Professional Documents
Culture Documents
2017 - e3 Poison - A. Smith(PDF)
2017 - e3 Poison - A. Smith(PDF)
Title Page
Structure of the Book 3
Bibliography 6
Key to symbols used & Thanks 8
PART 1 – Introduction
4 Sneaky Grünfeld 23
5 Reversed King’s Indian Attack 36
6 Poor Man’s Benoni 61
7 Anti-Benko Gambit 85
8 Queen’s Indian and Bogo-Indian 97
PART 4 – Junctions
11 Panov 142
12 Timid Tarrasch 174
13 Irregular Slavs 187
14 Chigorin 216
15 Dutch 226
PART 8 – Exercises
By
Axel Smith
Quality Chess
http://www.qualitychess.co.uk
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.
The big challenge with understanding how to use the e3 poison is gaining a feel for the many possible move orders, but
it would be rough to delve into them without first being familiar with the general ideas.
The book thus starts with an introduction (Part 1), discussing the overall concept and why a practical player should be
happy with an academic advantage.
The Indian Defences (Part 2) have one thing in common. Regardless of whether we start with 1.d4, 1.Nf3, 1.c4 or 1.e3,
we often reach the same tabiya. The Grünfeld, King’s Indian, Benoni and Queen’s Indian are, in that sense, easy to
handle. The Bogo-Indian and Benko Gambit, on the other hand, are often avoided, but there are antidotes covered in
Chapters 7 and 8.
After that we are ready for a first fight with the various move orders (Part 3), combined with some historical
background.
That leads to the ‘Junctions’ (Part 4), which are the openings that many opponents will transpose into when they meet
the e3 poison, even if these openings are outside their normal repertoire. The Panov, Tarrasch, Chigorin and Irregular
Slavs (Schlechter, Chebanenko and Meran) are all important to us. The Dutch is also covered in this part, since the
Dutch is almost a universal weapon (only starting with 1.e4 prevents it).
The choice between different first moves is most important against opponents who play 1...d5 and the Queen’s Gambit
(Accepted) or the Slav. There are three ways to deal with them:
The last section of the book (Part 8) is Exercises – a Final Test with Solutions to check that you have become an
effective e3 Poisoner!
Bibliography
Important Sources
Avrukh, Boris: Grandmaster Repertoire 1 – 1.d4 Volume One, Quality Chess 2008
ChessBase Magazine:
#32 (Viswanathan Anand)
#46 (Lubomir Ftacnik)
#57 (Rustem Dautov and Nigel Short)
#86 (Evgeniy Solozhenkin)
#120 (Mihai Marin)
#124 (Levon Aronian)
#164 (Martin Breutigam)
#168 (Milov Pavlovic)
#169 (Igor Stohl)
Delchev, Alexander: The Modern Reti, Chess Stars 2012
Emms, John: Starting Out: The King’s Indian Attack, Everyman Chess 2005
Karpov, Anatoly: Karpov’s Caro-Kann: Panov’s Attack, Batsford Chess Books 2006
Khalifman, Alexander: Opening for White according to Kramnik, Chess Stars 2002
Kornev, Alexei: A Practical White Repertoire with 1.d4 and 2.c4, Vol 1, Chess Stars 2013
Rudel, David: Zuke ’Em, Thinkers Press 2008
Summerscale, Aaron & Johnsen, Sverre: A Killer Chess Opening Repertoire, Gambit 2010
Other Sources
Aveskulov, Valery: Attack with Black, Gambit 2012
Avrukh, Boris: Grischuk vs. The Grunfeld, Internet 2013
Avrukh, Boris: Grandmaster Repertoire 17 – The Classical Slav, Quality Chess 2014
Bologan, Viktor: The Chebanenko Slav According to Bologan, New in Chess 2008
Bologan, Viktor: The Chebanenko: Still Improved (DVD), ChessBase 2012
Bronznik, Valery: The Chigorin Defence, Schachverlag Kania 2005
D’Costa, Lorin: The Panov-Botvinnik Attack, Everyman Chess 2013
Dunnington, Angus: The Ultimate King’s Indian Attack, Batsford 1998
Flear, Glenn: Starting Out: Slav & Semi-Slav, Everyman Chess 2005
Goldsmith, Jeffrey: The Last Human Chess Master, Wired 1995
Greet, Andrew: Play the Queen’s Indian, Everyman Chess 2009
Hansen, Carsten: The Gambit Guide to the English Opening 1...e5, Gambit 2009
Houska, Jovanka: Opening Repertoire: The Caro-Kann, Everyman Chess 2015
Johnsen, Sverre & Bern, Ivar: Winning with the Stonewall Dutch, Gambit 2009
Karpov, Anatoly: How to Play the English Opening, Batsford Chess Books 2007
Khalifman, Alexander: Opening for Black according to Karpov, Chess Stars 2001
Lakdawala, Cyrus: The Slav: Move by Move, Everyman Chess 2011
Morozevich, Alexander & Barskij, Vladimir: The Chigorin Defence According to Morozevich, New in Chess 2007
Moskalenko, Viktor: The Flexible French, New in Chess 2008
New in Chess: 2014/8 (Vladimir Kramnik)
Nielsen, Peter Heine: Intens skak i Sibirien, Skakbladet 2011/06
Palliser, Richard; Emms, John; Ward, Chris & Jones, Gawain: Dangerous Weapons: The Benoni and Benko,
Everyman Chess 2008
Raetsky, Alexander & Chetverik, Maxim: Starting Out: Queen’s Gambit Accepted, Everyman Chess 2007
Vitiugov, Nikita: The French Defence Reloaded, Chess Stars 2012
Watson, John: Play the French 2nd Edition, Cadogan Chess Books 1996
Williams, Simon: Attacking Chess: The French, Everyman Chess 2011
Williams, Simon: The Killer Dutch, Everyman Chess 2015
Wisnewski, Christoph: Play 1...Nc6!, Everyman Chess 2007
Checked Sources
Aagaard, Jacob: Starting Out – The Grünfeld, Everyman Chess 2003
Aagaard, Jacob: Stonewall II, Quality Chess 2007
Avrukh, Boris: Grandmaster Repertoire 1A – The Catalan, Quality Chess 2015
Avrukh, Boris: Grandmaster Repertoire 1B – The Queen’s Gambit, Quality Chess 2016
Avrukh, Boris: Grandmaster Repertoire 8 – The Grünfeld Volume One, Quality Chess 2011
Bezgodov, Alexey: The Liberated Bishop Defence, New In Chess 2014
Bologan, Viktor: English 1.c4 c5 for Black, ChessBase 2012
Collins, Sam: The Tarrasch Defence: Move by Move, Everyman Chess 2013
Cox, John: Declining the Queen’s Gambit, Everyman Chess 2011
Delchev, Alexander: Attacking the English/Reti, Chess Stars 2016
Delchev, Alexander & Semkov, Semko: Understanding the Queen’s Gambit Accepted, Chess Stars 2015
Dembo, Yelena: Play the Grünfeld, Everyman Chess 2008
Emms, John: Starting Out: The Queen’s Indian, Everyman Chess 2004
McDonald, Neil: Play the Dutch, Everyman Chess 2010
Mikhalevski, Victor: Grandmaster Repertoire 19 – Beating Minor Openings, Quality Chess 2016
Ntirlis, Nikolaos & Aagaard, Jacob: Grandmaster Repertoire 10 – The Tarrasch Defence, Quality Chess 2011
Raetsky, Alexander & Chetverik, Maxim: Starting Out: Benoni Systems, Everyman Chess 2005
Ribli, Zoltan & Kallai, Gabor: Winning with the Queen’s Indian, Batsford Chess Books 1987
Rowson, Jonathan: Understanding the Grünfeld, Gambit 1999
Schandorff, Lars: Grandmaster Repertoire 7 – The Caro-Kann, Quality Chess 2010
Smerdon, David: Queen’s Gambit Accepted in Yearbook 119, New in Chess 2016
Vegh, Endre: Starting Out: The Modern Benoni, Everyman Chess 2004
Vigorito, David: Play the Semi-Slav, Quality Chess 2008
Vigus, James: Play the Slav, Everyman Chess 2008
Key to symbols used
² White is slightly better
³ Black is slightly better
± White is better
µ Black is better
+– White has a decisive advantage
–+ Black has a decisive advantage
= equality
© with compensation
„ with counterplay
ƒ with an initiative
÷ unclear
? a weak move
?? a blunder
! a good move
!! an excellent move
!? a move worth considering
?! a move of doubtful value
™ only move
# mate
¥ Symbols next to diagrams show the pieces exchanged to reach that structure.
Thanks
I couldn’t imagine how much attention Pump Up Your Rating would get. At tournaments, I was often caught for a small
chat. One player used a line from the book when preparing against me. Another said that he was finally satisfied with
his playing strength. The book convinced him that he was not ready to put in the effort needed to reach the next level.
I am glad for all the support. It also made Quality Chess ask me to write another book on any subject I wanted –
except theoretical endings.
There is one word in the book I really regret, but so far nobody has mentioned it. I have however been criticized for one
thing: I didn’t dedicate the book to my wife.
So now I asked her what I should write.
“Oh, such things are ridiculous,” she said, “even though it’s the first thing I look at in any book. There is no reason to
thank me.”
I don’t agree, because it would have been harder without her support. Not least agreeing to have all her clothes, shoes
and jackets in a small wardrobe, because I tore down our clothes closet to create a writing studio. That’s where I
finished this book.
I am also happy that our daughter sometimes allows me to work after office hours – and that she more often does not.
Finally, thanks to all the friends who read and improved the draft: Björn Ahlander, Stellan Brynell, Torbjörn Ivarsson,
Martin Jogstad, Martin Lokander, Sebastian Mauritsson and Aditya Subramanian.
Axel Smith,
27th June 2017
ChessPublishing.com
In Revolution in the 70s, Garry Kasparov explained how opening theory exploded after 1972, under the influence of
Bobby Fischer. Information became more accessible and the players could, instead of searching for games, focus on
analysing. That suited a hard worker such as Kasparov.
A few decades later many openings were over-analysed. It became harder and harder to get a tangible advantage and
to avoid being neutralized, White repertoires had to be broader. Still, the top players played for an advantage.
Things changed again when the engines made their entrance. It was easier to find out how to defend, and preparation
had to be even deeper. A new move could yield better results than the objectively best move, and the main task was to
surprise the opponent. But after a single game, everybody knew how to react against the idea, and it was time to find
another novelty.
Then along came Magnus Carlsen.
Okay, this story is simplified. There are other views and other players, but there’s no doubt that Carlsen has changed
the general attitude towards openings. Rather than an advantage, he looks for interesting positions.
When the opponent plays a dubious line there is little point in avoiding the known refutation. But against a good line, it
may not be practical to use the main lines. Chess is after all a draw, and we use time and effort only to lose the surprise
effect, while still not getting anything. Theory has developed to such an extent that even players who work harder and
know more than their opponents have started to avoid the main lines.
And so we entered the post-theoretical era.
When I started to work on this book, Quality Chess proposed 1.Nf3 followed by 2.g3. That was for a while a good
choice, but it was taken up by more and more players, and today theory has developed heavily even there. The time has
come to move forward, and I think my repertoire is a good choice: 1.Nf3 and 2.e3 with options of varying the order
from the very first move.
The last variation I analysed for my first draft was the Anti-Queen’s Gambit with 1.Nf3 d5 2.e3 Nf6 3.c4 e6 4.b3 Be7
5.Bb2 0-0 6.Nc3.
A few hours after I finished, Sergey Karjakin played like that against Anand in the 2016 Candidates tournament, and
won a nice strategic game. I was happy, of course, but also worried. Please leave the theory untouched!
When annotating the game for New in Chess, Anish Giri summarized today’s attitude among top players towards
openings.
“I was surprised that even some decent players thought that this [2.e3] was a sign of bad preparation. In fact, this is
the modern approach, where surprise value and unpredictability are often the key to success. The game is evolving;
deal with it.”
One person who has done so is Vladimir Kramnik. After being a consistent analyst with deep novelties, he shifted gear
in the World Blitz and Rapid Championship in Berlin in October 2015. But the real fight was a week later when he
played the e3 system in the European Club Cup. The opponent was none other than his big rival, Veselin Topalov.
There was no handshake before the game; Kramnik even looked away when Topalov started the clock.
Kramnik in New in Chess: “It’s my new way of playing chess with White. Trying to get a game.”
With his new attitude, Kramnik’s drawing ratio dropped and he experienced a revival as a player. And it might not just
be by chance that those games were played soon after he had a training camp with Magnus Carlsen in Berlin.
3...c5
If Black wants to place his bishop on b7, I think it makes sense to keep flexible with the c- and d-pawns. After 3...b6
4.Bd3 Bb7 5.0-0 Black could consider playing 5...d5, or continue to postpone the decision with 5...Be7.
4.Bd3 b6
There are many sensible choices, of course, one being to play as if Black was White: 4...d5 We will return to this
position later.
But since c2-c4 was played before ...d7-d5, White captures first.
10...0-0 11.Qg4
White threatens to win with 12.Bh6 Bf6 13.Qe4 g6 14.Bxf8.
The main line runs 11...Nf6 12.Qh4 Ne4 13.Qh3 Qxd4 14.Bf4 Nf6 15.Ne2, but as Kramnik wrote in New in Chess,
Topalov was tricked into this position and not prepared to play it.
13...Re8
13...Nd7 14.Nc6! wins a pawn.
14.Rd1 Nd7
After 14...Nc6 15.Nxd5 exd5 16.f4!, White has an advantage due to the strong knight on e5. The bishop on c4 is
untouchable, and after 16...Qd6 Boris Avrukh gives 17.Ba6!. Exchanging Black’s bad bishop may seem paradoxical,
but if 17.Bb3?! Na5, Black continues with the knight to c4 and gives up a pawn to open the diagonal.
The tactical try 17...Nxd4? 18.Rxd4 Bxe5 19.fxe5 Rxe5 doesn’t work after 20.Qf2 Bxa6 21.Bf4. Black has enough
material for the exchange, but his bishop can’t challenge White’s control over the dark squares.
18...g6
For the second time, Topalov weakens the dark squares on the kingside.
If White was forced to retreat after 18...a6 then Black would have nothing to complain about, but there is 19.Bg5! which
wins on the spot.
Kramnik gives 18...Nxe5 19.Bg5 Bf3 20.gxf3 Qxg5† 21.Qxg5 Nxf3† 22.Kg2 Nxg5 23.Bxe8 Rxe8 and although Black
has enough material for the exchange, he is still a move short of consolidating. White can exploit this with 24.Rd7.
22...Qf7 23.b3 Qf8 24.Qf4 Rc2 25.h4 Rac8 26.h5 Qe8 27.Rd3 R2c3 28.Rad1 gxh5 29.Rxd5! exd5 30.e6
30...R3c7
Kramnik writes that Topalov probably missed that he has no defence after 30...R3c6 31.Rd3 Rxe6 32.Rg3† Kf7
33.Qh6!. For example: 33...Rxf6 34.Qg7† Ke6 35.Re3† and wins.
31.Rxd5 Qxe6 32.Qg5† Kf8 33.Rxf5 Rf7 34.Qh6† Ke8 35.Re5 Rc6 36.Qxh5
1–0
It must be said that Kramnik uses the e3-systems only when Black is committed to ...e7-e6, thus not being able to
develop the bishop to f5 or g4. But the attitude is clear. When he caught a big tasty fish on his hook, he illustrated that
it’s time for practical openings.
The repertoire in this book suits players who like to play chess. There will be fewer games where Black loses straight
out of the opening, but it also avoids the kind of dull positions that often arise from sharp lines.
Another plus is that it takes less time and effort to prepare. Over the last ten years, I have normally tried to remember
a few thousand moves before a single game. But when the first draft of this book was finished, I simply read what I had
written. And it was maybe no surprise that I suddenly had more energy during the games.
What is the best way to learn opening theory? Much has been written and I have probably given some advice myself.
But the question already in some ways signals the wrong attitude. Moves should not be remembered, they should be
understood. When memory artists remember long series of numbers, they create an artificial meaning by transforming
the digits into pictures, years or places. In chess, we do not have to do that because there already exists a true meaning.
(Or maybe that’s a philosophical question?)
Human memory is based on concepts. If we have understood the logic behind a move, it’s much easier to find it at the
board. This book contains a lot of material, but don’t check the lines too many times; take it slow and trust your brain to
organize it.
Talking about memory, I warmed up for the 2017 Swedish Championship by playing a lot of ‘Memory’ (also known
as ‘Concentration’, this is a card game where all the cards are placed face down, then flipped over two at a time before
being put face down again, and you have to remember where every card is). It was a way to get back into competitive
mode after a long break. Insufficient focus is punished much harder than in chess. To remember the images and places,
I transformed them into chess moves and openings. I gave them meaning...
To my ears, “a practical opening” and “a playable position” have both been negative phrases – synonyms for something
that doesn’t give an advantage. And it’s a valid question to ask if this isn’t just a second-rate repertoire that only gives
an academic advantage.
But there’s no reason to fear the answer.
Akademia was a grove in ancient Greece with sacred olive trees, enclosed by a wall. Plato used the garden to lecture in
philosophy, and his school became known as the Academy. It was given great importance and is the foundation for
much of today’s philosophy.
Two thousand years later, much bureaucracy has entered huge academies, and the word academic has a second
meaning: formalities that we pretend are important, but are only grandstanding.
But is an academic advantage in chess of only bureaucratic importance? Something we can use to claim that we
played better than the opponent, but that doesn’t give any extra points?
No. Even top players can’t be sure to make a draw just because the position is a draw with best play. If the position is
harder to play, they make more mistakes. And this is even more the case at lower levels.
When the Informant symbols appeared in the 60s, ‘²’ meant an advantage that’s not enough to win the game if the
opponent defends well. However, over time we have adapted to the computer’s evaluations and today ² is seen as an
advantage of between 0.3 and 0.7 pawns.
In go, on the contrary, the computers haven’t made much impact yet. Positions are “balanced” or “advantageous” – it
makes no sense to try to make a more exact evaluation. A Japanese pro has claimed that you win if you always make
the 80% move, a move that’s honte (proper).
Even AlphaGo’s victory over Lee Se-dol won’t change much. The computer’s knowledge is based on human games
and its playing style is surprisingly humanlike.
I think it makes sense to go back to human evaluations in chess also. I don’t make it a secret that Black is okay in many
positions in this book. But they are not equal either; we have a positive imbalance that often gives a tiny Carlsenesque
advantage. Maybe some space on the queenside. An extra pawn in the centre. Or the bishop pair.
In each case a lot of work is required from the opponent to keep the advantage academic. And such an advantage is
valuable.
Position 1
White has an isolated pawn, but it’s not the same kind as Black gets in the Tarrasch. There is no pressure against d4,
since the knight on d5 obstructs the queen, and the bishop is placed on e7 instead of g7.
After provoking the weakening move ...h7-h6, White continues by transferring the queen to e4.
12.Qe2
If it was White’s move again, 13.Qe4 Nf6 14.Qh4 would be a possibility. However, it’s probably better to develop
first. The rooks belong on d1 and e1 – White focuses on the kingside since Black’s bishop is on c8. (With the bishop on
g4, it would make more sense to use the c-file.)
Black’s plan is to complete development without losing control over d5, but that’s easier said than done: ...b7-b6 can be
met with Nxd5, and if ...Qxd5 then Be4.
Another way is ...Re8, ...Bf8 and ...Bd7, allowing the exchange on d5 with an equal position. But there’s nothing that
forces White to take.
Curiously, I reached this position two days after I had written about it. My opponent played ...Re8 and ...Bf8, which
fortified his king but gave him a passive position. I decided to play on the queenside with Be3, Rc1 and Ne4-c5 and
after exchanging two minor pieces (Qxe3 and d4xc5), I gained a queenside majority which made it all the way to the
eighth rank.
The game can be found in Chapter 11 about the Panov.
Position 2
When the bishop moved to b2, the diagonal was blocked and the pawn on d4 could move nowhere. But White relied
on the fact that Black had to play ...c7-c5 sooner or later in order to free himself.
Both side’s pawn structures will be identical after the exchange of Black’s c-pawn for White’s d-pawn. But White’s
pieces are better placed. He has two good bishops, a more active knight and it is easier to find a good square for the
queen. 12.Qe2 is a move to consider, or 12.Rc1 first. The other rook goes to the d-file.
And we should not forget about the possibility of winning a tempo with e3-e4.
Position 3
In a blitz game between Alexander Grischuk and Ian Nepomniachtchi in Riga 2013, Black tried to play the Grünfeld,
but since there was no knight on c3, he had to answer e3-e4 with the retreat ...Nd5-b6.
He has failed to get enough counterplay against the d-pawn (it should be noted that ...Bxd4 loses a piece) and he thus
has no compensation for the pawn centre.
10.h3!
A good move, getting the bishop pair and opening up for the bishop on c1. It has awaited its time, but not hindered
White from gaining a long-term advantage. Next he will simply complete development. Grischuk won after 55 moves.
Position 4
A messier position than the previous ones. In the King’s Indian, Black has been allowed to push the e-pawn to e4. It is
better there, but it has cost quite a few tempos with ...Nbd7 (to get in ...e7-e5), ...Re8 and ...Qe7 (to defend e4).
White has used the time to run with the a-pawn. One idea behind it is to play Nc3-d5 and, after ...Nxd5 cxd5, be able
to put pressure along the c-file with Ra1-a3-c3 and Rd1-c1. But if so, the pawn on d5 will be weak as well.
All in all, there are still quite a few moves before something real will be happening on the queenside. Instead, White
relies on the fact that he is actually defending himself quite well on the kingside. He has answered ...h5-h4 with h2-h3,
avoiding a back rank mate in the endgame (something that should not be underestimated!).
If Black had placed the knight on g5, instead of the pawn, Be2-f1 would be a standard way of defending against a
piece sacrifice on h3. Instead Black is trying to use the h3-pawn as a hook to open up the position, but his last move
(15...g5) has exposed the knight on h7 to some tactics.
16.a6
Exposing the rook on a8.
16...b6
Not the best move, but a logical one to keep the queenside closed.
It’s difficult to decide whether to label the sample positions ² or = isn’t it? But the important thing is that we appreciate
them, even though they are objectively unclear.
Unclear Positions
In a way, this book was more difficult to write than Pump Up Your Rating. In an opening book, it’s possible to question
the evaluations, and with that the foundation of the book.
But in my opinion many authors make it harder on themselves by striving too hard in their goal to label all positions
with ². In the long run, optimistic evaluations favour no one and sometimes we have to accept that the position is just
unclear. So Informant symbols will be rare in this book, but I will also avoid hiding behind vague evaluations such as
“with chances for both sides”.
From time to time, though, I am sure you will be frustrated by not getting a clear answer on the question if White is
better or not. But I am not hiding the answer to be evil; it’s just that chess isn’t black and white.
Some of the chapters can be read on their own, but given how the material is organized, with pawn structures,
instructive games and exercises, it’s not possible to find everything in the theoretical sections. For pedagogical reasons,
some lines are given in connection to a game, so you have to buy the whole package, reading the chapters from scratch.
This book asks for an effort from the reader, but if you put in the energy you will be rewarded with something you can
play for a long time.
But isn’t there a risk that the material will be too disorganized? It’s impossible to learn all the possible move orders.
It’s very human, at least for adults, to strive for control (or rather the feeling of control). But it’s doubly beneficial to
get rid of it. Organizing chess knowledge risks over-simplifying, which hampers intuition.
And don’t worry, I will not go as far as a recent opening book that arranged the chapters in random order.
So are you ready to strive for interesting (and often also advantageous) positions rather than entering a theoretical
battle? Go ahead! It took me ten years to understand that such an attitude also works for grandmasters.
A ‘practical repertoire’ and a ‘grandmaster repertoire’ is not a conflict. It is still about understanding the variations
you play.
This book is based on a set-up with the moves d4, Nf3, c4 and e3. They could be played in twenty-four different orders,
and it is possible to make a case for many of them. But before we discuss the move orders in Part Three, we have to
familiarize ourselves with the general ideas. (At first I wasn’t comfortable about using we, but then I understood that I
was writing the book for myself as well.)
Unfortunately, White cannot play according to a simple grand scheme. It depends on Black’s choice of opening, and
every chapter will thus begin with a summary of typical plans. They will be explained from a few diagrams showing
the common pawn structures, and the pieces are added only if they are necessary for the discussion. That’s often when
they stand in the way of a pawn lever (a pawn move that puts the pawn in contact with a pawn of the opponent).
Then there will follow a few illustrative games, before the lines are presented just like in an old-fashioned opening
book. And finally there are some exercises.
The following Q&A discusses the pros and cons of the e3 poison repertoire.
Q: Playing e2-e3 without being forced seems passive! What does White get in return for shutting in the bishop on c1?
A: The question almost answers itself. By keeping the bishop on c1, White avoids the early confrontation that might
give Black space or lead to simplifications. With the bishop on g5, Black might push ...h7-h6 and ...g7-g5 (e.g.
the Ragozin or the King’s Indian) or play ...Nf6-e4 as in the Lasker Defence. And a bishop on f4 can be hit by
...e7-e5 or ...Bd6.
Instead, we keep the tension. But of course, Black’s centre is not challenged as much as with the bishop on g5.
Q: Isn’t going for flexibility a strange choice as White? Black will have more information whatever we do.
A: Another negative question! Let me first explain the advantage of having the pawn on e3 rather than on e4. Just like
with the bishop on c1, it keeps the tension. White is able to pressure the centre with c2-c4, but still keeps control
over d4 and c4 (which would not have been the case with a kingside fianchetto).
One example that illustrates the difference is the King’s Indian. When White has played e2-e4, Black is normally
able to put enough pressure on the d4-pawn to force d4-d5. With the pawn on e3, that’s no longer possible.
Q: What if I am lazy and don’t read the whole book. Can you tell me when this system comes into its own and when it
struggles?
A: It’s good against the Grünfeld. It’s not for nothing that Hikaru Nakamura and Anish Giri have used it against the
Grünfeld expert Maxime Vachier-Lagrave. I also like it against the King’s Indian, because even though it usually
leads to positions where White attacks on the queenside and Black on the kingside, it’s a version that the Black
players are not familiar with. And if you start with a move other than 1.d4, it avoids the normal Queen’s Gambit
Accepted. We will see one game by Kasparov later. Finally, Nimzo-Indian players won’t get the chance to place
the bishop on b4.
Those are the positive sides.
One opening where Black gets a clearly better version is the Benoni. In the worst possible case it’s a full tempo,
but luckily we use a prophylactic strategy where e3-e4-e5 is spared for later. And with the pawn on e3 rather than
e4, Black can’t use some of his standard ways of creating counterplay.
Black can also play an improved Tarrasch, where White is no longer able to put maximum pressure on the
isolated pawn. Remember what I said earlier about this kind of position. But instead of accepting his fate, White
can try another idea.
White puts no immediate pressure on the centre, so Black is free to develop in any way he likes...
It’s not possible to memorize every single move, so you have to trust your understanding...
If you don’t play well, you might play just another Colle...
...but he can’t play against White’s dark-squared bishop or against the d4-pawn
...but they are still unfavourable, and if he plays a passive move we can use the development advantage to punish
him
...but if you play well, you have a dangerous repertoire while Black doesn’t smell the danger
Shameless Advertisement
In May 2016, I started to use the repertoire more often. Most of the analytical work was finished and it was a great
feeling to play without preparing much.
1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 Nc6 6.Nf3 Bg4 7.Bg5! 7...e6 8.cxd5 exd5 9.Qe2† Be6 10.Bxf6
10...gxf6
We will see that it’s possible to give up a pawn with 10...Qxf6.
11.Qc2 f5?
Grischuk played 11...Qb6. Moving the f-pawn is too slow. Goganov wanted to put his queen on f6, but allowing Nf3-
e5 is not a good idea.
Besides a great position, I had gained 45 minutes on the clock.
12.Bb5
The positional threat is Bxc6† followed by 0-0 and Na4, when Black is completely deprived of pawn levers.
12...Rc8
13.Bxc6†!?
Giving up the second bishop to win a tempo on the rook. White can’t spend too much time to force ...b7xc6. After
13.Ne5 Bd6 14.Bxc6† bxc6 15.0-0 0-0 16.Na4 White is one move away from complete control, but it’s not enough:
16...c5!
18...0-0
It’s not possible to capture the pawn: 18...Rxd4 19.Nxe6 fxe6 20.Qh5† and the bishop is lost.
Giving Black something to play for: the e4-square. I don’t play well in the following phase.
25...Ba6 26.Nb4 Bc4 27.Re1 Rde6 28.Rxe6 Rxe6 29.Re1 Qe8 30.Nc2?!
The last chance to keep the advantage was 30.Rxe6 Qxe6 31.Nd2 keeping the queen away from e4.
33...Bb3 34.Na1 Ba4 35.Nf5 Bf8 36.Ne3 Bc6 37.Kf2 h5 38.Nd1 Kf7 39.Nc3 f5 40.Nc2 Bg7 41.Ke3
½–½
Playing in the same tournament was Aditya Subramanian, who was one of the ‘test readers’ I mentioned at the start of
the book. For his White openings he followed this book.
“It felt a bit daunting to use the repertoire,” he said afterwards. “I had never played most of these lines before but I
was confident in them and my understanding of the ideas. I felt that compared to other repertoire books I have used I
understood the nuances of the positions a lot better. I played the repertoire 4 times out of 4 White games and had more
or less winning positions by move 15 in every game.”
He didn’t win any games with Black, but yes, the sample size could have been greater...
To nuance the picture, Subramanian didn’t like White’s position in the King’s Indian, even though he couldn’t come
up with a refutation.
4...0-0 32
4...d5 32
5.Be2 d5 32
5...b6 32
6.cxd5 Nxd5 7.e4 Nb6 8.0-0 Nc6! 32
8...f5 32
8...c5 32
8...Bg4 33
9.d5 Ne5 10.Nc3 e6! 33
10...c6 34
11.Bf4 34
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3! 0-0 5.Be2 d5 6.cxd5 Nxd5 7.e4 Nb6
It’s true that e2-e3-e4 costs a tempo, but Black’s dissatisfaction with the knight’s retreat is greater. In the Anti-Catalan
there is in fact an identical line with reversed colours: 1.d4 Nf6 2.c4 e6 3.g3 c5 4.Nf3 cxd4 5.Nxd4 d5 6.Bg2 e5 – even
with a tempo less Black can afford to move the e-pawn twice.
The name Sneaky Grünfeld was invented by Ken Smith and John Hall in Winning with the Colle System (1990) and
referred to 1.d4 d5 2.Nf3 Nf6 3.e3 g6, where Black tried to sneak into the Grünfeld. In this chapter, it is White who
tries to sneak away from it!
Pawn Structures
Structure 1
In the Grünfeld, White has normally recaptured b2xc3 when the knights were exchanged. It seems that it strengthens
the centre, but there are actually three reasons why it’s not so. In the diagram position, White can:
a) Answer ...c7-c5 with d4-d5 without leaving the pawn on c3 en prise (Black has a bishop on g7)
b) Capture on c5 without getting an isolated pawn on c3
c) Stop ...c7-c5 with Rac1
a) If Black attacks the centre with ...Bg4, White often plays Nbd2 (the exception is the development scheme outlined
above) followed by h2-h3.
b) If Black plays ...e7-e5, White can capture, win a tempo with f2-f4 and use the 4–3 majority on the kingside. Or he can
move past (d4-d5) and limit the scope of the bishop on g7.
c) The first question against ...c7-c5 is if it’s good to capture the pawn. If it is – go for it. Otherwise, d4-d5 is good as
well. Now a further e4-e5 would seriously cramp Black’s position, so usually Black hits first with ...e7-e6. And
then we reach the next diagram.
d) Sometimes Black plays ...Nb8-c6 and manages to provoke d4-d5 without moving either the c-pawn or the e-pawn. In
theory, he could then be more successful in challenging White’s d-pawn (both ...c7-c6 and ...e7-e6 are possible)
but the bad news is that the strategy involves moving the knight several times (just as with the other knight).
e) ...f7-f5 is discussed under Structure 3.
f) If Black decides just to sit passively, White makes use of the half-open c-file.
Structure 2
Strictly speaking, the d-pawn becomes passed only after the exchange on d5. But Black has no good alternatives; it is
unlikely that he will manage to undermine the pawn with ...f7-f5.
After ...e6xd5, White takes back with the e-pawn and keeps the knights on the board. It makes a great difference that
they are there.
White’s plan is to build up behind the passer with Bf4, Qd2 and Rd1. Finally, the pawn takes a step into what is now
safe territory.
Structure 3
White has three reactions against ...f7-f5.
a) Taking on f5, but that’s not a good idea if Black can take back with the bishop – it helps him to develop.
b) Defend with Nc3. If Black exchanges, it’s White who develops for free. Black has weakened the kingside, but
White’s isolated pawn is more exposed than Black’s. So on general grounds, it’s not possible to say who it
favours.
c) Play e4-e5, which is the standard reaction. Black gets an outpost on d5, but it is difficult to use. White develops with
Qb3 (check!) and Nc3, plays Rfd1 and might then consider a2-a4-a5. The big problem for Black is that he can’t
get in ...Bc8-e6. However, note that e4-e5 works badly if White has a bishop on e3 that then runs into ...f5-f4.
Games
On behalf of the Swedish chess magazine Tidskrift för Schack, I interviewed Rudolf Forsberg, a 98-year-old study
composer. Three years earlier, he had woken up in the middle of the night and vaguely recalled a position he had seen
in his dream. He immediately got up and tried to set up the position on a board, but failed. Since that day it has
occupied his time, and he has come closer to recreating the position.
His hand was sore and full of fragile veins, but it moved quickly. This, that, tic, tac, check, check, capture and Black
was stalemated. Next variation. This time it’s White who is stalemated. Then he removes the bishop and we have a new
problem. The bishop appears again when Black underpromotes – and so does White. It is a magical dance and I am
perplexed, missing almost every word. But the 98-year-old is not satisfied. Something is missing before the study is
perfect.
During the interview, Forsberg strongly criticized the chess magazine, not really understanding that he was sitting in
front of the editor-in-chief. As an active player, a few decades earlier, stress prevented Forsberg from enjoying his
games. Only as a composer was he able to experience the real beauty of chess. And while playing through games, he
didn’t want to be disturbed by the players’ comments. He just wanted a lot of unannotated games.
As was the case when he first subscribed to the magazine in 1935.
Now they were all gone.
While I might agree about the stress of playing chess, I don’t agree with his preference for bare game scores. So the
games in this section will be annotated – but only briefly after (if) they stop being relevant to our topic.
But I also feel that I have to defend my habit of showing complete games. When I was young (and naïve) I didn’t like
such opening books. “Just give me the moves to remember. I can study middlegames later.” However, in the long run
it’s quite clear to me that the payback is higher if understanding is given priority over memorizing.
Memory also works better if we have something to hang it on. It’s easy to remember that we should play like
Kramnik. Another facilitating feature is to add a few pedagogical exclamation marks (!) to important moves that aren’t
necessarily better than others, but which are in line with the repertoire.
But what happened to the dream problem? Forsberg estimated that he needs to work on it for a few more years.
Meanwhile we will start with the games.
10.d5 Bxf3
Nepomniachtchi played 10...Nb8 but was worse after 11.a4! followed by a4-a5 and h2-h3.
10...Ne5 11.Nxe5 Bxe2 12.Nxf7 is a well-known trick that collects a pawn, but Black gets some counterplay after
12...Bxd1 13.Nxd8 Raxd8 14.Rxd1 e6.
Instead, White does better with 12.Qxe2 Bxe5 when his advantage consists of his space in the centre. There will soon
be a rook on the d-file, so Black can’t play ...e7-e6 or ...c7-c6 easily. Still, this was Black’s best option.
14.a4! a5
Including the moves with the a-pawns has seriously weakened the b6-square.
18.a5
18...Nd7
18...Nc8 19.e5 f6 would fail to 20.Bg4! and Black can’t win the e5-pawn.
19.e5 b6 20.Bb5 Nc5 21.Bc6 Ra7 22.axb6 Rxa1 23.Rxa1 cxb6 24.Ra7
White has the bishop pair, a better pawn structure and an active rook. That’s dangerous in the hands of Ivanchuk.
24...e6 25.d6 f6 26.Bd4 fxe5 27.Bxe5 Bxe5 28.fxe5 Nd3 29.Bd7 Nc5
29...Nxe5 30.Bxe6† Kh8 31.Rc7 wins as there is no defence against 32.d7 and 33.Rc8.
With ...Bg4 Black tried to undermine White’s centre, just like he usually does in the Grünfeld. But he failed and it’s
fitting that the d-pawn had the final word.
In the next game we will see a more prudent strategy from Black.
Vachier-Lagrave, April 2016: “It’s a decent variation, there are some tries but it should be okay to play the Benoni with
a tempo up.”
Giri, April 2016: “For sure it’s a real variation, but Black could play the King’s Indian with e3.”
5...d6 6.Nc3 d5 7.cxd5 Nxd5 8.e4 Nxc3 9.bxc3 would be a funny way to transpose into a normal Grünfeld. Even
though White has a favourable version (Black has castled instead of putting immediate pressure against the centre),
there is no reason to enter it. 6.0-0 keeps the knight away from c3.
6.0-0 Bb7
7.Nc3
White would be happy to find another useful move, but it’s not easy.
7.d5?! was played by Hans Tikkanen against David Howell at the Tromsø Olympiad in 2014. The pawn move is
unjustified: it doesn’t win a tempo and Black can challenge the d5-pawn with both ...e7-e6 and ...c7-c6. After the game,
Tikkanen recommended 7.Nc3.
7.b3?! runs into 7...c5 and the d-pawn can’t move. But I like 7.b4!? as played in Arnaudov – Bok, Karpos 2014.
7...d5 8.cxd5
8.b4!? still has some venom. The idea is to answer ...c7-c5 with b4xc5 and Rb1, hitting the bishop on b7.
8...Nxd5 9.Bd2
9.Qb3 was played by Daniil Dubov against Vachier-Lagrave in Qatar 2014, but the bishop move makes more sense.
White’s first priority is to get a rook to the c-file.
Black has two fianchettoed bishops and that’s lovely, but to be happy he also has to play ...c7-c5.
