move. He has no plan to guide him as his dark squares would
become too weak if he pushed his pawns to e4 and f4 and tried to break with
e4-e5.
Well, at least the strong knight on c4 gives him some consolation. After
15 0-0 Rb8 16 a5!, Black doesn’t get a rolling pawn mass with ... b7-b5.
15 Nb6?
White compounds his 13th move error by exchanging off the knight. It’s
not as if Black’s bishop on c8 was attacking anything.
15 ... Rb8 16 Nxc8 Rfxc8
Black is even allowed to get his rook to a strong square with gain of time.
17 0-0

After the faulty exchange with 15 Nb6 and 16 Nxc8 there are no
complicating factors to obscure the difference in power between the two
bishops. Black’s bishop can act with the rooks to support a queenside pawn
advance. In contrast, White’s bishop can do nothing to bring his central pawn
majority to life.
 Question: How should Black energize his queenside pawns?

Answer: 17 ... c4!

Threatening 18 ... b5, when his queenside pawn front remains intact and
mobile. White has to ensure they are fractured after ... b7-b5 with his next
move.
18 a5 Ne4!
Finally unleashing the mighty bishop.
19 Nxe4 Qxe4 20 Bf3 Qe7 21 Ra2 b5
The moment has come to open the b-file.
22 axb6 Rxb6
Every single black piece – the queen, two rooks and bishop – can use
their power to attack the pawn on b2. If it falls Black will retain at least one
passed pawn.
23 Qe2 Qb7 24 Rc1 Rb4 25 Rc2
It’s never a good sign when the heavy pieces are all tied down to
defending a backward pawn.
25 ... Qb5 26 Qd2 a5!
Do not hurry: before winning the b2-pawn Malakhov advances his a-
pawn so it becomes a strong passed pawn.
27 Kh2 a4 28 Be2 Rb8
 This is what a perfectly co-ordinated chess army looks like.
29 g3 Bxb2 30 Bxc4
Hoping for simplification after 30 ... Rxc4 31 Rcxb2. Naturally, Black
keeps the bishops on the board.
30 ... Qb7! 31 Bf1 a3
A far-advanced passed pawn on the a-file often works well with the
bishop on g7 in ‘Indian’ set-ups. Here the bishop and a-pawn are mutually
defending each other and shut in the rook on a2. Black’s plan over the next
moves is to put his queen on the a-file to defend the passed pawn, and then,
with ... Be5 and ... Rb2, force White to exchange rooks on b2. The passed
pawn will be only one step from queening. At the same time Black will guard
against any white attempt to counterattack on the c-file or against d6.
32 Rc6 Rb6 33 Qc2 Qa7 34 Rc7 Qa5 35 Bg2 Be5 36 Rc8+ Kg7 37
Rxb8 Rxb8 38 Qc4 Rb2
The rook finally reaches b2.
39 Be4
 39 ... Qe1!

Question: Can you see a faster way to win?

If now 40 Rxa3 then 40 ... Qxf2+ wins at once. Remember how Black
won in Radovanovic-Saric above? When you have tied down the enemy
pieces on one side of the board, the key to winning is often to create a threat
on the other wing. The stressed defenders are unable to respond in time to the
change of front.
Answer: Nevertheless, although the concept is correct, the
implementation would have been faster after 39 ... Qd2! 40 Rxb2 Qxb2!,
when the pawn advances to a1.
40 Rxb2 axb2 41 f4
If 41 Qc2 then 41 ... f5! (even better than queening at once) 42 Bd3 b1Q
43 Qxb1 Qxf2+ 44 Kh1 Qxg3 wins.
41 ... Qf2+
Trivial was 41 ... Qd2+ 42 Kh1 (42 Bg2 lets the b-pawn queen) 42 ...
Qc1+ and White must resign.
42 Kh1 Bf6 43 Qc2 Qxg3 44 Qe2 Qxh3+ 45 Kg1 h5 46 Bd3 Qg3+ 47
Kf1 h4 48 Qf2 Qxf2+ 49 Kxf2 g5 0-1
White won’t be able to cope with passed pawns so far apart once Black
brings his king forward.
A fine positional win by Malakhov despite the rather long-winded finish.

Game 71
D.Navara-T.Pähtz
Gibraltar 2018

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 Nf3 0-0 6 h3 Na6 7 Bg5 c5 8 d5 e6 9

Bd3

9 ... Nb4?
Going against everything we have learnt about the need to co-ordinate the
action of the pieces.
The great 18th century French player Philidor claimed that the pieces
were the servants of the pawns. This is an exaggeration, but it is certainly true
that the big pieces have to play a supporting role when the position requires
the engineering of a pawn advance. The limelight goes to the pawns because
their expendability makes them plucky adventurers. They can be fearless
when used as a spearhead to undermine the opponent’s structure, as no harm
is done if they are exchanged for an enemy pawn. Therefore an individual
pawn can be both a hero and cannon fodder at the same time.
For these reasons Black’s 9 ... Nb4 is a senseless move. It looks
aggressive and threatens a favourable exchange on d3, but once the white
bishop moves out of the range the knight is out on a limb on b4. It does
nothing to help Black’s queenside build-up and is merely a target. It is too
valuable to sacrifice, too useless to help Black’s pawns.
In contrast, after 9 ... h6 10 Be3 Nc7 or the immediate 9 ... Nc7, the
knight can help support action on the queenside with ... b7-b5 (probably after
the exchange ... e6xd5 and recapture c4xd5; White might try to hinder this by
taking back with e4xd5, but then he has a smaller centre). The knight would
be safe from attack on c7 and not adrift from the other black pieces.
10 Be2 h6 11 Be3 exd5 12 cxd5 Re8 13 Nd2 Bd7 14 0-0 Rb8
 Black decides to advance ... b7-b5 despite the knight’s foray to b4, but
putting the rook on b8 is too stereotyped. A much better method would be 14
... Qb8!, which defends the d6-pawn and then, if allowed, 15 ... b5.

Question: How can you put your finger on Black’s main weakness?

Answer: 15 Bf4!
Exactly. The attack on d6 is more awkward with the black rook sitting on
b8.
15 ... Qe7
Another ‘obvious’ but dubious decision, as the queen is left awkwardly
placed on e7. It was better to go into deep defence with 15 ... Bf8.
16 Re1 b5
To stop Nc4. The d6-pawn is difficult to defend. Here is what might
happen if Black cleared e8 for his knight to help guard it: 16 ... Red8 17 a3
Na6 18 Nc4 Ne8 19 e5! (a central breakthrough to take advantage of the
black queen’s precarious post) 19 ... dxe5 20 Nxe5 and Black is being
squashed in the centre. 20 ... Bxe5 merely hands over the key dark-squared
bishop after 21 Bc4 (intending 22 Bxe5), as 21 ... f6 22 d6+ wins the queen.
17 a3 Na6 18 a4!
Navara wants to gain control of c4. With the knight on a6 blocking the
pawn move ... a7-a6 there is no way for Black to hold onto the key square.
18 ... Nc7
After 18 ... b4 19 Nb5 (another square from which to attack d6) 19 ...
Bxb5 White has the lovely choice between 20 Bxb5, winning the exchange,
or 20 axb5 Nc7 21 Rxa7.
19 axb5 Nxb5 20 Bxb5 Bxb5 21 Nxb5 Rxb5 22 Nc4

White’s opening play has ended in complete triumph.

22 ... Rb4
If 22 ... Nxe4, the only way to annotate 23 Nxd6 is “ouch!”
23 Qd3
There’s no need to get involved in tactics after 23 Nxd6 Rd8.
23 ... Rd8 24 Bd2 1-0
If 24 ... Rb5, 24 ... Rb7 or 24 ... Rbb8 then 25 Na5 and if necessary 26
Nc6 will win the exchange.

Solid Play and Then a Freeing Move

A lot of openings contain ideas which are a mixture of Plans A and B. Thus
in the Caro-Kann main line Black doesn’t maintain a pawn on d5, but neither
is he provoking a dynamic fight. He is aiming for solid development with a
view to freeing his game later on by dissolving the white centre. It is at that
later – and critical – point in the struggle that fireworks are possible.

Game 72
Ste.Mazur-A.Zahedifar
Ordu 2019

1 e4 c6
Black prepares an immediate challenge to the e4-pawn without (as is the
case in the French) blocking in the bishop on c8. In the main line White
maintains a space advantage, but Black is well entrenched with no
weaknesses.
2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bf5 5 Ng3 Bg6 6 Nf3 e6
WFM Anahita Zahedifar has developed her light-squared bishop to an
active square before closing the path from c8. She is on the way to achieving
an easy development, so White has to respond vigorously.
7 h4
The plan is to establish a space advantage by hounding the bishop on g6.
This move is taken for granted nowadays, as theory has taught players of
White that there is no need to worry about the apparent loosening of their
king’s future residence on the kingside: he is going to be whisked off to the
queenside.
7 ... h6
Here 7 ... Nd7 is more usual, but Black is willing to invite the white
knight to e5 and then undermine it.
8 Ne5
Accepting the offer. Also possible was 8 h5 Bh7 9 Bd3 Bxd3 10 Qxd3,
preparing to castle with 11 Bd2 and 12 0-0-0.
8 ... Bh7 9 Bd3
 If you play an opening you have to be aware of all the traps associated
with it.

Question: Why would taking the pawn on d4 now be a big mistake for
Black?

9 ... Bxd3
Answer: 9 ... Qxd4? looks like a useful extra pawn but is disastrous after
10 Nxf7!, when 10 ... Kxf7 11 Bg6+ wins the queen, while 10 ... Bxd3 11
Nxh8 is ruinous for Black.
10 Qxd3 Nd7 11 f4
Continuing his plan to gain space. Another approach was 11 Bf4 Nxe5 12
Bxe5, which would keep the bishop active.
11 ... Be7
A natural move; 11 ... Bb4+ will be examined in the next game.
12 Bd2 Ngf6
Sensible development, as White would have a strong initiative after 12 ...
Bxh4 13 0-0-0.
13 0-0-0 c5

Black’s standard freeing move. Everything else being equal it’s

appropriate to remove the d4-pawn to reduce the white centre and undermine
the knight on e5. The black horse on d7 will gain access to c5 or can be more
favourably exchanged on e5, while the bishop on e7 gets more scope.
Black plays it at once, as after 13 ... 0-0 White has the chance for 14 Qf3!,
evacuating his queen from the d3-square (why d3 can be an uncomfortable
post for the queen becomes apparent in the next game). But, jumping ahead,
this is how Black should have played.
A freeing move radically changes the pawn structure. A fixed or
restrained centre is converted into an open or a mobile one, requiring a re-
evaluation of the roles and effectiveness of pieces on both sides. If it is
sound, the player freeing his game has chances for equality or even the
initiative – such is the energy released by liberating the previously pent-up
pieces.
Both players therefore have to keep a constant watch for potential freeing
moves. Whether they are tactically and strategically watertight could well
decide the game. If you are contemplating a freeing move, ask yourself: has
my opponent gone wrong and let me equalize easily, or does he have a trick
up his sleeve?
In general, the pawn advance ... c6-c5 (or ... c7-c5) is a favourable idea as
long as:
1) White can’t respond with d4-d5 in an effective manner (that is, the d5-
square still has enough defenders, despite the removal of the barrier on c6,
and the defenders aren’t side-tracked from other duties).
2) White doesn’t have a pair of bishops, especially a light-squared one,
which can profit from the opening of lines, in particular the diagonal a8-h1.
3) The black king is castled or cannot otherwise be disturbed by a move
like Qb5+.
Point #2 doesn’t apply here, while an attempt to disturb the black king
with a check on b5 fails after 14 dxc5?! Nxe5. But White does have:
14 d5!
A powerful pawn sacrifice.
14 ... Nxd5
After 14 ... exd5 a gap appears in Black’s defences, allowing 15 Nf5!,
when the g7-pawn is hanging and White has a decisive initiative; e.g. 15 ... 0-
0 16 Qg3 Nh5 17 Nxh6+ Kh7 18 Qd3+ f5 (if 18 ... Kxh6 19 Ng4 mate) 19
Nef7 Qe8 20 Qxf5+ g6 21 Qxd5 etc. Greedy White is two pawns up with an
attack to boot.
Instead, 14 ... Nxe5 15 fxe5 c4! 16 Qf3 Nxd5 puts up a lot more
resistance, though 17 Nh5 keeps up the white attack.
15 Nh5!
And here we see a second breach in Black’s defensive wall caused by 14
d5.
15 ... Bf6
Dealing with the threat to g7, but Black doesn’t have time to consolidate
on the open d-file.
16 c4! N5b6 17 Qg3
 Now g7 hangs again, and after 17 ... 0-0 18 Nxd7 (also winning is 18
Ba5!, tying up the black pieces) 18 ... Nxd7 19 Bc3! (removing the defender
of g7) 19 ... Bxc3 20 bxc3 g6 21 Rxd7! Qxd7 22 Nf6+ White wins the queen.
17 ... Bxe5 18 fxe5 Qc7 19 Nxg7+ Kd8 20 Rhf1
With the black king stuck in the centre a breakthrough on f7 will decide
the game.
20 ... Kc8 21 Bc3 Rf8 22 Rxf7! Rxf7 23 Nxe6 1-0
It’s a pity that Black didn’t allow the elegant finish 23 ... Qc6 24 Qg8+
Rf8 25 Qxf8+! Nxf8 26 Rd8 mate.