9...c5
9...Nd7 prepares ...c7-c5, but it forces Black to take back on d5 with the bishop. After 10.Nxd5 Bxd5 11.Qc2, e3-e4
follows and 11...c5 12.e4 Bb7 13.d5 leads to one of the discussed pawn structures, but with Black having difficulties in
breaking with ...e7-e6.
12...Nd7!
A radical solution, but thematic in the Grünfeld.
12...bxc5 13.Qa4 allows White to disturb Black along the b- and d-files. As in many positions, e2-e3 was not a slow
move and White has managed to open the position with an advantage in development – one tempo better than when the
game started.
13.cxb6 axb6
Black argues that two open files against isolated pawns gives enough compensation for the sacrificed pawn. With
White’s pawn on e4 (another weakness) that would be true. It also favours White that the dark-squared bishops have
been exchanged, as the bishop on g7 would have exerted pressure against c3.
14.Qd4!
The extra pawn also gives White’s pieces more squares to choose between. Neither d4 nor b4 can be challenged by
the pawn on b6.
To me, it seems rather unnecessary to give up control over the d5-square; Giri had still not showed anything real. But
I am sure Giri wasn’t in a hurry to do so, as there are many useful moves left: doubling the rooks and bringing the king
to the centre.
After the text move, White has a clear plan: to bring any piece to d5. The rest of the game is only briefly annotated.
22.Nb3 Ra7 23.Nd2 Rc8 24.c4 Bc6 25.Rb2 Kg7 26.f3 Re8 27.Kf2
Next would be Nd2-b1-c3-d5, and Vachier-Lagrave decides that he can’t sit and wait.
27...e4
Giving up another square: d4.
28.Nb3 exf3 29.gxf3 Ba4 30.Rd1 Re5 31.Rd5 Rxd5 32.cxd5 Bxb3 33.axb3 Kf6 34.f4 g5 35.Rc2 gxf4 36.exf4
41.Kg3 h5 42.Kh4 Rf5 43.Rb4 Rd5 44.Rc4 Kf6 45.b4 Kg6 46.Rc6† f6 47.Rc2 Rb5 48.Rb2 Kf5 49.Kg3 Ke4 50.Rb1
50...Kd3
50...Ke3 was mentioned by Giri as a draw a few hours after the game.
51.Kf3 Kc2 52.Ra1 Rxb4 53.Ra5 Rb3† 54.Ke4 h4 55.Kf5 Kd3 56.Kxf6 Ke4 57.f5 h3 58.Ra4† Kf3 59.Kg5 Rb5
60.Ra2 Rb4 61.f6 Rg4† 62.Kf5 Rf4† 63.Ke6 Re4† 64.Kd6 Rd4† 65.Ke7 Re4† 66.Kf8 Rb4 67.f7 Rf4 68.Rb2 Ra4
69.Rb6
Black can’t stop the pawn from promoting while getting the h-pawn.
1–0
White again had the final word in this game, and he will have an overwhelming score throughout the book, since those
games illustrate our case best. But don’t worry, as there are also six victories for the opponent, two of which are in the
next chapter.
Recap
The Sneaky Grünfeld is not a pleasant story for Black. It’s easy for him to get frustrated when he can’t get the
counterplay he is used to. But he has to stay calm, not allowing d4-d5 with tempo and giving up the bishop pair, as
Tabatabaei did. The improvement 8...Nc6! will be analysed in the theoretical section.
The Grünfeld move ...d7-d5 is better played in the same way as Vachier-Lagrave used it (with ...b7-b6 and ...Bb7
first). Still, to get a pleasant position, White doesn’t have to do much more than develop slowly (Be2, Bd2).
This will also be the path in the Schlechter Slav. There, it’s important to know how to react against an early ...Bg4 or
...Nbd7 (c4xd5) and how to play if Black stays passive (Ne5 or b2-b4).
Theory
5.Be2 d5
5...b6 6.0-0 Bb7 7.Nc3 d5 8.cxd5 Nxd5 9.Bd2 was seen in Giri – Vachier-Lagrave, World Cup 2015. And as pointed
out in the annotations, 7.b4!? is an interesting possibility.
5...c6 should, after a later ...d7-d5, transpose to the Schlechter Slav (page 200).
7...Nb6
8.0-0
8.a4 is similar to Rudel’s idea, but without the possibility of Bb5†. The difference can be seen after 8...a5 9.0-0 Bg4
10.Nbd2 Nc6. This move would not have been possible if Black had to play ...c7-c6 to defend against the check on b5.
After 11.d5 Nb4 Black gets the use of the b4-square.
8...Nc6!
8...f5 is most precisely met by 9.Qb3† Kh8 10.e5 with an advantage.
White has nice compensation on the dark squares after 11...Bxc3 12.bxc3 exd5 13.e5!.
If 11...exd5 then it seems good enough to play 12.Nxd5 Nxd5 13.exd5 when it’s not easy for Black to develop and
blockade the d-pawn. After 13...Bxb2 14.Rb1 Bg7 15.Re1, White’s first threat is 16.Bb5 followed by 17.Be7.
8...Bg4 9.Nbd2!?
9...Nc6
9...e6 10.h3 was the previously mentioned blitz game between Alexander Grischuk and Ian Nepomniachtchi.
9...f5 10.e5 gives a pawn structure discussed in the beginning of the chapter, but with the bishop on g4 instead of
c8. That’s an improvement, but Black still has problems with his light squares. White continues with h2-h3 and
Qb3(†).
9...e5 is an interesting try, but after the simple 10.dxe5 Nc6 11.h3 Be6 12.Qc2 Nxe5 13.Nxe5 Bxe5 14.Nf3,
White will get a lot of tempos with Bf4, Rd1 and Nd4. Black has problems.
10.d5 Nb8
10...Ne5N 11.Nxe5 Bxe2 12.Qxe2 Bxe5 and Black is satisfied with the exchange of two minor pieces, but he still
lacks space. If White is quick to place his rooks in the centre, he will prevent Black from playing ...c7-c6 or ...e7-
e6 in a good way. 13.Rd1, 13.Nf3 and 13.a4 are all moves to consider, but 13.f4?! is too much: 13...Bd4† 14.Kh1
f5! And White’s centre collapses, with an unclear position.
11.a4! c6 12.a5 N6d7 13.h3 Bxf3 14.Nxf3 Nf6
Now Grischuk played 15.Qb3 in his second game against Nepomniachtchi. That might be good, but it was easier to
enjoy the bishop pair after 15.dxc6.
9.d5
9.Be3 is a mistake due to 9...f5!. White is in general happy with the position and doesn’t wish anything other than
peace and quiet until he has completed development. But with the bishop on e3, the ...f7-f5 lever is well motivated. The
reason is that 10.e5 is met by 10...f4! 11.Bc1 Bg4 and Black gets counterplay against d4.
9...Ne5 10.Nc3
10.Nxe5 Bxe5 11.Bh6? would be a mistake due to 11...Bxb2.
10...e6!
10...c6 allows 11.Nxe5 Bxe5 12.Bh6 when Black will have a weakened king whatever he plays.
11.Bf4
Instead 11.Nxe5 Bxe5 12.Bh6? fails to 12...Qh4 or 12...Bxh2†.
13...g5!
14.Bg3 f5! is the main point, harassing the dark-squared bishop. Instead I recommend a surprising move.
14.Bc1
The ambition is only to complete development, for example with Qd3/c2/b3, Be3 (if there’s no ...Nc4), Rad1 and
Rfe1. Black has a stable blockade on d6 and can put his knight there, but one day we hope he will regret advancing the
pawn to g5. One way to force further weaknesses is Qd3 followed by Bd1-c2.
Exercises
Some exercises are positions that are covered in the theoretical section. There are two ways to work with them.
a) Before reading the theoretical section. You digest the moves better if you have started by shaping an opinion of
your own.
b) At the end, as active repetition.
There is no scoring system other than your own conscience, so no temptation to cheat. After all, the real test takes place
only when you play the positions in tournament games.
Exercise 1
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 d5 6.cxd5 Nxd5 7.e4 Nb6 8.0-0 f5 9.Qb3† Kh8
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 d5 6.cxd5 Nxd5 7.e4 Nb6 8.0-0 Nc6 9.Be3
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 c6 6.0-0 d5 7.Qb3
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 4
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 d5 6.cxd5 Nxd5 7.e4 Nb6 8.0-0 Bg4 9.Nbd2! Nc6 10.d5 Nb8
White to move
Show/Hide Solution
Show/Hide Solution
6...Nbd7 52
6...Nc6 52
6...Qe8 52
6...c6 52
6...c5 52
7.0-0 e5 8.Qc2 Re8 52
8...b6 53
9.Rd1 e4 53
9...exd4 54
9...c6 54
9...Qe7!? 54
10.Nd2 Qe7 11.b4 Nf8 54
11...c6 54
12.a4(!) h5 54
12...a5?! 55
12...Bf5 55
13.a5N a6! 55
13...N8h7 55
13...h4 55
14.b5 55
A) 14...Bg4 56
B) 14...h4 56
C) 14...Bf5! 57
The King’s Indian is impossible to avoid, unless you play 1.e4 or the Trompowsky, and Black almost makes his first
five moves without taking notice about White’s set-up. So why on earth would White restrict himself to e2-e3 followed
by Be2? It doesn’t look threatening at all.
True, 4.e3 is a move designed to meet the Grünfeld and having to play it against the King’s Indian is something that
follows as part of the package. But it has its points as well, and leads to a different kind of position than in the normal
King’s Indian lines. Whether you like the positions at first sight probably depends on how you feel about defending
against an attack. But as we will see, there are several rules of thumb that facilitate the defensive task.
The main point behind 4.e3 in the King’s Indian is that Black can’t play ...e7-e5 without preparation. In the line 4.Nc3
0-0 5.e4 d6 6.Be2 e5, Black’s tactical point is 7.dxe5 dxe5 8.Qxd8 Rxd8 9.Nxe5 Nxe4! – but this is not possible
without a pawn on e4. (But not 5...e5?? as I played when I was 13 and was talked into playing the King’s Indian in a
crucial game. It took me ten years to try it again.)
So how is Black preparing ...e7-e5? 6...Re8 does not defend e5 properly (the rook has to take back on d8) and 6...Nc6
is met by 7.d5. So he has to start with 6...Nbd7 and that is already a small concession.
The next question arises after 6...Nbd7 7.0-0 e5 8.Qc2.
In the ...Nbd7 King’s Indians, Black usually follows up with ...e5xd4, ...Re8 and ...Nc5, hitting the pawn on e4. But
White will take back on d4 with the pawn, so that plan is no longer available.
Black should develop the light-squared bishop, and it has two diagonals to choose between.
a) Fianchetto: 8...b6 9.dxe5! Nxe5 10.Nxe5 dxe5 With an extra tempo, Black would be fine, but now White is fast with
annoying moves like Rd1, Ba3, Bf3 and Nd5.
b) Normal is ...e5-e4 followed by ...Nd7-f8 and ...Bf5. It closes the centre and allows both sides to start the wing attacks
that are typical in the King’s Indian. However, there are two important differences with the black pawn on e4.
1) White’s knight on f3 is kicked away, so he has fewer pieces defending the king.
2) Black has to spend two tempos defending the pawn: ...Re8 and ...Qe7.
The normal continuation is 8...Re8 9.Rd1 e4 10.Nd2 Qe7 11.b4 Nf8.
The position we get is actually a Reversed King’s Indian Attack. The original line goes 1.e4 e6 2.d3 d5 3.Nd2 c5
4.Ngf3 Nc6 5.g3 Nf6 6.Bg2 Be7 7.0-0 0-0 8.Re1 Qc7 9.e5 Nd7 10.Qe2 b5 11.Nf1.
It’s considered promising for White and he has about 60% in top games. And the only difference is that we have won
the semi-useless Rd1 (...Rd8). The engines consider that, on average, to be worth 0.07 pawns.
It’s right that the rook doesn’t do much, but it has vacated the f1-square. We will see many positions where a knight
or a bishop can use the square to defend against Black’s attack.
Alexander Grischuk believed in the white side of the Reversed King’s Indian Attack and played it in an important
game against Fabiano Caruana in the 2013 Paris Grand Prix. Does that convince you? I hope not. Trust the moves, not
the authorities.
The following examples navigate through the similarities and differences with the rook on d1, even though it’s like a
labyrinth – especially since the theory doesn’t reach a verdict in the original position.
However, it’s a relief that it’s possible to play according to a quite clear-cut plan.
Pawn Structures
Structure 1
The key is Black’s e-pawn. It will be attacked with Nc3, Nd2 and Qc2, and defended by ...Nf6, ...Re8, and ...Qe7.
Black can also consider overprotecting the pawn with ...Bf5 to make all of his pieces flexible.
A black pawn on e4 means no knight on f3, which in turn means that Black has good attacking possibilities. White
will not try to challenge the pawn with f2-f3, but put his hope on the queenside.
Black’s attack starts with ...h7-h5, ...Nb8-d7-f8-h7 and then one of the following plans:
a) ...h5-h4, which should be answered with h2-h3. Black continues with ...Ng5, to take on h3. The sacrifice is prevented
with Be2-f1.
b) ...h5-h4 followed by ...g6-g5-g4. But the pressure against the e-pawn makes it difficult for Black to recapture on g4
with a minor piece.
c) ...Ng4 with the idea of ...Qh4. White doesn’t want to give up the light-squared bishop and h2-h3 can often be met by
...Nxf2 or ...Nxe3. It’s better to try to stop ...Ng4 altogether, by keeping pressure on e4 or being ready with Nd5.
d) ...Bg4 followed by ...h5xg4 in case White exchanges.
e) ...Nh7-g5-(f3) However, it’s even better for Black to start with ...Bf5 and then manoeuvre the knight via e6. It creates
the tactical threat 1...Nxd4 2.exd4 e3 with a discovered attack against the queen on c2.
White has different ideas of how to get things going on the queenside. A simple plan will be recommended: running
with the a-pawn as far as possible. If it reaches a6, it undermines Black’s control over the c6- and d5-squares, may open
up for Ra1-a5 (a defensive move!) and creates tactical possibilities against the rook on a8.
Following in the footsteps of Fischer, Black should probably block with ...a7-a6.
White also likes to move the c-pawn. It may be an idea to start with d4-d5, to stop Black from blocking the position with
1.c5 d5, but that has its drawbacks as well. The d5-square is occupied and Black gets the e5-square for a knight or a
bishop. White gets the d4-square, but the manoeuvre Nd2-b3-d4 is often too slow.
Another way to open the queenside is Nc3-d5. In case of c4xd5, White triples on the c-file with Rdc1 and Ra3-c3. But
he also has to watch out so the d5-pawn doesn’t become too weak.
All in all, the structure is very complicated and even after the game it could be difficult to point out which moves were
good or bad. But I think those guidelines make the position relatively easy to play over the board.
Structure 2
It would be easy to conclude that White’s space gives an advantage, but that would be far from true. The pawn on d4
is slightly weak, and opens the diagonal for Black’s bishop if it moves. Black also has a square to play for (e4), while
White doesn’t have any similar option.
When the same kind of position is reached via the Petroff (1.e4 e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d4 Nf6) White
has an easy game. The difference is the placement of Black’s bishop. In the Petroff it doesn’t have time to go to g7
(6.Bd3 g6? 7.0-0 Bg7 8.Re1†).
The good news is that Black can’t reach this position without making a concession. Normally he has blocked the
bishop on c8 by ...Nbd7. It means that White will have enough time to control e4.
After developing, White’s plan is either pushing the queenside pawns, without anything particular in mind, or fighting
for the e-file. Black’s position is cramped and he doesn’t have a natural square for the queen, and thus has problems
connecting his rooks.
By now, we can with good conscience conclude that we have a preferable structure.
Games
I have not found any good examples of Black winning with a kingside attack, so we have to see an example with colours
reversed. It’s quite tricky to flip the board in our heads, but surely good practice.
And the good news is that we will see the following masterpiece, even though I don’t think it’s as one-sided as
considered in previous chess literature.
1.e4 e6 2.d3 d5 3.Nd2 Nf6 4.g3 c5 5.Bg2 Nc6 6.Ngf3 Be7 7.0-0 0-0 8.e5 Nd7 9.Re1
9...b5
9...Qc7 would have forced White to defend the e5-pawn with Qe2, where it could be hit later by ...Nd4. 10.Qe2 b5
11.h4 a5 12.Nf1 Nd4!? was played in Fischer – Uzi Geller, Netanya 1968. 13.Nxd4 cxd4 14.Bf4 Ra6 15.Nh2 Rc6
16.Rac1
After 16...Ba6, Geller was hit by 17.Bxd5! exd5 18.e6, which gave him an awful position when White’s knight
continued from h2 to e5.
He should have played 16...Bb4 17.Red1 Bb7.
18...Rc8 is a serious threat. Since the rook has moved from e1, White’s queen is restricted to the defence of the e-
pawn. He has two options:
a) 18.Bxd5 exd5 19.e6 Qc8 20.exd7 Qxd7 21.Nf3 Bc5 22.Ne5 The knight coming here is no longer the end of the
world, since the queen on e2 is undefended. Black is fine after 22...Qe7 with 23...Re6 coming next.
b) 18.a3 Bc5 19.Re1?! Reinforcing e5 and planning 20.Bxd5 or 20.Qg4. But White is too slow after 19...b4 20.a4 Rc8
21.Qg4 (21.Bxd5? exd5 22.e6 Rxe6! wins material when the queen on c7 is defended.) 21...Bf8 22.Re2 b3 and Black
breaks through.
10.Nf1 b4
It is too late for 10...Qc7 when White has 11.Bf4.
11.h4 a5 12.Bf4 a4
13.a3!
My intuition says that this is a mistake, helping Black to open the queenside. But the move became standard after this
game and remains so even today. One reason to stop ...a4-a3 is that it would have undermined the c3-square and thus
diminished White’s control over the d4-square.
15.Ne3 Ba6
Another piece moves to the a-file.
16.Bh3
Fischer immediately takes advantage of it, hitting the e6-pawn. His threat is 17.Ng5 h6 18.Nxe6 fxe6 19.Bxe6† Kh7
20.Qg4 Ra7 21.Nf5, winning.
According to the engine, 16.Ng5 was stronger because 16...h6 17.Nxe6! fxe6 18.Bh3 is enough to give an initiative.
16...d4 17.Nf1
17.Ng4 looks more natural, but stops the queen from joining the attack.
23.Bf6! Qe8
23...gxf6 24.exf6 Kh8 25.Nf3! Rg8 26.Ne5 hits d7 and f7.
The best defence is 25...Nd5 showing that the journey to c3 was a shot in the dark. After 26.Ne5 Nxf6 27.Qh6 Qe7
White has only one move, but it’s enough.
24.Ne4
24...g6
Weakening the dark squares.
24...Nxe4 25.Rxe4 and the rook continues to g4. After 25...Rc7 26.Rg4 g6 27.Qh6 Qf8 28.Qf4 White continues with
h4-h5, as in the game, and Black’s position collapses.
28...Rc7 defends against the sacrifice on h7, but White can win in many ways. One is the straightforward 29.Bxe6 fxe6
30.hxg6 Qxg6 31.Qxg6† hxg6 32.Rh8† Kf7 33.Rh7† Kf8 34.Rxc7 as pointed out by Don Maddox.
29.Bg2!
Not the only winning move, but the fastest. The bishop is redeployed to sacrifice itself on g6.
29...dxc2?
29...Bb7 neutralizes the bishop, but also blocks the rook on a7 and thus allows 30.hxg6 fxg6 31.Rxh7! with mate.
30.Qh6 Qf8
Black could delay his fate with 30...c1=Q† 31.Rxc1 Rxc1† but White doesn’t even have to take the rook.
31.Qxh7†!
Black resigned due to 31...Kxh7 32.hxg6† Kxg6 33.Be4 mate.
1–0
The game didn’t end well for our side, but even so there are many reasons to be optimistic. Even though Black
voluntarily moved his pieces to the queenside, and then delayed opening that side of the board, Fischer’s attack wasn’t
decisive. Even at move 20 it seems there are still defensive resources.
The engines have changed chess by showing that it’s possible to defend ugly positions. As a consequence, the top
players have become more concrete and less “positional” or, one might say, they have refined their positional intuition.
A justified question is if this matters to you and me. We are still humans and can’t calculate like Houdini or Grischuk.
But knowing the approximate evaluation helps a lot. After having seen this game, we may conclude that Black’s king is
not so easy to kill if we keep a knight on d5 protecting f6. Of course, White can try other attacking possibilities than
Bh3, but all the same our confidence helps us to look for the best moves with a constructive mind.
Before moving on, let’s see one more disaster with colours reversed.
Kaidanov, April 2016: “Sorry to disappoint, but I wasn’t inspired by Fischer when I played this line.”
1.e4 c5 2.Nf3 e6 3.d3 Nc6 4.g3 d5 5.Nbd2 Nf6 6.Bg2 Be7 7.0-0 b5 8.Re1 0-0 9.e5 Nd7 10.Nf1 a5 11.h4 b4 12.Bf4
a4 13.a3! Ba6!?
14.N1h2
Taking the offered pawn with 14.axb4?! cxb4 15.Rxa4 gives Black an initiative with 15...Nc5 16.Ra1 b3!,
undermining the d3-pawn.
14...c4
A logical move, trying to open the queenside by 15...c3, attacking White’s b2- and a3-pawns at the same time.
15.d4
White should keep the centre blocked.
17.Ng5
17...Nb6?
As a rule of thumb, Black should play 17...h6! as soon as the knight reaches g5.
a) In the last game, we saw the sacrifice 18.Nxe6 fxe6 19.Bh3 but that is not as good without a knight on e3 that can
reach f5.
b) 18.Qh5 hxg5 19.hxg5 with the idea of Nh2-g4-f6† doesn’t work after 19...Nxd4 20.Ng4 Be2!, as pointed out by
John Emms.
18.Qh5! Bxg5
18...h6 doesn’t work anymore: 19.Ng4! hxg5 20.hxg5 Nxd4
Before Black has played ...Be2, White uncorks 21.Nf6†! gxf6 22.gxf6 Bxf6 23.exf6 Qxf6 24.Be5 and the mating
threat on h8 wins the queen after 24...Be2 25.Rxe2 Nxe2† 26.Kh2.
19.Bxg5 Qe8?
Black’s idea is 20...f6, exchanging queens. However, pure logic warns against this move – White’s rook is pointing
towards Black’s queen.
20.Bf6! Nxd4
The problem with the queen move is seen after 20...gxf6 21.Ng4 Nd7 22.Bxd5! exd5 23.exf6 with the double threat
24.Rxe8 and 24.Qg5† with mate.
24...Qd8
With an extra move, Black could have defended with 25...Ne8. Now he has to give up his queen to avoid mate.
25.Qh6 Qxf6 26.Qxf6 Rae8 27.g4 Nd7 28.Qf4 Bc4 29.h5 Rc8 30.Rab1 f5 31.exf6
1–0
The winner gets the last word: “We beat a very strong Dutch team with a score of 4–0. I felt inspired thanks to my
teammates, we had a great time in Elista and were on the first place ahead of Russia by a point before the last round,
but ended up with a silver.”
18 years after these silver medals, the US team won their first team gold since 1976. Their success was aided by the
return of the player playing Black in the next game.
The Paris Grand Prix was played in the Chapel de la Villedieu (Chapel of the City of God) and was the last tournament
of the Grand Prix series, where the top two qualified for the Candidates Tournament. Caruana and Grischuk both had
everything in their hands.
Instead of the organizers’ luxury preference, Grischuk chose a simple chair for this game. Eventually he reached the
French set-up against the Reversed King’s Indian Attack that we saw in the previous games.
Grischuk, September 2016: “Obviously e3 is a way to combat the Grünfeld; it is rather marginal against the KID but
still it leads to a playable position.”
The first annotations to the game, on ChessBase, claimed that: “4.e3 is undesirable in a King’s Indian.” But soon
after, they concluded that “Grischuk did obtain a certain queenside initiative”. How could that happen?
6.Nc3 Nbd7
6...Re8 does not prepare 7...e5.
But 6...Qe8 is a move worth considering. 7.0-0 e5 8.b4 e4 9.Nd2 h5 10.Qc2 Bf5 justifies Black’s idea: he managed to
play ...Bf5 without having to move a knight from d7.
There are many ways to deviate, including 8.dxe5!?, but I recommend 10.a4. When Black can defend comfortably
with ...Bf5, 10.Qc2 makes no sense.
9.Rd1
Given that the rook won’t be too useful on d1, it would be natural to look for alternatives. 9.b4 suggests itself, since
it’s White’s plan as soon as the centre is closed. But I feel that attacking on the flank is premature, and Black can
switch plans with 9...exd4. This is not the usual choice in similar situations with colours reversed. The difference is that
Black isn’t sitting behind the white pieces, and thus feels no responsibility to play for an advantage.
The position is balanced.
12...h5 13.Rac1
A useful move if White plays Nd5 followed by c4xd5.
13...Bf5
13...N8h7 keeps the bishop flexible, but there is also a point to the immediate 13...Bf5: the knight on f8 could go to
e6 and threaten the pawn on d4 – just as Fischer did against Uzi Geller.
14.Nb3
Deciding to hit the bishop on f5 with d4-d5 and Nd4.
14...N8h7 15.d5
15...h4
15...Ng4 equalizes, as we will see in the theoretical section.
17...dxc5?!
When he showed the game at the press conference, Grischuk quickly scrolled on to this position and said that he had
no idea about the beginning. It’s understandable that a professional player doesn’t want to talk about the opening – his
professional secrets.
Grischuk thought that 17...dxc5 gave him a nice initiative, so he recommended 17...h3! when Black is not worse.
19...Qb6
19...Rec8 20.Nxe4 Qxd5 21.Nxf6† Nxf6 22.Bc4 is one line where it would have been good for Black to have
inserted 17...h3 18.g3.
19...Qa5 20.Ncb5 Rec8 21.a4! was a line that Grischuk showed. By delaying Bc3 and following the principle that the
threat is stronger than its execution, White threatens 22.Bc3 Qb6 23.a5. After 21...Qb6 he has 22.Nxc7 winning back
the pawn with an initiative.
20.Ncb5 c5
20...Rac8 still allows 21.Nxc7.
29.Nxf7!?
Grischuk: “Far away, when I calculated this position, I thought that I would simply play 29.Bd4 but then I didn’t like
29...Rd8 30.Rb1 Nh5.” But White has 31.Nb7 with a double threat on d8 and g6.
29...Nxf7 30.Rc7
30...Rf8
Grischuk: “Extremely dangerous for Black, but unfortunately I already had about one minute left.”
The problem was that the tournament was played without increment, so he only got more time at move 40.
Grischuk was afraid of 30...Rc8! 31.Bxf7† Kh7 32.Rxc8 Rxc8 33.Bd4 Rc7.
He thought this looks better for White than it really is. Black exchanges the dark-squared bishops and then returns
with the knight to f6. It will be difficult for White to make progress and Grischuk concluded that it’s almost a draw.
31.Bxf8 Rxf8 32.Rxa7 Ng4 33.Bd5 Bh6 34.Re7 Bg5 35.Rxe4 Nf6 36.Red4 Nxd5 37.Rxd5 h3 38.f4 Bd8 39.gxh3
The players thought that 39.g3 would have given a technically winning position.
41.Kf3
41.Rd6 Bxe3†! 42.Kxe3 Nf5† 43.Ke4 Nxd6† 44.Rxd6 Kg7 is also an easy draw.
41...Nf5 42.e4 Ne3 43.Rd6 Nxd1 44.Rxg6† Kf7 45.Rxb6 Rh8 46.Rb7† Kf6 47.e5† Kf5 48.Rf7† Ke6 49.Rf6† Ke7
50.Rd6 Nc3 51.f5 Rxh3† 52.Kf4 Rxh2 53.f6† Ke8 54.e6 Rf2† 55.Ke5 Re2† 56.Kf5 Rf2† 57.Ke5 Re2† 58.Kf5 Rf2†
59.Ke5
½–½
In the final standings, Caruana was half a point away from qualifying, while Grischuk didn’t have his best tournament.
The next game was less important, as it’s only blitz, but nonetheless it illustrates how White can break through on the
queenside.
8.Qc2 Re8 9.Rd1 e4 10.Nd2 Qe7 11.b4 Nf8 12.Bb2 h5 13.Rac1 Bf5
14.d5
Evacuating the d4-square to put a knight there, just as Grischuk did. But it also gives Black the possibility to use the
e5-square for one of his knights.
14...h4?!
It was better to keep ...Bg4 as an option when the bishop is hit by the knight.
The best way seems to be 14...a5 15.a3 N8d7 16.Nb3 Ne5 which threatens 17...Nf3†. There is no time to capture the
pawn on a5, and White has to play 17.Nd4 Bg4 18.Nxe4. We will see why it was important to include the moves with
the a-pawn after the forced sequence: 18...Nxe4 19.Qxe4 axb4 20.axb4
20...Nf3†! 21.gxf3 Qxe4 22.fxe4 Bxe2 23.Nxe2 Bxb2 24.Rc2 Bf6 25.f3 When Black has compensation for the pawn
due to 25...Ra3!.
15.h3!
Still and always so.
15...N8h7
16.c5!
Opening the queenside is more important than counting pawns.
16...Ng5
16...dxc5 17.bxc5 Qxc5 18.Nb3 Qe7 19.Nb5 and White is clearly first.
17.Bf1
The standard defence against the sacrifice on h3.
19...Rad8
According to the engine, Black’s best chance was the long line 19...Nd3 20.Nxf5 gxf5 21.Bxg7 Nxc1 22.Qb2 Nxh3†
23.gxh3 Qg5† 24.Kh1 Qxg7, but to me this looks winning for White if he keeps the queen with 25.Qxc1 and slowly
improves his minor pieces.
21...cxd6?
An old saying is that Black’s attack in the King’s Indian won’t succeed without the light-squared bishop, but Korobov
could have played 21...Rxd6! and White can’t go 22.Qxc7? due to 22...Nxh3†! 23.gxh3 Rg6† 24.Kh1 Qg5 with mate
to follow.
23...Qf6 24.Nc4
24.Qxb7 allows 24...Nef3† with the bishop on b2 undefended.
24...f4
32...Rg8 33.Ke1
It looks shaky, but Dreev easily keeps his position together for the rest of the game.
33...Bh6 34.Rc3 e3 35.fxe3 Rd7 36.Bf3 Bg7 37.Rd2 Ng5 38.Bxg7 Rgxg7 39.Kd1 Nf7 40.Qc4 Qf6 41.Rcd3 Ne5
42.Qd4 Rg6 43.Rc3 Rxd6 44.Rc7† Kh6?
This speeds up the end, but Black was lost anyway.
Recap
At the press conference after his game against Caruana, Alexander Grischuk said that it’s “a very risky line for both
players”. That’s not an understatement! If two unprepared players face each other, I think Black is the favourite – it’s
true that it’s easier to attack than to defend.
But if the White player has studied the lines and thus taken advantage of today’s engines’ bold defence, I believe he
can be optimistic. More top games are needed to reach a verdict, but it would be cowardly of me to hide behind such a
statement. I believe in White!
Theory
6...c6 is an example of Black’s slow development. Logical moves are 7.0-0 a6 8.b4 b5 9.a4 bxc4 10.Bxc4 and I think
White is slightly better. If Black breaks with ...c6-c5, White will be first to make use of the b-file and the a3-f8
diagonal. And with the pawn on e3, there’s no ...Nxe4 trick.
After 6...c5, White has two ways to avoid the Poor Man’s Benoni.
7.d5 e6 8.e4 exd5 9.exd5 is a worse version of what Mamedyarov played against Gelfand in the Benoni chapter, because
Black can prepare ...Ne4 with ...Bf5 rather than ...Re8.
7.0-0 is possible. 7...cxd4 8.exd4 d5 gives Black the ideal set-up against the isolated pawn, but being Black and
giving a tempo is a high price. After 9.cxd5 Nxd5 10.Qb3, Black would have had 10...Be6 11.Qxb7 Nxd4 – if only the
queen’s knight were on c6. But it’s still on b8, and he has to choose between 10...Nxc3 11.bxc3 and 10...Nb6 11.d5!
The last move is a sacrifice, but the dark squares compensate for the pawn after 11...Bxc3 12.bxc3 Qxd5 13.c4.
7.0-0
Trying to delay castling to avoid Black’s attack would be thinking far too deeply.
7...e5 8.Qc2
This is the move I like, but of course it is not the only option. Almost equally common is 8.b4.
8...Re8
8...b6 was mentioned in the discussion about the pawn structure. 9.dxe5 Nxe5 10.Nxe5 dxe5 11.Rd1 gives a position
where developing moves put pressure on Black.
9.Rd1
White attacks on the queenside only when the centre is closed, and the rook tries to make a semi-useful move while
waiting for ...e5-e4. There are other tries with the same idea.
Mamedyarov and Ni Hua have both been successful with 9.Re1!?. If Black plays ...exd4 at some point then exd4 will be
the reply, when White’s rook may be useful on e1.
9.b3 e4 10.Nd2 Qe7 11.Bb2 h5 12.b4 was given in a theoretical source, which I think I am doing a favour if I don’t
reveal. With b2-b3-b4, White has transposed to the King’s Indian Attack, but without the extra tempo.
9.b4 exd4! and Black is okay, for example 10.exd4 a5!? 11.b5 d5 12.c5.
White would have a clear advantage if he had time for Bd3 to reinforce the e4-square, but 12...Ne4! 13.Bb2 Ndxc5 is
killing: 14.dxc5 Nxc3 15.Bxc3 Bxc3 16.Qxc3 Rxe2 and Black has won a pawn.
9.h3 is a brainchild of my own. It’s clearly useful in case of 9...exd4 and after 9...e4 10.Nd2 Qe7 11.b4 it stops ...Ng4,
...Bg4 and ...h5-h4-h3. There are three arguments against the pawn on h3.
White has managed to keep control over e4, but has been forced to exchange the bishops. After 13...d5 14.c5 Nbd7
the space on the queenside would give a slight advantage, but Black plays ...Nf8-e6 and is very solid.
b) 9...c6 is a move that’s played from time to time (not too often) with opposite colours. Giving up the d6-square is
something Black can live with in the King’s Indian, but it gives him the task of defending patiently. If you are satisfied
with that, I suggest 10.dxe5 dxe5 11.Rb1 to play b2-b4 next.
c) 9...Qe7!? was played by Vidit Santosh Gujrathi in the 2015 World Cup. Black keeps the option of choosing between
...exd4 and ...e5-e4. If White wants to be consistent he has to play 10.b4 after which Black gets an improved version of
the ...exd4 lines.
Grischuk’s and Caruana’s choice was 12.Bb2 h5 13.Rac1 N8h7 14.Nb3 (14.Nd5? is too early as long as Black’s second
knight can go to f6. 14...Nxd5 15.cxd5 Nf6 16.Qxc7 Qxc7 17.Rxc7 Nxd5 and White loses the b4-pawn.) But honestly,
I do not like 14.Nb3, at least not before ...Bf5. Keeping the pressure on the e4-pawn discourages Black from playing
...Ng4.
a) Caruana played 14...Bf5, but I don’t think that was necessary. However, after 15.d5 he could have equalized with
15...Ng4 and 16.h3 Nxf2 or 16...Qh4.
b) For a long time I was suspicious of putting the rook on c1 – why isn’t White playing a faster move? But finally I
found the long line 14...Ng4 15.h3 Nxf2 16.Kxf2 Qh4† 17.g3 Qf6† 18.Kg2 Bxh3† 19.Kxh3 Qf2 when Black
threatens mate in four. But there is a fantastic resource in 20.Bxh5! which works only because the queen is defended by
the rook. The line continues 20...Ng5† 21.Kh4 Qf5 22.g4 Qf3 23.Qh2 Qxe3 24.Qg3 Nf3† 25.Kh3 gxh5 26.gxh5 with
an edge for White (I guess you have understood that it’s the work of the computer).
c) 14...h4! 15.h3 g5 looks promising for Black, although White won in Hera – Schreiner, Austria 2013.
12...h5
12...a5?! is never played in similar positions. White is happy to capture and open the queenside.
13.a5N 13...a6!
Playing like Fischer.
a) 13...N8h7 14.a6 b6 15.c5! gives an initiative and 14...bxa6 15.Ra5! is a nice square for the rook. The missing pawn is
nothing to worry about.
b) 13...h4 14.h3 g5 is a straightforward line. Note that 15...N8h7 is not threatened, as it would allow the tactic 16.Ndxe4
Nxe4 17.Nxe4 Qxe4 18.Bd3 with a skewer. However, Black’s problem is that he also isn’t threatening ...g5-g4, since
e4 would be hanging after the recaptures on g4.
A logical line is 15.b5 Bf5 16.a6 b6 17.Ba3 g4 18.hxg4 Nxg4.
White can give up the exchange for fine compensation: 19.Bxg4 Bxg4 20.Nd5 Qd8 21.Nxe4 Bxd1 22.Rxd1 and even
though 9.Rd1 turned out to be bad, White is all the same much better.
14.b5
Black can now choose to play with or without ...h5-h4, ...Bf5 and ...N8h7. It’s impossible to cover everything and I
focus on the lines I think are logical.
14...N8h7 is a probable move order, but will most likely transpose.
14...axb5 15.cxb5 was not the intention with 13...a6. White continues with Ba3 and b5-b6.
A) 14...Bg4
As soon as Black has challenged the bishop on e2, he can consider recapturing on a6 with the rook. The standard
reaction is to capture, to avoid losing time.
16...N8h7 17.c5 Qe6 18.cxd6 cxd6 19.Rab1 Ng5 20.bxa6 bxa6 21.Rb6 Nf3† 22.Kh1!
According to my engine, Black’s different sacrifices don’t work. If that’s true, White is better.
B) 14...h4 15.h3
Black has two ways to continue the attack.
b1) 15...N8h7 16.bxa6 bxa6 17.Ba3 Ng5 threatens to sacrifice, but White plays the standard 18.Bf1.