In the example above it was the white pieces, rather than the black pieces,
which benefited from the opening of lines. Let’s see what happens if Black
plays in a more subtle style.

Game 73
V.Ivic-A.Predke
European Championship, Skopje 2019
1 e4 c6 2 d4 d5 3 Nd2 dxe4 4 Nxe4 Bf5 5 Ng3 Bg6 6 h4 h6 7 Nf3 e6 8 Ne5
Bh7 9 Bd3 Bxd3 10 Qxd3 Nd7 11 f4 Bb4+
This time Black gives a check before putting the bishop on e7. Cajoling
White into playing his next move doesn’t seem to make much difference, but
it will prove vital – such is the sophistication of modern opening systems.
12 c3 Be7 13 Bd2 Ngf6 14 0-0-0 c5
The freeing move which turned out to be a disaster in the previous game.
But this time it is White who has to be wary. Alexandr Predke, a rapidly
improving 25-year-old Russian grandmaster, had undoubtedly studied this
position with a computer. So why would he allow the next move if it is good
for White? But Ivic is fearless:
15 d5?!

Question: Can you see the riposte Black has against this move which
wouldn’t be available if White’s pawn were still on c2?
 Answer: 15 ... Nxe5! 16 fxe5 Qxd5!
This is a key resource. With the pawn on c3, rather than c2, the white
queen is hanging, which means there is no time for White to win a piece with
17 exf6.
17 Qb5+
Still we must remember #3 in our list of possible drawbacks to the ... c6-
c5 freeing move. Black must tread carefully after the check on b5.
17 ... Nd7!
After 17 ... Qd7 18 Qe2! Nd5 19 c4 White has the initiative.
18 Bxh6
This looks strong at first glance, as White wins a pawn back and threatens
both the black queen and g7, but Predke has everything under control.
18 ... a6! 19 Qa4
White loses material after 19 Rxd5 axb5 29 Bxg7 Rg8.
19 ... b5 20 Qf4 Qxa2
The only move, but it more than keeps Black in the game. The safety of
both kings is now an issue.
21 Bxg7 Rg8 22 Nh5 b4!
 Utilizing his resources. Black’s basic idea is to play ... b4xc3 to force
b3xc3, then ... c5-c4 and finally ... Ba3 mate.
23 Rhf1?
White cracks. He makes the fundamental mistake of attacking the black
king where it is, rather than where it is going to be. Therefore he loses a
tempo which is fatal in such a double-edged position.

Question: So where is the black king going to be in the future and how
can
we immediately plan action against him?

Answer: 23 Qe4! anticipates Black castling queenside. Then it seems the

mutual attacking chances balance out to a draw; e.g. 23 ... 0-0-0 (or 23 ...
Rd8, when 24 Qb1 Qa4 25 Qe4 Qa1+ 26 Qb1 is another way to draw) 24
Qc6+ Kb8 25 Nf6 (White should avoid 25 Rxd7 Rxd7 26 Qxd7 Qa1+ 27 Kc2
Qxh1 28 Qxe7 Qxg2+ etc) 25 ... Nxf6 26 Qb6+ Ka8 27 Qc6+ Ka7 28 Qc7+
Ka8 with perpetual check.
 23 ... 0-0-0!
Startling castling. With the white knight and bishop stranded on the
kingside, any let-up in the attack on the black king is bound to leave the
white king at the mercy of Black’s initiative.
24 Qe4
One move too late, and it makes all the difference. In a double-edged
battle every tempo matters. A lot of players would rather have an objectively
smaller advantage in a quiet position, where you can afford to make some
second best moves, to a bigger advantage in a dynamic struggle with the
attendant risk of being punished severely for one in-exactitude.
24 ... bxc3 25 bxc3
25 Qa8+ Nb8 is hopeless.
25 ... c4!

Black has just enough time to carry out the plan outlined above. The
threat of mate is crushing.
26 Qc6+ Kb8 27 Rxd7 Qa3+ 28 Kb1
Or 28 Kc2 Qb3+, while 28 Kd2 Rxd7+ 29 Qxd7 Rd8 wins the queen
while keeping an overwhelming attack.
28 ... Qb3+ 29 Kc1 Qxc3+ 30 Kb1 Rxd7 31 Qxd7 Qb3+ 32 Kc1 c3! 0-1
A neat finish. Here we see the attacking power of a queen supported by
an advanced pawn. The main threat is mate on a3 again as d2 is denied to the
white king, or if 33 Qxe7 then 33 ... Qb2+ 34 Kd1 Qd2 mate.

A Theory Arms-race
We saw in the Caro-Kann games above how theory becomes ever more
refined: 11 ... Be7 gets replaced by the nuance 11 ... Bb4+! 12 c3 Be7, when
the onus is on White to find a better idea. Some openings come with a huge
amount of theory attached, none more so than the Sicilian main lines (well,
maybe the Botvinnik System of the Semi-Slav is a contender). By all means
play these sharp lines if they suit your style, but remember that you’ll have to
do some deep preparation and be “standing on the shoulders of giants”, so to
speak: that is, having to follow long lines discovered by other players rather
than thinking for yourself (which might sound great until you remember that
your opponent might be standing on even higher shoulders!). Still, all the
effort feels worth it when you get to play some nice attacks with sacrifices, as
in the following game.