18...Bf5 (18...Bxh3 19.gxh3 Qd7 20.Kh2 Qf5 would have been interesting were it not for 21.f4!) 19.Rdc1! is a good
move to remember. The point is that d1 is evacuated and can be used by the queen after: 19...Qd7 20.Qd1 (Instead
20.Nd5 may look a logical way to open the c-file, but there is a tactical flaw: 20...Nxd5 21.cxd5 Bxd4! is OK for
Black.) White has the immediate sacs on h3 well covered, as eventually White’s queen or knight could go to f1 to foil
the attack. Instead a devious try is 20...c5!? with the idea: 21.dxc5 dxc5 22.Bxc5 Bxh3! But after the calm 23.gxh3
Qxd2 24.Kg2! Black must exchange queens, when White’s bishop pair is worth a slight advantage.
b2) 15...g5 16.Ba3 Bf5 Necessary to play ...g5-g4. After 17.c5 g4 18.cxd6 cxd6, the engine prefers the cool 19.Nc4
gxh3. But the human choice is: 19.hxg4 Nxg4 20.bxa6 bxa6 21.Bxg4 Bxg4 22.Ndxe4! A typical sacrifice we have
seen before.
C) 14...Bf5!
Black plays ...Ne6 next, which threatens d4 and wins a tempo. Heinz Polsterer has, in similar positions with colours
reversed, used the set-up twice. But it originates from Fischer’s game against Uzi Geller.
15.Ba3
15.Nd5?! runs into a trick: 15...Nxd5 16.cxd5 Bxd4!
15...Ne6 16.Qb3
It seems odd to move the queen for the second time, but the problem with 16.Nb3 is that 16...Ng5 wins another
tempo due to the threat of 17...Nf3† 18.gxf3 exf3 19.Bd3 Ng4 with a mating attack.
16...Ng4
Black has three typical alternatives.
16...Ng5?! is better after ...h5-h4 first. Now, White is faster after 17.c5.
16...h4 17.h3!
17...Ng5 gives White serious problems on the kingside, but it appears to me that he is just in time with: 18.c5 (The
standard 18.Bf1 doesn’t help here. Black plays 18...Qd7 and the sacrifice is inevitable.) 18...Bxh3 19.cxd6! cxd6
20.Nd5! (20.gxh3 Qd7! gives enough counterplay) 20...Nxd5 21.Qxd5 This wins back the pawn and tries to exchange
queens with the better pawn structure.
Aditya Subramanian pointed out that 17...Nxd4 18.exd4 e3 doesn’t lead to a forced draw: 19.Nf1! exf2† 20.Kxf2
Ne4† 21.Nxe4 Qxe4 22.Qb2 and White consolidates.
16...c5 tries to be clever, but only helps White after 17.bxc6 bxc6 18.d5. However, Black has the sacrifice 18...Nf4!
19.exf4 e3. For the curious, the engine gives: 20.dxc6 Rab8 21.Bb4 exf2† 22.Kh1 Ng4 23.Bf1 The only move.
23...Rxb4 24.Qxb4 Nxh2 25.g3 Nxf1 26.Nxf1 h4 When White is happy to survive with 27.Qxd6 Qxd6 28.Rxd6 Bxc3
29.Rad1 with compensation.
With the text move, Black threatens 17...Nxf2 18.Kxf2 Nxd4 19.exd4 e3† 20.Kg1 exd2 with ...Bxd4 and ...Qe3
coming as winning moves.
18...axb5 19.cxb5
19...Ng5
Black can take a pawn with 19...Rxa5 but I think he should focus on the kingside. White has compensation after
20.Rac1.
23...Nh3†
Or 23...Bxd5 24.Qxd5 Qxc7 25.Qxg5 and with Qd5 next, White has more than enough for the exchange.
The Averbakh/Modern
With 1.d4 g6 2.Nf3 Bg7 3.c4 d6, Black leaves the knight on g8 and can play ...f7-f5 before ...Nf6. But he also has the
possibility of transposing back to the King’s Indian. After 4.Nc3 e5, exchanges on e5 and d8 deprive Black of the right
to castle, but the structure is favourable for him, as explained in the English, Structure 4 on page 324.
However, Black can also play 4.Nc3 Nd7 5.e3 e5 6.Be2 Ne7 7.0-0 0-0, when I see no good moves for White while
Black plays ...h7-h6 and ...f7-f5. It’s hard to argue for e2-e3 in the Averbakh.
So, after 1.d4 g6, I recommend taking a step away from our desired set-up and playing 2.e4, a move that avoids the
King’s Indian for sure. The following set-up can be played after a one-minute lesson: 2...Bg7 3.Nf3 d6 4.Bc4 Nf6
5.Qe2
White continues with 0-0 and c2-c3. Bc4-b3 is played as a response to ...c7-c6, and ...Bg4 is answered by Nbd2 and
h2-h3. Whenever Black goes ...e7-e5, White captures and plays against the restricted bishop on g7. And if not, White
goes e4-e5 himself.
Exercise 1
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 d6 6.Nc3 Nbd7 7.0-0 e5 8.Qc2 Re8 9.Rd1 e4 10.Nd2 Qe7 11.b4 Nf8
12.a4 h5 13.a5 a6! 14.b5 h4 15.h3 N8h7 16.bxa6 bxa6 17.Ba3 Ng5 18.Bf1 Bf5
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 d6 6.Nc3 Nbd7 7.0-0 e5 8.Qc2 Re8 9.Rd1 e4 10.Nd2 Qe7 11.b4 Nf8
12.a4 h5 13.a5 h4 14.h3 g5 15.b5 Bf5 16.a6 b6 17.Ba3 g4 18.hxg4 Nxg4 19.Bxg4 Bxg4 20.Nd5 Qd8
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 d6 6.Nc3 Nbd7 7.0-0 e5 8.Qc2 Re8 9.Rd1 e4 10.Nd2 Qe7 11.b4 Nf8
12.a4 h5 13.a5
Black to move
Show/Hide Solution
Show/Hide Solution
Sideline 1: Black delays ...Nf6
1.d4 c5 2.d5 e5 3.e4 d6 4.Nc3 Be7 (4...g6) 5.Nf3 75
a) 5...Bg4 76
b) 5...a6 76
c) 5...Nf6 76
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3! 4...c5 5.d5 0-0 6.Nc3 e6 7.Be2 exd5 8.cxd5 d6 9.0-0
White’s long-term plan is to create a passed d-pawn with e3-e4-e5. The stop on e3 loses a tempo, but does that mean
that we will be playing a Poor Man’s Benoni? Or is it possible to take advantage of it?
An English version of an old fable tells us one possible answer. There was once a fox that wanted to eat a tasty-
looking cluster of grapes hanging high up on a vine. But he couldn’t find a way to reach them, and finally he convinced
himself: “The grapes are sour anyway!”
To avoid feeling uncomfortable, he changed his perception of reality. A very human (and possible foxy) trait – we are
much more subjective than we want to admit to ourselves. We accept lies more easily than failure. When I started to
write this chapter, I had to convince myself that the e-pawn would have been sour on e4. It was quite difficult.
It’s possible to argue that the Benoni is a bad opening and that we are happy to face it even a tempo down. Actually,
this seems like an argument that top players accept. Hikaru Nakamura pushed e3-e4 very early against Maxime
Vachier-Lagrave in the 2015 London Chess Classic. And I am quite sure he didn’t feel like a confused fox.
But I think it makes sense to keep the pawn on e3. In Sweden, the fox is actually quite clever when he convinces
himself that the fruit is sour. Since it’s too cold to grow grapes, our foxes instead try to reach rowans. And they really
are sour!
In the original Greek fable, the fox doesn’t conclude that the grapes are sour. He says they are unripe, so he will come
back to collect them later.
The Benoni is a counterattacking opening and Black does whatever he can to play ...b7-b5, in order to get his majority
moving. Often it works only because White is overloaded defending the e4-pawn (Nc3xb5 can be met with ...Nf6xe4).
The same goes for ...c5-c4, an ugly-looking move that helps Black to activate with ...b7-b5 and/or ...Nd7-c5 (again
Nc3xb5 can be met with ...Nf6xe4).
With the pawn on e3 instead of e4, neither ...b7-b5 nor ...c5-c4 is possible. Black has lost his main tactical trumps.
So first we stop Black’s counterplay, then we improve our pieces and finally, when the grapes are ripe, we push e3-
e4-e5. And it will come in like the tide, slow but unstoppable.
I shouldn’t add too much sugar to the story – sometimes White really does get a Poor Man’s Benoni. The following
examples and games will show us how to avoid it.
Pawn Structures
Structure 1
White’s first aim is to stop ...b7-b5, which will be easy with Nc3, Be2 and a2-a4. The last of these moves weakens
the b4-square, so it should preferably be played only after ...a7-a6, when ...Na6-b4 is no longer possible.
The second aim is to transfer the knight to c4. It is placed ideally there, hitting d6 and supporting e4-e5. If Black
chases the knight with ...b7-b5, it could jump ahead to c6 via a5. (With the pawn on e4, our set-up with Be2 and Nf3-
d2-c4 is named the Knight’s Tour in some Benoni books.)
Finally, White chooses between two different plans. One is the standard e3-e4-e5-break. In addition to Nc4, it’s
prepared with Re1, h2-h3 and Bf4. If it’s successful, it gives a passed d-pawn while liberating the pieces: the rook on
e1, the bishop on f4 and the knight on c3, which could jump on to e4 or d5.
But if Black is careful he should be able to prevent e4-e5. Then White could switch to b2-b4. I think it’s an
underestimated pawn lever, probably since it seems odd to attack the well-protected c5-square. The exchange of
White’s b-pawn against Black’s c-pawn gives the c5-square for one of the black knights.
So why is the pawn lever b2-b4 all the same strong? First of all, there’s the b-file.
a) In case of ...b6xc5, we get an open file. That’s not good news for Black, since the queenside is supposed to be his
wing.
b) In case of ...Nxc5, we can harass the pawn on b6 with a4-a5. And we also get access to the b4- and d4-squares, which
gives the knights a route to c6.
(With that said, b2-b4 is more effective with a black knight on c7 than with a knight on d7.)
Black will not wait for his execution; he will strive to exchange a pair of knights. That could be done either by
challenging White’s knight on c4 with ...Nb6 or ...Ne5, or by playing ...Re8 (or ...Bf5) and ...Ne4. White should avoid
this.
Black can also get rid of a minor piece with ...Bg4xf3. However, White isn’t unhappy with that. Our light-squared
bishop returns to e2 and controls the vital squares c4 and b5 without having any counterpart.
So, all in all, Black is doomed just to piece play, and that’s not easy if White knows what to do. I don’t think Black will
manage to get in ...b7-b5, and pushing pawns on the kingside (...f7-f5, ...g6-g5) will expose his own king more than
White’s.
Lastly, a few words about playing a4-a5 after Black’s ...a7-a6. Even though it’s optically nice, it’s not a good option in
many cases. Black pushes ...b7-b5 anyway, getting an isolated pawn on a6, but the b2-pawn might be just as weak. The
fact that Black has three pawn islands against White’s two is much less important than that something is happening on
Black’s wing.
In addition, a4-a5 allows Black to use the b5-square for the knight or the bishop.
Structure 2
White could have placed the pawn on c4, but has chosen to keep it on c2. Instead he keeps control over the b5-square
with Nc3, a2-a4, Bc4 and Qd3 (when necessary). But if Black is persistent, it is not possible to stop ...b7-b5 in the long
run. With ...Na6-c7, ...a7-a6, ...Bd7 and ...Rb8, he already has enough pieces. And if he succeeds, he usually takes over
the initiative.
In the meantime White plays for e4-e5 with the preparatory moves (Nf3), Re1, possibly h2-h3, and Bf4. The pawn
lever will give more active pieces than normal – one advantage with the queen on d3 is the possibility of placing the
queenside rook on d1. And c4 is a good square for the bishop.
Structure 3
This is a mix of the Schmid Benoni (White’s set-up) and the Czech Benoni (Black’s set-up). After a first thought, it
seems clever for Black. He has stopped White’s main plan (e4-e5) while it’s still impossible to stop ...b7-b5 in the long
run.
Instead, White puts his faith to the pawn lever f2-f4, which is best prepared with Nf3-d2-c4. As a consequence, the
light-squared bishop stays on e2.
Except if Black has a knight on g6, White normally doesn’t prepare f2-f4 with g2-g3. Black’s dream – to play ...e5xf4
followed by controlling the e5-square with a knight on e5, a bishop on g7 and a rook on e8 – will stay a dream. The
main downside is that if he places a knight on d7, he is unable to push ...b7-b5. For that, he needs a bishop on d7.
White thus can prepare f2-f4 for a long time (possibly even with g2-g3).
What Black usually does is exchange the dark-squared bishop with ...Bg5. That’s a good positional plan, both since it’s
the bad bishop and since Black has less space. The downside is that it allows White concrete possibilities, but more
about that later.
Games
Already by the next move, Kramnik signalled that he wanted a complicated game to increase his winning chances. He
was one point behind Viswanathan Anand in the World Championship, an eight-player round robin. And since
Morozevich shared last place, he was an intended victim.
3...Bg7
3...Nf6 4.d4 cxd4 5.exd4 d5 would have been the safe theoretical choice, leading to a Panov.
4.d4
Black is suddenly unable to stop d4-d5.
4...Nf6 5.d5(!)
After 5.Nf3 0-0, White should play 6.d5, because 6.Be2?! cxd4 7.exd4 d5 leads to a Panov where a slow move like
Be2 doesn’t fit.
Black can also go 5...cxd4 6.exd4 d5 immediately. White would have liked to capture the pawn on d5 instead of
spending a tempo on Ng1-f3.
9...Bg4
9...Re8 was played in Smith – Istvan Almasi on page 121.
10.h3
With less space, Black is happy to exchange a minor piece and, without the light-squared bishop, there is no traffic
jam on the d7-square; next, he develops with ...Nbd7.
At the same time, White’s bishop pair is a long-term asset.
In some Benoni positions, White plays 10.Nd2 in order to exchange his restricted bishop, but that makes more sense
with the other bishop on f4. With a pawn on e3, it will not reach there easily, so we have to forget about that grape –
it’s sour, for sure! Black can play 10...Bxe2 11.Qxe2 Nbd7, which would not have been possible with the bishop on f4.
After 12.Nc4 Nb6 White can’t retreat to e3, so this is simply a failure.
12.a4
Not obligatory to stop ...b7-b5, but the main point is that White can attack a black pawn on c4 with a4-a5 and Ra4.
12...a6
Honestly, I don’t understand why this is played here.
13.g4?!
g2-g4 is a thematic move in the Benoni – when it hits Black’s knight on h5. But I think it’s too early when there are
still several pieces to develop.
13.Be2 was played in Tikkanen – Nybäck, Göteborg 2012. It’s a good move, but my recommendation is 13.Bd2N.
13...c4
Mihai Suba is a great expert in the Benoni and he says that Black has to find an active option. And if it’s not possible?
Suba’s motto is to go for it anyway.
14.Be2!
With White’s e-pawn on e4, Black could have played 14...Nc5 15.f3 Rc8 16.Bxc4 Ncxe4. So the grape is not ripe
yet!
14...Rc8?!
In his annotations to the game, Mihail Marin thinks that Kramnik’s lack of experience with the Benoni stopped him
from executing the following thematic sacrifice: 14...Ne5!? 15.f4 Nd3! 16.Bxd3 cxd3 17.Qxd3 h5 18.g5 Nd7
Black is happy to have exchanged two minor pieces, and the two remaining are very active. It more than compensates
for the sacrificed pawn.
Marin’s conclusion is that: “From a psychological point of view, Morozevich’s choice [3.e3] was simply brilliant.”
16...Qe7
Black’s two active options (...Nc5 and ...b7-b5) both lose a pawn, but Suba’s motto should still have been followed.
16...Nc5! 17.Bxc4 and ...f7-f6, ...Qd7 or ...Bxc3 and ...Ng7 in any order gives decent compensation.
17.Ra3 Rc5?!
A creative manoeuvre, but the rook will simply be exposed.
20.Qe2 Nb3
21.Ne4!
As Marin notes, White can use the e4-square since the pawn is still on e3. However, that’s a curious fact more than an
argument for an early e2-e3.
21...Nxd2
21...Rxd5 22.Qxc4 leaves White controlling the light squares after 22...Nxd2 23.Qxd5 Nxf1 24.Kxf1 or 22...Rxd2
23.Rxb3.
23...b5 24.axb5 Rxb5 25.Qxc4 Qb6 26.Qc6! Bxb2 27.Qxb6 Rxb6 28.Ra2 Bg7
It is not possible to activate the knight and rook: 28...Nc7 29.Nxd6! and the rook on b6 has one task too many.
29.Rc1 h6 30.h4 hxg5 31.hxg5 f6 32.Rc6 Rxc6 33.dxc6 fxg5 34.Nxg5 Nc7
Morozevich’s way of playing was not the only one, but it was good enough. Black has to dedicate all his efforts to
maintaining the blockade of the c-pawn.
39.e4
Finally, Morozevich pushes e3-e4.
39...d4 40.e5 Bb6 41.Rb3 Rb8 42.Rh3 Ba5 43.Rh6 Rb2† 44.Kg3 Be1† 45.Kf3 d3 46.Rxg6† Kf8 47.Rd6 d2 48.Ke4
Kramnik gave up the light-squared bishop and now he has to pay the price: he can’t take control over d1.
1–0
It’s nice to see that the repertoire worked well at the highest level. In the end, one point was exactly what Kramnik
lacked in order to make it to a tiebreak against Anand.
In the Bundesliga, a few months after he had to relinquish the title of World Champion, Anand reached the same
position after 11 moves.
6...d6 7.Be2 e6 8.0-0 exd5 9.cxd5 Bg4 10.h3 Bxf3 11.Bxf3 Na6
Ragger deviates from Kramnik’s play. However, the move is not logical: ...Bg4xf3 was played partly to get the d7-
square for the knight. If it’s placed on c7, Black would have wanted to keep the light-squared bishop in order to prepare
...b7-b5.
12.Bd2
Not a threatening move, but nonetheless developing. The c1-square is available for the rook and the knight on c3 gets
extra protection, which is good in case White goes b2-b4.
12...Rc8
12...Nc7 13.a4 a6 14.a5 Nb5 15.Qb3 shows one advantage of keeping the pawn on e3 – the d4-square is not available
for Black’s knight.
13.Qb3
With his move, Anand invites Black to a dance...
A more prudent move is 13.Rb1N, leaving the long diagonal and preparing b2-b4 in the future.
13...c4!
...and Ragger does not step away – he may have heard of Suba’s rule.
The manoeuvring phase started only a few moves ago, and now we leave it for a tactical fight. That’s modern chess in
a nutshell.
24.Na4 cxb3 25.axb3 Rb8 26.Qc7 Ra5 27.Rfd1 Rab5 28.Qxa7 Be5 29.f4
Anand avoids getting back rank mated after the tactical sequence that follows after his next move.
29...Bf6 30.Nb2!?
Anand solves the problem of his worst piece in a radical way: he allows Black to win all his material back.
30...Qxb3 31.Nd3
31...Qxb1
Instead 31...Qc2! 32.Qxb8† Rxb8 33.Rxb8† Kg7 gives a position where White is materially winning, but his
coordination is bad and the rooks have no available targets. Black is better and it gets even worse if White tries to hold
on to the material: 34.Nf2 Qe2 35.Rb3 Bh4 36.Rf1 Bxf2† 37.Rxf2 Qd1† picking up the rook on b3.
35...Rbe7 36.Qxd6 Rxe3 37.Qxf6 Rxd3 and White has to be careful to get the draw.
36.Nf2 Rdd8 37.Qd3 Bg7 38.g4 Rd7 39.g5 h5 40.gxh6 Bxh6 41.Ng4 Bg7 42.f5 gxf5 43.Qxf5 Rde7
Threatens to take on e3 with a draw.
Ragger probably thought that this was a draw. And with the king on g7, it would have been so.
53...Rf6† 54.Ke4 Rf1 55.Kd5 Rf5† 56.Kxd6 Rf6† 57.Kd7 Rf5 58.Qd6 Rg5 59.Qf4 Kg6 60.Ke7 Rf5 61.Qg4† Rg5
62.Qe4† Rf5 63.Qd3 Kg5 64.Qg3† Kh5 65.Ke8 Kh6 66.Qg4 Rf6 67.Kf8
1–0
The end is coming closer, but I would not have resigned yet.
I don’t want to give the impression that Black always plays ...Bg4xf3, but it would not be fair to take away the pleasure
of seeing the next game.
Tiger Hillarp at a seminar in Lund: “It’s one of the best games I have seen during the latest decade. Stellan can play
really well, and if he only knew it himself he would do it more often.”
12...Nbd7
12...Re8! would have stopped the manoeuvre Brynell plays in the game.
13.Be2!
The bishop is transferred to the best diagonal.
15.a5 h5?!
It is not easy to just wait. This is a weakening move that rules out ...f7-f5 once and for all – in that case g6 would be
left without protection.
20...Bg7 21.Nb1! Qd8 22.Nd2 Nhf6 23.Raa1 Rc8 24.Bg5 Qc7 25.Nc4 Nh7 26.Bd2 Re7
When I told Stellan Brynell about Hillarp’s quote, he said in his usual self-deprecating manner: “But I did nothing.”
But that’s the point! Remember the tide? It inevitably comes in, and after the next move, we will see it closing in.
27.f4 Ree8 28.Be2 Bd4† 29.Kh1 Rb8 30.Bf3 Bg7 31.Rad1 Rbc8 32.Re2 Qb8 33.Qb3 Bf6 34.Rde1 Bh4 35.Rf1 Bd8
36.e5
Here comes the tide! Black’s position immediately gets flooded.
38.Rfe1?
This allows a neat escape attempt. Several moves were winning, among them 38.d6 Bf6 39.Nxe5 Rxe5 40.Rxe5 Bxe5
41.Bd5 with too many threats against f7.
38...Bc7?
38...Bg5! keeps the piece after 39.Bxg5 Nxf3 or 39.Rxe5 Bxd2 40.Rxe8† Rxe8 41.Rxe8† Qxe8 42.Nxd2 Qe1†.
However, White can deviate with 41.Rf1 followed by 42.d6, with a clear advantage despite the missing pawn. The
difference is the active minor pieces.
41...c4 42.Qxb7 f6 43.Bd5† Kh8 44.Rxe5! fxe5 45.Rxe5 Rxe5 46.Qxc8† Kg7 47.Qe6
1–0
We will finish where we started, with a game from the World Championship cycle. This time it’s the quarterfinal of the
Candidates match between Garry Kasparov and Alexander Beliavsky. Kasparov was 19 years young, and had won the
Interzonal convincingly – except for a draw offer against Ulf Andersson. He was two pawns down, but with some
drawing chances. When Andersson had been thinking for three minutes, Kasparov saw his chance and offered a draw,
which was accepted.
Asked by a journalist if his behaviour was dubious, he answered: “Try to understand me! It was in the middle of a
highly unusual situation: mutual time trouble, the tension had reached its climax. Beliavsky said that he was convinced
that the result [against Andersson] was not yet decided. The sport ethics was thus not broken. I only made practical
considerations...”
And who could blame him? The saved draw was his turning point, and he followed up by winning four games and the
tournament. Andersson, on the other hand, was half a point away from reaching a tiebreak for a spot in the Candidates
matches.
4.Nc3 g6 5.e4
It doesn’t make any sense to keep the pawn on e3.
5...Bg7 6.Bb5†!
A precise check which disrupts Black’s development.
6...Bd7
6...Nbd7 7.a4! stops ...a7-a6 and ...b7-b5. Black should play 7...0-0 8.0-0 a6! before the bishop could retreat to f1.
White has lost a tempo after 9.Be2 but has avoided ...Bg4 and ...Na6-c7.
9.Re1 Nb4?!
The beginning of a good plan that doesn’t work.
10.h3 e6
Black threatens ...e6xd5 followed by ...Bf5.
11.Bf4!
Kasparov continues to make natural developing moves. Put the queen on d2 and the a-rook on d1 and he has a perfect
position.
11...e5
11...exd5 12.Bxd6 wins material.
And if 11...Bxb5 12.Nxb5! exd5 13.Bxd6 Black has to give up an exchange, since 13...Re8 is met by 14.e5 followed
by 15.Nc7.
12.Bg5
The game has transformed to a Czech Benoni position. In the discussion about the pawn structures, Nd2-c4 followed
by f2-f4 was pointed out as White’s plan.
19...Rg8 20.c3 Na6 21.Ng3 Qf8 22.Ndf1 Nh7 23.Ne3 Bf6 24.Bxf6† Nxf6 25.Ngf5
It’s almost over; Kasparov just has to attack the pawn on h6 with his heavy pieces.
Kasparov continued his bumpy journey towards becoming World Champion. He withdrew from the semi-final against
Viktor Korchnoi, since the match was due to take place in the USA, but Korchnoi agreed to move it to London.
Kasparov won, and in the final he convincingly beat Vasily Smyslov, who had qualified via a roulette wheel.
Then came 48 games against Anatoly Karpov in the match for the World Championship. We all know that match was
stopped, but later in 1985 Kasparov finally became the champion. It was Tigran Petrosian’s wife Rona who then said:
“Garry, I feel sorry for you, because the happiest day of your life is over.”
Recap
Are you convinced that there is a point to leaving the pawn on e3, or do you think that Black has tricked us? Somewhere
in between, I hope. There are other good choices against regular Benoni players, but I also think this line has a right to
exist. Instead of the English’s sour and the Greek’s unripe, we may conclude that the e4-pawn would have been
poisoned.
The motto you should keep in mind is patience. Don’t push e3-e4 until it’s good. Small developing moves like Rb1
and Be2-f1 may not seem much to the world, but they are useful even if we don’t see a concrete line.
However, in the Old Benoni (1.d4 c5 2.d5 e5) and in the Schmid Benoni (1.d4 Nf6 2.Nf3 c5 3.d5 d6 4.Nc3 g6), we
push the pawn to e4 in one go. More about that in the next section.
Theory
1.d4 Nf6 2.Nf3 c5 3.d5 e6 will be covered in the next chapter; as well as 1.d4 Nf6 2.Nf3 c5 3.d5 b5. But before moving
on to the main line, we start with four sidelines.
5.Nf3
a) 5...Bg4 Exchanging the dark-squared bishops runs into a problem. 6.h3 Bxf3 7.Qxf3 Bg5 8.Bxg5 Qxg5 9.Nb5! and
Black is forced to enter the following line: 9...Qd8 10.Qg4 Kf8 11.Nxd6! Nf6! 12.Qc8 Qxc8 13.Nxc8 Nxe4 14.Bd3
Nd7 15.Bxe4 Rxc8 16.0-0-0
With the king on d6, Black would have been better, but now White has easy play with d5-d6 and Rhe1.
b) 5...a6 6.a4 Bg4 is an improved version for Black, since there’s no Nc3-b5. Instead, White plays 7.h3 Bxf3 8.Qxf3
Bg5 9.Bxg5 Qxg5 10.Qg4!? and Black is forced to exchange. 10...Qxg4 11.hxg4 improves White’s pawn structure.
12.a5 is a threat, with Na4, Rb1 and b2-b4 coming when development is finished. 11...b6 12.Nd1!
The knight is ideally placed on e3, looking at both c4 and f5. The h-file is annoying for Black as well.
c) 5...Nf6 6.Be2 keeps the c4-square for the knight. Normal moves are: 6...0-0 7.0-0 Ne8 8.Nd2 Nd7 9.a4
White continues with Nc4 and prepares f2-f4.
a) 6...Nfd7 makes it difficult for Black to prepare ...b7-b5. 7.a4 Na6 8.0-0 Nc7 9.Bc4 0-0 10.Re1 a6 11.Bf4 Rb8
12.Qd3 and Black is not in time for ...Nf6 plus ...Bd7 and ...b7-b5 before e4-e5 comes.
This should be played before White can retreat to f1. Black wins a tempo, but the price is that he will not be allowed
to play either ...Bg4 or ...Na6-c7. A couple of replies are worth considering:
c1) 9.Be2 is the common move. White continues with Re1, Bf4, Nd2-c4 and possibly even Qd2 and Rad1 before
pushing e4-e5.
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 c5 6.d5 d6 7.Nc3 e5
Black threatens 8...e4, and when Jean-Marc Degraeve played like this against me in the French League in 2016, I
swallowed my pride and moved the e-pawn for a second time.
8.e4
It looks like White is a tempo down, but he has one argument in his favour: in the Czech Benoni structure, Black
normally wants to go ...Be7-g5. Instead what we have looks like a blocked version of the KID. However, as strong a
player as Rauf Mamedov actually goes for this position time after time even when it’s White to move.
The only decent alternative I see is to stop ...e5-e4 with 8.Nd2 Qe7 9.Qc2 Re8 10.e4. Now White has got something for
the tempo: ...Nh5 isn’t possible and the rook would have been better on f8. Both 0-0 intending a2-a3 and b2-b4, and
h4-h5 with 0-0-0 are possible plans.
Black does better to prepare ...f7-f5 with 8...Ne8. The engines always hate the King’s Indian and they are wild about
White’s position after 9.h4 f5 10.e4. I don’t see what’s so special, but at least the lost tempos means that the king is still
on e1 and the rook on h1 – two pieces of good news.
10...gxf5 11.g3!?
A surprising move that’s actually quite often a good antidote to Black’s ...f7-f5 in the King’s Indian. When Black has
a knight on d7, White attacks the pawn on f5 with Nh4 and Qc2. And ...e5-e4 is not a long-term defence due to f2-f3.
Another plan for White is f2-f4, to attack the f-pawn with the rook. And in case of ...e5-e4, White ideally manoeuvres
Nc3-d1-e3 and Bd2-c3. But that’s usually only a dream.
11...Nd7
If Black tries to attack on the light squares with 11...f4 12.gxf4 Bh3 he is in for a surprise: 13.Ng5! Bxf1 14.Qc2 Nf6
15.fxe5 dxe5 16.Bxf1 And it’s White who is winning on the light squares.
In Sweden, 1.d4 Nf6 2.c4 c5 3.d5 b5 is called both the Benko Gambit and Volga Gambit. I am not trying to make a case
for either name, however note that Pal Benko himself initially called the opening the Benoni Counter-Gambit.
And for us, it’s logical to cover the different tries here, since they arise from a Benoni move order. The next chapter
handles the Anti-Benko, 1.d4 Nf6 2.Nf3 c5 3.d5 b5 or 3...e6.
Martin Breutigam proposes 7.cxb5 a6 8.Nc3 axb5 9.Bxb5 d6 10.0-0 Ba6 11.Bxa6 Nxa6 12.e4 with an advantage.
The main line
The tabiya comes after nine moves, where Black chooses between exchanging his bishop on f3 (as we have seen three
games with), or 9...Re8, or the clever 9...Na6 followed by 10...Rb8, being ready for ...b7-b5 straight away.
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3(!) 4...0-0 5.Be2 c5 6.d5 e6 7.Nc3 exd5
7...d6 8.0-0 Re8 9.e4 exd5 10.exd5 was how Shakhriyar Mamedyarov beat Boris Gelfand in the 2014 Grand Prix in
Baku. Black is allowed to play ...Ne4, but he would have preferred to prepare it with ...Bf5 instead of ...Re8. After
10...Bf5 11.Bd3 Ne4 12.Nxe4 Bxe4 13.Bxe4 Rxe4 14.Qc2 Re8 15.Bf4 Black still has a few problems to solve
regarding his development.
The straightforward moves run into a problem: 15...Qb6 16.Rfe1 Nd7 17.Re2! And White wins the e-file since
17...Rxe2 18.Qxe2 Nf6 19.Re1 Re8?? 20.Qxe8† delivers mate.
8.cxd5 d6 9.0-0
The line splits into A) 9...Bg4, B) 9...Re8 and C) 9...Na6. 9...a6 10.a4 is likely to transpose to one of the other lines.
12.a4
12...c4N
Since I criticized 12...a6, we have to look at an alternative.
12...a6 was Kramnik’s choice. I prefer 13.Bd2 (to Morozevich’s 13.g4) and a few logical moves are 13...Rc8 14.Be2
c4 15.a5 Re8 16.Ra4 when White won’t win the pawn, but will open up the light squares.
13.Be2 Rc8 14.Bd2
14.a5 followed by Ra4 is what White wants to play, but first he has to protect the a-pawn indirectly since 14...Rc5
15.Qa4 Nxd5 exchanges the c-pawn for the d-pawn. This looks nice for the light-squared bishop, but we want to win
the pawn without giving anything in exchange! A plausible line is the following:
17...Re5!
Black has to create a counter-threat before White has played Ra4.
18.Bf3
White will still try to get the pawn on c4 for free.
B) 9...Re8
10.Nd2 Na6
10...a6 11.a4 Nbd7 is an example of Black’s simple development. We could play the standard 12.Nc4 Ne5 13.Na3
and continue developing with h2-h3, e3-e4, Be3 and so on. And f2-f4 is spared until we have improved the heavy
pieces.
11.Nc4
11.e4 accepts the loss of a tempo. It is not that stupid, considering that Levon Aronian and Hikaru Nakamura have
entered those positions with White.
11...Nc7
11...Ne4N has not been played to this date, but is a better option. 12.Nb5! threatens 13.f3 and after, for example,
12...Nc7, then 13.Nxc7! is not given in Attack with Black (2012). After 13...Qxc7 White should play 14.f3 Nf6 15.a4
when Black doesn’t have the counterplay he strives for in the Benoni.
12.a4
12...Ne4N
This has to be played immediately, as otherwise it’s stopped by 13.f3.
Clever, because White doesn’t want to play either 11.Nc4 b5 or 11.a4 Nb4.
15.f3N
Freeing the knight. Now there are two moves to consider: 15...Nb5 and 15...Bb5.
a) After 15...Nb5 16.Nc4 it’s not so easy for Black to create counterplay. Note that 16...Nh5 17.g4! holds everything
together:
a1) 17...Bxc3 18.bxc3 Nxc3 19.Qd3 Nxe2† 20.Qxe2 and White regains the pawn on d6.
a3) 17...Nxc3 18.bxc3 Bxc3 19.Ra3 Bxe1 20.Qxe1 Ng7 21.Nxd6 and the dark squares and the pawns in the centre
more than compensate for the exchange.
b) 15...Bb5 16.Nc4 Bxc4 17.Bxc4 b5 18.axb6 Rxb6 A standard exchange. With two vulnerable pawns on the queenside,
it’s difficult to organize e4-e5. A year after finding 15.f3, I reached this position in the Swedish league against Jonathan
Westerberg.
I evacuated the long diagonal with the prophylactic 19.Ra2. White has many useful moves (Bd2, Kh1 and f3-f4), but
Black can also play actively (...Nd7, ...Qf6 and ...Nb5). I would have wanted to say otherwise, but I think the position
is balanced.
The engines suggest the quickest way is 19.e5!? with decent compensation.
After 12...Bd7! I played 13.Bf1 against Sophie Milliet in the French League in 2016, but 13...Ng4 14.h3 Nxf2 15.Kxf2
Bd4† 16.Kg3 Be5† would have given a perpetual. Instead, Stockfish suggests a surprising move.
16...Bxe3† 17.Nxe3
Black has several dark squares (b4, f4 and e5) but White slowly recovers some of them with Bf1 and g2-g3.
However, it would be strange if Black wasn’t okay here and that can also be the conclusion regarding the whole
chapter. It’s strange, but it is.
Exercise 1
1.e3 c5 2.d4 Nf6 3.Nf3 g6 4.c4 Bg7 5.d5 0-0 6.Nc3 e6 7.Be2 exd5 8.cxd5 d6 9.0-0 Bg4 10.h3 Bxf3 11.Bxf3 Nbd7
12.a4 a6 13.g4 c4 14.Be2 Ne5 15.f4
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 c5 6.d5 e6 7.Nc3 exd5 8.cxd5 d6 9.0-0 Re8 10.Nd2 Na6 11.Nc4 Ne4
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.e3 c5 2.d4 Nf6 3.Nf3 g6 4.c4 Bg7 5.d5 0-0 6.Nc3 e6 7.Be2 exd5 8.cxd5 d6 9.0-0 Bg4
White to move
Show/Hide Solution
Show/Hide Solution
A) 3...b5 4.Bg5 91
a) 4...Qb6 92
b) 4...d6!? 92
c) 4...Bb7 92
d) 4...Qa5† 92
e) 4...Ne4 92
For a long time, I considered 1.d4 Nf6 2.Nf3 to be a move order for chickens, designed to avoid the Budapest Gambit
(1.d4 Nf6 2.c4 e5). Then I faced it – in Budapest – and suddenly realized that it stops the Benko Gambit as well.
A few days later I visited Pal Benko in his house on the outskirts of the city. For hours he told amazing stories,
leaving his coffee untouched until it had cooled. Escaping war, sitting in jail, defecting from the east, getting amnesty,
fighting Fischer physically before becoming a close friend to him.
He also said that he started to play the Benko Gambit in order to avoid theory and get a fighting game. Studying
openings made him fall asleep.
Inspired by his pep talk, I went straight to the round, played 1.e3 and won with the ideas in this book. And the next
day I was lucky enough to be allowed to play the Benko Gambit.
Pal Benko never had a reliable antidote to 2.Nf3 and usually transposed into other openings. Nonetheless, Black can try
1.d4 Nf6 2.Nf3 c5 3.d5 b5, where he will be behind in development, or 3...e6 4.Nc3 exd5 5.Nxd5, where he has to live
with a backward pawn. But they have their pluses as well.
The previous chapter saw how to handle Black’s Benko tries in the Benoni.
Pawn Structures
Structure 1
White has a structural advantage. The half-open d-file points towards a backward pawn, while Black’s e-file only
attacks a pawn that’s easily defended (by f2-f3 if necessary). In addition, d5 is a permanent outpost while e5 is not.
White plays c2-c3 to control the d4-square and develops naturally to keep d5 under firm control. He is happy to
exchange every minor black piece except for the dark-squared bishop.
But avoid Bc4xe6 – that would allow Black to re-establish control over d5 with ...f7xe6. (However, it can be an
option if it’s possible to follow up with e4-e5, threatening a knight on f6. The pawn exchange creates an isolated pawn
on e6, but it’s easy to defend and I would say that the structural advantage is bigger in the diagram position.)
It should also be noted that a piece exchange on e5 with ...d6xe5 doesn’t help Black much. There is still a weak
square on d5.
If everything goes according to the plan, the next step would be to increase the pressure on the d6-pawn.