Game 74
N.Doghri-Z.Ilincic
Istanbul Olympiad 2000

1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Bc4

The Sozin Attack. The bishop aims itself immediately at the black
kingside.
6 ... e6 7 Bb3 b5
Black has made some positionally desirable pawn advances at the cost of
falling behind in development. He trusts that his apparently solid shell of
pawns on d6 and e6 will blunt White’s attacking aspirations. This will allow
Black to catch up with his development and then enjoy the strategic benefits
of having expanded on the queenside.
8 Qf3
 Another white piece springs into action with the first direct threat: 9 e5,
winning the exchange. More prosaically the queen lends a hand to the
defence of e4 and is ready to go to g3 where she will attack g7 once Black
has developed with ... Be7.
8 ... Qb6
The obvious way to meet White’s threat was 8 ... Bb7. But Black wants to
keep his bishop on c8 for the time being as it helps to bolster the e6-pawn –
yes, the possibility of a Nxe6 or Bxe6 sacrifice is already in both players’
minds. Therefore Ilincic prefers to attack the white knight on d4 and, after
White’s next move, retreat the queen to b7. He is falling even further behind
in development. On the other hand, if things work out for him his queen will
be well placed to attack e4 in conjunction with ... b5-b4, chasing away the
knight which defends it.
9 Be3
Note that 9 e5? fails to 9 ... Bb7.
Now Black has to move his queen again to avoid being splattered after 9
... Nbd7 10 Nxe6! (that’s one sacrifice we can have no doubts about!).
Therefore White’s bishop has got to e3 as a kind of ‘free developing move’.
But it isn’t all gain for White as, given a choice, he would probably have
preferred to put the bishop on g5 at some point, rather than on e3.
9 ... Qb7 10 0-0-0 Nbd7
Hunting the e4-pawn is too risky: 10 ... b4 11 Na4 Nxe4 12 Nb6! (forcing
the queen back into range of the bishop on e3) 12 ... Qxb6 13 Nxe6! Qb7 14
Nxf8 Rxf8 15 Bd5 and Black is busted.
11 Rhe1 Be7
In a later game Doghri built up an attack after 11 ... b4 12 Nd5! exd5 13
exd5, when he successfully exploited the open lines around the black king.
Ilincic’s move is much more solid. Indeed, it feels like White’s lead in
development hasn’t amounted to much and is being dissipated, after which
Black will gain the edge due to having the superior layout of pawns on the
queenside. But Doghri has one very important trick in his book.
12 Qg3!

12 ... b4?
It seems that Black has only to play 12 ... 0-0 to secure his king and have
a decent game, as 13 Bh6 can be met by 13 ... Nh5 (with a possible repetition
after 14 Qf3 Nhf6). However, there is a nasty surprise in store: 13 Bd5!!,
when 13 ... Nxd5 (13 ... exd5 14 Nf5 Ne8 15 Nxe7+ Kh8 16 Ncxd5 is
crushing) 14 Bh6! Bf6 (after 14 ... g6 15 exd5 either f8 or d5 falls) 15 Nxd5
exd5 (15 ... Be5 16 f4 is not much different) 16 Bxg7! Bh4 (or 16 ... Bxg7 17
Nf5 and mates) 17 Qxh4 Kxg7 18 Nf5+ Kh8 19 exd5 is catastrophic for
Black. There’s no good defence against threats such as 20 Qh6 Rg8 21 Re8!
planning to mate on g7, for if 19 ... Rg8 20 Re7 will break through on f7.
Therefore Black can’t escape his problems by castling. The 13 Bd5!! trick
is the sort of thing you need to know in advance before you play a sharp
opening. When I looked in my database there were no games in which 12 ...
0-0 was played. Anyone studying the Sozin without making the effort to
discover the refutation of plausible, but bad moves could find themselves
facing 12 ... 0-0 in a tournament and needing to find 13 Bd5 with their clock
ticking. Not an easy task. Of course, the chances are that you don’t play the
Sozin Attack and never intend to. Fair enough. But there are the equivalent of
natural, but bad, moves like 12 ... 0-0 in every opening which can only be
refuted by something difficult to see. Forewarned is forearmed: the more
work you do on your openings, the fewer nasty surprises you will face (or the
more opportunity you will give yourself to unleash nasty surprises on the
opponent).
To return to our game, Black’s choice of 12 ... b4 is also inadequate.
Grandmaster Ilincic not only allows his opponent to make a strong sacrifice,
he in effect goads him into doing so. It is clear that he has underestimated the
danger (perhaps being rated over 200 Elo points above his opponent had
made him in too much of a hurry to complicate matters).
Rather than spend a precious tempo on this provocation, Black should
play 12 ... Nc5!, when the chance to exchange off White’s bishop on b3
deters any sacrifice. Play might then proceed more quietly with 13 Kb1 (after
13 Qxg7 Rg8 14 Qh6 Nxb3+ 15 axb3 b4 16 Na4 Rg6 17 Qh4 Bd7 18 Kb1
Nxe4 19 Qxh7 Nf6 Black has plenty of play for a pawn) 13 ... Nxb3 14 axb3
0-0 15 Bh6 Ne8 16 f4 and only now is 16 ... b4 a good idea. Note that 13 ...
Nfxe4? would be a mistake in this sequence due to another sacrifice: 14 Nxe4
Qxe4 (14 ... Nxe4 15 Qxg7 is good for White) 15 Nxe6! Nxe6 (15 ... Bxe6 16
Bxc5 is no better for Black) 16 Bd5 Qa4 17 Bxa8 and White is winning.
 Question: It’s your chance to carry out a sacrifice. Where should it be?

Answer: 13 Nf5!
Also crushing was 13 Nd5! exd5 (declining the sacrifice doesn’t save
Black; e.g. 13 ... Bf8 14 Bg5! h6 15 Bxf6 Nxf6 16 Nxf6+ gxf6 17 Nf5! exf5
18 exf5+ Kd8 19 Rxd6+ Bxd6 20 Qxd6+ Bd7 21 Bd5 Qb8 22 Qxf6+ Kc7 23
Re4! and the threat of 24 Rc4+ is decisive) 14 Bxd5 Nxd5 15 exd5 Nf6 16
Nc6 0-0 17 Bg5! and the threat to e7 is decisive: after 17 ... Re8 18 Bxf6
Black can’t recapture on f6 without allowing mate, while 17 ... Be6 (the best
try) 18 Nxe7+ Qxe7 19 dxe6 fxe6 20 Rxd6 sees White a pawn up with a
dominant game.
13 ... bxc3
If 13 ... exf5 then 14 Qxg7 attacks f7 as well as h8. Following 14 ... bxc3
15 Qxh8+ Nf8 16 exf5 or 14 ... Rf8 15 Nd5 Nxd5 16 Bxd5 Qb8 17 exf5, the
attack down the e-file is unstoppable.
14 Nxe7
 14 ... cxb2+
The knight can’t be recaptured: 14 ... Kxe7 15 Qxd6+ Ke8 16 Bc5!
(threatening mate on e7) 16 ... Ng8 (or 16 ... Nxc5 17 Qd8 mate) 17 bxc3!
(Black was hoping for 17 Ba4 when he is totally lost, apart from the fact that
he has 17 ... Qxb2 mate!) 17 ... Qb8! (a tough way to resist, but White’s
initiative is too strong even after the exchange of queens) 18 Ba4 Qxd6 19
Rxd6 Rb8 20 c4! (to stop 20 ... Rb5 and also control the d5-square to prevent
Black playing ... Ngf6-d5 in some lines) 20 ... Kd8 (if 20 ... Ngf6 then 21 e5
and 22 Red1 wins) 21 Bxd7 Bxd7 22 Red1 Rb7 23 Rxa6 Ne7 24 Bb6+ and
now, as 24 ... Ke8 loses to 25 Ra8+ Nc8 26 Rxc8+ Bxc8 27 Rd8+ Ke7 28
Rxh8 Rxb6 29 Rxc8, Black has to enter an endgame with 24 ... Rxb6 25
Rxb6 where White has the winning advantage of a (very active) rook and
three pawns for two minor pieces.
That was a lot of moves! It reminds us that we can’t calculate everything:
we have to trust our intuition when it tells us our initiative isn’t going to
vanish, despite a major change such as a transition from a middlegame to an
endgame.
 15 Kb1 Nxe4 16 Qxg7 Rf8
After 16 ... Kxe7 there is a safe way to win with 17 Qxh8 Ndf6 18 Bd4,
but another sacrifice also does the trick: 17 Bxe6! Kxe6 18 Bf4 d5 19 f3 and
the black king will soon be at the mercy of the white rooks.
17 Nd5 a5
After 17 ... exd5 18 Bxd5 Black is poleaxed.
18 Bh6
An aggressive move which also clears the way for the rook on e1.
18 ... a4
A last desperate counterattack.
19 Rxe4 axb3

Question: Black is actually threatening mate in one, but how can we get
at
the black king first?

Answer: 20 Rxe6+! Kd8

 Or 20 ... fxe6 21 Qe7 mate.
21 Bg5+ 1-0
If 21 ... Nf6 then 22 Qxf8+ and mate next move.

Not all contributions to theory turn out to be a spectacular success. I played

the hero of the game above, Tunisian FM Nabil Doghri, some years ago at a
First Saturday tournament in Budapest. I prepared what I thought was an
opening novelty involving the sacrifice of a pawn. Later I discovered it had
been played in other games, but I didn’t have a database back then.

Game 75
N.McDonald-N.Doghri
Budapest 1996

1 e4 c5 2 Nf3 Nc6 3 d4 cxd4 4 Nxd4 e5 5 Nb5 Nf6 6 N1c3 d6 7 a4 a6 8 Na3

Be6 9 Bc4 Rc8 10 0-0 Nd4 11 Qd3

Here is the idea. White gives up a pawn to get mastery over the d5-
square. It seems good enough for equality, but no more than that.
11 ... Bxc4 12 Nxc4 Nxc2 13 Qxc2 Rxc4 14 Bg5 Be7 15 Bxf6 Bxf6 16
Qb3 Qc8 17 Rfd1 Qc6 18 Rd3 0-0 19 Rad1 Be7 20 Nd5 Bg5 21 Nc3 Be7
22 Nd5 Bg5 23 Rg3
I should have called it a draw with 23 Nc3.
23 ... Bd8 24 Qe3 Qxa4 25 Re1 Qd7
It turns out that 25 ... f5!, preparing to answer 26 b3 with 26 ... f4, is good
for Black.

Question: Can you see how White escapes his dubious position by
forcing
a draw?

Answer: 26 Nb6! Bxb6 27 Rxg7+ Kxg7 28 Qg5+ Kh8 ½-½

After the game my opponent said to me: “of course you didn’t see after
11 Qd3 I can win a pawn?” I was embarrassed and nodded in agreement, not
wanting to admit my preparation was so bad it could be mistaken for a
blunder!

From the Opening Into the Endgame

Some players are content with a small but risk-free advantage from the
opening as White. They are happy to exchange off queens to avoid the danger
of a counterattack on their king. They also want to get down to the business
of outplaying their opponent with a minimum intrusion from theory.
You shouldn’t underestimate the danger of this approach if you face it as
Black. It might seem that the opponent isn’t ambitious, or even half asleep,
but in fact he is simply guiding the course of play along channels that suit his
style and temperament. If you are careless you’ll be in for a lot of suffering.
If you play like this as White, well, you have to be patient and accept that
a lot of the time you won’t get much from the opening. But you are safe and
giving full rein to the stronger parts of your game, while preventing an
aggressive opponent from expressing his own style.
A sharp eye for tactical opportunities is an important quality for anyone
who aspires to be a strong positional player. The following example shows
that you can’t switch off your tactical radar in even the most peaceful-looking
scenarios.

Game 76
V.Artemiev-Z.Izoria
World Team Championship, Astana 2019

1 Nf3 Nf6 2 c4 e6 3 Nc3 d5 4 e3 c5 5 d4 dxc4 6 Bxc4 a6 7 0-0 b5 8 Bb3

Bb7 9 e4 cxd4 10 Nxd4 Nc6 11 Nxc6 Qxd1 12 Rxd1 Bxc6 13 f3 Bc5+ 14
Kf1 Ke7
In his treatise Chess Fundamentals Capablanca has a section called “the
danger of a safe position”. There doesn’t seem to be much going on either
tactically or strategically in the diagram position. The queens have been
exchanged and the pawn structure is symmetrical. Artemiev adds to the
impression of torpidity with a seemingly lethargic knight retreat.
 15 Ne2 Rhd8?

Question: The obvious response, but a serious mistake. Can you see
why?

Izoria stops 16 Nd4 and prepares the exchange of rooks on the d-file. He
must have thought he would soon be shaking hands to agree a draw.
And yet common sense should have warned him: “I’m up against a player
who is on fire at the moment, crushing almost of all his opponents with
White! He’s almost in the top ten in the world now. Did Artemiev really
come today without a plan for advantage? And this in a key match between
Russia and the United States ... ? No, let me look again at the position. What
does he have in mind?”
Alas for Black, no such checking took place.
Answer: 16 Bg5!
One careless move has put Black in trouble. The threat is 17 e5, and if 17
... h6 then 18 Rdc1! hxg5 19 Rxc5 wins a pawn as both c6 and g5 are
hanging. Now we see that 15 Ne2 may have looked passive but it cleared the
c-file for the white rooks to exploit the vulnerable black bishops. Had he been
more alert Izoria would have preferred 15 ... Rac8!, defending the bishop on
c6 and so avoiding trouble on the c-file.
16 ... Rxd1+ 17 Rxd1 Bd6

It appears at first glance that Izoria has found a way to deal with the
double threat of 18 Rc1 and 18 e5, but now comes a neat forcing sequence.

Question: How can we punish Black tactically?