So how does Black react? One plan is fighting for d5 with ...Be6 and ...Nc6-b4 or ...Nd7-b6. However, that will not be
successful if White knows what he is doing. Another possibility is to challenge White’s e-pawn with ...f7-f5, in order to
be able to play ...d6-d5 later. But I can’t see how that should work either.
Structure 2
Black has grabbed space on the queenside and hindered White from defending the pawn on d5 with c2-c4 or Nc3. He
will follow up by attacking that pawn with ...Nf6 and ...Bb7.
1) Defend the pawn with 1.c4 bxc4 2.Nc3. The sacrifice will probably be temporary, but in the meantime Black could
challenge the d-pawn with ...e7-e6 and force White to give up the centre.
2) Play for development with an early e2-e4, even if it exchanges the centre pawn for the b-pawn.
All in all, the position is quite messy, and piece play and unorthodox solutions are more common than standard moves.
Games
There is one game each for 1.d4 Nf6 2.Nf3 c5 3.d5 e6 and 1.d4 Nf6 2.Nf3 c5 3.d5 b5. In the first, White gets the ideal
situation and a smooth positional advantage.
7.e4 Nc6
But now 7...0-0! is more flexible – Black postpones the decision between ...Nc6 and ...Nd7-b6.
Black threatens 10...Be6 11.Qd3 Nb4 12.Qe2 Bxc4 13.Qxc4 d5 with full equality, but since he spent a tempo on
6...Be7, White has time to stop this plan.
This is the kind of position White strives for. He has managed to prevent the ...d6-d5 break once and for all, and can
double on the d-file.
Black can defend the pawn by exchanging light-squared bishops, placing the queen on e6, and doubling behind the
pawn. However, with a flexible pawn structure, it’s clear that White will have many ways to increase the pressure. We
don’t really have to think about the next step yet.
But I want to enjoy the position a bit more, so I suggest tripling on the d-file followed by Be3. The pawn on c5 hangs
and Black has to defend it with ...b7-b6, which weakens the light squares c6, b5 and a6. Also remember that if Black
exchanges his knight he can forget about all of those squares.
Instead of sitting passively, Miezis tries to challenge White’s control over d5 with ...f7-f5, but it falls short. It’s no
surprise considering that all White’s pieces are well-placed.
13...Kh8
13...f5 allows White to move on to the e6-square with 14.Bxe6† Qxe6 15.Qd5! Qxd5 16.exd5, as given by Dautov. If
the knight was able to retreat to d8, he could still have some hopes, but that square is occupied.
14.Rfe1 f5?!
Black threatens 15...fxe4 16.Qxe4 Bxc4 17.Qxc4 d5 18.Qa4 b5, winning a piece. But the threat is easy to parry and it
would have been better to stay passive.
15.Ng5!?
The same reaction as against 13...f5 still works fine: 15.Bxe6 Qxe6 16.Qd5 Black could (must!) avoid the exchange,
but it simply leaves White in control.
15...Bxc4
15...Bxg5 16.Bxg5 Ne5? looks tempting, but 17.Bxe6 Nxd3 18.Bxd7 Nxe1 19.Bxd8 wins. The knight is trapped
after 19...Rxd8 20.Bxf5 Nc2 21.Rc1.
16.Qxc4 Bxg5
Or 16...fxe4 17.Ne6 with a fork.
17.Bxg5 Rde8
17...Ne5 18.Qd5 Rde8 19.Bf4 will probably win a pawn quite soon. But not 19.Qxd6? Qxd6 20.Rxd6 Nf7 forking
the rook and bishop.
18.Qxc5
18...Re5?
Black actually gets the pawn back after 18...fxe4 19.Rxd6 Qf7 with a double threat, but 20.Be3 Qxa2 21.Qb5 leaves
him with every single piece worse than its white counterpart.
19.Qxd6 Qxd6 20.Rxd6 f4 21.Rd5 h6 22.Rxe5 Nxe5 23.Be7 Re8 24.Bd6 Nd3 25.Re2 Nc1
25...Rd8 tries to get in a back rank mate. White’s most precise reply is 26.Bc7 Rd7 27.Rd2, winning.
26.Rc2
1–0
White has too many pawns.
Instead of being positionally squeezed, Miezis went all in for counterplay. That’s the modern approach, which is why
old games are usually more instructive.
The next game is certainly old, but the mess starts after just a few moves. Even so, there is a positional lesson to learn
– about development.
4...Qa5†
The queen defends the pawn in advance and it seems like it doesn’t lose a tempo, as White blocks the check with a
pawn move. But the queen will be exposed when White plays a2-a4, so there’s a good reason why 4...Qa5† isn’t
common.
4...Ne4 and 4...Qb6 are the main moves, while 4...e6 5.e4 creates more problems for Black.
6...Nxg5 7.Nxg5
7...h6?!
Black was better advised to use White’s knight in order to make the developing move ...Bh6 with tempo. The
difference can be seen after 7...g6 8.a4 b4 9.Nc4 Qd8 10.cxb4 Bh6! 11.Nf3 cxb4. With the bishop still on f8, Black
would be in great trouble after 12.Qd4.
11.Rxa4?!
11.Nc4 Qc7 12.Qxa4 is more exact, again with e4-e5 as the killer.
15...Nc8 16.0-0 e5
17.dxe6
A no-brainer, opening the position with Black’s king in the centre.
17...fxe6 18.Nh4
Not only hitting the weak square g6, but also making space for f4-f5.
22.Qa2†
A winning move even when Black can block with the queen.
22...Qe6
23.Ne5†! Ke7 24.Nc6† Kd7 25.Qa4 Ke8 26.exf5 Qe3† 27.Kh1
Black’s king can’t escape.
1–0
Stahlberg was in great form and the organizers worried that a non-Soviet player was going to win the tournament, which
would have been embarrassing. They tried to give him free drinks, as he was known to appreciate those, but he
continued winning. Finally they offered Pal Benko a reward if he managed to beat the Swede. He started an early
attack, but it was hampered. “I was on the way to offering a draw,” says Benko 64 years later. “But then I thought about
the money and continued.” Stahlberg blundered in time-trouble.
The money didn’t serve him well. The secret police confiscated it when he was thrown in jail after trying to defect
from Hungary to West Germany.
Recap
The two illustrative games were very different. Against 3...b5, Gideon Stahlberg played forcefully to make use of the
fact that Black spent a tempo on the pawn move while placing it on an undefended square. After 3...e6, Rustem Dautov
on the contrary kept the position solid to put pressure on a backward pawn.
The similarity is that White doesn’t play c2-c4 in either game.
It is Black who chooses the line, but in any case he won’t be able to play his favourite Benko System. Remember that
even Pal Benko didn’t find a good solution to 2.Nf3 and transposed to other openings.
Theory
3...g6 4.Nc3 transposes to the chapter about the Benoni (page 76). It also makes sense to play 4.c4 b5 5.cxb5 a6 6.b6,
because in the b6-line Black’s most reliable antidote is to break in the centre with an immediate ...e7-e6. But this is
outside our repertoire.
A) 3...b5 4.Bg5
4.e4 Nxe4 5.Bxb5 is an interesting trade to win another tempo when threatening the knight on e4, but there is a
problem: 5...Qa5†
It was not easy to decide what to play here. 6.e4 is logical, remembering White’s first priority is to use his
development advantage. But after 6...Qxb2 7.Nbd2 g6 8.Rb1 Qxa2 9.Bxb5 Bg7 it doesn’t seem to offer more than
compensation.
6.c3 is the safe choice, recommended by Summerscale and played in almost every game in the database. After 6...d6,
I like the rare 7.Nbd2, leaving the e4-square free for at least another move. It stops Black from playing ...g7-g5, and a
piece move could be more useful than a pawn move if the position opens up with 8.a4 b4 9.cxb4.
b) 4...d6!? is a reasonable move. Then 5.e4!? is an interesting attempt to fight for the initiative.
5.Bxf6 will be our main focus though, leading to a thematic structure, even though 5...exf6 is much better for Black
than after 4...Bb7, since the bishop on c8 supports ...f6-f5.
6.e4 a6 7.a4 b4 8.Bd3 White has to put the bishop here to stop ...f6-f5. The following line is illustrative: 8...g6 9.Nbd2
Bg7 10.Nc4 0-0 11.0-0 Bb7!? What’s this? Wasn’t the bishop worse here? The difference is the placement of White’s
bishop: d5 is weaker when White’s bishop blocks the d-file and Black has been doing okay after 12.Qd2 f5. An
improvement that suggests itself is 12.a5 since 12...f5 13.exf5 Bxd5 14.Nb6 forces Black to give up the light squares.
e) 4...Ne4 5.Bf4 (It’s not good to play like Stahlberg with 5.Nbd2. With the queen still on d8, Black wins a tempo on
the knight after 5...Nxg5 6.Nxg5 e6.) 5...Bb7
The main move is 6.a4, but it’s not easy for White to get rid of the knight on e4 after 6...b4!. 6.e3!? isn’t common, but
was played in Daniel Gurevich – Ray Robson, Spice Cup 2015. White’s plan is 7.c3, threatening the pawn on b5. After
7...Qa5 8.Nbd2! Bxd5 9.Nb3 Bxb3 10.axb3 he wins back the pawn and ends up in a position with control over the
light squares. Robson played the logical 6...Qa5† 7.c3 Nf6 but 8.d6! gave White an initiative and later a fine scalp.
B) 3...e6 4.Nc3
White’s choice is a matter of taste and 4.c4 works for those who aren’t afraid of the Blumenfeld: 4...b5
4...exd5
4...b5 challenges White’s centre. It’s time for action: 5.e4! b4 6.dxe6! and Black has three alternatives.
a) 6...bxc3 7.exf7† Kxf7 8.e5 which wins back the piece with an attack.
b) 6...fxe6 7.Nb5 Nxe4 is a typical pawn sacrifice. Black is far behind in development after 8.Bf4 Na6 9.Ne5 Nf6
10.Qf3 and I don’t think Black will ever manage to castle.
This is recommended in Dangerous Weapons: The Benoni and Benko by Richard Palliser. He writes that White’s set-
up fails to impress, and he doesn’t even find a way to equalize. However 8.Ne5! makes Nc3-b5 stronger if Black
defends the pawn, and 8...Rg8 9.Nb1 gives a structural advantage with equal material.
6...Be7 was Garry Kasparov’s choice in a simultaneous display. He managed to play ...d6, ...Nd7-b6 and ...d6-d5. 7.e4
was Dautov – Miezis, but better is 7.Bf4!, a move that’s only been played in one serious game. Black has no way to
keep control over d6, for example 7...Nc6 8.0-0-0 0-0 9.Bd6.
7.e4 d6
7...Be7 is a clever move order that avoids the concrete Ng5-lines, as well as Bf4-d6. 8.Bc4 0-0 9.0-0 d6 10.Qd3 is
similar to the main line, but without White winning one or two tempos. However, even so Black has no way to
challenge the d5-square.
8.Ng5
The knight will be forced back to f3, but White and Black will not lose the same number of tempos. We will count
after White’s twelfth move.
8.Be2 was recommended in A Killer Chess Opening Repertoire (Summerscale and Johnsen, 2010) but the text move is
stronger.
8...Qe7
8...Qc7 would be more natural but allows White to develop smoothly. The text move blocks the bishop on f8, but
there is a tactical point after 9.Bc4 f6! 10.Nf7 Nb4 when White can’t defend both c2 and e4.
9.Bb5!
The Serbian grandmaster Dragan Kosic told me that this move was his invention. White forces Black to put the
bishop on d7.
13.Qd3N
13.Bf4 was played by Dragan Kosic against me in Budapest in 2015. Up to this point I had been following my
preparation, but that didn’t make me appreciate the position. However, now I should have played 13...Be6 14.Qd3
Bxc4 15.Qxc4 Qe6 16.Qd3 0-0-0! with decent chances. The rook on a1 has to defend a2, ...Qf6xb2 is a possibility and
Black also has ...g5-g4.
13...Be6 14.Bd5
It will not be convenient for Black to castle long after this move.
14...Qd7
Normal development would be 15.Bf4 followed by Rad1 and c2-c3, but White has another interesting possibility as
well.
Exercise 1
1.d4 Nf6 2.Nf3 c5 3.d5 b5 4.Bg5 Bb7 5.Bxf6 exf6 6.e4 a6 7.a4 b4 8.Nbd2
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 Nf6 2.Nf3 c5 3.d5 e6 4.Nc3 exd5 5.Nxd5 Nxd5 6.Qxd5 Be7
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 Nf6 2.Nf3 c5 3.d5 e6 4.Nc3 exd5 5.Nxd5 Nxd5 6.Qxd5 Nc6 7.e4 d6 8.Ng5 Qe7 9.Bb5 Bd7 10.Bc4 f6
White to move
Show/Hide Solution
Show/Hide Solution
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 Bb4† 5.Bd2 107
B) 5...Be7! 108
5...Qe7 107
5...a5 107
6.Nc3 0-0 7.Bd3 109
1.d4 Nf6 2.Nf3 e6 3.e3(!) b6 4.Bd3 Bb7 5.0-0 109 5.c4 112)
A) 5...c5 6.c4 Be7 109 (6...d6; 6...g6) 7.Nc3 cxd4 110 (7...0-0?; 7...d5?!; 7...a6!?) 8.exd4 d5 (8...0-0; 8...d6?!) 9.cxd5
Nxd5 10.Ne5 0-0 11.Qg4 Nf6 (11...f5) 12.Qh4 Ne4 (12...Nc6?; 12...Nbd7) 13.Qh3 Qxd4 14.Bf4 Nf6 (14...f5?)
15.Ne2 Qa4! (15...Qd8?!) 16.Rfc1 Ba6 (16...Na6) 17.Bc2 111
B) 5...d5 6.c4 Bd6 112 (6...Be7; 6...Nbd7; 6...dxc4) 7.b3 0-0 8.Nc3 Nbd7 9.Bb2 a6 (9...c5) 10.Rc1 Qe7 (10...Ne4)
11.Na4 dxc4(11...c5?!; 11...Ne4) 12.bxc4 Ne4 113 (12...c5) 13.c5 bxc5 14.Ne5! Rfd8! 114 (14...Qh4; 14...cxd4?)
15.Nxd7 Rxd7 16.Qc2N cxd4! (16...f5; 16...Qh4) 17.Bxe4 115
1.d4 Nf6 2.Nf3 e6 3.e3(!) 3...b6 4.Bd3
In the Queen’s Indian, we meet the isolated queen’s pawn for the first but not the last time in the book. Black has
blocked the pawn firmly, but on the other hand left his kingside without much protection. The game we will see is won
by a trap or a blunder, but it’s actually quite telling about the position.
A related structure is the hanging pawn couple. In the position we will see, White has space and “the freer game”, but
nothing concrete. But neither has Black: the pawns only become weak if we misplay.
The first things to learn about the Bogo-Indian are the different move orders.
a) 1.Nf3 Nf6 2.c4 e6 3.e3(!) avoids the Bogo-Indian.
b) The check on b4 is also avoided after 1.d4 Nf6 2.Nf3 e6 3.e3(!). Black has to play another move before White goes
c2-c4. However, 3...d5 4.c4 Bb4† is not the same as the original Bogo-Indian. After the exchange of the dark-
squared bishops, Black would have preferred to arrange his pawns on dark squares, which is no longer possible.
We will see one example of this, but from a strange move order.
c) When Black keeps the possibility of developing the light-squared bishop outside the pawn chain, White should think
twice before playing e2-e3. However, in the next part we will see that 1.d4 d5 2.Nf3 Nf6 3.e3 is actually
playable, but after 3...e6 we don’t have anything better than 4.c4, leading to the previous line (not 4.Bd3?! c5 or
4.b3?! c5).
Pawn Structures
Structure 1
There are some rules of thumb regarding exchanges in positions with isolated pawns:
– White wants to keep everything on the board, but can exchange heavy pieces if he is playing for a draw.
– Black’s dream is to exchange all minor pieces and triple on the d-file. The pawn would probably be doomed; Black
has ...e6-e5 if White defends from behind.
– The only minor piece that White should consider exchanging without getting anything in return is the restricted
dark-squared bishop.
So most exchanges are bad news. Instead, the advantage with the isolated pawn is the outpost in the opponent’s camp
(e5). In order to control it fully, Black has to play ...f7-f6, which weakens the e-pawn and the b1-h7 diagonal.
This pawn structure will appear again in two more chapters, with only slight differences. In this version, Black already
has a bishop on b7. That makes White’s standard plan impossible: placing his pieces more actively than Black (who has
less space) and pushing d4-d5, releasing his forces.
Instead White will play with the pieces on the kingside: Bd3, Qg4-(h3), Bh6/g5. However, there is no reason for
White to panic. White doesn’t need any sparkling magic to justify the isolated pawn. It’s not a real weakness as long as
unfavourable exchanges are avoided.
Structure 2
The minor pieces are arranged like this: Nc3, Nf3, Bd3, Bb2, 0-0 (and Rc1), ...Bb7, ...Bd6, ...Nf6, ...Nbd7 and ...0-0.
White’s plan is to threaten c4-c5 and force a reaction from Black. That’s done with Nc3-a4, which also opens up for the
bishop on b2 and allows Nf3-e5. Black can keep control over c5 with ...Qe7, but sometimes White could play c4-c5 as
a pawn sacrifice, opening the diagonal b2-f6. In case of ...c7-c5, White takes on d5 and Black has to choose between
giving up the centre (...Nxd5) or getting a slightly weak pawn (...e6xd5).
Structure 3
Black has weakened the c6-square and if he doesn’t get in ...c7-c5, White will build up pressure along the c-file with
Qa4 (tying the rook to a8), Rac1-c2 and Rfc1. Black has two ways to defend.
a) ...c7-c6, blocking his light-squared bishop. White can open the position on the other side of the board with e3-e4.
b) If Black tries to do without ...c7-c6, White can consider b4-b5 or a2-a4-a5, though the latter is most efficient if the
rook has never left a1.
And even if Black gets in ...c7-c5, he gets a worse structure with an isolated pawn on d5. What he is hoping for is
instead to play ...Ne4 and ...f7-f5, followed by an attack on the kingside. However, he is usually too busy defending to
have time for that.
Games
We already saw a Queen’s Indian game in the introduction: Kramnik – Topalov. Now we will start with the Bogo-
Indian.
Harikrishna, April 2016: “I was actually playing for advantage in the game, but probably Black missed chances
somewhere.”
1.d4
1.Nf3 Nf6 2.c4 e6 3.e3 d5 4.d4 Bb4† 5.Bd2 Bxd2† 6.Nbxd2 transposes to the game.
1...e6 2.c4
It’s not possible to castle before c2-c4: 2.Nf3 Nf6 3.e3 d5 4.Bd3 c5! and with a move spent on Bd3, White can’t
convert to a Tarrasch.
7.Bd3
7.Be2! is accurate. The bishop doesn’t step into an attack with ...c7-c5 and ...Nxc5.
7...0-0 8.0-0 b6
8...c5! would have taken advantage of White’s inaccuracy.
9.cxd5!?
This exchange is usually inoffensive, since it allows Black’s bishop to develop to e6 or g4(-h5-g6). But things are
different with ...b7-b6 having been played. Black has weakened the c6-square and the pawns can’t defend each other in
the chain b7-c6-d5.
Instead, he should play for ...c7-c5 with the knight on d7. As a consequence, the bishop has no other choice than the
b7-square.
9...exd5 10.Rc1?!
10.b4 forces Black to accept an isolated pawn if he goes ...c7-c5.
10...Bb7?!
But now, Black had the chance to get a hanging pawn couple with 10...c5 11.dxc5 bxc5, even though 12.e4! splits the
pawns and still gives White an advantage.
11.b4!
This makes the positional advantage stable, but what’s the next step? Now a2-a4-a5 isn’t inviting without a rook on
a1. The best way to put pressure on the queenside is instead b4-b5 followed by Qa4, Rc2 and Rfc1. If Black defends
passively, the turn could eventually come for the a-pawn.
11...c6 12.Ne5!?
12.e4 is always a move to consider when Black has restricted the bishop on b7. Black should not allow the pawn to
advance another step and 12...dxe4 13.Nxe4 activates White’s minor pieces.
12...Nxe5
There is no other good way of defending the c-pawn as 12...Rc8 13.Qa4 hits two pawns.
14...Qe7!
Without a pawn on d4, Black has 14...c5. However, it’s clever by Arkell to start with a developing move that wins a
tempo due to the threat against the pawn on b4.
15.Nf3
15...f6?!
Of course 15...Qxb4? allows the threat: 16.Bxh7†! Kxh7 17.Ng5† Kg6 18.Qg4 And White wins due to the double
threat of f4-f5† and Ne6†. For example, 18...f5 19.exf6 Nxf6 20.f5† and the black queen drops.
15...g6 was a safer way to defend against 16.Bxh7†. White can’t stop both ...Qxb4 and ...c6-c5.
16.Qc2
16.Nh4! would have been strong, pinpointing that the fianchettoed bishop on b7 doesn’t defend f5.
16...fxe5?
Giving up the h-pawn is clearly a mistake. With only one pawn left on the kingside, Black will always have problems
with his king.
16...g6 17.e6 almost wins for White after the forced variation (from Black’s perspective): 17...Qxe6 18.Bxg6 hxg6
19.Qxg6† Kh8 20.Ng5 Qxe3† 21.Kh1 fxg5 22.Rce1
But there is 22...Rxf4! and White has to go for a perpetual.
21.Bxd7!?
Removing a potential defender.
21...Qxd7 22.f5
Strategically, advancing the kingside pawns undresses the black king and wins the game. However, Black has a tricky
way to achieve his desired ...c6-c5.
22...a5
There’s a sneaky threat: 23...Ba6
23.Qb2?
23.b5! c5 threatens 24...d4 but 24.Rcd1 Re8 25.Rf4 still stops Black’s pawns from advancing: 25...d4? 26.exd4 cxd4
27.Qc4 with a double threat on d4 and f7.
23...Qe7!
28.Kh1 d4 29.Rf7
29...Bd5?
There was a draw after 29...Qxe3! 30.Rxb7 Rxa2 31.Qxa2 Qxc1† 32.Kg2 Qg5† with too many checks.
34.Ra7?
The draw was there with 34.e7 e3 35.Qxa8 Rxa8 36.Rd8† Kh7 37.e8=Q Rxd8 38.Qxd8 and a repetition after
38...Qc6† 39.Rg2 Qc1† 40.Rg1 Qc6†.
White loses the queen and can’t stop the black pawns. After 38.Rg3 Bxg2† 39.Kxg2 Qxa2† 40.Kf3 White’s rooks
will create some mating threats, but the queen has a lot of checks and defends against the threats with tempo.
The players agreed a draw after 37...Kh7, but initially I thought that the result was wrong in ChessBase. But trying to
convince Harikrishna that he had lost was an impossible task! He actually believed that the final position was drawn,
and that confidence probably made it easier for Arkell to believe the same.
White gained a pleasant positional edge from the opening and decided to make more of it with 12.Ne5!?, at the same
time allowing Black some counterplay. Harikrishna followed the right path for quite a while before losing the thread.
The following two games are much smoother; both are miniatures decided straight out of the opening.
However, the semi-waiting move 7...a6 is interesting. White should use the same strategy with 8.Re1.
11.Qh5 is just as common as the text move, but theory gives 11...g6 12.Qh3 Nc6 13.Bh6 Re8 14.Bb5 Nxd4! as a strong
exchange sacrifice.
11...Nf6
11...f5 12.Qe2 was seen in Kramnik – Topalov in the introduction (page 10).
12.Qh4
White sacrifices the pawn on d4 to spoil the coordination of Black’s pieces.
12...Ne4
12...Nc6? 13.Bg5! leads to two tactical exercises after 13...g6 and 13...h6. Can you solve them?
Black has to give up an exchange after 13...Nxe5 14.Bxf6 Nxd3 15.Bxe7 and 16.Bxf8.
15...Rf6 (15...Nc5 defends the pawn on e6, but not for long after 16.Qg3N followed by Rfd1 and b2-b4.) 16.Rad1! Qc5
17.Nxe4! Bxe4 18.Bxe6†! Rxe6 19.Qb3 Qc8 20.Rd7!! Black can’t defend the rook on e6 and after 20...Nxd7
21.Qxe6† Kh8 22.Nf7† Kg8 23.Nd6† it was time to resign, since she loses her queen.
So this is a good example of a threat being stronger than its execution. White should have avoided 16.Bg5, and in the
theoretical section we prefer 16.Rfc1.
17.Bxh6
Braun is the first grandmaster to compete in chess-boxing, and this is his punch.
17...gxh6 18.Qxh6
The threat is Ne2-g3/f4-h5 followed by mate on g7 or h7.
18...Ne4
19.Rae1!
Accurate. The queen is kept away from d4 for another move. Instead 19.Ng3 Qd4! would not have been as clear.
19...Qe8
19...Bg5 20.Qh5 doesn’t change anything.
The isolated pawn positions in the Queen’s Indian are perfect ground for creating a quick attack against the king. Black
was lost even earlier in the next game, but that fate is not logical considering the position – it’s more of an opening trap.
Flear, April 2016: “It’s perhaps more satisfying to win a tough game where one has to make several good moves to
make the difference, but I was still happy with this one!”
Glenn Flear – Yannick Pelletier
Montpellier (rapid) 2015
6...Nbd7 7.b3
Or 7.Nc3 first.
7...Bd6
7...Be7 transposes to something that will be covered in the chapter about the Queen’s Gambit (page 288).
11...c5
We will see more of the main move 11...dxc4 later.
12.c5 was not yet a threat: 12...bxc5 13.dxc5 Nxc5 14.Nxc5 Bxc5 15.Bxf6 gxf6 16.Bxh7† Kg7! and Black is better
with the pair of bishops. But 11.Na4 has another idea as well: making Nf3-e5 possible.
12.cxd5 exd5?
12...Nxd5 defends the pawn on b6. White’s advantage in the centre is mostly optical – I think Black is okay.
13.Nxb6!
Despite being a rapid tournament, Pelletier took his time in the opening and Flear had time to find and calculate the
sacrifice. “Although I wasn’t then totally sure it was 100% sound, I just couldn’t resist!” he says.
16.Nh4!
The problem for Black is that he can be mated on either g7 or h7.
16...Rfc8 17.Nf5
17...Qe5
17...Qf8 18.Qh5 Kh8 defends for a moment, but White can take the piece back with 19.Nd6 and it would soon be
over.
Exactly a month later Glenn Flear had a chance to create another miniature against Nicolas Brunner in the French
League. That time, he could play the moves quickly and won almost without having used any time. Brunner played
17...Qe6 18.Qh5 Ba3 19.Qh6 Qxf5 20.Bxf5 Rxc1 21.Bxh7† and resigned three moves before being mated.
18.Qg4† Kf8 19.Rxc5! Rxc5 20.Qg7† Ke8 21.Qg8† Kd7 22.Qxf7† Kc8 23.Qf8†
1–0
Black loses the rook on c5 and resigned, since he will be two pawns down and will probably lose more material.
Recap
The illustrative games showed one loss for White and two traps Black fell into. Miniatures actually appear more often in
the Queen’s Indian than in any other chapter. Black gets a solid position on the queenside, but has to pay the price with
a (temporarily) unprotected kingside.
Theory
The Bogo-Indian
5...Qe7 and 5...a5 are also possible, though in these variations of the Bogo-Indian, Black usually places his d-pawn on
d6. White can develop normally or try the engine’s 6.c5 – it’s not as stupid as it looks. White is happy to force the
exchange of the dark-squared bishops, and with ...a5 played, ...b7-b6 followed by ...a7xb6 is no longer possible.
A) 5...Bxd2†
Exchanging the dark-squared bishops is normally followed up by arranging the pawns on dark squares (...d7-d6 and
...e6-e5). In this position, Black has to develop the light-squared bishop to b7, but whichever move order he uses, he
gets a weak pawn.
6.Nbxd2
From d2, the knight put less pressure on d5 than it would have done from c3. Instead, it discourages ...d5xc4.
6...0-0 7.Be2!
Staying away from threats with ...Nbd7xc5.
Black has three options:
b) 7...dxc4 8.Nxc4 b6 9.0-0 Bb7 10.b4 If Black ever manages to play ...c7-c5, White has at the very least d4xc5 plus b4-
b5, with a queenside majority.
c) 7...Nbd7 8.0-0 c5 9.dxc5 Nxc5 The downside of having the knight on d2 is shown: it’s not possible to give Black an
isolated pawn on d5. But Black’s problem is that he has opened the position two moves before he has completed
development. 10.b4 Ncd7 11.c5
White has a nice majority, while Black’s centre isn’t strong without the dark-squared bishops.
B) 5...Be7!
This is an improved line as White can’t play Bd2-g5 (but in a game Epishin – Drozdovskij in 2009, it was Black’s
move after the remarkable Bc1-g5-d2 followed by e2-e3). I can’t say I have a great suggestion about how to handle this
move.
Instead 6...d6 7.Nc3 Nbd7 is in Hedgehog style. 8.b3 (without a discovered attack, White can’t play 8.d5) 8...Be7
9.Bb2 cxd4 (9...0-0?! allows 10.d5!) 10.exd4 0-0 11.Re1 Yusupov – Teske, Germany 1995, showed an interesting plan
for White: h2-h3 followed by Re3, Qe2, Rae1 and Ng5 – and all of a sudden Black couldn’t defend against a sacrifice
on e6 or f7.
6...g6 allows 7.d5! (White usually starts with 7.Nc3 Bg7 before going 8.d5, but the text move gives Black fewer
options) 7...exd5 8.cxd5 and:
a) 8...Bg7 9.e4 is a good Benoni for White. The lost tempo e2-e3-e4 is compensated by the two dubious moves ...b7-b6
and ...Bb7.
b) 8...Nxd5 has only been played once. Without the knight on c3, White has 9.Be4 Nxe3 10.Bxb7 Nxd1 11.Bxa8 Nxf2
12.Rxf2 when the rook and two pieces will eventually prove better than the queen and pawns.
c) 8...Bxd5 9.e4 Bc6 Compared to a similar pawn sacrifice after 4.g3 in the Queen’s Indian, White’s bishop is better on
d3. It gives good compensation with normal moves: Re1, e4-e5, Nc3-e4. It’s most precise to start with 10.Re1 and
delay e4-e5 until Black has played ...Bg7. The reason is that White has Bg5 as a follow-up without the possibility of
...Be7.
7.Nc3
7...cxd4
A typical variation is: 7...0-0? 8.d5 exd5 9.cxd5 Nxd5 10.Nxd5 Bxd5 11.Bxh7† Kxh7 12.Qxd5
7...d5?! allows 8.cxd5 exd5 9.Bb5†! Bc6 10.Qa4 Bxb5 11.Qxb5† Qd7 12.Ne5! with an initiative, as in Petrosian –
Keres, Moscow 1951.
7...a6!? is played from time to time. White isn’t ready for d4-d5 as long as Black delays castling (no Bxh7†). And
8.Re1!? 0-0 9.d5! allows Black to show his idea with 9...exd5 10.cxd5 b5 but even though ...b6-b5 is a good move
against the knight on c3, Black is unhappy with his light-squared bishop after 11.e4! d6 12.a4! b4 13.Nb1. In Malaniuk
– Conquest, Oviedo (rapid) 1993, White jumped around to c4.
8.exd4 d5
There are two dubious ways to allow d4-d5: 8...0-0 9.d5 and 8...d6?! 9.d5 e5 10.Nh4!, which is a typical move with
the bishop on b7. 10...g6 11.Bh6 is much better for White.
11...Nf6
11...f5 12.Qe2 was Kramnik – Topalov in the introduction (page 10).
12.Qh4 Ne4
We have already seen that 12...Nc6? 13.Bg5 wins for White.
12...Nbd7 releases the pressure against d4, so White can play: 13.Qh3 Re8 14.Re1
The threat is 15.Nxf7 Kxf7 16.Qxe6† Kf8 17.Bc4. 14...Rc8! defends the c4-square. I recommend: 15.Bd2N The
bishop is normally developed to g5, but that allows the simplifying manoeuvre ...Nf8 followed by ...N6d7. 15...Nf8
can now be met by 16.Bb5! N6d7 17.Nxd7 Nxd7 18.Rxe6! fxe6 19.Qxe6† Kh8 20.Bxd7 with two pawns more.
15.Ne2 Qa4!
Black prepares ...Ba6.
15...Qd8?! 16.Rfd1 Bd5 is not safe: 17.Bg5 g6 and 18.Bh6 wins the exchange due to 18...Re8 19.Bb5, but it’s even
stronger to keep up the pressure with 18.Qh4.
16.Rfc1
Braun played 16.Bg5, using up his threat against h7.
16...Ba6
16...Na6 17.Rc4 Qe8 18.Ng4 is another way to make use of the threat. 18...g6 19.Nh6† Kg7 20.Be5 gives a strong
initiative, since the knight on a6 can’t challenge the bishop on e5.
19...Rc8! 20.Qf3
But not 20.Ra2? Rxc3! 21.Rxb2 Rxh3 22.gxh3 Nd5 which wins for Black with the double threat 23...Nxf4 and
23...Bxa3.
The point behind threatening the rook on a8 is 20...Nbd7 21.Nxd7 Nxd7 22.Ra2 Rxc3 23.Qxa8† winning two heavy
pieces.
20...Bd6! 21.Qxa8
21...Nc6 traps the queen, but White gets too many pieces after 22.Nxc6! Rxa8 23.Bxd6.
23.Bb8! Ng4
A 2001 correspondence game, Borrelli – Gellert, was drawn after Rab1-a1. The engines suggest a way to play on.
B) 5...d5 6.c4
6.b3 allows 6...c5 7.c4 cxd4 8.exd4. Black captured first and White has to play with hanging pawns, like in Franco
Ocampos – Goldin later in the book (page 291). This position was also reached in a blitz game Kramnik – Kasparov,
Moscow 2001.
6...Bd6
6...Be7 transposes to the Queen’s Gambit.
6...Nbd7 will transpose elsewhere based on what Black does next.
6...dxc4 7.Bxc4 a6
This gives a Queen’s Gambit Accepted where White has lost a tempo with the bishop. However, Black is restricted to
...b7-b6, so it’s logical for White to choose a set-up without a2-a4. One such is 8.b3 Nbd7 9.Bb2 Bd6 with Qe2, Rfd1,
Nbd2 and Rac1.
The difficult choices are:
a) Should White go e3-e4-(e5)? Probably not, as Black would easily counterattack against the centre.
b) How should White handle ...c7-c5? The best would be to pass by with d4-d5 (and a knight on c3), but it’s not likely
that it will work. The second option is to allow ...c5xd4, take back with the knight, and try to harass Black’s
queen. White’s advantage is that he has a good square for the queen on e2, while Black doesn’t have a similar
option.
For that reason, it’s interesting to discuss the move order 5.c4 (instead of 5.0-0) even though it allows the Bogo-Indian
without the pawn on d5 (5...Bb4†). After 5.c4 d5 we have:
a) 6.0-0 transposes and allows Black to take on c4.
b) 6.b3 also allows 6...Bb4†. It’s not a common move, but I think Black is okay after 7.Nbd2 0-0 8.0-0 Nbd7 9.Bb2 and
even though White gets a2-a3 for free, he would have wanted the knight on c3, as in Glenn Flear’s game.
Also interesting is 6.b3 c5!? when the normal move would be to capture first with 7.cxd5. However, b2-b3 is
superfluous after 7...Nxd5! when White would have preferred to develop with 0-0, e3-e4, Nb1-c3 and Bc1-f4.
c) 6.cxd5 exd5 7.b3 Bd6 8.Ba3 was played in an impressive rapid game Kramnik – Pelletier, Geneva 2013, where
White’s pressure along the c-file finally prevailed.
Frode Urkedal improved with 7...Bb4†! against Johan Salomon in Tromsø 2016. White can’t go Bc1-a3 any more.
And if White avoids the check with 7.0-0 Bd6 8.b3, Black has 8...Qe7, again avoiding the exchange.
So what is the conclusion? I faced 6...dxc4 against Urkedal in the Swedish League in October 2016. A few days before
the game, I had broken a bone behind my eye, and so I wore a pirate patch to avoid seeing every piece twice. So I was
happy to play solidly, exchange all the pieces, and make a draw.
However, next time I will probably play 5.c4 d5 6.b3.
10.Rc1
10...Qe7
10...Ne4 11.Ne2!? Qf6 stops 12.Ne5. 12.Qc2N defends the bishop on b2 and threatens 13.c5. After 12...Qh6
13.Ne5!? White plans 14.Nf4.
11.Na4 dxc4
11...c5?! happened in Glenn Flear’s games, as we saw.
After 11...Ne4 12.Ne5! it’s too late for Black to take on c4 – White will take back with the knight.
12.bxc4 Ne4
The main theoretical move which leads to concrete lines; first comes a pawn sacrifice.
12...c5 is playable when White can’t open the c-file. I played 13.Ne5 Rfd8 14.f4 against Vladimir Sveshnikov in the
2016 Olympiad in Baku, but he could have equalized with 14...b5! 15.cxb5 axb5 16.Bxb5 Nxe5 17.fxe5 Bxe5 18.Nxc5
Rxa2 – another example where rook against queen is an important factor.
Logically, White should step off the file with 14.Qe2. It’s a nice-looking version of hanging pawns, since Black has
no pressure against the pawns. He wants to exchange pieces, but 14...cxd4 15.exd4 Ba3 fails to 16.Nxd7 Nxd7
17.Bxh7† Kxh7 18.Qd3† Kg8 19.Bxa3.
Black should start by defending the b6-pawn. After 14...Rab8 15.f4 Ba8, he has two threats.
1) To exchange bishops on a3 (since he can take back on d7 with the rook and on h7 with the knight).
2) ...Ne4 (since there’s no threat of Nxd7 plus Nxb6).
White can defend against the first threat with 16.Rb1, a useful move with Ba1 in mind. However, against Sveshnikov I
experienced that a knight on e4 is a great obstacle, so the first priority is 16.Nc3. After 16...cxd4 17.exd4 Ba3 18.Bxa3
Qxa3 19.Qe3, Black has at least made a concession by exchanging on d4. White is slightly better, even though not as
much as the engines believe (0.55). Part of the advantage is only visual – there’s no clear way forward.
13.c5 bxc5 14.Ne5!
The threat is to win a piece with 15.Nxd7 Qxd7 16.Bxe4 Bxe4 17.dxc5.
14...Rfd8!
Rook against queen.
14...Qh4 15.g3 Qh3 doesn’t deliver mate: 16.Bxe4! Bxe4 17.f3 Bxe5 18.dxe5 Bc6 19.Nxc5 and Black has problems
with the c-pawn and the offside queen.
14...cxd4? 15.Nxd7 Qxd7 16.Bxe4 Bxe4 17.Qxd4 creates a double threat, but 17...f5 defends. White has a much better
pawn structure after 18.Nc5 and taking twice on e4.
15.Nxd7 Rxd7
16.Qc2N
16.dxc5 Bxh2†! 17.Kxh2 Bc6 is given in Play the Queen’s Indian (Andrew Greet 2009). The threats against a4, d3
and g2 give Black a good position.
16...cxd4!
16...f5 17.f3 and the knight cannot move due to d4xc5-c6. Black should try 17...Qh4 18.g3 Nxg3 19.hxg3 Qg5! but
he has insufficient compensation after 20.Bc4.
16...Qh4 leaves the protection of the c5-pawn, which is unfortunate after 17.f4! f5 18.dxc5 Nxc5 19.Nxc5 Bxc5
20.Qxc5 Rxd3 21.Qxc7 with threats against b7 and g7. Better is: 20...Qg4! 21.g3 Rxd3
This time, g7 is defended. But White has 22.Qb4! Rb8! 23.Rxc7 with a double threat. After 23...Rd1! 24.Rxg7† Qxg7
25.Rxd1 Black has to give up a second pawn to avoid mate.
17.Bxe4 Bxh2† 18.Kxh2 Qh4† 19.Kg1 Bxe4 20.Qd1
The queen defends against ...Bxg2, which would secure an immediate draw.
Suddenly it’s White who wants to keep the queens on, as an exchange would allow Black to advance his pawns. But
now Black has to worry about his king, and with opposite-coloured bishops, an attack seems promising for White –
even though it hasn’t started yet.
Exercise 1
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 Bb4† 5.Bd2 Bxd2† 6.Nbxd2 0-0 7.Be2 b6 8.0-0 Bb7 9.cxd5 exd5
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 Nf6 2.Nf3 e6 3.e3 b6 4.Bd3 Bb7 5.0-0 c5 6.c4 Be7 7.Nc3 cxd4 8.exd4 d5 9.cxd5 Nxd5 10.Ne5 0-0 11.Qg4 Nf6
12.Qh4 Ne4 13.Qh3 Qxd4 14.Bf4
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 Nf6 2.Nf3 e6 3.e3 b6 4.Bd3 Bb7 5.0-0 d5 6.c4 Bd6 7.Nc3 0-0 8.b3 Nbd7 9.Bb2 a6 10.Rc1 Qe7 11.Na4 dxc4
12.bxc4 c5 13.Ne5 Rac8
White to move
Show/Hide Solution
Show/Hide Solution
Tamaz Gelashvili
Many players have repeatedly played several of the
lines in this chapter, but not many have been as
consistent as the Georgian grandmaster Tamaz
Gelashvili. He has 65 games with 1.Nf3 and 2.e3 in the
databases, with a performance of 2600+.
But there are also famous players who have played with
the e3 poison. One is Alexander Grischuk. Sometimes
he seems to care more about playing well than scoring
points. He can spend a huge amount of time to find a
small detail, and when it’s clear that he has outplayed
his opponent, he might be in nasty time trouble and
sometimes misplays. And preparing all night until
breakfast is anything but practical, as he did in the
2011 World Cup.
But Grischuk also has a strong practical side, and he
has contributed to many of the lines in this book. There
are three of his games and many references.
Garry Kasparov
Alexander Grischuk
In 1994, Fritz 3 was allowed to enter the Intel World
Chess Express Challenge, a blitz tournament in
Munich. Up to that point, Garry Kasparov had never
lost a tournament game to a computer and he still felt
that he proved humanity’s superiority against the
machine’s brute force.
Before the tournament, he was allowed to practise
with Fritz, and he found a weak spot in its opening
repertoire.
1.e3
Kasparov had played 1.e4 and 1.d4 against his grandmaster colleagues, but he had found a way to outsmart the
computer.
1...d5 2.c4 dxc4 3.Bxc4 e5
3...e5 is regarded as a good move after 1.d4 d5 2.c4 dxc4 3.e3. But here, White can delay d2-d4 to put immediate
pressure against f7.
4.d4
In the final, Kasparov played 4.Nc3 and won a nice
game where he sacrificed a piece for the initiative. But
this move order allows Black to transpose into a
normal Queen’s Gambit Accepted.
b) 4...Nc6 5.d4 exd4 (5...e4 runs into 6.Qb3! Nh6 7.Nfd2 when 7...Qg5 8.Bf1!? f5 9.Nc3 has been played. Black’s
position is too airy, and the weakness on c7 probably forces him to retreat with the queen and allow White to develop
the bishop from f1.) 6.0-0!? White avoids the check on b4. Black has two tries.
b1) 6...Nf6 7.exd4 Bd6 is Black’s ideal piece placement, but since he has been forced to play ...Nc6 instead of castling,
there is 8.Re1†. 7...Be7 is of course viable but passive.
b2) 6...dxe3 is critical. The line goes 7.Qb3! exf2† 8.Rxf2! Bc5 9.Bxf7† Kf8 10.Be3! Bxe3 11.Qxe3 and the bishop on
f7 is untouchable due to 11...Kxf7 12.Ne5† followed by a discovered attack, winning the queen.
Black is still material up after 11...Nf6 12.Bb3 and can force a queen exchange with 12...Ng4 13.Qc5† Qe7, but
White’s initiative is worth more than the pawn: 14.Qxe7† Kxe7 15.Re2† and the knights will do the job.
15.Bb3??
15.Bxb6 axb6 16.f5 would have won a piece. Commentator Maurice Ashley said: “For him to have missed this move
is like Michael Jordan missing a lay-up totally open under the basket.”
17...c6 18.f5 Qe7 19.f6 Qxe5 20.fxg7 Kxg7 21.Ne4 Nd5 22.Bxd5 cxd5 23.Ng3 Kg8 24.Nf5 Rac8 25.Qf2 Rc4
26.Nh6† Kh8 27.Bxa7 f6 28.Nf5 Re8 29.a3 Be1 30.Qg2 Re4 31.Nh6 Re7 32.Rf5 Re2 33.Rxe5 Rxg2† 34.Kxg2 fxe5
35.Bb8 e4 36.Be5† Rxe5 37.Nf7† Kg7 38.Nxe5 Bd2 39.Kf1 Bc1 40.b3 Bxa3 41.g5 d4 42.Ke2 d3† 43.Kd2 Bd6
44.Nc4 Bf4† 45.Kc3 b5 46.Nb6 b4† 47.Kb2 e3
0–1
Fritz 3 qualified for the final, where Kasparov repeated 1.e3. It brought him two victories, leading to a convincing 4-1
win. But he could no longer claim that it was impossible for him to lose against a computer.
“The real fight is 25-minute games,” he said. And a few months later: “In serious classical chess, computers do not
have a chance in this century.”
The rest of the story is well known, but let’s stay here, happy that our hero’s 1.e3 caught the machine with its pants
down.
French Attack
Before I played my first tournament in 1997 as an 11-year-old kid, the club organized a simul with Stellan Brynell. A
friend told me that he played the French Defence and I decided to give it a try.
Brynell: “The 1200 player collapsed in the end.”
1.e4 e6
My knowledge of theory ended here.
2.d4 c5 3.d5
Probably not a move that I had considered. At that time, I thought that you had to exchange if you could. At least I did
so.
3...d6
The closed position allowed me to survive a bit longer than some of my friends. It took until move 18 before I lost
material.
Later I came up with 3...Bd6 4.Nf3 f6, called it Smith’s Defence, and even wrote a summary about it.
There’s no doubt we are beyond the part of the chapter about heroes, and it was a good decision not to show the
pamphlet to anyone other than myself. But my love for the French lasted ten years, and in my first tournament I started
all my White games with 1.e3. But it was a French Attack, not a poisonous repertoire.
In December 2015 I travelled to Hungary with a one-way ticket to chase my final GM norm. Preparing for the trip, I
asked Brynell to help with the practical side of chess: time management, goals and attitude.
Suddenly I recalled how 18 years earlier he had influenced me to play 1.e3, and thought that the move might be worth
another try. The idea behind Poisoned Brynell was sown.
After a month’s work I really liked it. 1.e3 is useful and flexible, even though the bishops wouldn’t agree: it restricts
the dark-squared bishop and rules out a kingside fianchetto for the other.
One of the ideas behind the set-up and the name is that it looks innocent, but contains more poison than one might
think. Black can easily be tricked into an opening he normally doesn’t play. In one game, my opponent even said
“What!” when he saw the first move, but then fell for a trap and lost straight out of the opening. He believed he was
facing just another Colle.
It also worked well in Hungary.
3...Bg7
Black shouldn’t think that he can do whatever he wants, as 3...d5?! 4.c4! gives him problems after 4...cxd4 5.cxd5.
3...cxd4 4.exd4 probably leads to a Panov. Black can be tricky and delay ...d7-d5 with 4...Bg7 5.c4 Nf6. There is no
logical move other than 6.d5, but Black is okay playing this position with ...0-0, ...e7-e6xd5, ...Re8 and ...d7-d6.
4.c4
Black gets a last chance to take on d4. It’s possible to avoid the Panov with 4.d5, as we will see in the next game.
4...Nf6
5.d5(!)
The pawn structure has settled and we have a Benoni with the not-so-dangerous-looking e2-e3. It may not be the best
choice against a Benoni player, but many opponents will not have any experience with this opening. Istvan Almasi had
never played it before.
5.Nc3 allows 5...cxd4 6.exd4 d5. If Black fianchettoes his bishop in the Panov, White wants to capture and keep the d5-
pawn, but that’s no longer possible.
10.Nd2 Na6
A logical move. The knight doesn’t need to be on d7 since White is far away from threatening e4-e5.
11.e4
Black is a tempo up, but it’s not so important considering that White is playing a prophylactic game: overprotecting
e4 and defending against ...b7-b5. Only in the long run will White try to push e4-e5 or b2-b4.
13.f3
Freeing the knight on d2 from its duty of defending the e4-pawn. An easier way would have been to keep the pawn on
e3!
13...Rb8?!
Once again a move that prepares ...a7-a6 followed by ...b6-b5. One easy remedy to that plan is to place the knight on
c4, answering ...b6-b5 with Nc4-a5-c6. Another is to get in our pawn break first – with b2-b4.
15...Bxc4?
Exchanging the light-squared bishop for a knight is a common idea in the Benoni, but then Black plays for the dark
squares with the knight on d7. In combination with the knight on c7, it’s simply illogical.
16.Bxc4 a6 17.Qd3
17...Nh5
17...Qc8 defends the pawn, and it looks like Black is preparing ...Qb7 followed by ...b6-b5. But it’s an illusion: as
soon as the queen steps onto the b-file, White opens up for his rook with b2-b4. In addition, Black would have
problems defending the d6-pawn after 18.Bf4.
18.g4
A standard move in the Benoni. It might be that Almasi was not too familiar with that, having been tricked into an
opening he doesn’t play.
22.exf5! Qh4?
22...gxf5 was better. After 23.Bf4 fxg4 24.fxg4 White is still a pawn up, but the weak g4-pawn allows Black to
dream.
23.fxg6
The knight on g7 is almost dead, and the only way Black can revive it is as in the game: by exchanging his dark-
squared bishop for the knight.
26...Qxg4
White is still a pawn up, but he has to play actively to avoid getting a bad bishop against a strong knight.
27.Rg1 Qh5
27...Qf5 28.Qxf5 Nxf5 29.Rxg6† Kf7 is Black’s best chance, with compensation for at least one of the two pawns.
28.Rb2
The rook enters the game and it’s soon over.
28...Nf5 29.Rbg2 Kh7 30.c4 Rf8 31.Rg5 Qh6 32.Qe2 Rbe8 33.Qg4 Nh4 34.f5 Nxf5 35.Rxg6
1–0
1.e3 might lull your opponent into a false sense of security, but against many players there’s no theoretical value in the
move compared to 1.d4 and 2.Nf3, with e2-e3 to come on move 3 or 4.
A New Trend
The openings from Chapters 4 to 8 can arise from several move orders, but the Slav and the Queen’s Gambit give
different variations depending on White’s first moves. I will recommend 1.Nf3 d5 2.e3 as our main move order, but
also give a repertoire for 1.d4, 1.c4 or 1.e3.
Capablanca and Botvinnik are two world champions who have played 1.Nf3 followed by 2.e3. But the move order
was really brought into the limelight when Sergey Karjakin used it to beat Viswanathan Anand in the 2016 Candidates.
The game is quoted in the chapter about the Anti-Queen’s Gambit (page 251).
Thinking of Karjakin and Grischuk reminds me of a great conversation they had a few years ago. Karjakin said that a
man who doesn’t have Facebook and Twitter can’t be a professional sportsman today and Grischuk replied: “It was of
course hard for me to get to sleep that night. In general it’s unpleasant to find out after 30 years that you’re not
actually a professional... So I played cautiously yesterday, though in the end everything worked out okay.”
He won the rapid tournament. Anyway, it feels like we have started to leave our topic, so let’s move on to a recent
game with 1.Nf3 and 2.e3. It shows that a cautious start doesn’t necessarily imply cautious intentions.
Kramnik, September 2016: “I think it makes a lot of sense nowadays to have some decent sidelines in your repertoire
with White and those e3 lines are rather interesting and have some venom for sure.”
2...Nf6 3.d4 g6
3...cxd4 4.exd4 avoids the bind White has in the game, but opens up the bishop on c1 and in a way justifies e2-e3.
I reached the game position against Isan Reynaldo Ortiz Suarez at the 2016 Olympiad in Baku, but with a different
move order.
Instead after 1.Nf3 c5 2.e3 g6 3.d4 Bg7 4.d5 the possibility of 4...e5 5.e4 d6 6.Nc3 intrigued me.
Can Black play for ...f7-f5 before he places a knight on f6? In a normal Schmid Benoni, it’s a big no-no, but maybe
the lost tempo with e3-e4 makes a difference? It’s always difficult to evaluate such positions, but here is a try:
a) 6...f5? 7.exf5 gxf5 8.Ng5 and Black faces disaster on the light squares.
c) 6...Ne7 7.Be2 0-0 8.0-0 h6 9.Nd2 is similar, manoeuvring the knight only when the g5-square is defended. After
9...f5 10.Nc4 then f2-f4 eyes the weak pawns on d6 and h6. The reason to post the bishop on e2 is to meet ...f5-f4 with
Be2-g4, which is possible as long as White keeps the f-pawn on f2.
Ortiz Suarez played 7...e5 8.e4 Ne8. When the c4-square is not available for the knight, I thought I had to find another
way to meet ...f7-f5. But with hindsight, I prefer to copy the set-up with 9.a4 h6 10.Nd2 f5 11.Be2. It’s not really
giving away a tempo, since Black’s knight has also moved twice. However, the knight is better on e8 since it defends
the d6-pawn against an attack with Nc4 plus Nb5.
So the knight should stay on e8 and 11...Nd7 is a good move, to answer 12.Nc4 with 12...Nb6. White should perhaps
try 12.Rb1!?.
The rook makes no sense as long as Black has a knight on d7, but after 12...Ndf6 (keeping the knight on e8 to defend
the d6-pawn13.f3 (liberating the d2-knight) 13...f4, White has to open the queenside with b2-b4 as quickly as possible,
before being overrun by ...h6-h5 and ...g6-g5-g4.
I like 12.Rb1, but maybe it’s too deep.
7...Bg4 8.h3 Bxf3 9.Qxf3 Nbd7 10.Be2 and I evaluated the position as slightly better for White.
However, Tiger Hillarp writes: “Maybe Peter Heine Nielsen and Vladimir Kramnik would agree, but to me it sounds
like a far too optimistic evaluation. After 10...Ne8 11.e4 Nc7, Black is two tempos up against a position that’s just
marginally better for White. Against a computer, it’s maybe Black who should equalize (with emphasis on maybe), but
between two humans it’s rather Black who holds the initiative. If the a-pawns are exchanged, the influence of the
bishop pair is reduced by almost fifty percent. The critical continuation may be 12.a4 b6 13.Bd2 with the idea to meet
...b6-b5 with b2-b4.”
Hillarp considers 7...Na6 strongest, preparing ...b7-b5 while keeping the option to play ...Bg4.
10...Nc6
10...Nxe4? runs into 11.Ng5! Nxg5 12.Bxg5 Rxe1† 13.Qxe1 Qd7 14.Nf6† Bxf6 15.Bxf6 with a winning attack.
13...Qxg5 14.Qe8† Bf8 15.Qxc8 and it’s not possible to save the rook on a8.
18...Ne5
18...b5 doesn’t work due to 19.Nxd6! Qxd6 20.Bxf7† Kg7 21.Bxe8 Rxe8 22.axb5 axb5 23.Qxb5 with too many
pawns.
25...Rxb2 26.Rxb2 Rxb2 27.Qa4 Qd8 28.Ra7 Be7 29.Rd7 Qe8 30.Qa7 Kf8 31.Kg2
A nice move. Black can’t do anything.
31...h6 32.Bc6
32...Bg5?
A mistake in a lost position. After 32...h5 33.h4
White continues with Qc7 and e4-e5, when the queen
is on the way to h8.
Juggling with the move order will one time or another lead you to end up in a position you haven’t planned. Maybe you
are paired against a hard-core Slav player and want to reach the Panov by 1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4, but you are
surprised by 1...c5. You try 2.Nf3 Nc6 3.e3 (when 3...Nf6 4.d4 cxd4 5.exd4 d5 is the Panov), but he plays 3...e5. Is this
the end of the world?
If yes is your answer, I am sorry to say that there are other books that will suit you better – or that you need to stop
aiming for control. With this repertoire, you have to believe in your ability to find good (or rather honte, to use go-
terminology) moves. It’s constructed to give positions where you can outplay your opponents.
When you prepare for an opponent, start by checking what he plays against 1.d4 – even if you prefer to begin with
another first move. That tells us what he is trying to reach.
The move order doesn’t matter if he plays an Indian defence: the King’s Indian, the Queen’s Indian or the Grünfeld.
White is able to play d2-d4, c2-c4, Nf3 and e2-e3 before there’s a confrontation.
But it’s different if the opponent:
a) ...plays ...c7-c5 before White has played d2-d4
b) ...has a classical repertoire (1...d5)
In our repertoire, Black can play 1...d5 and develop his light-squared bishop to f5 or g4 before setting up a pawn chain
with ...e7-e6. It would not have been possible without e2-e3:
1.d4 d5 2.c4 c6 3.Nc3 Nf6 4.Nf3 Bf5? 5.cxd5 cxd5 6.Qb3 b6? 7.e4! dxe4 8.Ne5 e6 9.Bb5† Nbd7 10.Bg5, winning a
piece. The quick attack against b7 is the trademark when Black’s bishop leaves c8.
However, a single slow move justifies ...Bf5, and it would therefore have been logical to postpone e2-e3 until Black
has played one of the following moves:
However, that’s not possible. After 1.d4 d5 2.Nf3 Nf6 3.c4 c6 we only have 4.e3 left and Black can enter the Slow Slav
with 4...Bf5 or 4...Bg4.
The main question when choosing White’s move order is how to tackle the Slav. Which of our repertoire’s four moves
does least in attacking b7? Let’s review the line above when White won a piece.
We will now discuss the different first moves, with our main move order last: 1.Nf3 followed by 2.e3(!).
1.d4
1.d4 d5 2.Nf3
On principle, White should not be afraid of 2.c4 e5. But starting with the knight makes some difference in the
Queen’s Gambit Accepted. Black can’t play 2.c4 dxc4 3.Nf3 a6 or 3.e3 e5.
2...Nf6
Black has four ways to use the fact that White delayed c2-c4.
3.c4
3.e3 with the intention of 4.c4 is possible. 3...Bg4 4.c4 e6 5.Nc3 c6 gives a Slow Slav. 3...Bf5 4.c4 e6 seems like a
way to punish the move order, but after 5.Nc3 Black’s best move is 5...c6, again transposing to a Slow Slav. The
difference compared to the Baltic Defence, 1.d4 d5 2.Nf3 Bf5 3.c4 e6 4.Nc3 (we go 4.Qb3) 4...c6 5.e3 is that Black
plays 5...Nd7 and without a knight on f6, Black can’t win the bishop pair with 6.Nh4.
With that said, 3.e3 doesn’t actually avoid anything! Black can still go for the Gambit Accepted with 3...e6 4.c4 dxc4 or
3...Bg4 4.c4 dxc4.
a) 3...c6 4.e3 Bf5 or 4...Bg4 is the Slow Slav, one of the few main lines we go for in this repertoire.
b) 3...c6 3.e3 e6 5.Nc3 is the Semi-Slav.
c) 3...e6 4.e3 avoids the normal Queen’s Gambit. It’s well-known that an early e2-e3 gives Black the possibility of
entering a favourable Tarrasch Defence, however that’s one of the cores in our repertoire!
d) 3...dxc4 4.e3 with a Queen’s Gambit Accepted.
1.d4
Avoids: 1...c5 (we would go 2.d5 Nf6 3.Nc3) and 1...e5
Allows: The Slav, Queen’s Gambit Accepted and the Queen’s Gambit in their normal versions
a) A good reason to play 1.c4 is to avoid the Slav and steer the game into a Panov after 1...c6 2.e4! d5 3.exd5 cxd5
4.d4.
b) 1.c4 c5 2.e3(!)
Our hero Alexander Grischuk is one of the few who have used this move order. After d2-d4, it will, as far as I can see,
transpose to the Panov, the Benoni or the Tarrasch. The exception is 2...e5 as Edward Loewe played three times in
London 1851. This is covered in the chapter about 1.c4 e5 2.e3 (page 321).
2.Nf3 hopes for 2...Nc6 which is indeed the most common move. After 3.e3 and 4.d4, Black will be forced to react
since he can’t allow White to kick the knight with d4-d5.
c) 1.c4 e6 2.Nf3
2.e4 d5 3.exd5 exd5 4.d4 gives the Exchange French, but we want to wait with c2-c4 until Black has developed his
dark-squared bishop. If we like the positions with an isolated pawn, we can play 4.cxd5 Nf6 5.Bb5† Nbd7 6.Nc3 and
Black will later take back the pawn with ...Nb6xd5, kick the bishop with ...c7-c6 and reach a position from Chapter 23
with an extra tempo. So we really have to like the structure to play like this.
d) 1.c4 e5
I have played the English in a hundred games, mainly to avoid the Nimzo-Indian, the Queen’s Indian, the Slav and the
Grünfeld. But when my opponents had 1...e5 in their repertoire, I usually chose something else. I simply didn’t like the
idea of playing something as timid as the reversed Sicilian.
2.e3(!)
If I had won against Bengt Lindberg in the final round of the 2012 Swedish Championship, I would have taken the
title. The game started with 1.c4 Nf6 and was eventually drawn, and in a chess magazine I wrote: “1...e5 is one
disadvantage with 1.c4 and since my opponent never played it before it was the move I expected.”
This was not at all true: I just wanted to discourage future opponents from playing that move. I hope you believe that I
am more honest in this book.
Of course, Lindberg surprised me with 1...e5 the next year. After half a minute of regret, I decided to avoid the g3-
lines I usually play and went for e2-e3. It was a first step away from the heavily-prepared concrete player I was; a
journey that finally led to this book.
When Alexander Grischuk played White against Viswanathan Anand in the 2015 London Chess Classic, he started with
1.c4 e5 2.d3. Grischuk has used many of the lines in this book, and his idea with 2.d3 was to get a game without too
much theory. It turned out well. Anand thought for three minutes and was outplayed. However, I feel that 2.d3 requires
more knowledge about the Sicilian and we will thus stay with 2.e3.
1.c4
Avoids: 1...d5 with the Slav or the Queen’s Gambit Accepted
Allows: 1...e5, the Symmetrical English and the Panov (1...c5 2.e3 Nf6 3.d4 cxd4 4.exd4 d5)
1.e3
This is sometimes called Van’t Kruijs opening, after the Dutchman Maarten van’t Kruijs (1813-85). However, his
concept was completely different, like 1...e5 2.Bc4 d5 3.Bb3.
a) 1.e3 d5 2.c4(!) and Kasparov – Fritz 3 showed that the Queen’s Gambit Accepted isn’t working well. Instead, Black
chooses between the Slav and the Queen’s Gambit, but White is able to play anti-lines in both cases.
b) 1.e3 c5 2.d4(!) will probably give the Panov after 2...cxd4 3.exd4 d5 4.c4. If Black avoids ...c5xd4, we reach a
Benoni: 2...Nf6 3.Nf3 g6 4.d5
c) After 1.e3 Nf6, White can make any of the three moves in our set-up, but it’s also possible to steer the game into a
Nimzowitsch-Larsen with 2.b3!?, avoiding the lines where Black occupies the centre; more about that in the next
paragraph.
2.c4 with the English or 2.d4 exd4 3.exd4 and Black has to play ...d7-d5, transposing to an Exchange French.
When I finished a second draft in late summer 2016, I considered sending it to a strong player to test it on a high level.
Magnus Carlsen came to my mind, but I didn’t dare. Guess my surprise when he played 1.e3 in the Olympiad in Baku
against Enamul Hossain (Bangladesh). After 1.e3 Nf6 2.Nf3 e6 3.c4 b6 4.Nc3 Bb7 5.d4 Bb4, the game transposed to a
Nimzo-Indian and he won in crushing style. However, I prefer 3.d4 b6 4.Bd3, to prevent the bishop from reaching b4.
After that game, I did indeed send a few variations to one of his seconds, and Carlsen used an e3 system in one of his
match games against Karjakin: 1.d4 Nf6 2.Nf3 d5 3.e3 e6 But the system can’t be held responsible for his only loss
since he played 4.Bd3 c5 5.b3 instead of 4.c4.
And I am pretty sure he got his inspiration elsewhere.
Grischuk, September 2016: “I have never really considered 1.e3, but Magnus played it very recently :). In fact it is a
completely OK move, but it doesn’t really have independent value. It will transpose to something else in a matter of a
couple of moves.”
Magnus’s idea behind 1.e3 in the Olympiad was to reach an Exchange French after 1.e3 e5 2.d4 exd4 3.exd4 d5 4.Nf3.
A few rounds later, he managed to get that position by starting with 1.e4.
1.e3
Avoids: Queen’s Gambit Accepted
Allows: 1...e5 with the Exchange French (2.d4) or the English (2.c4)
1.Nf3
I think 1.Nf3 followed by 2.e3 does a good job of moving our opponents out of their comfort zones. Maybe it will do
the same to the readers as well, but hopefully only initially, before reading all of this book!
We now have A) 1.Nf3 c5 2.e3, B) 1.Nf3 Nf6 2.e3, C) 1.Nf3 d5 2.e3 and D) 1.Nf3 d5 2.e3 c5 3.d4.
A) 1.Nf3 c5 2.e3
1.Nf3 c5 2.c4 is also possible, but compared to 1.c4 c5 2.e3, ...e7-e5 is better since ...e5-e4 hits the knight. For example
2...Nc6 3.e3 e5 4.Nc3 f5!? 5.d4 e4. White gets a better structure after: 6.d5 exf3 7.dxc6 dxc6 (7...fxg2 8.cxd7† Qxd7
and White should not help Black to develop, but instead play 9.Bxg2 with an advantage due to the d5-square) 8.Qxd8†
Kxd8 9.gxf3
Both players have doubled pawns, but Black’s majority is more restricted. Normal development is b2-b3, Bb2, 0-0-0,
Bd3 and f3-f4, as Anish Giri played against Wesley So in the 2015 Sinquefield Cup. The game ended in a draw, but not
without a fight.
After 1.Nf3 c5 2.e3(!), White plays 3.d4 next. Black has six main options.
1) Panov: 2...Nf6 3.d4 cxd4 4.exd4 d5 5.c4 Black could also play 2...Nc6 or 2...g6.
4) 2...g6 3.d4 Bg7 4.d5 Nf6 when White chooses between the Benoni (5.c4) or playing like Kramnik did against
Sjugirov (5.Nc3).
5a) 5.c4 Bg7 6.Nc3 0-0 and if White plays something like 7.Be2, Black has 7...d5! with a preferable Panov. And 7.d5
isn’t dangerous when Black has exchanged on d4 and has the c5-square for a knight.
5b) 5.Bd3! Bg7 6.0-0 0-0 7.Re1 Nc6 8.c3 d5 gives an Exchange Caro-Kann, where Black’s main plan is the minority
attack ...b5-b4. But the bishop is misplaced on g7; White uses the e5-square.
8...d6 to play for ...e7-e5 is an alternative, but here we see the downside by exchanging on d4. White is able to control
e5 not only with the rook but also after 9.Bf4.
6) 2...Nf6 3.d4 cxd4 4.exd4 e6 5.c4 b6 with a Queen’s Indian. We will discuss the move orders in connection to Eljanov
– Grischuk later in this chapter.
Make sure you understand all of them by heart! (I will not use exclamations mark often, but one is definitely deserved
here.)
2.e3(!) is a move that Aaron Nimzowitsch played in several games, but sometimes with a more passive set-up than
proposed here. But as already mentioned, the player who really deserves to be mentioned is the Georgian grandmaster
Tamaz Gelashvili with his 39 games in this position.
The idea behind the flexible 2.e3 is to choose between 3.c4 and 3.d4, based on which move challenges a black pawn.
2...d5 3.c4 or 2...c5 3.d4.
2.c4 is of course also possible, but it deprives White of the chance to follow Kramnik – Sjugirov.
If Black wants to force White to play the Poor Man’s Benoni, he should go 1.Nf3 Nf6 2.e3 g6 3.d4 Bg7 4.c4. The only
move that fits White’s set-up against the King’s Indian and the Grünfeld. 4...0-0 5.Be2 c5 6.d5 Otherwise Black
captures on d4 and plays ...d7-d5, when Be2 is not the forceful move that the position requires.
Players who have the Nimzo-Indian and the Bogo-Indian or the Queen’s Indian in their repertoire will probably play
1.Nf3 Nf6 2.e3 e6. Neither of the first two are possible after 3.c4 so they have to go 3...b6 4.b3 with a Queen’s Indian.
We don’t allow 4.d4 Bb4† before Black has played ...d7-d5.
a) The diagonal is not secure yet: White could still play d2-d3 and e3-e4. But this is more a theoretical than a practical
argument.
b) White is faster in creating an initiative on the queenside with Nc3, Qb3 and Bb5. It makes a huge difference.
The lines will be covered in the chapter about the Slav Nirvana (page 257). 2...Bf5 3.c4 c6 or 2...Bg4 3.c4 c6 also give
lines from that chapter.
I haven’t found any game between strong players with 2...Bg4 3.c4 e6 or 2...Bf5 3.c4 e6. 4.Qb3 is strong in both
cases.
1.Nf3 d5 2.e3(!) 2...Nf6 3.c4 dxc4 4.Bxc4 is a good antidote to the Queen’s Gambit Accepted if White delays d2-d4.
A short section, highlighting the independent options Black gets against 1.Nf3 followed by 2.e3.
4.exd4
What we have now is a Panov where we would have preferred 4.c4 instead of starting with the knight. Black can
make use of that.
4...Bg4 5.c4
5...Nf6!
5...e6 allows 6.cxd5(!) before Black can take back on d5 with the knight.
a) 6...Qxd5 7.Be2 Nf6 is a c3 Sicilian where Black has exchanged on d4 too early. White develops with 0-0, Nc3, a2-
a3, Be3, h2-h3 and Qb3. Since Black’s light-squared bishop is on the kingside, White’s should play on the other wing
and therefore the rooks belong on c1 and d1.
Black should start with this knight, so he has ...Bd6 plus ...Nge7 in case White plays Bb5. 8.Be2 Nf6 9.0-0 Be7 10.h3
Bh5 11.Qb3!
Just in time to create a threat before Black has castled – we use the tempo we started with as White. It’s not easy to
defend b7 in a convenient way. We have:
a) 11...Qd7? 12.Ne5 is horrible for Black.
b) 11...Bxf3 12.Bxf3 Nxd4 13.Qa4† Nc6 14.Rd1 with the bishop pair in a wide-open position.
e) 11...a6 12.Bf4, renewing the threat against b7. But not 12.Qxb7?? Na5.
6.Nc3 e6
I haven’t found a GM game with this move, but it is difficult to crack. Black delays ...Nc6 and tries to castle as soon
as possible, but as we will see, he is not in time anyway.
9.Ne5
9...N5b6
The only move.
1.Nf3
Avoids: 1...e5, Queen’s Gambit Accepted, the Slav
Allows: The Symmetrical English and the Panov
Summary
From now on, every chapter starts with a discussion of different move orders, but even though I will do my best to show
possible transpositions, it’s impossible to cover everything.
The next part covers openings that can be reached via several move orders. That includes Line A above with:
Openings after 1.Nf3 d5 2.e3 are covered in Part 5 and openings after 1.d4 d5 2.Nf3 in Part 6.
Deciding on the first move for a single game is much about avoiding the opponent’s favourite opening.
a) Search your opponent’s games to check what he wants to play
b) Pick the move that fits best against that opening
In Practice
The following game from the 2015 FIDE World Cup shows how difficult it is to adjust the move order to take into
account all the opponent’s intentions. Grischuk had to win with Black to equalize the match, so he tried hard to get
unorthodox play. But Eljanov was clever enough to use the e3 poison.
Eljanov, September 2016: “The system is unpleasant for Black in a must-win situation.”
Pavel Eljanov – Alexander Grischuk
Baku (3.2) 2015
1.d4 e6
Black opens up for ...Bb4, as in the Nimzo and Bogo-Indian. Some players transpose with 2.c4 Nf6 and have thus
avoided the Trompowsky, but normally Black has other aims.
However, Black must be ready to play the French: 2.e4 d5
2.Nf3
Avoiding the check on b4, but the downside is that White can’t play d4-d5 after Black’s next move.
One idea behind delaying 1...Nf6 is 1.d4 e6 2.c4 Bb4† 3.Nc3 b6 when ...Bb7, ...f7-f5 and ...Nf6 all fight for the e4-
square. Another is that White has less time to fortify his centre after 3...c5.
2...c5 3.e3(!)
3.e4 is a Sicilian, so Black needs to be able to play that opening as well.
3...Nf6
This position could also be reached via 1.Nf3 c5 2.e3 Nf6 3.d4 e6. Black still has a wide choice: the Panov (...c5xd4
plus ...d7-d5), the Tarrasch (...d7-d5), the Queen’s Gambit Accepted (...d7-d5xc4) and the Queen’s Indian (...b7-b6).
Grischuk chooses the latter, and continues to fiddle with the move order.
4.c4 a6
If White goes for the Benoni with 5.d5, a pawn on a6 makes more sense than one on b6. Another point is that White
must be prepared for a Tarrasch, since all Black’s moves so far fit into that scheme.
Eljanov gets ready by choosing a move in accordance with our repertoire.
5.a3
A month later, Eljanov played 5.Nc3 against Grischuk in the World Blitz Championship.
5...b6
5...d5 6.dxc5 is how we meet the Tarrasch.
6.Nc3 cxd4
Avoiding 7.d5, which would have been Eljanov’s move in case of 6...Bb7?! – not a good diagonal for the bishop in
the Benoni.
7.exd4 d6
Black has a Hedgehog set-up, but in contrast to the normal lines, White has a pawn on d4 instead of e4 – that’s an
advantage. The pawn structure is less susceptible to the typical counter-breaks ...b6-b5 and ...d6-d5, and can advance
more easily. The game is a good example.
7...d5 gives a Queen’s Indian with the a-pawns moved one step each. I feel that it’s not a big problem for Black, since I
can’t see how White could make use of the b6-square.
14...Qc7 15.Rc1
This time it’s easier: rook against queen. Grischuk hurries to move away from the c-file.
Black has a nice-looking Hedgehog, but as White’s plan of advancing on the queenside has more power with the
pawn on d4, Black can’t use the standard plan: to wait while pretending to do something. Grischuk tries to create play
along the long diagonal. It’s logical in one way, since there’s no pawn on e4, but with Black’s set-up, it’s impossible to
transfer additional pieces to the kingside.
However, I have no better suggestion.
19...Qb7
The threat is 20...Rxc4 21.Bxc4 Qxg2 mate.
24...e5?
It would have been better to wait. Eljanov could of course have done the same, since a draw was enough to win the
match, but positionally speaking he should place both rooks on the queenside and open up with a4-a5 or c4-c5.
25.d5
Black has the f4-square and a pawn majority on the kingside, but it’s of little value with several pieces stuck on the
queenside.
32.Ng3 Qd8 33.h5 Bf8 34.Bd3 Nh7 35.Bf5 Ng5 36.Kg2 Rc7 37.Ne4 Be7 38.c5
Now when it’s possible with an easy tactic, Eljanov directs some attention to the queenside.
40.c6
If the a8-bishop dreamt about manoeuvring via b7 to c8 and then attacking the king from g4 or h3, it’s time to wake
up.
40...Rc8 41.Qd3 Nxe4 42.Qxe4 Bg5 43.Be3!
Every piece exchange increases the significance of the dead bishop on a8.
45...Rb8 46.Nc4!
The knight is on the way to the kingside.
51.g5
Black is up in material, but with no possibility of defending his king.
The tournament was over for our hero Grischuk, and with that the chance to reach a match for the World Championship.
But he is still a soldier with ambitions.
1) Classical
1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 e6 6.Nf3 158
1a) 6...Nc6?! 159
1b) 6...Be7 160
1c) 6...Bb4 161
Move Orders
In the Panov, White’s e-pawn is exchanged for Black’s c-pawn. This can be done in two different ways.