Answer: 18 Rxd6!! Kxd6 19 e5+ Kxe5 20 Bf4+

Black was hoping his king would be an asset in the endgame and so left
him in the centre. White has only three minor pieces left but they swarm
around His Majesty, making up in energy what they lack in numbers.
20 ... Kf5 21 Bc2+ Ne4
There’s no option but to hand back a piece.
22 fxe4+ Kf6
 If 22 ... Bxe4 then 23 g4+! Kxg4 24 Bxe4 wins.
23 Kf2

Question: What is your assessment of the endgame? (see the comment

below and after the next move)

Answer: The tactics have come to an end. Nominally material is

balanced, but the black rook doesn’t have anything to attack: White’s pawns
are secure and he can stop any attempt by the rook to invade down the c-file
or d-file. Also, Black’s bishop has little scope, being blocked in by its own
pawn on b5 and White’s pawn on e4. And taking the black rook, bishop and
king as a whole, how can they co-ordinate their action?
Meanwhile, as we shall see, White can improve the layout of his pieces
and get them working together.
23 ... e5
On the negative side this leaves a hole on d5, which the white knight can
exploit, and in general weakens Black on the a2-g8 diagonal. However, if he
doesn’t try to utilize his kingside pawn majority, he is going to be slowly
squeezed. Therefore he pushes the e-pawn to gain space and in the hope of
arranging ... f7-f5 at some point. He also prevents a future e4-e5+, though
White would think hard before opening the diagonal for the black bishop.
24 Bd2 Ke7 25 Bb4+ Ke8

Black hopes to advance 26 ... a5 and then 27 ... b4. This would energize
his queenside pawns, clear the light squares for the bishop, and guard the c3-
square.

Question: So what should we do to stop him?

Answer: 26 Ba5!
A simple prophylactic move which keeps the black queenside tied up.
26 ... Bd7
The bishop tries to become active on the a2-g8 diagonal, but White is
ready again with a preventive measure.
27 Bb3!
 One by one the white pieces take up good posts, while 27 ... Be6 28 Bxe6
fxe6 is gruesome for Black’s pawn structure.
27 ... Rc8 28 Nc3
And now it is the turn of the knight which heads for its dream post on d5.
28 ... Rc6
Black hopes to activate his rook along the third rank, as there is nothing
to gain on the c-file.
29 Nd5 Rh6
Here too if 29 ... Be6 then 30 Nc7+ and 31 Nxe6 inflicts doubled pawns.
30 Bc3!

The bishop has done its duty on a5 and is rerouted. If now 30 ... Rxh2
then 31 Bxe5 attacks the rook and will win the g7-pawn as well, while 30 ...
f6 31 h3 leaves the rook cut off and Black facing 32 Nc7+ and 33 Nxa6.
30 ... Kf8 31 h3!
It would be entirely unnecessary to activate the black rook with 31 Bxe5?
Re6 32 Bc3 Rxe4.
31 ... f5
 If 31 ... f6 then 32 Bb4+ Ke8 (or 32 ... Kf7 33 Nc7+) 33 Nc7+ wins the
a6-pawn. Black can’t stand the slow death any longer and so, instead of
returning his rook to c6, gives up a pawn to try to activate his position on the
queenside.
32 exf5 Bxf5 33 Bxe5
Quite good enough, but the neat 33 Nc7! would be a high-class
prophylactic move. It threatens mate in two beginning with 34 Bb4+. Then
after 33 ... Ke7 34 Bxe5 Black wouldn’t have the chance to advance his
queenside pawns.
33 ... a5 34 Ke3 Rc6 35 Nc3
White’s task now is to restrain the energy of Black’s queenside pawns
and then pick them off.
35 ... b4 36 Bd5 Rc8 37 Nb5 Rc1
At last the black rook gets to enter the enemy camp, but he is confounded
by the enormous power of an aligned pair of bishops. Between them they are
guarding the a2, b2- and g2-pawns.
38 Nd6 Bb1 39 Nb7 Re1+ 40 Kd4 Re2
 Izoria has put up a tough and excellent defence. If only he had been more
alert back at move 15 he would have avoided all this suffering.
41 a4!
A clinical winning move. Artemiev sees that he can hold back the
resulting black b-pawn and queen his own passed a-pawn.
41 ... Rxb2 42 Nxa5 Bc2 43 Nc4 Ra2 44 Bd6+ Ke8 45 Bxb4 1-0
The black rook drops for nothing after 45 ... Bxa4 46 Nd6+, while 45 ...
Rxa4 46 Bc6+ Kd8 47 Bxa4 Bxa4 leaves Black a piece down.

Flank Openings
White isn’t obliged to try to exploit his right to move first by building a pawn
centre. He could, for example, play 1 b3 or 1 g3 or, less modestly, 1 c4 or 1
Nf3 and then try to attack any centre that Black creates. He might not get
much initiative, but if he is familiar with the structure, ideas and theory, then
it promises chances of success. Basically, White is exchanging the advantage
of moving first for the chance to adopt a set-up he is comfortable with. Such
an approach is also attractive to players trying to avoid critical opening
theory.

Game 77
V.Artemiev-H.Nakamura
Gibraltar 2019

1 Nf3 Nf6 2 g3 d5 3 Bg2 e6 4 0-0 Be7 5 c4 0-0 6 b3 c5 7 Bb2 Nc6 8 e3 b6 9

Nc3 dxc4 10 bxc4 Bb7 11 Qe2 Rc8
 Question: You might like to evaluate the position. What features give
White a slight edge? What pawn advance should he be trying for?
(Look beyond the obvious one.)

Answer: Black has every piece developed, a safe king and no pawn
weaknesses. But he isn’t quite equal. White is superior in two or three minor
ways:
1) His bishop on b2 has more scope than its counterpart on e7.
2) His queen has a comfortable and secure observatory post on e2 from
where she can look in all directions. The black queen’s view on c7 (where it
is natural to develop her) will be more circumscribed.
3) His centre has the potential to expand thanks to his extra pawn there. In
contrast, if Black tries to utilize his queenside majority with ... a7-a6, and
ever manages to threaten ... b6-b5, White can clamp down on it with a2-a4.
Well, this talk of White’s extra centre pawn and potential mobility might
make you conclude that he should deploy his rooks to c1 and d1 and then
advance d2-d4. But this would be against the spirit of the opening.
Black’s pieces would be well placed to put pressure on the hanging
pawns which result after the exchange ... c5xd4 and recapture e3xd4. Black
already has his rook on c8 ready for the job and could add moves like ... Qc7,
... Rfd8, ... Na5 and ... Ba6 to the pile. Notice how, in the assessment above,
we said that the black queen has a limited view on c7. That’s no longer true
after White breaks open the c-file and lets her eye the c4-pawn. And the meek
bishop on e7 would also be presented with an open diagonal.
In conclusion, Artemiev needs a plan to exploit his centre’s potential to
expand, but he mustn’t allow Black to activate his game by bombarding it. A
quick look around shows us that there are other ways to mobilize the pawns
besides d2-d4.
12 Rad1 Qc7 13 Ne1!
This modest retreat epitomes White’s plan. Black has three pieces lined
up on the c-file waiting for an attack that never comes. Meanwhile, White is
planning the advance of his f-pawn.
13 ... Ne8!
 Question: Can you give some reasons for this move? (see below and the
note to 14 ... Nd6)