White exchanges on d5
Slav: 1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4
Alapin: 1.e4 c5 2.c3 g6 3.d4 cxd4 4.cxd4 d5 5.exd5 Nf6 (there has been a second pawn exchange here)
Black exchanges on d4
Symmetrical English: 1.c4 c5 2.e3(!) 2...Nf6 3.d4 cxd4 4.exd4 d5
Anti-Benoni: 1.d4 Nf6 2.Nf3 c5 3.e3 cxd4 4.exd4 d5 5.c4
The Tarrasch: 1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 c5 5.a3 cxd4 6.exd4
Early e2-e3: 1.e3 c5 2.d4 cxd4 3.exd4 d5 4.c4 or:
1.Nf3 d5 2.e3(!) 2...c5 3.d4 cxd4 4.exd4 Nf6 5.c4
Besides the exchange, the Panov is characterized by tension in the centre as in the diagram below.
As we will see, there are circumstances where c4-c5 is strong, but normally White exchanges with c4xd5. We will
then reach a reversed Tarrasch.
The theoretically threatening set-up against the Tarrasch is to use all our pieces to put pressure on the isolated pawn:
Nc3, Bg2 and Bg5. Consequently, Black wants to develop with ...Nc6, ...Bg7 and ...Bg4 in the diagram above, but the
extra tempo makes a difference. After 1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 g6 6.Qb3! dxc4 7.Bxc4, Black is
not in time to castle.
But can’t Black delay ...d7-d5 to avoid White’s quick attack in the line above? No.
However, after our main move order, it’s White who is slightly tricked.
So against the fianchetto, White prefers to start with the queen’s knight. But unfortunately, the knight may be on f3
already before the Panov arises.
Both cases are covered in the theoretical section.
What White strives for, generally, is a position with an Isolated Queen Pawn (IQP). There are some obstacles along the
way and one is the dull rook and bishop ending that many see as a downside with the Panov: 1.Nf3 d5 2.e3 c5 3.d4
cxd4 4.exd4 Nf6 5.c4 Nc6 6.Nc3 Bg4 7.cxd5 Nxd5 8.Qb3 Bxf3 9.gxf3 e6 10.Qxb7 Nxd4 11.Bb5† Nxb5 12.Qc6†
Ke7 13.Qxb5 Qd7 14.Nxd5† Qxd5 15.Qxd5 exd5
Some players try to claim that White has something with the queenside majority, but why learn long lines to get an
advantage that’s so small that is has given only 5 wins in 79 GM games? (And 9 losses!) The good news is that we
avoid the ending altogether with 7.Bg5(!).
Pawn Structures
Several different structures are possible, but the illustrative part focuses only on the typical ones – not structures that
appear “by chance” after a concrete line. However, to examine Black’s different possibilities, there will be more games
than in other chapters.
Structure 1
This position is similar to Structure 1 in the Queen’s Indian, but with the difference that there’s no bishop on b7. And
it’s too late: ...b7-b6 at any point will be answered with 1.Nc3xd5 Qxd5 2.Be4 (or Re1 first, if necessary).
The good news for Black is that he can place his knight on c6, where it puts pressure on d4 and stops Nf3-e5. The bad
news is that White gets another square, on c5, and an exchange there with d4xc5 gives a strong pawn majority.
Standard moves are Nc3, Nf3, 0-0 and Re1 – all these pieces know where they want to be. The light-squared bishop is
normally happier on d3 than on c4, but if Black leaves e6 less defended (...Bb7/c6 and ...Nxc3), it could be a killer
from c4 after a sacrifice on f7. The dark-squared bishop is a joker, biding his time and sometimes landing on g5,
sometimes on h6 (after ...g6) or even e3 or f4. The queen could be directed to the queenside, but the main idea is Bc2
plus Qd3 (with a2-a3 to stop ...Nb4).
White often plays at a slow manoeuvring pace. He places his pieces on good squares and looks for opportunities. If
they don’t appear, he continues to look – while avoiding doing anything he will regret.
Structure 2
One of the players usually has the advantage in this structure. It’s not often balanced. The reason is that the c4-c5-
advance gives a clear plan, but at the same time it’s very committal.
When White plays c4-c5, he releases the pressure against the d5-pawn. That allows Black to play for ...e6-e5, which
undermines the c-pawn and takes control over the centre. White’s first aim is thus to keep the e5-square under control
with moves such as Bb5(xc6), Re1 and Bf4.
If he succeeds in consolidating, the queenside majority can start to warm up, while Black will have difficulties in
finding counterplay.
a) There is a black knight on c6 (otherwise he can play ...b7-b6 without worrying about Bb5 or b4-b5).
b) Black has played ...e7-e6 and can’t play ...Bg4xf3, a manoeuvre that would have diminished White’s control over e5.
Structure 3
Black has given up the d5-pawn to:
a) Get the optimal set-up against the isolated pawn, with a fianchettoed bishop and the possibility of playing ...Bg4 later.
b) Avoid the initiative White would get if Black had taken back on d5 with the knight: Bc4, Qb3 and Nc3 attacks d5,
but also the f7-pawn before Black had time to castle.
Black’s dream is to recapture the pawn with ...Nbd7-b6xd5, (...Qd6 and ...Rfd8). However, White will not be
cooperative and moves that defend the pawn are Qb3, Be2-f3 (if the knight is not already on f3), Nge2-f4, Bg5xf6 and
a2-a4-a5 to kick away the knight from b6. It may seem like a knight on f4 isn’t stable, since Black can play ...g6-g5, but
such a move is weakening (initially the pawn is even undefended) and gives White something else.
Instead Black’s task is to find a useful job for the bishop on c8.
a) ...Bg4 to exchange White’s bishop on f3, even though it loses a tempo to play ...Nf6xg4-f6.
b) ...Bf5-d3-c4, attacking the pawn from the side.
As soon as the bishop moves, White has to find a concrete reply – usually something related to the hanging pawn on b7.
Two ideas are d5-d6 (opening up for Bf3xb7) and a4-a5.
Games
Even though the isolated pawn is a central feature in the Panov, quite often events along the way give rise to other
structures. We will see three such examples, but start with three games with the isolated pawn.
7.a3 transposes to a position we reach in the next chapter, about the Tarrasch.
10...Ncb4?!
Black’s third most common move, but I doubt that it’s sound to leave c6, where the knight stopped White from
playing Nf3-e5. And since White hasn’t played a2-a4, the knight isn’t even stable on b4.
13...Rc8
Best was 13...Nc6, admitting that 10...Ncb4 was a mistake.
14.Re3
White threatens 15.Bxf6 Bxf6 16.Bxh7† Kxh7 17.Qh5† Kg8 18.Rh3 with mate.
14...g6
14...h6 15.Rh3! does not give Black much hope. 16.Bxh6 is a threat and 15...hxg5 16.Bh7†! Kh8 (16...Nxh7 17.Qh5
with unavoidable mate) 17.Ne4! gives the queen access to h5 and decides the game.
15.Rg3
There is no good way to avoid 16.Bh6 followed by a sacrifice on g6. Best is to give up the exchange and defend with
...Be8, but instead Black chooses to go down in style.
17.a3!
Instead 17.Nxg6 hxg6 18.Bxg6 fxg6 19.Qb1 would allow Black’s knight on b4 to shine with a hara-kiri move:
19...Nc2!! 20.Qxc2 Be4 and Black defends and wins.
With 10...Ncb4?!, Black released the pressure against the d-pawn and White was allowed to build up a decisive attack.
In the next game, Kamsky deviates with 10...Nxc3?! but that’s also a move that leaves White with free hands on the
kingside. The fate is similar, even if it takes more moves.
1.e4 c6 2.c4 d5 3.cxd5 cxd5 4.exd5 Nf6 5.Nc3 Nxd5 6.Nf3 e6 7.d4 Nc6 8.Bd3 Be7 9.0-0 0-0 10.Re1
We have reached the same position as in the previous game.
10...Nxc3?!
As 10...b6? is not possible due to 11.Nxd5 Qxd5 12.Be4 Qd6 13.Ne5 Bb7 14.Bf4 Nxe5 15.Bxb7, winning an
exchange, Kamsky starts by exchanging on c3. But it’s too high a price for developing the bishop to b7. The exchange
on c3 should at least be delayed until White has spent a tempo on a2-a3.
The two most common moves are 10...Nf6 and 10...Bf6.
11.bxc3 b6 12.Qc2 g6
After 12...h6 White rearranges the battery on the diagonal with 13.Qe2 and Black will be forced to play the ugly ...f7-
f5.
14...Bf8
Kamsky tries to get his queen into the defence. It was not possible to take the pawn: 14...Bxh4? 15.Bb5 Bb7 16.Qe4
and the double threat picks up a piece.
15.Bg5 Be7?!
Maybe he hoped for a repetition since he was much higher rated. However, the bishop should have stayed on f8 to
keep the seventh rank open for the queen.
16.Bf4!
16...Bb7
16...Bxh4 is still not possible: 17.d5! and the black queen is overloaded to the defence of h4, e8 and d5. 17...Na5
18.Bb5 wins an exchange.
The game continuation also gives a winning attack, but without the end in sight.
18...Qd5 19.Be4?!
This is given as strong in one source, but today’s engines think that Black can escape without being punished as he
deserves. An alternative was to keep the attack going with 19.hxg6 hxg6 20.Bh6† Kg8 21.Qf4.
19...Qxh5 20.Ne5
20...Rac8?
20...g5! is the move that was overlooked. Black’s position is ugly after 21.Nxc6 gxf4 22.Nxe7 Bxe4 23.Rxe4 Rxe7
24.Qxf4 but material is still equal and his king is not mated.
21.Re3!
Black loses his queen and the rest is one-sided, even though there were better moves for both players.
21...f6 22.Rh3 Qxh3 23.gxh3 fxe5 24.dxe5 Rc7 25.Kh1 Na5 26.Bxb7 Nxb7 27.Bh6† Kg8 28.Rd1 Rec8 29.Qe2 Rxc3
30.Qg4 Kf7 31.Rd4 R3c6 32.Bg5 Bxg5 33.Qxg5 Nc5 34.h4 Re8 35.h5 Rc7 36.Qf6† Kg8 37.Rd8 Rc8 38.h6
1–0
In both games so far, White has used the e5-outpost for a knight to create a kingside attack. In the next game he plays on
the queenside with the other outpost that the isolated pawn controls, c5.
1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 e6 6.Nf3 Bb4 7.cxd5
As a curiosity, 7.Bd3 dxc4 8.Bxc4 transposes to a line in the Nimzo-Indian. But White does better to take the chance
to force Black’s knight to d5.
12...Re8 13.Rd1
The rook makes sense on f1 if I play Be3 and f2xe3. However, Black should probably avoid taking on e3, so it
doesn’t make any difference.
16.Rac1 Nf4?!
After 16...Nxe3 both captures are good. I planned 17.Qxe3 with Ne4-c5, Bd3-b5 and b2-b4.
19.dxc5
The pawn structure has transformed and White has a clear-cut plan with b2-b4-b5. Lindgren makes the best of the
situation by activating his majority.
19...e5 20.Qe4!
20...Qxe4?!
During the game I was sure that this was a mistake, since it opens the d-file for my rook and drags the bishop to a
better diagonal (remember that exchanges often help the opponent to develop).
I expected 20...Bg4 21.Qh7† Kf8 when White doesn’t have time to take the pawn on g7. However, 22.Rc4! Nd4
23.Nxd4 exd4 24.Qh8† Ke7 allows White to go for a forcing line: 25.Re1† Kd7 26.Rxe8 Rxe8 27.Rxd4†! Qxd4
28.Bb5† Kc7 29.Qxe8 Black can’t make use of the back rank, so White is happy with the extra pawn.
21.Bxe4 Re7?!
It was a better try to play 21...Bg4, take on f3, and get a strong knight on d4 – although even then, White’s bishop is
even stronger.
22.Rd6 f5?
Strategically a good move, but it works badly with all White’s pieces so well placed.
Black runs with his pawn majority. I tried to calculate how to push my queenside pawns all the way to the eighth
rank, but they were not fast enough when Black goes for counterplay on the second rank.
Finally I asked the positional question: which piece do I want to keep on the board? Black’s bad bishop!
35.Kf1?!
I was looking for more than 35.b5 but in vain.
45.Rxd5! cxd5
After 45...exf2† 46.Kxf2 cxd5 47.b5 Rc3 I had planned: 48.b6 Rxc5 49.b7 Rc2† 50.Ke3 Rb2 51.Kd4 Kg6 52.Kc5
Rxb7 53.Nxb7 Kf5
White wins by a tempo: 54.Kxd5 Kg4 55.Nd6 Kh3 56.Ke5 h5 57.Ne4 Kxh2 58.Kf5 h4 59.g4 h3 60.Nxg5 Kg2
61.Nxh3
With 18.Nc5 and 19.dxc5, White created a pawn majority on the queenside that eventually decided the game. An easier
way to get the majority is c4-c5, but that’s a move White should be careful with as long as Black has ...Bg4. As noted
in pawn structure 2, the advance c4-c5 is mainly played when:
a) Black has a knight on c6 – the undermining ...b7-b6 can be met by b4-b5 or Bb5
b) Black has played ...e7-e6 – ruling out ...Bg4
The following game shows that it’s also possible when Black has lost a tempo.
1.e4 c6 2.d4 d5 3.exd5 cxd5 4.c4 Nf6 5.Nc3 Nc6 6.Nf3 Be6
6...Bg4 is the main theoretical move, and then we go 7.Bg5.
7.c5!
Black’s bishop has to move a second time to reach g4, and the lost tempo gives White enough time to consolidate:
7...Bg4 8.Bb5! and ...e7-e5 is stopped.
To divert focus from the queenside, Miles starts to push pawns on the other wing.
23.Ree1?!
23.Re2 keeps the pawn.
27.Bd6 Rfe8
28.Bxe7! Rxe7 29.b5! axb5 30.axb5 Rxa1† 31.Nxa1 cxb5 32.Nxd5 Rf7 33.Nb6
Black’s problems with the e-pawn are not easy to solve. The best he can hope for is to give it away with ...e6-e5, but
the problem is that White doesn’t have to take.
33...e5
The last chance was 33...g3! 34.fxg3 e5 when 35.d5? Bg4 gives some counterplay. However, White wins with:
35.Nxd7 Rxd7 36.c6! Rd8 (36...Rxd4? allows 37.c7 Rc4 38.Rc2 queening) 37.d5!
34.d5
It has not been a faultless game, but even so it’s logical that the queenside pawns decide the result.
34...Bf5 35.Ra2 Bf8 36.Nb3 Rb7 37.Ra6 Kg7 38.d6 Rf7 39.Na5 b4 40.Nbc4 Bd3 41.Rb6 Ra7 42.c6 Ra8 43.c7 Bf5
44.Rb8 Ra6 45.Ne3 Be6 46.Re8
1–0
Next there follows a game where Black enters a position with the isolated pawn, but quickly uses concrete means to
change the pawn structure. It’s an important theoretical line.
Topalov, April 2016: “He introduced this idea to sacrifice the queen and had very good compensation, but maybe the
computer finds something now.”
10...Nxc3 11.bxc3
11...Nxd4! 12.Nxd4 Qxd4 13.Bb5† Bd7 14.0-0! Rc8!?
This was almost a novelty when it was played, but after Topalov’s powerful play it didn’t get many followers.
14...Qd5 is the main move and was played against Topalov a year earlier. We will return to it in the theoretical section.
Black had to give up his queen, but with a rook, bishop, pawn and above all a healthy pawn structure he would be
okay if it weren’t for the pin along the fifth rank. He will not be given time to coordinate his pieces.
17...a6 18.Bb2!
Threatening c3-c4 followed by Bxg7.
It seems that White should be able to move the bishop and keep the pin, but every square has a disadvantage.
21.Bc3 Rxa4! 22.Rxa4 Bxc3 and Black is a pawn up compared to the game.
21.axb5 Rxb2 22.bxa6 is also tempting, but the pawn won’t survive after 22...Bb6 23.axb7 Bc7! 24.Qf3 Rb8 and Black
will easily create a fortress.
21.Bc1 Re4! and White has to defend the first rank. 22.Be3? Bc3 creates another threat and after 23.Rc1 Bb2 24.Rb1
Rxa4 it’s time for White to ask why he gave away his queenside pawns.
In How I Beat Fischer’s Record, Polgar comments: “The hero of the initial struggle was my centralized knight. It was
worth giving up a pawn and even exchanging queens to make the most of its overwhelming activity.”
However, if White had played an early Nf3, Black would have had the time needed to castle without giving up a pawn.
7.cxd5 0-0
8.Be2!
Two ways to defend the pawn on d5 are Ne2-f4 and Be2-f3. Starting with the bishop allows White to return to normal
development (Nf3) if Black sacrifices with ...b7-b5 or ...e7-e6.
13.0-0 Qd6
Black delays ...Nf6 until White has played Bf4 and taken the f4-square from the knight. And Ne2-f4 is not possible
as long as the bishop on g7 has free sight towards d4.
14.Bf4 Qb4
15.Qd1
15.Ra3! looks clumsy, but is the recommendation in the theoretical section. However, Polgar’s choice is also good.
15...Rfd8
15...Nf6 16.d6! exd6 17.Nb5 threatens not only on d6, but also b2-b3, trapping the queen.
Polgar has said that the knight is her favourite piece and in this position it’s no wonder. Black has four (!) undefended
pieces.
19...Qa7?!
This ultimately loses an exchange.
19...Qd8 does not run into a fork, but the knight on d5 is strong nevertheless. 20.Nec3 Nh6!? is interesting to keep both
White’s knights fighting for the same square, while Black’s knight continues to f5 and cooperates with the bishop.
However, if White finds 21.Re1 Nf5 22.Re4 Black’s activity has petered out and, unlike White, he can’t develop the
queen and rooks to active squares.
So it’s easy to understand Skembris’s decision to give up the exchange in order to change the pace of the game, but it
was just not enough.
A fork on b6 or f6 is unavoidable. Initially Black has two pawns for the exchange, but the one on d6 will be lost.
23...Bxb2 24.Ra2! Bg7 25.Nb6 Rad8 26.Nxd7 Rxd7 27.Rd1 Bf8 28.Rad2 Nf5
The strategy to fight for a draw is exchanging the knight and keeping the rook, but a pawn storm on the kingside
should decide in the long run.
35.Rd8 Rxa4??
36.Rxf8†!
1–0
With her active style, the Panov was perfect for Judit Polgar. In this game her opponent cleverly chose a line that
doesn’t give White attacking possibilities, but even so he couldn’t keep White’s initiative at bay.
Recap
Six games and in none of them did Black conquer White’s isolated pawn and win. Did I hide the whole picture?
There is no guarantee that it won’t happen, but the IQP positions arise in their best versions and Black is usually far
away from creating pressure against the pawn. That’s why he often steers away from the kind of position by going for
concrete lines, like in Topalov – Magem Badals and in some sense even Polgar – Skembris.
In the next section, we will see another way with 6...Bg4 7.Bg5, a new move that’s not mentioned in Grandmaster
Repertoire 7 – The Caro-Kann (Lars Schandorff, 2010).
Theory
We have four different lines: 1) Classical, 2) The main line 6...Bg4 7.Bg5, 3) Fianchetto without Nf3, 4) Fianchetto
with Nf3.
1) Classical
6...dxc4 is logical, since it keeps the knight on f6. But it’s better to use the move order in the Queen’s Gambit Accepted,
where Black delays the exchange on d4.
Instead, he has three main alternatives on the sixth move: a) 6...Nc6?!, b) the calm 6...Be7 and c) the aggressive
6...Bb4.
A standard move in its best version. If White is allowed to play Bb5, 0-0, Re1 and Bf4, the queenside pawns will do
the job like a self-playing piano. So Black should try to disrupt matters before it’s too late.
A similar move was played in Anand – Miles after 1.Nf3 c5 2.c4 Nf6 3.e3 Nc6 4.d4 cxd4 5.exd4 d5 6.Nc3 Be6 7.c5
Anand – Miles. When the bishop makes a stop on e6, Black doesn’t get time for the undermining ...e7-e5.
7...Be7
7...Ne4 8.Qc2 Qa5 and White has the consolidating 9.a3!. If 9...Nxc3 10.Bd2!, he wins back the piece with an intact
pawn structure. And 9...Bd7 10.Rb1! is also safe, but not 9...Bd7 10.Bd3? Nxc3 11.Bd2 Nb4.
10...f5
Black prepares counterplay on the kingside.
11.Ne2!?
The knight protects d4 and White threatens 12.Bxc6 bxc6 13.Ne5 Bf6 14.Nxc6 Qc7 15.Nb4, which would have been
bad if the d-pawn was hanging.
Black is not without counterplay after 11...g5 but 12.Bxc6 bxc6 13.Ne5 Qc7 14.b4 and f2-f3 looks good for White.
The e5-square should be worth something. But not 14.f3? Nxc5.
7.a3 0-0 8.c5 transposes to the Tarrasch – see the next chapter.
7...Nxd5
7...exd5 is a solid but passive option – White develops his bishops more actively than Black.
8.Bd3
A classical IQP position. The difference to the Queen’s Gambit Accepted is that White doesn’t have to answer ...a7-
a6 with a2-a4, and that the knight is on d5 instead of f6. Both are to Black’s disadvantage.
a) 10...Nxc3?! was played in Winants – Kamsky. Black should not take on c3 before White has played a2-a3.
b) 10...Ncb4?! 11.Bb1 Nf6 is played from time to time, but Black’s strategy is too passive. 12.Ne5 was Shariyazdanov
– Petukhov.
Another way to play is 12.a3 Nbd5 13.Qd3, when a typical win arises after 13...b6?! 14.Nxd5! Qxd5 15.Bg5 g6
16.Ba2. With ...g7-g6 played, the bishop has often done its job on the b1-h7 diagonal (even though Shariyazdanov’s
piece sacrifice on g6 showed that’s not always the case). 16...Qd8 17.Ne5 Bb7 18.Bh6 Re8 19.Nxf7 Kxf7 20.Bxe6#
c) 10...Bf6 is the most common move. I like the unusual 11.a3!?, to prepare Bc2 plus Qd3. The reason 11.a3 is not
chosen more often may be the possibility of 11...Nxd4?! 12.Nxd4 Bxd4 13.Bxh7† Kxh7 14.Qxd4, but the exchange of
two or three minor pieces doesn’t change the fact that Black’s king is weak. At grandmaster level Black has lost one
game and made one draw – agreed immediately in this position!
10...Nf6
A retreat that clearly shows that Black would have preferred to play ...d5xc4 to keep the knight on f6.
11.a3 b6 12.Bc2!?
An ambitious move that keeps the bishop on c1 to go straight to h6 without a stop on g5. The downside is Black’s
next move.
14.d5!?
A pawn sacrifice with the idea of chasing the rook with Bc2-f5.
14...exd5
14...Nxd5 leads to a nice line: 15.Nxd5 Qxd5 16.Qxd5 exd5 17.Bf5 Rcd8 18.Bc7 Ra8 19.Bd7 It seems that Black is
caught, but he has the resource: 19...Bf6! 20.Bxc6 Rac8 And after 21.Bd6 Rfd8 22.Be7 Bxe7 23.Rxe7 Rxc6 24.Rxa7
White holds a positional advantage due to the minor pieces. The knight will protect both entry squares on the second
rank when it goes to d4 on the next move.
1c) 6...Bb4
7.cxd5 Nxd5
Just as after 6...Be7, 7...exd5 gives Black worse bishops. The only thing to note is that 8.Bd3 Qe7†?! 9.Ne5 Nc6
10.0-0 is a pawn sacrifice that White doesn’t have to think twice about before playing.
8.Qc2
8.Bd2 is the most common, but we try to manage with only moves that fit into our set-up. There is a concrete
downside though.
8...Nc6
8...0-0?! 9.Bd3! forces a weakness.
9.a3 Ba5!
The move we would really enjoy facing is:
9...Be7
Compared to 6...Be7, White has been forced to play an early Qc2, but has got a2-a3 for free. That’s a deal to
appreciate!
10.Bd3 Bf6
10...g6 11.Bh6 is standard. After 11...Bf8 12.Bxf8 Kxf8 13.Nxd5 we get a positional advantage if Black takes
with the pawn or an initiative on the dark squares if he takes with the queen.
10...h6 11.0-0 0-0 12.Qe2 was Smith – Lindgren.
11.0-0 h6
We have a normal isolated pawn position; note the following pawn sacrifice:
12.Qe2
This has given White 4½/5 in GM practice.
10.Bd3 Nxc3!
10...Nxd4 leads to the same.
One standard move to remember is: 10...h6 11.0-0 0-0 12.Be3!? In case of fxe3, White gets a centre that controls
squares but doesn’t advance. Black will probably have to give back the bishop after Ne4-c5.
14.0-0!
The tabiya of this line.
a) 14...Qxc3?! is too greedy and allows White’s initiative to develop with tempo: 15.Bxd7† Kxd7 16.Qa4† Ke7
17.Bg5† Kf8 18.Rac1 Qe5 19.Be3 b6
b) 14...Rc8 was Topalov – Magem Badals. The first moves are: 15.Rd1! Qxd1†! 16.Qxd1 Bxb5 17.Qh5! a6 18.Bb2 0-
0 19.c4! Rxc4 20.a4!
c) 14...Qe5 15.Bxd7† Kxd7 16.Qa4† Ke7 was played three times in a match between Ghaem Maghami and Karpov in
Teheran in 2009. They won one time each, but the engine prefers White after: 17.Rb1 (but not 17.Bf4 b5! with
simplifications) 17...b6 18.Bf4N
d) 14...Qd5 15.c4 Qf5! 16.Bxd7† Kxd7
Unlike 14...Qe5, White won’t win a tempo on Black’s queen.
17.Qb3
The double threat with 17.Qb2 is useless as 17...b6! 18.Qxg7 Rhg8 turns the initiative around.
White’s history of 4/5 in GM games tells us how the position should be evaluated in a practical game. The engine
gives one acceptable line for Black.
23...e5!
Only this isn’t mentioned in Lorin D’Costa’s book about the Panov.
24.Rd5 f6 25.Rf3 Qe6 26.Qc2† Kg7 27.Qe4 Rac8 28.Rh3 h6 29.Rhd3 f5! 30.Qe2 e4!
Even if Black finds all these moves, White is better and won in Parkkinen – Poli, corr. 2008, after an undermining
move.
31.g4!
6...Bg4 7.Bg5(!)
7.cxd5 Nxd5 8.Qb3 Bxf3 9.gxf3 is the start of a very theoretical line.
The text move was played in Yugoslavia in 1977 but has not yet become popular, even though we might expect a
change, since Levon Aronian tried it against Alexander Grischuk in the 2015 Sinquefield Cup.
a) The pin with 7...Qa5 is not a problem after: 8.Bxf6 exf6 9.h3! Bh5 10.Be2 It’s important to include h2-h3, so the
bishop can’t retreat to e6 after 10...dxc4? 11.Nd2!.
Better is 9...Bxf3 10.Qxf3 dxc4 when 11.Qe3†N 11...Be7 12.Bxc4 gives a structure that would be winning in a pawn
ending. However, the opposite-coloured bishops make it difficult for the passed pawn to get past d6.
b) 7...Ne4 8.Be3 was played in Lagno – A. Muzychuk in 2015. Games played by top women are more theoretically
interesting than games by other grandmasters of the same strength, since the women often have seconds much stronger
than themselves. The game continued: 8...e6 9.cxd5 Nxc3 (9...exd5 10.Bd3 Bb4 11.0-0! is a strong pawn sacrifice that
Black shouldn’t accept) 10.bxc3 Qxd5 11.Be2 Be7 12.0-0 0-0 13.h3 Bxf3? 14.Bxf3 And the bishop pair led to a
comfortable win.
Instead, Black should retreat with her bishop with an okay position.
c) 7...Bxf3 8.gxf3 is not a great problem since we castle long. White has had a few victories after: 8...dxc4
9.d5 (9.Bxc4 is equal) 9...Ne5 10.f4 Nd3† 11.Bxd3 cxd3 12.Qxd3 And now Black should prioritize development with
12...g6 13.d6 Bg7 14.0-0-0N 0-0 15.d7 but Black will have to live on the eighth rank for the rest of the game.
d) 7...dxc4 8.Bxc4 Bxf3 is never played. 9.gxf3?! leads to a position we avoided after 7...Bxf3, so 9.Qxf3 is logical.
Black doesn’t have time to pick up the rook on a1 after 9...Nxd4? 10.Qxb7. He should capture with the queen, when
White has compensation for the pawn after 9...Qxd4 10.Bb5 Qe5† 11.Be3 Qc7 12.0-0. Black is still three moves from
castling and has to find 12...a6!.
9.Qe2†! Be6 10.Bxf6 gxf6 was played by Grischuk, and also in Smith – Goganov from the introduction. After 11.Qc2
Qb6 I suggest deviating from Grischuk’s play with 12.Bb5 0-0-0 13.0-0. One point of castling immediately is that it’s
possible to take the h-pawn if Black plays 13...Rg8. However, I am not sure how to evaluate the position after 13...h5.
In ChessBase Magazine, Milos Pavlovic analysed 10...Qxf6!? 11.Nxd5 Qd8 12.Nc3 Nxd4 13.Nxd4 Qxd4 but I think
there is a logical deviation in: 14.Rd1 Qb6 15.Qe5N
White keeps on the queens since Black will have a hard time developing the bishop from f8 and castling short.
It’s good to fianchetto the bishop against an IQP, and if White plays normal moves he will reach a Tarrasch with
reversed colours. But the extra tempo gives White an additional possibility.
6.Qb3! Bg7
The problem for Black is that he is not in time to castle after 6...dxc4 7.Bxc4.
7.cxd5 0-0
Without a knight on c6 Black can leave the pawn on d5, castle and then try to win it back with ...Nbd7-b6. Blocking it
with ...Ne8-d6 doesn’t give equality. The doubled pawn is strong, even though it can’t advance – just look at the
squares on c5 and e5. An exchange on one of those would be fatal for Black.
8.Be2!
White prepares to defend the pawn with 9.Bf3, but he is also ready for quick castling with 9.Nf3. Therefore he
nullifies the sacrifices 8...b5 or 8...e6.
8...Nbd7
The other knight route starts with: 8...Na6 9.Bf3 Qb6! The queen evacuates d8 for the rook and allows the knight to
jump to b4. Black wins back the pawn if White exchanges, but there is another idea. 10.Nge2 Qxb3 11.axb3 Nb4 12.0-
0 Rd8 13.Ra5!? The point of allowing the exchange on b3. The rook is ready to create a double threat on b5 in case
Black develops his light-squared bishop. 13...a6 14.Na4 This move is not mentioned in Karpov’s book about the Caro-
Kann, but was played in the EU Woman’s Championship in 2010.
14...Nbxd5? 15.Rxd5! wins material due to the fork on b6, so Black should play 14...Nd3 15.Nb6 Nxc1 16.Rxc1 Rb8
17.Rc7 when Black is still far away from winning back the pawn.
Now 10...a5 is not as dangerous as in Tikkanen – Smith (page 168, with the knight on f3). The difference is that
without ...exf6, Black can’t harass the queen with a quick ...Bf8 after 11.0-0 a4 12.Qb5 Bd7 13.Qb4.
Instead, he has two ways of increasing the pressure against d5: either manoeuvring his bishop to c4 (10...Bf5) or
exchanging the defender on f3 (10...Bg4!).
The main move after 11...Qd7 is 12.a4, but I propose something else.
12.Re1!?
Black can give 12...Bd3 a new try now that b7 is defended, but 13.d6 exd6 14.Na4! hits b6 instead. White is better.
12...Rad8 is a logical try, considering that the e-file opens up after 13.d6. But with the rook on d8 and f8, White can
defend d5 indirectly with 13.Bg5! since 13...Nbxd5? 14.Bxd5 Nxd5 15.Nxd5 Qxd5 16.Bxe7 forks the rooks.
B) 10...Bg4!
Exchanging the bishop that not only defends d5 but also looks towards b7.
11.Bxg4 Nxg4 12.a4! a5
12...Nf6 was played by the well-prepared Evgeny Tomashevsky in the 2015 Qatar Masters. I can’t see that it has any
independent value since 13.Nf4 g5 doesn’t win back the pawn (13...a5 transposes to the normal line) and after 14.Nfe2
h6 15.h4 Black has only managed to destroy his own structure.
13.0-0 Qd6
Previously the main move, delaying ...Nf6 to stop Ne2-f4.
13...Nf6 transposes to Tomashevsky’s game. Unfortunately his opponent didn’t go for the critical move, so we can’t
know what he planned. I recommend following Horwitz – Landero Luna, corr. 2014:
14.Nf4 Qd6 15.Nb5 Qd7 16.d6 exd6 17.Be3 White is better, since Black’s pawns are easier to attack. In the game,
Horwitz won after taking the pawn on a5 – on move 36!
14.Bf4 Qb4 15.Ra3!
Polgar played 15.Qd1 against Skembris.
Before I started working on this book, I thought I had a lot of novelties. But Quality Chess sent me a database with
rapid and blitz games included, and I learned that almost everything has been played. This is such a case. The strange
piece configuration on the queenside works out well for White after a few concrete lines.
15...Bxd4
15...Nc4?! doesn’t trap the rook after 16.Na2! – the only move, but a strong one.
15...Rac8 16.Qd1 Qc4 (only move) 17.d6 exd6 18.Nb5 wins material.
16.h3!
16.Nxd4 Qxd4 17.Ne2 is also favourable, but the text is stronger.
After 6...Bg7, 7.Bb5† is the main move, but I think Black is okay after the clever: 7...Nbd7 8.d6 a6!
With a knight on c3 instead of f3, White would have had 9.dxe7 Qxe7† 10.Be2, but now he loses after 9.dxe7? Qa5†.
9.Be2 is nothing to worry about, since White’s point was to give a check on e2.
And 9.Bd3 0-0! 10.dxe7 Qxe7† 11.Be2 is better than the known sacrifice 8...0-0. Black has got ...a7-a6 for free and
that makes a difference when he plays ...b7-b5.
Instead, I recommend meeting 6...Bg7 with 7.Qb3. There could follow: 7...0-0 8.Nc3 Nbd7 9.Bg5 Nb6 10.Bxf6
Hans Tikkanen followed the repertoire in Zuke ’Em when we played in the Swedish Championship play-off in 2012.
But he had not seen: 10...exf6!N (10...Bxf6 11.Bd3 stops ...Bf5 and gives White a pleasant position after 11...e6 12.d6!
even though my second Sebastian Mauritsson thinks that Black can hold with 12...Bd7!) 11.Be2 a5! (I played 11...Re8
and lost without a fight) 12.a4 Qd6 With ...Rd8 and ...Qb4, Black should win back the pawn with an unclear position.
8...e6
This is not a move Black wants to play.
1.Nf3 c5 2.c4 Nf6 3.e3 Nc6 4.d4 cxd4 5.exd4 d5 6.Nc3 g6 7.cxd5 Nxd5
With the knight on c6, Black is forced to take back on d5. This position is normally reached after the popular line 1.c4
c5 2.Nf3 Nc6 3.Nc3 g6 4.e3 Nf6 5.d4 cxd4 6.exd4 d5 7.cxd5 Nxd5. I will only give a brief summary since it only
arises in about 2% of Panov games – although it was recommended by Jovanka Houska in Opening Repertoire: The
Caro-Kann (2015).
8.Qb3 Nxc3
8...e6 is a concession, but in contrast to the other g6-lines, White can’t force Black to take on c3 due to 9.Bc4 Na5.
The main line goes 9.Bb5 Bg7 10.Bg5 Qb6 11.Qa3 threatening mate and leading to immense complications that
eventually end with perpetual check.
A safer choice is 9.Bg5 Be7 10.Bh6 Bf8 11.Bxf8 Kxf8 12.Bb5.
Black could have avoided the repetition, but not without getting a bad position. After the exchange on f8, he is fine
but the venom in the IQP-position should not be underestimated even after further exchanges: 12...Nce7 13.0-0 Kg7
14.Rfe1 And then Black has two alternatives.
This was agreed drawn in Donchenko – Sutovsky, Gibraltar 2015, but Black still has some work left after 18.Bxb3
Nb4 19.Ne5 Nc6 20.Nxc6 bxc6 21.Red1 since the c6-pawn is the weakest.
When I played against Bjørn Møller Ochsner in the 2016 Nordic Championship, he chose another square for the knight
on the 12th move. It was the first time I felt out-prepared with the e3 poison.
12...Na5
16...Qd6!
a) 16...Nc6 17.Nxd5 Qxd5 (17...exd5 is a worse version of 16...Qd6) 18.b5! Nxd4 19.Nxd4 Qxd4 This is better for
White after the precise: 20.Rac1! The point of delaying b5xa6 is that Black isn’t allowed to play ...Qe5 plus ...b6-b5 to
round up the pawn on a6.
b) 16...Nb7 17.Nxd5 exd5 tries to manoeuvre the knight to c4, but there are a lot of obstacles along the journey: the
weak d5-pawn, the c6-square.
17.Nxd5 exd5 18.Qc3 Nc6
18...Nc4! 19.Bxc4 dxc4 looks bad after 20.d5†! Qf6 21.Qxc4. However, the engine destroys many good-looking
positions these days, and it claims enough compensation for a draw after 21...Bb7. But that’s theoretically speaking – as
a matter of fact, Ochsner didn’t care to play it even after he had looked at it.
19.Rfc1!
A small tactical detail. With the bishop on d3, Black could have taken on b4. Now 19...Nxb4? 20.a3 traps the knight.
19...Bd7 20.Ne5
Black has to defend patiently. White has the better minor pieces and a trick involving Bxa6 plus b4-b5, but according to
the engine, Black can hold on with:
20...Rhe8! 21.Bg4 Nb8
Not very easy moves to find over the board, and White can still play on.
If Black tries to exchange the heavy pieces on the c-file, it’s possible to abandon the file since there are no good entry
squares. A long-term plan could be to slowly advance the kingside pawns.
9.Bc4!
As in other lines, White strikes before Black is ready to castle.
9...Nd5!
Black should avoid: 9...e6 10.bxc3 Bg7 11.Ba3!
12...Qd5
12...Bg7 13.Qa3 Qe7 14.b3! is a nice way to keep control.
13.Be3!
Houska writes that this is “not the most challenging move as the bishop has a purely defensive role on e3”. I think it’s
strong, and that’s not only based on White’s impressive 8½/10 in the 2450 database – I have myself helped to raise the
statistics by twice being a victim with Black. Anish Giri beat me in Reykjavik 2013, but the move had of course been
played already – kudos to Heinz Wirthensohn.