Answer: I’ve been praising White’s position but in fact the situation is
very close to equality. Nakamura knows a thing or two about strategy and
isn’t going to twiddle his thumbs while his opponent builds up pressure on
the f-file.
By mirroring the knight retreat of his opponent, he clears the way for the
advance of his own f-pawn and the activation of his dark-squared bishop on
f6.
14 f4
Intending 15 f5, when Black is liable to have an isolated pawn after an
exchange on e6, as the black e-pawn can’t abandon its duty of guarding the
d5-square from invasion by a white knight.
14 ... Nd6
We see another point to 13 ... Ne8: on d6 the knight holds up the advance
f4-f5 and is able to lend a hand in any queenside activity Black can drum up.
15 Nf3
Philidor would have approved of the courtesy shown by the white knight
to its f-pawn (compare the note to 9 ... Nb4 in Navara-Pähtz above). The
Frenchman viewed the pieces as the servants of the pawns, and this
relationship is exactly what we see in the knight ushering the pawn forward
and then following in its wake.
15 ... a6
Hinting at a possible expansion with ... b6-b5, so White puts a stop to it at
once.
16 a4 f5 17 d3 Bf6
Nakamura has completed his manoeuvre ... Nd6, ... f7-f5 and ... Bf6. His
pieces are well centralized, deterring White from expanding there. In addition
his pawns on e6 and f5 are keeping the white pieces out of the d5- and e4-
squares. Still, compared with White’s solid structure, the e6-pawn is kind of
‘floating’ with no supporting pillar. Meanwhile, the f5-pawn is vulnerable to
a flanking action on the kingside. Artemiev wastes no time in carrying it out.
18 h3
For the moment the d3- and e3-pawns hold the centre while the kingside
pawns get to have the fun. But their turn will soon come.
 18 ... Nb4

The knight uncovers the bishop on b7 and so prepares a tactical response

to a white advance with e3-e4, as begins at move 20.
Furthermore, the prospect has emerged of Black utilizing his queenside
pawns with 19 ... Bc6 and 20 ... b5. This might have helped persuade
Artemiev that the time was ripe for his own pawn advances.
19 g4
Not only putting pressure on f5, but also with the idea of 20 g5 to drive
the black bishop from the a1-h8 diagonal and gain ascendancy over the e5-
square.
19 ... g6
Widening his defensive wall and so holding onto the f5-point, while
giving the bishop a retreat square on g7.
20 e4
Artemiev finally judges it to be the right moment to roll forwards in the
centre. We can’t talk of an objective advantage for White, but the
defensive/counterattacking task will prove too taxing even for Nakamura.
 20 ... fxe4
A solid alternative was 20 ... Bg7. Instead, Nakamura intends to use
tactics to eliminate the e4-pawn.
21 dxe4 Nxe4?
Black misses the strength of 21 ... Bd4+!. We could blame time trouble
or, more accurately, the speeding up of a player’s moves as time trouble
began to loom. But the main barrier to discerning the bishop move is its
inherent oddity: offering to exchange off the dark-squared bishop makes a
strange impression. Nonetheless, the tactics work in Black’s favour: after 22
Nxd4? cxd4 23 Rxd4 Qc5 the pin spells big trouble for White; for example,
24 Qf2 Rxf4! 25 Qxf4 Qxd4+, followed by 26 ... Nd3, is terrible for him. So
White should reply with 22 Kh2 with an unclear position.
22 Nxe4 Bxb2

Question: Can you see a way for White to create an outpost in the centre
in
good old Chapter Five style?
 Answer: 23 Neg5!
Nakamura must have underestimated the strength of this move which
targets the isolated e6-pawn. Instead, 23 Qxb2 Bxe4 is fine for Black.
23 ... Bxf3
In order to win the d4-square for the bishop.
24 Rxf3 Bd4+ 25 Kh1 Rce8
Black’s problem isn’t so much the loss of the e6-pawn (after all the pawn
count remains equal) but the fact that the white knight will be able to take
over the e6-square as an outpost which can be supported by f4-f5.
26 Nxe6 Qc6 27 f5

Mission accomplished. The white knight dominates the centre. Only now
has White got a ‘real’ advantage in contrast to the vague promises of
‘superior mobility’ or ‘the initiative’. A knight on an outpost protected by a
pawn is something concrete that we can all readily understand, even see on
the board in front of us – if so inclined we can even tap the little horse on the
back and say “well done, fellow!”
 27 ... Qxa4
In such situations you should grab material and hope for the best.
28 fxg6!
Well, the pawn on f5 didn’t stay defending the knight for long! But White
is converting one advantage – an outpost square supported by a pawn – into a
better one: the destruction of the black king’s pawn defences.
28 ... Rxf3
If 28 ... hxg6 then 29 Rxf8+ Rxf8 30 Nxf8 wins.
29 gxh7+ Kh8 30 Bxf3?
This is a natural recapture, but it might have allowed Black to escape.
Instead, 30 Qxf3! gives White a decisive attack; e.g. 30 ... Rxe6 31 Qf8+
Kxh7 32 Rf1 Bg7 (or 32 ... Rg6 33 Be4) 33 Qf5+ Rg6 34 Be4 (the pin on g6
is fatal) 34 ... Qe8 35 Qh5+ Bh6 36 Bxg6+ Qxg6 37 Rf7+ and Black loses the
queen.
Notice how the presence of opposite-coloured bishops gives White’s
initiative extra potency, as Black’s bishop can do nothing to guard against an
attack on the light squares. It has even been said that “when there are
opposite-coloured bishops the player with the attack has an extra piece”. A
wild exaggeration of course, but it’s worth remembering if we need
encouragement to play a move like 30 Qxf3.
 30 ...
Nc6?

Question: What is the tactical refutation of this move and can you make
the strange-looking 30 ... Nd3 work for Black?