So why is the bishop good on e3? True, it is passive for the moment, but it defends a strong pawn which controls the
c5- and e5-squares. White has only one weakness, whereas Black has both his c- and a-pawns to worry about.
Meanwhile the bishop stays away from being threatened with ...f7-f6 and ...g6-g5. White’s plan is to build up on the
queenside with Rfc1, Nd2, Qc2/c3 and Nb3/e4-c5 (or Ne5 if Black refrains from ...f6).
13...Be7
Houska recommends that Black refrains from exchanging queens, instead continuing with the text move followed by
...f7-f6, as in the game quoted below. If Black exchanges on b3, he opens the a-file and makes the a-pawn vulnerable.
Black can try to give up the c6-pawn and make a draw, but we can allow him to try, don’t you think?
13...f6 led to the same type of game in Balajayeva – Houska, Riga 2017. Black was slightly worse although she was
always within the margins and managed to outplay her opponent and win in the end.
The text move maintains White’s slight advantage, and he can think about Nc5 next.
Exercise 1
1.Nf3 c5 2.c4 Nf6 3.e3 Nc6 4.d4 cxd4 5.exd4 d5 6.Nc3 Bg4 7.Bg5 Bxf3
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 e6 6.Nf3 Nc6 7.c5 Ne4 8.Qc2 Qa5 9.a3 Bd7
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.Nf3 c5 2.c4 Nf6 3.e3 g6 4.d4 cxd4 5.exd4 d5 6.cxd5 Bg7 7.Qb3 0-0 8.Nc3 Nbd7 9.Bg5 Nb6 10.Bxf6 exf6 11.Be2
a5 12.a4
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 4
1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 g6 6.Qb3 Bg7 7.cxd5 0-0 8.Be2 Nbd7 9.Bf3 Nb6 10.Nge2 Bg4 11.Bxg4
Nxg4 12.a4 a5 13.0-0
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 5
1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 e6 6.Nf3 Bb4 7.cxd5 Nxd5 8.Qc2 Nc6 9.a3 Ba5 10.Bd3 Nxc3! 11.bxc3
Nxd4 12.Nxd4 Qxd4 13.Bb5† Bd7
White to move
Show/Hide Solution
Show/Hide Solution
5...a6 183
a) 5...Nc6?! 183
b) 5...dxc4! 183
c) 5...cxd4 183
6.dxc5 Bxc5 7.b4 Be7 184
7...Ba7?! 184
8.Bb2 0-0 184
8...dxc4 184
9.Nbd2 a5 184
9...b6 184
9...dxc4 184
10.b5 184
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 a6(!) 5.a3 dxc4! 6.Bxc4 b5 185
6...c5 185
7.Be2(!) 185
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 c5
Move Orders
The Tarrasch is a universal opening that can be played against almost any move order. The only ways I know to stop
Black from playing ...d7-d5, ...e7-e6, (...Nf6) and ...c7-c5 are:
a) 1.c4 e6 2.e4 d5 3.exd5 exd5, entering a worse version for White of the Exchange French
b) 1.e4
c) 1.b4
It’s even difficult for White to avoid d2-d4. After 1.c4 e6 2.Nf3 d5 3.e3 c5, the threat of 4...d4 “forces” White to play
4.d4 (or 4.cxd5 exd5 5.d4).
However, (after 1.c4 e6 2.Nf3 d5 3.e3) if Black starts with 3...Nf6, there is 4.b3 c5 5.Bb2 Nc6 6.cxd5 exd5 7.Be2, an
uncommon and good move that’s examined in the chapter about the Anti-Queen’s Gambit (page 243).
Pawn Structures
In a way, the Tarrasch is a reversed Panov – it’s Black who tries to get an isolated queen’s pawn. But the extra move
allows White to delay c4xd5, so Black won’t get the chance to play aggressively with ...Bg4.
Still, if Black gets an isolated pawn without being punished, he will soon have a fine position. And to punish him, we
need the most aggressive set-up: bishops on g2 and g5, and a knight on c3. They would all put pressure on the isolated
d5-pawn.
Theory considers the Tarrasch to be a sound opening as soon as White deviates from the above scheme. Our e2-e3 is
one such way to deviate.
I agree with that view. But why play into Black’s hands and give him the isolated pawn? 1.d4 d5 2.Nf3 Nf6 3.c4 e6
4.e3 c5 5.a3(!) is a timid-looking move that isn’t even mentioned in many repertoire books about the Tarrasch.
Delaying Nc3 is clever, as we will see, and a2-a3 is useful in many situations.
a) It’s a normal move if Black decides to give White an isolated pawn with 5...cxd4 6.exd4.
b) We can play a reversed Meran with 5...Nc6 6.dxc5 Bxc5 7.b4, when the extra tempo we get by being White makes
a huge difference.
Actually, one of Black’s best moves is also to be clever with 5...a6. And soon the question will be which side was
cleverer.
Structure 1
White develops with Bb2, 0-0, Nbd2, Bd3 and Rc1. It’s the Meran with opposite colours where:
It’s an important tempo. After developing with ...Nc6, ...Nf6 and ...Bd6, Black is not able to play both ...d5xc4 and
...e6-e5-e4.
1) If he starts with ...e6-e5, White takes on d5 and forces the knight to leave f6 and lose control over the e4-square.
2) If he starts with ...d5xc4, White recaptures with the knight, which controls e5 and hits the bishop on d6. When Black
has prepared ...e6-e5 with let’s say ...Bc7, ...a7-a6 and ...Re8, White has already completed development and
taken control over e4.
More about this will be seen in the game Krasenkow – Grandelius (page 181).
It’s better for Black to develop with ...Nbd7. The knight doesn’t block a bishop on b7, and could have a future on c5
after ...a7(a6)-a5. But White has a freer position and uses the a1-h8 and b1-h7 diagonals, with Qb1 and possibly Bd4
plus Qb2.
Structure 2
Black’s advanced b-pawn has left weak squares behind on the queenside. White can start by positioning a knight on
c4 and maybe continue to c6 in the endgame. In order to increase his control over the light squares, he will try to
exchange the light-squared bishops with Be2-f3. Other normal developing moves are 0-0, Bb2, Rfd1, Rac1 and Nf3-e5.
The queen will eventually (hopefully) find its place on e2.
One question is whether to play a4-a5. It fixes the black pawns on a6 and isolates the pawn on b4, but I would say that
it’s not so important. Black can be allowed to play ...a6-a5 and create another light-squared weakness.
However, a4-a5 is useful for another reason – it stops ...Nd7-b6, challenging the knight on c4. But time is also a
factor, and White should start by completing development.
So far, the position seems too good to be true, but ...b5-b4 is not just weakening. If Black takes control, the pawn on b3
could be the one that’s weak after ...c5xd4, ...Nd7-c5 and ...Bb7-d5. To avoid that, White can consider taking back on
d4 with the e-pawn, keeping the knight on d7 at bay and defending his own knight on e5. It’s not a big problem that the
pawn on d4 is isolated – Black doesn’t really have time to attack it with things happening on the queenside. But even
so, I would generally advise against e3xd4, since it limits the bishop on b2 and gives a stable square on d5 for a black
knight.
Games
Despite losing only twice in 12 games, Antoni Schön finished last in the ICCF Grandmaster Norm Tournament 44. In
this game, he gained a nice-looking position that few opponents would have enjoyed defending over the board. But
more is needed to win in correspondence chess.
8.Be2(!)
This is actually the first structure with reversed colours (back to normal!) where Black was recommended to develop
with ...Nbd7 and ...Bb7. The idea behind Be2 is to weaken the light squares on Black’s queenside with a3-a4 and then
exchange the light-squared bishops on f3.
8.Ba2 and 8.Bd3 are the main moves, but not much fun is happening there. White can’t push d4-d5 (in the first case) or
e3-e4-e5 (in the second case).
8...Nbd7
8...cxd4 9.Qxd4!? makes sense after Black has weakened his queenside with ...b7-b5. In the event of a queen
exchange, Black loses protection of his queenside. For example, 9...Qxd4 10.Nxd4 Bb7 11.Bd2 Nbd7 12.Rc1 and
Nb3-a5, Nc3-a2-b4-c6 or Rc7 are three ways to gain access.
Black does better to avoid the exchange.
9.a4 b4
White has a first square under control: c4.
I will not try to claim that White is better here, since Black usually accepts the same kind of position in a worse
version.
1) 1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.b3 (a move played by Grischuk) 7...b5 8.Be2 Bb7 9.a4
b4 10.Nbd2 Nbd7 and it’s White to move.
2) 1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 b5 (Evgeny Romanov’s invention) 5.a4 b4 6.Bxc4 e6 7.0-0 Bb7 8.Nbd2 Nbd7
9.b3 c5 10.Be2?! And the pawn is on a7. I think it’s better on a6, not only to keep b5 under control but also to avoid a
bishop exchange on a6.
However, even though it’s not theoretically threatening, I still appreciate how Schön methodically takes square after
square under control.
In the Tarrasch, it’s always interesting to compare the position with what we play with reversed colours. Let’s go back
and see a line after 5...a6: 6.dxc5 Bxc5 7.b4 Be7 8.Bb2 0-0 9.Nbd2 a5 10.b5 Nbd7 11.Be2 b6 We have reached the
same position, with reversed colours, but with the difference that White has the extra tempo Be2. The theoretical
section recommends 12.Nd4 Bb7 13.Nc6 Bxc6 14.bxc6 Nc5 15.0-0.
The pawn on c6 can be defended with c4xd5 then Bb5. But isn’t that giving away the tempo? What happens if Black
tries to do the same?
11...Nd5 12.Bb2 Nc3 13.Bxc3 bxc3 14.Nc4 cxd4 15.Qxd4 and the bishop can’t move from f8. 14...Be7 15.Qd3 rounds
up the pawn for the same reason, so Black has to play 14...Nf6 15.Rc1.
He can now defend the pawn with the knight or the bishop.
a) 15...cxd4 16.Nxd4 Bb4 17.Bf3 and Ne2 wins the pawn.
b) 15...Nd5 and the knight is not stable on d5. After 16.Nfe5 Be7 17.dxc5 Bxc5 18.Nd3 Be7 19.Bf3 White wins the
pawn with Qe2 followed by e3-e4.
11...Bd6
Black’s queen has few decent squares, so the bishop leaves e7 for her and manoeuvres to c7, protecting the queenside.
15.Nfe5
In general, 15.a5 is a move to consider, not particularly to fix the weaknesses on a6 and b4, but to stop Black from
challenging the knight on c4 with ...Nb6. But first White has to create a square for the queen. Otherwise 15...Rfd8
would be annoying, with the plan ...e6-e5.
So 15.Nfe5 prepares 16.Bf3.
15...cxd4
15...Rfd8 is no longer a problem. White has 16.Nxd7 Rxd7 17.Bf3 and the queen escapes.
16.exd4!?
White keeps the knight stable on e5 and continues to control c5. But on the other hand it blocks the bishop on b2 and
gives Black a square on d5.
It’s a fair exchange, but more unbalanced than the alternative. The fact that the pawn becomes isolated is less
important – Black is not in a position to attack it.
24...Ra8 25.Nb5 Bd8 26.Rc2 Qe7 27.Rec1 Nd7 28.h3 N7f6 29.Ned6 Bb6 30.Kh2 Rfd8 31.Rc6 Rab8 32.Nc4
White’s position has reached its optical maximum. Look at the pieces on the queenside! However, with the bad
bishop on b2, it’s difficult to make progress. In the game, Schön eventually advances on the kingside and tries to
activate the bishop via c1, but his nice-looking pieces don’t really contribute to that plan.
32...Qe8 33.Ne5 Qf8 34.R6c4 Ne7 35.Kg2 Nfd5 36.R1c2 Rb7 37.Bc1 Qe8 38.Qg4 Nf5 39.Nc6 Ra8 40.Qe4 Rd7
41.Bf4
Offering Black a chance to take the bad bishop.
46.Kf1
Stepping out of the fork on h4 and finally creating a real threat. If White gets another move he has 47.g4! when Black
has to take the bishop to avoid Bf4-d6. After 47...Nxf4 48.Qxf4 Nd6 49.Nxd6 Rxd6 (49...Qxd6 doesn’t work due to
50.Qxd6 Rxd6 51.Ne7† with a fork on c8) 50.Qf3 the light squares will finally create something real. White would
continue with Re2-e5.
48.Ne5 Rd5 49.g4 Ne7 50.Re1 Red8 51.Qe3 Rb8 52.f4 Rbd8 53.Kg2 Qe8 54.Re2 Qf8 55.Kh3 Qe8 56.Re1 Qf8
57.Kg3 Re8 58.Kg2 Ra8 59.Qd3 Rad8 60.f5 exf5 61.gxf5 Kh8 62.Re2 Ng8
Suddenly the game ends in a typical correspondence way: with a tactical sequence that’s not really forced and thus
very difficult to find in a real game.
63.Nc7 Rxe5! 64.Rxe5 Qd6 65.Rd5 Qf6 66.Rxd8 Qxd8 67.Nb5 Qg5† 68.Kf2 Nf6
½–½
I can’t say that I see the draw, but I understand that the open white king makes the final result unavoidable.
It’s different to watch or annotate games that have no tactical mistakes, and I didn’t dare to mark any move with a
question mark. That’s not the case in the next game.
Krasenkow, April 2016: “It is just a way to play chess. I think there are more chances 1.Nf3 d5 2.e3 will become trendy
after Karjakin’s win against Anand :-)”
7...Bd6 8.Bb2
I prefer 8.Nbd2, to be able to capture on c4 with the knight.
10...dxc4!
10...e5?! frees the bishop, but it makes a big difference that White takes on d5 instead of Black taking on c4. The
differences after 11.cxd5 Nxd5 are:
a) h7 is undefended
b) There is no knight on f6 to support ...e5-e4
c) Neither the bishop on d3 nor the knight on d2 had to take on c4 and thus give up control over e4
The e4-square is so important that it completely turns the evaluation around. White would have a nice position after the
simple 12.0-0.
11.Nxc4 Bc7
Black must avoid 11...e5?? as 12.Nxd6 Qxd6 13.Bxh7† wins his queen.
12.0-0 a6
Or 12...e5 13.Qc2 Re8 (13...e4 14.Bxf6! wins a pawn or an exchange) 14.Ncd2! and White keeps control. Black also
has to watch out for b4-b5, which explains the reason for 12...a6.
14...Nd5
Black would like to question the knight on e5, but 14...Nd7? 15.Bxh7†! Kxh7 16.Qc2† Kg8 17.Qxc7 wins a pawn.
15.Rc1 h6 16.Qf3
Even better was 16.Bb1 followed by 17.Qc2, which forces ...f7-f5.
16...Qg5
Black is finally close to getting rid of the knight (and then eventually developing the bishop from c8) but Krasenkow
gives up a pawn to keep the knight on e5.
19...Bb8 is not fun to play, but it is the engines’ choice. White has a strong attack for the pawn after 20.Rg4 with 21.Bd4
next.
24...Ra7?
The best chance was 24...Be6! 25.Bxc7 Ra7 since surprisingly White is unable to keep the material. 26.Be5 Rb7
27.Bd4 Rfb8 traps the knight and 26.Rc1 Bc4! threatens both bishops.
Instead 25.Nxa8! gives a big plus with the bishop pair.
25.Rc1
The knight can’t move, again due to the poor bishop on c8.
25...Be6 26.Rxc7
Black’s move order gave Krasenkow the chance to take with the rook. The game is over.
26...Rxc7 27.Bxc7 f6 28.a4 Rf7 29.Bg3 bxa4 30.Nxa4 Bd7 31.Nc3
White threatens both 32.Bxa6 and 32.Bc4.
1–0
Recap
As we have seen, Black has a lot to worry about, even though White satisfies himself with semi-passive moves like a2-
a3 and e2-e3. In the correspondence game, Gerola held his own, but was undoubtedly under pressure.
What I like most is that Black has no straightforward way to get a simple position. Whatever he plays, he has to face a
long-time plan, be it in a position with an isolated pawn on d4 or with a queenside majority. Let’s see how it works in
theory.
Theory
1.d4
1.c4 c5 2.Nf3 Nc6 3.e3 e6 4.d4 d5 is a possible move order. After 5.a3 a6!? 6.dxc5 Bxc5 7.b4 Be7 8.Bb2 Black can
try to make use of the knight being on g8 by 8...Bf6, but Black has nothing to boast about. After 9.Nc3 Nge7 10.Qc2
then Rad1 will come, with a great position.
a) 5...Nc6?! was Krasenkow – Grandelius. White has a pleasant position after 6.dxc5 Bxc5 7.b4.
b) 5...dxc4!, with the Queen’s Gambit Accepted, is a clever option. White has painted himself into the a3-corner – see
the lines given later after 4...a6.
c) 5...cxd4 6.exd4 gives a Panov where White has played a2-a3 earlier than normal, but ...e7-e6 is not a way to exploit
that. 6...Nc6 7.c5 is a standard move once the knight is on c6. White continues with Bb5, takes control over e5 and
pushes the queenside pawns.
But it’s possible to push c4-c5 even without the knight on c6: 6...Be7 7.Nc3 0-0 8.c5!? Ne4 9.Bd3! This was played
in a regional league in the Czech Republic in 2003, and suggested by David Rudel in Zuke ’Em (2008). White’s pawn
majority is going nowhere and he will stay with a weak pawn, but in the meantime he gets a dangerous attack. After
9...Nxc3? 10.bxc3 b6 11.h4!, Black has problems defending against 12.Bxh7†.
Black should avoid 11...g6?! 12.h5, which opens the h-file.
11...Ba6 stops the sacrifice, because the bishop can defend h7 from d3. But 12.Ng5! Bxd3 13.Qxd3 g6 14.Qg3 gives an
attacking position, with h4-h5 next and possible sacrifices on h7 or e6.
Instead, Black should play 9...f5 and after Rudel’s 10.Ne2, White advances his queenside pawns while Black has
difficulties in creating anything.
9.Nbd2
9...a5
9...b6 was my own choice against Victor Mikhalevski in the 2015 Rilton Cup. Without ...a6-a5 and b4-b5, White
doesn’t get control over the c6-square. 10.cxd5 Nxd5 11.Qb1!?N would have given White a position with strong
bishops. In the game Mikhalevski played 11.Nc4? but ran into a tactical trick with: 11...b5! The planned 12.Na5 loses a
pawn to 12...Nxb4! 13.Qxd8 Nc2†! 14.Kd2 Bxd8 (or two pawns after 13.axb4 Bxb4† 14.Ke2 Bxa5).
9...dxc4 10.Nxc4! is a good move when White can take back on d1 with the rook. That’s the reason why 8.Nbd2 would
have been too early – the rook was still blocked by the bishop on c1.
10.b5 Nbd7
It’s usually Black’s move here, after ...a7-a5 in one go. But I think he is okay anyway, even though he still has to do
some work to activate his bishops. White can continue developing.
Black keeps the option of going for the Tarrasch (without Bb5†) and the Queen’s Gambit Accepted. The first is
logical against passive moves like 5.Nbd2 or 5.b3, while the second makes sense after 5.Nc3 since ...b7-b5 increases in
strength (thus White could play 5.Nc3 dxc4 6.a4 with a dull variation of the Vienna, but that’s another story).
5.a3
It is logical to wait with 5.b3 as long as Black can go for the Tarrasch. We don’t want to play a2-a3 and b2-b3-b4 in
three moves. But 4...Be7 5.b3 and 4...c6 5.b3 are perfectly possible.
5...dxc4!
The Queen’s Gambit Accepted makes sense when White can’t play a2-a4. The risk that we will reach this position
after 4.e3 is, according to my database, only 12%, but that’s not a convincing argument for White.
6.Bxc4 b5
6...c5 leads to the same.
7.Be2(!)
A rare move with the idea of creating light-squared weaknesses in Black’s queenside (a3-a4) and then exchanging the
light-squared bishops (Be2-f3). A similar idea exists in the Queen’s Gambit Accepted: 1.d4 d5 2.Nf3 Nf6 3.c4 dxc4
4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.b3!? – played by none other than our hero Alexander Grischuk (in a blitz game). After
7...b5 8.Be2 Bb7 9.Bb2 Nbd7 10.a4 White has quite a good score. This could just as well have been the
recommendation in the QGA chapter.
However, with our move order, Black’s clever 4...a6! made us lose a tempo with a2-a3-a4. So we are more playing a
position than challenging Black theoretically.
As a curiosity, 7.Bd3 c5 8.dxc5 Bxc5 9.0-0 0-0 10.b4 contains a trap: 10...Bd6?! 11.Bb2 Bb7 12.Bxf6! Qxf6?
13.Bxh7†! Kxh7 14.Qxd6 Qxa1 15.Qxf8 gives White an extra pawn. Instead, Black should play 10...Bb6.
Exercise 1
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 c5 5.a3 a6 6.dxc5 Bxc5 7.b4 Be7 8.Bb2 0-0 9.Nbd2 dxc4
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 c5 5.a3 a6 6.dxc5 Bxc5 7.b4 Be7 8.Bb2 0-0 9.Nbd2
Black to move
Show/Hide Solution
Show/Hide Solution
...a6 Meran
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 e6 5.Nc3 a6 6.a3(!) Nbd7 203 (6...dxc4) 7.b4 Bd6 8.Bb2 0-0 9.Qc2!? Qe7! (a)
9...e5?; b) 9...Re8) 10.c5 204
Miss Meran
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 e6 5.Nc3 Nbd7 6.Bd2(!) 204 (6...a6)
A) 6...b6! 7.Bd3 Bb7 8.0-0 Be7! (8 ...Bd6?!) 9.Qe2 205
B) 6...Bd6 7.Bd3 dxc4 8.Bxc4 b5 9.Bd3 Bb7 (9...a6; 9...b4) 10.Rc1 Rc8 11.0-0 0-0 (11...a6?!) 12.a3 a5! 13.e4
e5 14.dxe5 Nxe5 15.Nxe5 Bxe5 16.Qe2 Re8 17.Rfd1 Qd6! (17...Qe7) 18.g3 206
C) 6...Bd6 7.Bd3 0-0 8.0-0 dxc4 207 (8...e5?!) 9.Bxc4 e5 (9...a6) 10.Qc2! Qe7 (10... exd4; 10...h6) 11.h3 Bc7
12.Bb3! h6 (12...e4?!) 13.Nh4 208
...a6 Slav
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 a6 5.Nc3 209
A) 5...Bf5 209 (5...Bg4) 6.cxd5 cxd5 7.Ne5 (a) 7...e6?; b) 7...Nfd7?; c) 7...Nbd7) 210
A1) 7...h6 8.Qb3 Ra7 9.Bd2 e6 10.Rc1 Nbd7 210 (10...Be7) 11.Bb5!? Bd6 12.Nxd7 Nxd7 13.Bxd7†
Kxd7 (13...Qxd7) 14.0-0 211
A2) 7...Nc6! 8.Bd2(!) Nxe5 9.dxe5 Nd7 211 (9...Ne4) 10.Nxd5 e6 11.Nf4! Nxe5 12.Bc3 Bd6 13.Qd4
Qe7 14.Be2N Rd8! (14...f6; 14...0-0?) 15.Qa4† 212
B) 5...b5 6.b3 Bg4 (6...Bf5) 7.Bd2(!) 212
B1) 7...e6 8.h3 Bxf3 212 (8...Bh5) 9.Qxf3 Ba3 (9...Bb4) 10.Bd3 0-0 11.0-0 Nbd7 12.Rad1 Rc8
(12...Bd6) 13.e4N e5?! (13...dxc4) 14.exd5 213
B2) 7...Nbd7! 8.h3 Bxf3 9.Qxf3 b4 213 (9...e5?) 10.Na4 e5 11.cxd5 cxd5 12.dxe5 Nxe5 13.Qf4!?
Bd6 14.Qd4 0-0 (14...Nc6) 15.Rc1! Ne4?! (15...Qe7) 16.Qxd5 Qh4 17.g3 Qh5 (17...Nxg3) 18.Qxe4
214
Move Orders
1.c4 or 1.Nf3 are good moves against Slav-players, but the Meran, the ...a6 Slav and the Schlechter Slav are not
avoided. After 1.Nf3 d5 2.e3 Nf6 3.c4 c6 4.Nc3 we have:
a) Miss Meran: 4...e6 5.d4, when White plays to restrict the light-squared bishop on c8.
b) ...a6 Slav: 4...a6 5.d4, harassing the bishop when it lands on f5 or g4.
c) Schlechter Slav: 4...g6 5.d4, a solid but passive line.
The same positions can be reached after 1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 and
a) 4...e6 5.Nc3
b) 4...a6 5.Nc3
c) 4...g6 5.Nc3
Pawn Structures
Structure 1
Black has four ways to develop the light-squared bishop. In order of probability:
c) ...e6-e5: White gives Black an isolated pawn by taking on d5 and e5, and exchanges minor pieces – especially the
dark-squared bishop.
d) ...b7-b6 plus ...Bb7 and ...c6-c5. White should be ready to take on d5. If Black exchanges on c4 first, White takes
back with the b-pawn and answers ...c6-c5 with d4-d5.
White’s only pawn lever is e3-e4, but it’s not automatically good. Some possibilities:
a) If Black takes on e4 and can’t break free, he gets a passive position. Then c4-c5 is a standard move that seals the
fate of the bishop on b7. So if Black takes on e4, he wants to follow up with ...e6-e5 or ...c6-c5.
b) Black’s standard reaction is ...d5xc4 plus ...e6-e5. That undermines the pawn on d4 (no longer defended from e3)
and wins the c5- and/or e5-squares for the knight. Usually, White goes e3-e4 only if this option isn’t possible.
Structure 2
Black plays ...a7-a6 to threaten ...c6-c5, and is fine if he succeeds. However, playing four or five pawn moves on the
queenside takes time, and White will try to prevent the lever in one of the following ways:
1) a2-a3, threatening b2-b4. Black has three ways to react.
1a) Stopping b2-b4 with ...b5-b4 plus ...Bxb4 – probably not good if the bishop has to move twice.
1b) Stopping b2-b4 with ...a7-a5. This makes it harder to move the c6-pawn.
1c) Allow b2-b4 and attack the pawn with ...a7-a5, ...Bd6 and ...Qe7. When White defends (Rb1 and Qb3), Black plays
in the centre with ...e6-e5.
2) White plays a quick e3-e4-(e5) and uses his lead in development (due to Black’s ...a7-a6).
3) Bd2 plus Rc1, when c5 could be defended either with b2-b4 (if it works tactically) or Nc3-e4. This is the option
that’s recommended in the theoretical section.
Structure 3
If White manages to stop ...e5-e4 (and he will), there is only one reasonable way to develop the bishop on c8: Black
exchanges on d4 and moves the knight from d7 (probably to b6).
The recapture e3xd4 gives an isolated pawn: a position discussed in more detail in the chapter about the Exchange
French (page 345). The main plan is to cover all the squares of Black’s light-squared bishop:
White has several useful moves if Black delays the exchange on d4. Except the three moves mentioned above, there are
Re1, a2-a3, Bd2, and Rad1.
Structure 4
The Schlechter Slav is a solid but passive system that’s used by players who want to avoid theory. Nothing wrong
with that, but we are also happy!
With a restrictive move such as ...c7-c6, it’s less likely that Black will achieve a successful ...e7-e5 or ...c6-c5, like in
the Grünfeld. But it’s those pawn levers that White should watch out for anyway.
The bishop belongs on e2, since White doesn’t want to take back on f3 with the queen in case of ...Bg4xf3. The queen
want to attack the light squares on the queenside, more specifically the b7-pawn. In fact, White can meet both ...Bg4
and ...Nbd7 with c4xd5 plus Qb3, since in such a position Black’s best set-up is one with the bishop on c8 and the
knight on c6. If the structure and piece configuration is still intact when White has completed development (don’t
forget Bd2!), there are two plans to choose between.
1) Ne5 followed by f2-f4 is quite annoying for Black; it is difficult to get rid of the knight without making a
positional commitment.
2) Grabbing space on the queenside with b2-b4 (and a2-a4) is a good idea – that’s a reason why the queen shouldn’t
go to b3 too early.
However, White should not combine both plans. And he should not prepare e3-e4, since that would make the bishop on
g7 much stronger.
Games
The Swedish coach Emil Hermansson has realized that in order to understand an opening well, it’s important to focus on
the very first moves. Understanding move orders and why a piece belongs on a certain square makes it much easier to
find the best move in a new position. After one hour in one of his sessions, they had not yet reached move six.
I will not challenge your patience that much, but I will at least try to explain the reasoning behind 6.Bd2(!) in the
Meran.
6.Bd2
A move Ellinor Frisk suggested a couple of months before we married in 2012 – hence the name Miss Meran.
The main lines in the Meran start with 6.Bd3 dxc4 7.Bxc4 b5. Delaying castling is justified; the first priority is to
activate the light-squared bishop with ...Bb7, ...a7-a6 and ...c6-c5. If it’s done immediately, White doesn’t get time for
b2-b4.
The semi-waiting 6.Qc2 is common and only after 6...Bd6 does White play 7.Bd3. The idea is that the queen takes part
in stopping ...c6-c5, while Black’s bishop doesn’t contribute more than before. However, it’s not always convenient to
develop the queen too early and that is beautifully shown by: 7...0-0 8.0-0 dxc4 9.Bxc4 b5 10.Bd3 Bb7 (10...a6?
11.Ne4! wins a pawn and shows why the queen is good on c2) 11.a3 Rc8 12.b4
12...c5!! A novelty in Topalov – Kasimdzhanov, London 2012, and used with success several times thereafter. The
complications end with a drawish position.
So why not play another semi-waiting move? 6.a3 solves the problem, but the move is too meek if Black plays for an
isolated pawn with 6...Bd6 7.Bd3 0-0 8.0-0 e5! 9.cxd5 cxd5 10.dxe5 Nxe5 11.Nxe5 Bxe5.
Therefore: 6.Bd2 It vacates c1 for the rook and may back up b2-b4.
White will take back on d4 with the e-pawn and get an isolated pawn. His main aim is to limit Black’s light-squared
bishop.
10.Qc2
10.h3 stops ...Bg4 and was my intention before the game, which was the last in the Swedish Grand Prix series of
2014-15. A few years ago, I decided to never play for prize-money – a game of chess is in itself much more important
than money. But since this game could bring anything from 1,000 to 2,500 euros, I started to care. When the round was
delayed when one player continued for an hour in a dead-drawn position, I examined the position again and realized
that White’s first priority is to stop ...e5-e4.
However, that’s mainly a threat with Black’s bishop on c7, as the following illustrative line shows: 10...Bc7 11.a3? e4
12.Ng5 Nb6 13.Ba2 Qd6! And now 14.f4 exf3 15.Nxf3 allows Black to draw with: 15...Bxh3! 16.gxh3 Qg3† And
Black is also happy without taking on h3.
10...exd4
White has several useful moves if Black delays the capture: Re1, Bb3 and a2-a3.
Lindgren solved the problem with the bishop by accepting an isolated pawn on e6. It’s weaker than White’s, since
Black’s knights are far away from attacking the d4-pawn. However, Lindgren argues that it’s a fair exchange since
White’s remaining bishop is worse than his. But add the space and a safer king, and White has every reason to be
happy. The first step is to double the rooks.
Better was 19.Nd3 and if 19...Nhf4 20.Bxf4 Nxf4 21.Nxf4 Rxf4 22.Rxe6 Rxe6 23.Rxe6 Rxd4 24.Re8† White has an
attack.
19...Nhf6!
The knight has fulfilled its task.
20.Bg5
Trying to exchange the bad bishop. Lindgren decides to take concrete measures.
20...Nxc3 21.bxc3 c5
22.Bxf6!
This destroys Black’s coordination.
24.Kg2
24.Ng4!? wins the e6-pawn, but Black gets active with 24...Rf7! 25.Rxe6 Raf8 and 26...Bxg3 is a threat.
26...Rf7 27.Rc2?!
27.Rd1 and 28.Rd6 still gave a better position.
27...Qd7 28.Qe4 Rd8 29.Rec1 Qd5 30.Qxd5 Rxd5 31.Rc8† Rf8 32.R1c7 Rxe5 33.Rxf8†
½–½
The next game is an illustrative example of how to block, pressurize and conquer an isolated pawn. But again, I take the
chance to delve a bit deeper into the choice of recommending 6.a3(!) in the theoretical section.
Aronian, April 2016: “I love playing in Yerevan. People are interested in chess, and neighbours can come and watch.”
6.b3
The main line, as played a few times in the World Championship match between Anand and Gelfand. The Anand
team considered 6...Bb4 7.Bd2 0-0 8.Bd3 Nbd7 9.0-0 Bd6 10.Rc1 e5! to be the key move. By forcing the bishop to d2,
Black became ready to go for an isolated pawn. The last move stopped White from placing a knight on e5.
The position looks better for White after 11.cxd5 cxd5 12.dxe5 Nxe5 13.Nxe5 Bxe5 14.Ne2.
Black can’t really avoid Bc3, exchanging his good bishop. However, the advantage was considered to be academic.
But we are happy with that as well, and in GM practice White has scored +2 =7 –0.
The reason to choose something other than 6.b3 is rather that Black has several over-analysed possibilities along the
way. Just to mention one that is not a problem: 10...Re8 11.e4! dxc4 12.Bxc4 e5 From this position I prepared
13.Ng5N many years ago, with exciting lines. In Pump Up Your Rating I asked for an opponent who wanted to go into:
13...Re7 14.f4 exd4 15.e5 dxc3 16.exd6 cxd2 17.dxe7 Qb6† 18.Kh1 dxc1=Q
19.Qh5!! White wins: 19...g6 20.Qxh7† Nxh7 21.e8=Q† Ndf8 22.Qxf7† Kh8 23.Qg8 mate or 19...Qxf1† 20.Bxf1
Ne5 21.fxe5 Be6 22.exf6 h6 23.Nxf7 Bxf7 24.Bc4 Bxc4 25.Qg6:
It’s mate on the next move.
There are deviations along the way that keep the game unclear, but it should have been a hard task for an unprepared
opponent. But after three years of waiting, no volunteer had appeared, and I gave up trying to remember the 500 moves
of analysis (and regret that I didn’t do so earlier).
The theoretical section recommends 6.a3(!), meeting the ...a7-a6 move in the same way as in the Tarrasch and the
Queen’s Gambit. Black is ready to push ...e6-e5 after 6...Nbd7 7.b4 Bd6 8.Bb2 0-0 and White has to decide how to
react.
a) The main line goes 9.c5 Bc7 10.Na4, intending to take twice on e5 and playing for the b6- and d4-squares. However,
Black has proved to be solid. The alternative is to give Black an IQP position.
b) 9.Bd3?! would be a mistake though, due to 9...dxc4 10.Bxc4 e5. The tempo makes a huge difference, as White can’t
stop ...e5-e4 with 11.Qc2 Qe7 12.Bd3, since he would have wanted to take back with the e-pawn after 12...exd4.
c) 9.Qc2!? Re8 10.Bd3 e5 (with the rook on e8, Black can’t play 10...dxc4 11.Bxc4 e5 due to 12.Ng5!) 11.cxd5 cxd5
12.dxe5 Nxe5 13.Nxe5 Bxe5 14.Ne2
Compared to the position that Anand reached, we can see the following differences:
a) Considering piece development, Black is one tempo up
b) White has taken some space on the queenside
c) The exchange of dark-squared bishops is prepared by White’s queen rather than the rook
I think the second (slightly) and the third point (mainly) makes up for the tempo. White’s queen keeps some influence
over the dark squares. Still, I am pretty sure that Anand’s team would have been happy to defend this position to a
draw.
We should also note the engine’s suggestion 14...d4 15.exd4 Bd6. White’s extra pawn limits the bishop on b2 and the
knight on e2, but if we manage to develop, a pawn is still a pawn. And fortunately 16.0-0 Bxh2† 17.Kxh2 Ng4†
18.Kg1 Qh4 is not mate due to 19.Qc7!.
Also note that 15...Bxd4? 16.Rd1 wins material.
6...Bb4 7.Bd2 0-0 8.Bd3 Nbd7 9.0-0 Bd6 10.Rc1 e5 11.cxd5 cxd5 12.dxe5
Gelfand played 12.e4 against Anand, so he didn’t believe that he could win the position with an isolated pawn that
appears in the game.
14...Bg4
Exchanging the bad bishop or forcing a weakness.
14...Bxh2† seldom works without a black pawn on e4. 15.Kxh2 Ng4† 16.Kg3 Qg5
17.f4! Black could have taken en passant if he had a pawn on e4.
15.f3
15.Bc3 allows 15...Bxe2 16.Bxe2 Qd6 17.Bxe5 Qxe5 when a confident player would be sure to hold this with Black,
while I would still be nervous.
15...Bh5?!
15...Bd7 keeps the possibility of defending the isolated pawn.
18...Rfe8
18...Rac8 forces Black to sacrifice a piece after 19.Rxc8 Rxc8 20.g4 due to f3-f4-f5.
19.Rc3! Bg6
Black exchanges bishops while he can, although it doesn’t solve his problems.
20.Bxg6 hxg6
Aronian pointed out that the doubled g-pawns are a disadvantage, since the black king lacks a convenient post on g7.
21.Qd4
A queen exchange would allow White to defend the e-pawn with Kf2, without any risk.
21...Qe7
Aronian recommends 21...Qe6! with ...Rac8 to follow. If White wins a tempo with 22.Nf4 Qe7 he runs into ...g6-g5-
g4.
25...Rde7 26.R1c5 g5
Aronian: “This move is quite out of place. Now Black’s position becomes technically lost. He had to stick with waiting
tactics.”
27.Qc3 Qe6 28.Rxe8† Rxe8 29.Nd4 Qe5 30.Rc8?!
Getting a winning position without much effort caused Aronian to relax. He should have played 30.Rc7 b5 31.Rc8 to
create another weakness.
34.a4 Qd8
With the knight on h7, Black can’t allow the queen exchange after 34...f6 35.Qe6†, since he loses one of the
queenside pawns.
Aronian considers this position to be winning, with the plan being to conquer the a6-pawn with the king.
51.e6! Qc7†
Or 51...Qxe6 52.Qxe6 fxe6 53.Kg3 Kg7 54.Kg4 Kf6 55.f4 and White wins with an outside passed pawn.