Answer: Things are grim for the black king due to the absence of the
queen and knight from the defence. However, he had a spectacular chance to
remedy both ills with 30 ... Nd3!!. Upon 31 Qxd3 Rxe6 Black is ready to play
32 ... Qe8 to staunch the defence with equal chances, while after 31 Rxd3?!
Qxc4 surprisingly White doesn’t have a good way to stop 32 ... Qxe6, when it
would be Black for preference. In either variation an unexpected co-
ordination of Black’s queen and rook saves his king.
It would have been very difficult to see 30 ... Nd3 in time pressure, even
for a player of Nakamura’s class.
Alas, computers have distorted what should be perceived as an excellent,
controlled performance by White. If this game had been played in, say, 1985
everyone would have congratulated White on a smooth win. Nowadays, as
soon as a top player in time pressure misses a bizarre defence you get online
kibitzers yelling “30 ... Nd3!. He missed 30 ... Nd3!” across seven continents.
31 Nxc5! 1-0
The white knight gets to strike the final blow, but it is unexpectedly
backwards and to the left rather than against the black king. The game ends
abruptly, as after 31 ... Rxe2 32 Nxa4 Black is already two pawns down and
has a rook and knight hanging. If he saves them with 32 ... Re6 then the
bishop is the victim: 33 Bxc6 Rxc6 34 Rxd4.
Index of Games
Anand.V-Giri.A, Shamkir 2019
Anand.V-Nakamura.H, London Chess Classic 2010
Aronian.L-Carlsen.M, Stavanger 2017
Aronian.L-Rapport.R, Wijk aan Zee 2017
Aronian.L-Vachier-Lagrave.M, London Chess Classic 2018
V.Artemiev-Z.Izoria, World Team Championship, Astana 2019
V.Artemiev-H.Nakamura, Gibraltar 2019
Banusz.T-Vajda.L, Szentgotthard 2010
Belov.V-Socko.B, Hastings 2004/05
Buhr.C-Malakhov.V, European Cup, Kallithea 2008
Carlsen.M-Anand.V, Grand Chess Tour, Zagreb 2019
Carlsen.M-Grischuk.A, Shamkir 2019
Carlsen.M-Hossain.E, Baku Olympiad 2016
Carlsen.M-Karjakin.S, Sinquefield Cup, St. Louis 2017
Carlsen.M-Yu Yangyi, Stavanger 2019
Caruana.F-Carlsen.M, Grenke Classic, Karlsruhe/Baden Baden 2019
Caruana.F-Carlsen.M, Tromsø Olympiad 2014
Caruana.F-Karjakin.S, Sinquefield Cup, St. Louis 2018
Caruana.F-Karjakin.S, Stavanger 2018
Caruana.F-Shankland.S, US Championship, St. Louis 2016
Demchenko.A-Gukesh.D, Ho Chi Minh City 2019
Doghri.N-Ilincic.Z, Istanbul Olympiad 2000
Duda.J-Navara.D, Prague 2019
Duda.J-Rapport.R, Wijk aan Zee 2019
Fedoseev.V-Giri.A, Wijk aan Zee 2019
Fedoseev.V-Shankland.S, Wijk aan Zee 2019
Giri.A-Caruana.F, Wijk aan Zee 2013
Giri.A-Harikrishna.P, Shenzhen 2019
Grischuk.A-Gelfand.B, FIDE Grand Prix, Khanty-Mansiysk 2015
Grischuk.A-Volokitin.A, Baku Olympiad 2016
Habu.Y-Kasparov.G, Rapid Match, Tokyo 2014
Hou Yifan-Toufighi.H, Subic Bay 2009
Ivanisevic.I-Ivanchuk.V, European Championship, Batumi 2018
Ivic.V-Predke.A, European Championship, Skopje 2019
Karjakin.S-Esipenko.A, World Rapid Championship, Riyadh 2017
Karjakin.S-So.W, FIDE Candidates, Berlin 2018
Korobov.A-Anand.V, German Bundesliga 2019
Kramnik.V-Shankland.S, Wijk aan Zee 2019
Lagno.K-Ivanchuk.V, Cap d'Agde (rapid) 2008
Lalic.B-Vicas.M, Dublin 2019
Ledger.A-Kosten.A, British League 2001
Mamedyarov.S-Ding Liren, FIDE Candidates, Berlin 2018
Matlakov.M-Aronian.L, FIDE World Cup, Tbilisi 2017
Mazur.S-Zahedifar.A, Ordu 2019
McDonald.N-Doghri.N, Budapest 1996
Meier.G-Vachier-Lagrave.M, Grenke Classic, Karlsruhe/Baden Baden
2019
Melkumyan.H-Shankland.S, Batumi Olympiad 2018
N.Ross.C-Spinu.E, IBCA Olympiad, Ohrid 2017
Nabaty.T-McDonald.N, London Chess Classic 2017
Naiditsch.A-Svidler.P, Grenke Classic, Karlsruhe/Baden Baden 2019
Navara.D-Pähtz.T, Gibraltar 2018
Nepomniachtchi.I-Karjakin.S, Grand Chess Tour, Zagreb 2019
Nikac.P-Dukaczewski.P, IBCA Championship, Cagliari 2019
Nisipeanu.L-Radjabov.T, Medias 2010
Paravyan.D-Golubov.S, Korchnoi Memorial, St. Petersburg 2018
Petrosian.T-Martirosyan.H, Aeroflot Open, Moscow 2019
Phillips.O-Ross.C, Solihull 2017
Radjabov.T-Vidit.S, Wijk aan Zee 2019
Radovanovic.N-Saric.I, European Championship, Skopje 2019
Rapport.R-Giri.A, Wijk aan Zee 2019
Rasik.V-Laznicka.V, Czech Championship, Karlovy Vary 2004
Robson.R-Caruana.F, US Championship, St. Louis 2018
Sandipan.C-Ivanchuk.V, Gibraltar 2018
Serper.G-Becerra Rivero.J, Ledyard 2000
So.W-Sadzikowski.D, Gibraltar 2019
Sutovsky.E-Eljanov.P, Poikovsky 2014
Topalov.V-Wojtaszek.R, Shamkir 2018
Tuka.O-Nikac.P, IBCA Championship, Cagliari 2019
Vachier-Lagrave.M-Caruana.F, Stavanger 2018
Vachier-Lagrave.M-Navara.D, Biel 2018
Vachier-Lagrave.M-Tari.A, Gibraltar 2019
Valmana Canto.J-Harikrishna.P, Spanish League 2006
Van Foreest.J-Lewtak.D, European Championship, Skopje 2019
Volokitin.A-Mamedyarov.S, European Cup, Eilat 2012
Wang Yue-Radjabov.T, FIDE Grand Prix, Sochi 2008
Winants.L-Eliet.N, Belgian League 2014
Yu Yangyi-Artemiev.V, Gibraltar 2019