52.f4 gxf4 53.exf7† Qxf7 54.Qxa6 Kg7 55.Qc6 Qb3?
Morozevich speeds up the end by blundering a second pawn.
56.Qc7†
1–0
Lindgren developed the light-squared bishop at the price of getting an isolated pawn on e6, while Morozevich got an
isolated pawn on d5. In the next game, he plays ...Bf5 before he closes the pawn structure with ...e7-e6. But it’s not
without suffering. He has to move both flank pawns and he weakens the light squares on the queenside.
Grischuk, September 2016: “It was a very interesting game, but at some point our captains agreed to a draw in the
match, so our game was abandoned.”
5.Nf3 b5
Or if 5...Bf5 then 6.cxd5 is one of several moves, when 6...cxd5 7.Ne5 and 8.Qb3 gives some initiative. With 5...b5,
Black closes that possibility once and for all.
6.b3 Bf5
Black will manage to hold on to the queenside without the light-squared bishop. But he should avoid a voluntary
...b5-b4, since it would allow Na4 followed by c4xd5 and penetration down the c-file.
11.Bb2
The theoretical section recommends continuing in aggressive style with 11.f4, when g4-g5-g6 may kill the bishop and
fight for the light squares.
13...Bc7 14.Ne2
A few months later, Natalia Zhukova played 14.h3, but it’s not necessary to defend the g-pawn since 14.Ne2 Bxe5
15.dxe5 Nxg4 16.h3 traps the knight.
17.Nd3!
White has the space advantage, so he keeps the pieces on.
23.e4!?
The logical follow-up, but Morozevich solves the problems with some care. With hindsight, White should have
avoided 20.Qd2, and focused on opening the position before Black had time to coordinate.
26.Bf2?!
½–½
Grischuk’s intention must have been 26.Bf2 Bxb3 27.Nd6 Ndxc5! and then 28.Nxe8 without the bishop on d4
hanging with check, but 28...Rd7! still wins for Black.
Better is 28.Nxc5 Nxc5 29.Bxc5 but White can’t save the piece after 29...Rd7:
a) 30.Qb2 Bxd6 31.Qxb3 Bxc5† 32.Rxc5 Qb6 33.Rfc1 Re5 34.Qc4 b3 and the b-pawn wins a rook.
b) 30.Qb2 Bxd6 31.Bxd6 Bxa4 when the three connected passed pawns are stronger than the piece.
c) 30.Bxc6 Rxd6 31.Bxd6 Bxd6 with more than enough compensation for the exchange, since White can’t take on e8.
The final game starts with a Schlechter Slav. The lines will not be examined further in the theoretical section, so make it
a careful read.
Nielsen in Skakbladet #6/2011: “I refrain from the sharp lines. The plan was to play a long game, not a short intensive
fight with a quick end.”
1.d4 d5
Or 1...Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 c6 followed by ...d7-d5 is another route to the game.
6...0-0 7.0-0
7...dxc4
Black has many alternatives.
a) 7...Bg4 8.cxd5! (7...Nbd7 would have been met in the same way) 8...cxd5 9.Qb3 b6 10.h3 Bc8
The light squares on the queenside need protection. There are many decent ways of continuing, one being: 11.Bd2
Nc6 12.Ne5 Bb7 13.Nxc6 Bxc6 14.Rfc1 Qd7 15.Ba6N (Anand played 15.a4 against Gashimov in Nanjing 2010.
Black held a comfortable draw.) 15...Bb7 16.Bb5 Qd6 17.Be2 White plans Nb5 and Bb4, and Black will lose the fight
for the c-file.
b) 7...a6 was played a few times by Wang Yue at the time when he could draw with anyone. The easiest way for White
to keep the game going is 8.Bd2 and after 8...dxc4 9.Bxc4 Bg4 10.Be2!? it resembles the main game.
c) 7...e6 prepares 8...Nbd7, since the e-pawn can take back on d5. White’s best move is 8.b4. Compared to a normal
Meran, Black’s bishop doesn’t control b4.
d) 7...Re8!? is a semi-waiting move that I played against Radoslaw Wojtaszek in Le Grau-du-Roi 2015. The inspiration
came from Nigel Short’s 11...Re8 in a completely different position against Tomi Nybäck at Sigeman & Co in 2009.
Short won, and after the game he explained that 11...Re8 was a “F***-You move”, telling the opponent that he could do
nothing even if Short passed over the move.
Well, I didn’t think the same, and one point of 7...Re8 is that 8.b4 can be met with 8...Be6, when White would have
wanted to defend the c-pawn with b2-b3.
Another reason behind my choice was to avoid the main lines against one of Anand’s seconds. It was a nice feeling to
be 20 minutes up on the clock, even though I didn’t make the F***-you sign.
However, I don’t think the move has any theoretical value. Without b2-b4 working, White should use the other plan
outlined on page 190: Ne5 followed by f2-f4.
8.Bxc4 Bg4
Black tries to lure White’s queen to f3 and after 9.h3 Bxf3 10.Qxf3 Nbd7 11.Rd1 e5! there has been some debate as
to whether he has enough compensation for the bishop pair.
9.Be2!
Played by Evgeny Tomashevsky against Gashimov in the previous match-up. It’s logical to hold on to the idea behind
6.Be2: to keep the queen for the queenside.
12.Bxf3 e5 13.dxe5
Nielsen in Skakbladet: “I discussed this move with Jon Ludvig Hammer the day before, and liked White’s chances.
Magnus Carlsen reminded me that six years earlier we had discussed the kind of position that appears. During the
European Championship in 2005 I had, without luck, tried to convince him that White simply has a solid advantage
with the pair of bishops.”
13...Nxe5 14.Be2
White is passive, but since Black is going nowhere, White is not in a hurry. The bishop pair is a long-term advantage.
14...Rfe8 15.h3 h6 16.Bd2 Qc5 17.Rac1 Rad8 18.a3 Ned7 19.b4 Qe7
20.a4!?
Not the only way to make progress, but a sign that Nielsen is ready for confrontation. In the game he transforms his
advantage into another form: a better pawn structure.
20.Be1 is a standard move.
27.Qd2 Qe7 28.a5 Nc5 29.Qc2 Nb7 30.Qa4 Qc5 31.a6 Nd8 32.Na2 Qb6 33.Qb4 Qc7 34.Nc3 Qe7 35.Qb3 h5
36.Bf3 Qc5 37.Ne2 Qd6 38.Rd1 Qc7 39.Nf4 Kh8 40.Qd3 Rf8 41.Qd7 Be5 42.Be2 Kg7 43.h4 Qb6 44.Bc4 Qb4
45.Bf1 Qb6 46.Qe7 Bf6 47.Qe4 Qc7 48.Bc4 c5 49.Rd5 Qc6 50.g3 Qa4 51.Kg2
Instead of taking the pawn on c5, Nielsen spends time moving his king away from nonexistent threats. It’s illustrative
of the nature of White’s play in this game – with food on the table you don’t need to run.
72.Rh8 Rc6 73.Rh7 Rf6† 74.Ke4 Bb4 75.Bc4 Be1 76.Kf3 Bd2 77.Rh8 Rc6 78.Bb5 Rb6 79.Re8† Kf6 80.Bc4 Rc6
81.Bd3 Ba5 82.e4 Bb6 83.e5† Kg7 84.f5 f6 85.e6 Rc7 86.Rb8 Bc5 87.Be4 Bd6 88.Ra8 Bc5 89.Bd5 Be7 90.Ke4 Bf8
91.Bb7 Bc5
92.Kd5 Kh6 93.Bc6 Bxf2 94.Rg8 Be3 95.Rg6† Kh7 96.Rxf6 Bg5 97.Rg6 Be7 98.Rg2 Rc8 99.Rb2 Rd8† 100.Bd7 Rf8
101.Ke4 Kg7 102.Rg2† Kh7 103.Ke5 Bb4 104.Rg3 Be7 105.Rg2 Bb4 106.Rc2 Be7 107.Rc7 Kh6 108.Bc6 Bd8
109.Rb7 Kg5 110.Rg7† Kh6 111.Rxa7 Bf6† 112.Ke4 Rc8 113.Rf7
1–0
Nielsen had never beaten such a strong player before; he forced a rapid tiebreak which he eventually lost.
Recap
The first two games gave quite a good idea of how to play in the Meran with a better pawn structure. On the other hand,
Grischuk’s game was chaotic, even in the final position.
A clearer picture is hopefully given by the following lines.
Theory
The theory is divided into three sections: the ...a6 Meran, Miss Meran and the ...a6 Slav.
...a6 Meran
The advantage with 9...Qe7! is that it keeps the rook on f8, defending the f7-pawn.
10.c5
10.Bd3 stills runs into 10...dxc4 11.Bxc4 e5.
Miss Meran
5...Nbd7 6.Bd2(!)
Curiously, just as you will see in the Chigorin (page 219), we go for a move that was played between Wilhelm
Steinitz and Mikhail Chigorin in the 1800s.
We have A) 6...b6, B) 6...Bd6 7.Bd3 dxc4 and C) 6...Bd6 7.Bd3 0-0.
After 7...e5 8.Nxe5 Nxe5 9.dxe5 Nd7 10.b4 a5 11.a3 we see the value of the Bd2 move. 11...Nxe5 12.Na4! and
White is better.
Black’s second pawn lever is 7...b6. Then 8.b4 is logical, as well as 8.cxb6. The latter transposes to a line in the ...a6
Meran where White fights for the c5-square with Na4 and Rc1. Black can take with either piece on b6, but should
avoid 8...Bd6?! 9.Na4, as I faced in a rapid tournament. White has full control over the desired square after 9...Nxb6
10.Ba5 Bc7 11.Rc1.
A) 6...b6!
This looks like a passive move but it contains a lot of logic. Black doesn’t justify 6.Bd2 with ...d5xc4 and ...b7-b5, but
develops slowly with ...c6-c5 in mind. White’s bishop would then have preferred to be on b2, since the long diagonal
opens up.
It’s hard to stop Black from equalizing with ...dxc4 followed by ...c6-c5.
9.Qe2
9.e4 dxe4 10.Nxe4 c5 doesn’t give White anything, neither now, nor later.
9...0-0 10.Rfd1 dxc4 11.Bxc4 c5 12.Ba6 Qc8 13.Bxb7 Qxb7 14.e4 cxd4 15.Nxd4
White has slightly more active pieces, but Black should be okay.
8.Bxc4 b5 9.Bd3
9...Bb7
9...a6 10.Ne4! Nxe4 11.Bxe4 Bb7 12.Rc1 Rc8 13.b4 illustrates White’s dream.
9...b4 lead to a fairly short draw in Wojtaszek – Motylev, Bundesliga 2016, but White could have posed more problems
with 10.Na4!? since Black is not ready for 10...c5 11.Qc2 cxd4? due to the double threat 12.Qc6!.
10.Rc1
Now ...b5-b4 can be met by Na4 with control over c5.
12.a3
I played 12.Ne2 against Marie Sebag in 2016, but the reason was only to vary. At the same time, 13.b4 is a threat.
However, Black is okay after 12...b4! 13.e4 Be7! (the trick was 13...e5? 14.dxe5 Nxe5 15.Nxe5 Bxe5 and the pawn on
b4 hangs to 16.Bxb4).
12...a5!
The only good way to stop b2-b4, but making it hard to achieve ...c6-c5. The following change in the pawn structure
creates a pawn majority on the kingside.
17.Rfd1
17.f4 does not work: 17...Bd4† 18.Kh1 b4! 19.axb4 axb4 20.Nd1 Nxe4 21.Bxe4 c5 and the bishop on b7 wakes up.
White can hold on to the material, but with a clearly worse position.
17...Qd6!
17...Qe7 18.f3 is also possible.
19...h5!
Brodsky – Kharitonov, Cappelle-la-Grande 2012, ended in a draw after three more moves. The engines insist on an
ugly move.
8.0-0 dxc4
8...e5?! worked well against 6.a3, but with the bishop on d2 White has 9.cxd5 cxd5 10.Nb5 Bb8 11.dxe5 Nxe5
12.Nxe5 Bxe5 13.Bc3 when he has made two good exchanges in the IQP-position.
9.Bxc4 e5
9...a6 is played in the similar positions with Qc2 instead of Bd2. Black is flexible and keeps both ...c6-c5 or ...e6-e5
as options. 10.Rc1 c5 11.Bd3!? plans 11...b6 (or 11...b5) 12.Ne4 Nxe4 13.Bxe4.
10.Qc2!
White stops ...e5-e4; his next moves are Bb3, Rae1 and a2-a3.
10...Qe7
10...exd4 11.exd4 Nb6 12.Bb3 Bg4 13.Ne5 was Smith – Lindgren.
11.h3
Again the first priority is to stop 11...e4 12.Ng5 Bxh2†.
11...Bc7
Moving the bishop plans 12...e4, when the possibility of ...Qd6 at the end of the line changes the evaluation of the
position. 11...e4?! 12.Ng5 Nb6 13.Bb3 Bf5 14.f3 is better for White.
12.Bb3!
A prophylactic move, stepping away from ...Nb6 being a threat.
12...h6
12...e4?! does not work after 13.Ng5 Qd6 14.f4! exf3 15.Nxf3 with e3-e4 coming sooner or later. But if 15...Nb6 had
hit a bishop on c4, Black could have taken on h3 with the bishop.
13.Nh4
A standard move that is often played against ...h7-h6, threatening a fork on g6.
13...Re8 14.Nf5
14.f4 would be strong – if White had had time for Rae1.
16.Nxh6†!? gxh6 17.Qg6† Qg7 18.Bxf7† Kf8 19.Qxg7†! Kxg7 20.Bxe8 Nxe8 21.dxe5 Bxe5 22.f4
Now e4-e5 is coming, but I am not sure how to evaluate this position.
After 10.Qc2! Erik Blomqvist played 10...h6 against me in the 2016 Nordic Championship.
I had not seen the move before, but it’s easy to understand the logic: he stops Nf3-g5 and thus prepares ...Re8
followed by ...e5-e4. And without a queen on e7, White’s Nf3-h4-f5/g6 loses in strength.
After far too long a think, I went for a sequence that I find logical: 11.Rae1 It will soon be clear why the other rook
stays on f1. 11...Re8 12.Bd3 (with ...h6 played, there is less point keeping the bishop pointing towards f7) 12...Qe7
13.dxe5 Nxe5 14.Nxe5
White continues with f2-f4 regardless of how Black takes back. The engine prefers Black slightly, but I disagree.
White initially has some coordination problems to solve, but when the e- and f-pawns start moving, they kick away
everything in their path. The more minor pieces there are on the board, the better it is for White, but it’s still strong after
some exchanges. The Swedish GM Tiger Hillarp even said that he once based his whole repertoire around that pawn
majority.
In the game, I managed to run the f-pawn to f5 and the e-pawn to e6 with a great initiative.
White to move. 25.Ne4 would have attacked the knight on f6 and prepared to take another step with the f-pawn. I
expected Blomqvist to collapse, but instead I did. However, that has little to do with the pawn structure.
Even though a knight manoeuvre to f5 doesn’t attack the queen, White can allow ...e5-e4 and jump with the knight to h4
followed by a quick f2-f3. If you want to do it, prepare with the following moves, in order of importance:
You will probably not have time for all of them before Black plays ...e5-e4. In any case, I think the position is promising
for White, although the variations are very complicated.
...a6 Slav
4...g6 is the Schlechter Slav, where White should play like Peter Heine Nielsen did against Vugar Gashimov.
5.Nc3
Now we have A) 5...Bf5 and B) 5...b5.
A) 5...Bf5
5...Bg4 6.h3 Bxf3 7.Qxf3 b5 is dubious for Black – he should not weaken his light squares when he has exchanged the
light-squared bishop. White can even play 8.cxd5 cxd5 9.e4!.
5...Bg4 6.h3 Bh5? is worse: 7.g4 Bg6 8.Qb3 b5 9.cxd5 cxd5 10.g5 picks up the pawn on d5.
6.cxd5
Another logical move is 6.Qb3, but Black has found ways to play against it, one interesting option being 6...b5
intending 7.cxd5 cxd5 8.a4 b4! 9.Qxb4 Nc6.
6...cxd5 7.Ne5
One idea is to prepare Qb3 under improved circumstances, but it’s not the only purpose of White’s set-up.
There are a number of options:
a) 7...e6? 8.g4 Bg6 9.h4 is the main point behind 7.Ne5. Black has to give up a piece or accept a fatally weakened
structure with ...f7xg6.
b) 7...Nfd7? was played in Smith – Jirka, Pardubice 2009, and now 8.Nxf7! Kxf7 9.Qf3 wins back the piece (or the
king) with a disaster on the light squares.
c) 7...Nbd7 8.Qb3! b5 9.a4 b4 is still an interesting sacrifice. White’s best is 10.Nxd5 Nxe5 11.Nxf6† exf6 12.dxe5
fxe5 13.Bc4 Qf6.
Instead, Black has two main moves: A1) 7...h6 and A2) 7...Nc6!.
A1) 7...h6 8.Qb3 Ra7
A common move among those who believe that the ...a6 Slav gives a fireproof fortress.
11.Bb5!?
Black can’t take the piece due to the double threat against a7 and c7.
14.0-0
14.Na4 has been played with good results. But a few times Black has managed to escape with ...Re8 and ...Ke7-f8
before White has opened up with f2-f3 and e3-e4.
The engines want to save a tempo by staying on c3 with the knight. The next moves could be 15.e4 dxe4 16.Rfe1 with
an initiative.
Part of the concept is also to neglect the h2-pawn after 14.0-0 Qb8 15.f3!.
A2) 7...Nc6!
Taking on c6 does not yield anything. Black will manage to play either ...c6-c5 or ...e6-e5.
8.Bd2(!)
The idea is to play 9.Qb3 Qc7 10.Rc1.
I played 8.Be2 against Dmitry Andreikin in the 2016 Hasselbacken Open, with the belief that it was the line in the
book.
8...Nxe5
Black can take on e5 since the bishop on d2 stops Qd1xd5.
Against 8.Be2 Bologan recommends 8...Rc8 in his DVD about the ...a6 Slav. But here it’s dubious due to 9.Qb3.
8...e6 9.g4! Bg6 10.h4 is still strong, even though Black’s bishop can escape to d3 after 10...Nb4 11.Rc1 Nd7.
However, he loses a pawn after 12.Nxd7 Qxd7 13.Na4 Rc8 14.Nc5 Bxc5 15.dxc5 Nd3† 16.Bxd3 Bxd3 17.Bc3. There
is a double threat against d3 and g7.
9.dxe5 Nd7
A key line is: 9...Ne4 10.Nxe4 Bxe4 11.e6!
Black had problems with his development in Khruschiov – Looshnikov, Chelyabinsk 2008. He never solved the
problem of his king and was practically mated.
12...Bd6
Boris Avrukh’s improvement on Berkes – Balogh, Hungary 2014, in The Classical Slav (2014).
14...Rd8!
14...0-0? 15.g4 leads to a massive attack.
B) 5...b5 6.b3
There are two bishop moves to consider:
6...Bg4
Or 6...Bf5 is met in the usual way: 7.Ne5! e6 8.g4 Bg6 9.h4
The following sequence is also typical: 9...Ne4 10.Nxe4 Bxe4 11.f3 f6 12.fxe4 fxe5 13.cxd5 cxd5 14.Bg2 with a
clear advantage as the position and the light squares open up.
After 6...Bf5 7.Ne5! we also have to consider: 7...h6 8.g4 Bh7 9.Bg2 e6 10.0-0 Bd6 11.f4 (Grischuk played 11.Bb2
against Morozevich) 11...0-0 12.Bd2 White’s next moves are likely to be Qe2 and Rac1, and he can also consider g4-
g5-g6.
7.Bd2(!)
7.h3 Bxf3 8.Qxf3 e5! gives Black counterplay.
8...Bh5 9.g4 Bg6 10.Ne5 Nfd7 11.Nxg6 hxg6 12.Bg2 leaves White with an unopposed light-squared bishop. In The
Chebanenko Slav According to Bologan, the author writes that “White’s advantage is not great, but Black is practically
without counterplay and must await his opponent’s further action.”
That’s fine with us!
9.Qxf3
According to Peter Heine Nielsen, combining ...b7-b5 and ...Bxf3 creates weaknesses on the queenside. There is no
bishop to defend the b5- and c6-squares and White will try to use them by c4xd5 followed by a2-a4.
9...Ba3
Recommended in The Slav, Move by Move (2011).
Black can try to be disruptive with 9...Bb4 but the prophylactic 10.Rc1! allows White to meet 10...Qa5 with 11.Rc2!.
12...Rc8
My suggestion is 12...Bd6 with the idea 13.e4?! dxc4! 14.bxc4 c5 when Black again gets the e5-square. The point
with the bishop on d6 is 15.e5 Bxe5! 16.dxe5 Nxe5 winning back the piece. However, 12...Bd6 was hardly Black’s
idea behind 9...Ba3.
13.e4N
13...e5?!
Better is 13...dxc4 14.bxc4 e5 15.d5 but Black is deprived of the e5-square, while White will use the d5-square.
Black has enough pieces for the queen, but the passer on c6 makes it difficult to use the pieces actively. White is
better.
B2) 7...Nbd7!
Nielsen showed this move at a training camp in Lund, and it was later played by Ruslan Ponomariov. White’s plan is
to hold the blockade on d4, complete development, and finally transfer the dark-squared bishop to b2.
15...Ne4?!
The critical move, but it does not work properly.
If 15...Qe7 16.Be2 Ned7 17.0-0, then Rfd1 and Be1 are the next moves.
17...Qh5
17...Nxg3 18.fxg3 Qxg3† 19.Ke2 does not give enough compensation – Qg2 comes next.
Exercise 1
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 e6 5.Nc3 a6 6.a3 Nbd7 7.b4 Bd6 8.Bb2 0-0 9.Qc2 Qe7
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 e6 5.Nc3 Nbd7 6.Bd2 Bd6 7.Bd3 dxc4 8.Bxc4 b5 9.Bd3 Bb7 10.Rc1 Rc8 11.0-0 0-0
12.a3 a5 13.e4 e5 14.dxe5 Nxe5 15.Nxe5 Bxe5 16.Qe2 Re8 17.Rfd1 Qd6
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 a6 5.Nc3 b5 6.b3 Bg4 7.Bd2 e6 8.h3 Bxf3 9.Qxf3 Bb4
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 4
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 g6 5.Nc3 Bg7 6.Be2 0-0 7.0-0 Bg4
White to move
Show/Hide Solution
Show/Hide Solution
With e2-e3
1.d4 d5 2.Nf3 Nc6 3.c4 Bg4 4.Nc3 e6 (4...Bxf3) 5.cxd5 exd5 6.Qb3 Bxf3 223
6...Bb4 224
7.exf3 Nxd4? 224
a) 7...Qe7† 224
b) 7...Bb4 224
8.Qa4† Nc6 9.Ba6! 224
a) 9...Qb8 224
b) 9...Qc8 224
c) 9...Qe7†! 224
Move Orders
1.Nf3 d5 2.e3 is yummy for Chigorin adherents: e2-e3 is hardly a refutation against their favourite set-up and 2...Nc6!
is a good move.
Fortunately, most players aren’t very flexible and it’s not likely that a Slav enthusiast will put the knight on c6.
2...Nc6 is actually played in only 2% of the games.
It’s often no secret who plays the Chigorin, and it’s possible to wait with e2-e3 against those opponents. Otherwise,
they often disclose themselves with 1.Nf3 Nc6 (to avoid 1...d5 2.c4), when we can go 2.d4 d5 3.c4.
This chapter contains two unusual systems against the Chigorin, with or without e2-e3. One of them contains a
poisonous trap, as the first game shows, and the second is dug up from a World Championship match in 1889.
Finally, 1.c4 avoids the Chigorin, as does 1.e3 d5 2.c4.
Pawn Structures
A relevant question is if 2...Nc6 needs to be refuted. Isn’t the move positionally refuting itself by placing the knight in
front of the c-pawn?
The task lies with Black – he has to justify his knight move by active play.
Structure 1
The pawn structure is great from Black’s perspective. White doesn’t have a natural pawn lever and a minority attack
is difficult to organize with the weak pawn on d4. True, Black’s plan is not any more ambitious than exchanging some
pieces while keeping pressure on d4, but that’s enough.
Thankfully, White only allows the exchange on f3 when he gets a lot of play. Such as Qd1-b3 with threats against b7
and d5, and Bf1-b5 to pin a knight on c6. But he has to be careful to use his initiative before it ebbs away.
Structure 2
In the Ragozin and the Bogo-Indian, White often spends a tempo with Qd1-a4-c2 to force the black knight to c6 (the
check on a4 threatened a bishop on b4).
When Black goes there voluntarily in the Chigorin, he gets the following in return:
a) He can develop the light-squared bishop to g4 before he plays ...e7-e6.
b) Pressure against the d-pawn, and if it’s defended with e2-e3, White’s bishop is stuck on c1. But we are used to that!
The price Black pays is not only that the knight on c6 hinders him from moving the c-pawn, but also that he can’t easily
play ...b7-b6 since the knight would become unstable. White’s long-term strategy is to make use of that by exchanging
on d5, putting a rook on c1 and a knight on c5, hitting the undefended b7-pawn.
If Black defends patiently with ...Ne7 and ...c7-c6, White can start a minority attack with b2-b4.
Instead, Black tries to become active. The bishop goes to b4 and he attempts to push ...e6-e5 (maybe with ...d5xc4
first). If White stops this by exchanging on d5, he uses the e4-square with ...Re7, ...Bxc3 and ...Ne4, followed by an
attack on the kingside.
Games
We start with a miniature that includes a beautiful trap and a premature resignation.
Teschke: “I felt it was too quick and I felt a bit bad winning. It was like I had betrayed my opponent.”
6.Qb3!?
Teschke had learned about this move when he, six years earlier, played Black and fell for the same trap.
6...Bxf3
A tricky and unusual move. Black should probably defend b7 with 6...Bb4.
8...Nc6
9.Ba6!
When Reyer saw the move, he didn’t think long before resigning. 9...Rb8 10.Bxb7! is a disaster and 9...Qc8 10.0-0
gives an attack against the king in the centre.
However, 9...Qe7† 10.Kf1 Qb4 only loses a pawn, so he should have played on. Actually, Teschke later tried this
with the black pieces (!) against the same opponent who took on b7 the first time, but Teschke lost badly.
1–0
Since learning about 9.Ba6! in 2004, I have dreamt about playing it in a game. I have been unsuccessful so far, but the
variation has nevertheless given nice positions. But it’s not possible with an early e2-e3, which leads us to the next
game.
Mikhail Chigorin (1850-1908) learned the rules at 16 but didn’t start to play seriously until his mid 20s. But he quickly
became a chess star in Russia, and in 1889 and 1892 he played two matches for the World Championship against
Wilhelm Steinitz in Havana – with a young Capablanca among the spectators. He lost both, but his understanding was
actually ahead of his time. Steinitz stubbornly followed his positional school, which heavily overvalued material and
weak pawns. Chigorin was the first serious player to break the symmetry, and during those days it was actually 1.d4
Nf6 that was called the Chigorin Defence.
One of the most interesting chess events ever was a telegraph between Steinitz and Chigorin in 1890. They had agreed
to play two positions which they evaluated differently – Steinitz preferred an extra pawn with a passive position. About
one of these he wrote the following (cited from Isaac Lipnitsky): “Six of my pawns are on their starting squares, which
according to my theory is a big advantage, specially in the endgame when it’s good to be able to choose between
moving one or two steps.”
Only a real optimist thinks about the endgame while on the way to getting mated. And Chigorin duly stated that since
he didn’t belong to a school, he didn’t have to think theoretically and abstractly, but could use his energy to look at the
present position.
Steinitz thought that his king could “manage on its own” and the games in the telegraph match were played with slow
time controls (two days per move) to get rid of tactical mistakes. Chigorin won both games brilliantly.
However, Steinitz was a stronger practical player and won both World Championship matches (with 6 of 40 games
drawn). In the first, he beat Chigorin in five of the six games that started with what we call today the Chigorin Defence.
5.Qb3!?
This queen move has, more than a century later, been forgotten.
5...Bxf3 6.gxf3
6...exd4?!
The theoretical section examines the alternatives 6...Bb4† and 6...dxc4.
7.cxd5 Ne5
After 7...Nb4 8.e4 the knight makes no sense.
16.Bh6!
Black has no defence against the heavy pieces lining up on the e-file.
16...b5 17.Bb3 Nb6 18.Rge1 Kd7 19.Bf4
The threat is 20.Bxc7 or 20.Nxb5.
21...Ng8 22.Bxe7 Nxe7 23.Ne4 Rb8 24.Nf6† Kd8 25.Rxe7 Kxe7 26.Qxc7†
26...Nd7
Or 26...Kxf6 27.Qe5 mate.
27.Qxa5
1–0
Recap
In both games, Black played actively, in accordance with the general strategy in the Chigorin. It works well in other
lines, but seems to be too much against our solid set-up, at least in the first game!
Safer are calm options with ...Bb4 and ...e7-e6, respectively.
Theory
We start with the line where White is committed to e2-e3 before moving on to the Ba6 line.
With e2-e3
3...Bg4 4.c4
Now we have A) 4...e6 and B) 4...e5.
A) 4...e6 5.Nc3
This is a position that’s usually reached when Black didn’t have the option of ...e7-e5: after 1.d4 d5 2.Nf3 Nc6 3.c4
Bg4 4.Nc3 e6 (4...e5? 5.cxd5 wins a pawn) 5.e3.
White has no great ambitions; he simply develops with h2-h3, Bd2, c4xd5, Bd3 and 0-0. As usual in the Chigorin,
Black often gives up one or two of his bishops.
Since there are only a few games, a correct move order has not been carved out. The following is a rough guide.
5...Bb4
6.h3
As often, it is clever to force Black to show his hand early on.
6...Bh5
White’s ideal game looks like this: 6...Bxf3 7.Qxf3 Nf6 8.cxd5 exd5 9.Bd3 0-0 10.0-0 Re8 11.Bd2 Bxc3 12.Bxc3
Ne4 13.Rfc1 With the knight on e4, Black wants to take the next step on the kingside, but as we can see, there is not
even a shadow of an attack. Instead he has to take into account that White plans 14.Be1 to keep both bishops. 13...Nxc3
14.Rxc3 Ne7 finally allows him to play the move he neglected early on, ...c7-c6, but only to face the classical minority
attack: 15.b4
7.Be2
Without ...Bxf3 having been played, White should not open the e-file for Black with c4xd5.
7...Nf6
Natural moves fail: 8.0-0 0-0 9.Bd2 and Black makes a typical counterattack in the centre: 9...dxc4! 10.Bxc4 e5 11.d5
e4!
8.Bd2
This is not given in Bronznik’s The Chigorin Defence.
B) 4...e5 5.Qb3!?
Striking against b7 is always critical when Black develops the light-squared bishop before his kingside.
b) 6...Bb4† is the most common move. The idea is to force a white piece to d2 and get a threat after 7.Bd2 Bxd2†
8.Nxd2 exd4 9.cxd5 dxe3. However, practice has shown that the following march isn’t dangerous for White’s king:
10.fxe3 Qh4† 11.Kd1 Nd8
The king has a short path to b1 after 12.Rc1. Instead, it’s Black who will have problems castling after Bb5†.
c) 6...dxc4 is not as stupid as it looks. White develops for free with 7.Bxc4 Qd7 8.dxe5, but the engine considers that
Black has compensation for two-thirds of the pawn after: 8...Na5 9.Bxf7†!? Qxf7 10.Qb5† c6 11.Qxa5 Qxf3 12.Rg1N
The final move deviates from old analysis by Jonathan Tait, given in The Chigorin Defence.
9.Bd2!
White takes control over a5 and prepares to castle long. It’s up to Black to show that he has something. Steinitz
played 9.e4 against Chigorin in the 14th game of the match, but that gives Black the same kind of position as after
8...Na5, but with about a tempo more.
9...b5!
9...dxe3 10.Qxe3! and the pin decides.
4...e6
If 4...Bxf3 then 5.exf3 – generally, we are happy to take with this pawn and speed up development.
a) 7...Bxc3† 8.bxc3! Nxe5 9.dxe5 b6 (9...Ne7!? sacrifices a pawn, which is normal for a Chigorin player. White should
decline with 10.Ba3 and now 10...0-0?? 11.Qb4 wins a piece.) 10.e4!N
I used this novelty in the Rolf Martens Memorial rapid tournament in 2009, and won quite easily. The bishop on f1
becomes active after 10...dxe4 11.Bb5† Bd7 12.Bg5! when Black loses his a8-rook if he takes both bishops.
b) 7...Nxd4! 8.Qd1 Nc6 9.Nxg4 d4 (9...f5!? 10.Ne3 d4 11.Qb3 gives White the bishop pair and a sound structure. The
light-squared bishop may be developed to h3.) 10.a3 Be7 This was a recommendation in Wisnewski’s Play 1...Nc6!
(2007). Black wins back his piece, but after 11.Bd2 dxc3 12.Bxc3 White has the bishop pair for free. 11...h5 12.Ne3 is
similar.
7.exf3
7...Nxd4?
We know this move is bad, so let’s look more closely at the alternatives:
a) Black can defend b7 with 7...Qe7† 8.Be3 0-0-0 since 9.Nxd5 Qd7 wins back the pawn, but 9.Rc1 probably forces
him to exchange queens on b4, when the bishop pair starts to get warm (even though the structure is not yet fluid). In
the same Rolf Martens Memorial, this brought me another victory.
b) Another way is 7...Bb4 8.Bb5 Qd6! as in Smith – Søbjerg, Denmark 2008, but Black hasn’t solved his coordination
problems after 9.0-0 Nge7 10.g3 with Bc1-f4 coming.
c) 9...Qe7†! 10.Kf1 Qb4 11.Qxb4 Nxb4 12.Bxb7 Rd8 and Black has defended the pawn on d5, but will not hold on to
the material. 13.a3! Nc2 14.Bc6†! Ke7 15.Rb1 The initiative continues and White wins material with normal
developing moves.
Exercise 1
1.Nf3 d5 2.e3 Nc6 3.d4 Bg4 4.c4 e6 5.Nc3 Bb4 6.h3 Bh5 7.Bd3 Nf6 8.0-0 0-0 9.cxd5 exd5 10.Bd2 Re8 11.a3 Bxc3
12.Bxc3 Ne4 13.Rc1
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 d5 2.Nf3 Nc6 3.c4 Bg4 4.Nc3 e6 5.cxd5 exd5 6.Qb3 Bxf3 7.exf3 Nxd4? 8.Qa4† Nc6
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.Nf3 d5 2.e3 Nc6 3.d4 Bg4 4.c4 e5 5.Qb3 Bxf3 6.gxf3 Nge7 7.Nc3 exd4 8.Nxd5 Rb8
White to move
Show/Hide Solution
Show/Hide Solution
Leningrad
1.Nf3 f5 (1...d6) 2.c4 Nf6 3.e3(!) g6 4.b4 Bg7 5.Bb2 0-0 6.Be2 d6 7.d4 e6 (7...Ne4; 7...e5!?) 8.Nc3 237
Classical Dutch
1.Nf3 f5 2.c4 Nf6 3.e3(!) e6 4.b3 Be7 (4...b6) 5.Be2 0-0 6.0-0 d6 (6...c5; 6...b6) 7.d4 Ne4 (7...b6) 8.Bb2 b6 9.Nbd2
Bb7 (9...Nxd2) 10.Nxe4 238
1.Nf3 f5 2.c4 Nf6 3.e3(!) e6 4.b3 d5 5.Be2 c6 6.0-0 Bd6 7.Ba3! 238
Stonewall
1.d4 f5 (1...d5 2.Nf3 c6 3.c4 e6 4.e3 f5) 2.c4 Nf6 3.e3(!) e6 239 (3...d6) 4.Bd3 d5 (4...Bb4†) 5.Nc3 c6 6.Nge2 Bd6
7.f3 0-0 8.Qc2!? Kh8! 240 (8...dxc4; 8...Nbd7) 9.Bd2 Qe7 (9...dxc4) 10.0-0 dxc4 (10...Nbd7) 11.Bxc4 e5 (11...b5)
12.dxe5 241
Classical
1.d4 f5 2.c4 Nf6 3.e3 e6 4.Bd3 Be7 (4...b6) 5.Nc3 0-0 6.Nge2 d6 7.0-0 e5 8.f3 241
1.d4, 1.Nf3, 1.c4 or 1.e3 can all be answered by 1...f5.
Move Orders
Black has three main set-ups in the Dutch: the Leningrad (...g7-g6), the Classical (...e7-e6 and ...d7-d6) and the
Stonewall (...d7-d5). Against the last, White doesn’t need a knight on f3 to stop ...e7-e5. Keeping it on g1 gives more
possibilities, as does keeping the e-pawn on its initial square.
1.Nf3 followed by 2.e3 is therefore not the best way to meet the Dutch. However, it’s not necessarily easy to choose
something else against Dutch players. Black can delay ...f7-f5 until White has made a compromising move. One way is
1.Nf3 e6 2.c4 d5 3.e3 c6 4.b3 f5.
If you already have a favourite line against the Dutch, keep it when you know your opponent is a Dutchy. But if you
stumble into the Dutch when following the e3 poison, you may need one of the two set-ups that are examined in this
chapter:
a) Without the knight on f3
b) Without d2-d4
Normally we play our standard moves, and as soon as Black has played ...f7-f5, we see which of these options is still
available. However, there is one way that he can trick us and avoid both: 1.d4 d5 2.Nf3 c6 3.c4 e6 4.e3 f5.
But the probability is not high and if we know that our opponent uses that move order, we can deviate along the way,
for example by 2.c4 c6 3.Nc3 e6 4.e3 and either 4...Nf6 5.Nf3 with a Meran, or 4...f5 5.Bd3 with the line proposed
below.
My preference against the Classical Dutch and the Stonewall is: 1.d4 f5 2.c4 Nf6 3.e3(!) 3...e6 4.Bd3
Both 4...d5 and 4...Be7 are met with Nc3, Nge2, 0-0 and f2-f3.
If Black plays the Leningrad, we go for b2-b4 regardless of move order, a set-up played several time by the Polish
grandmaster Robert Kempinski.
Pawn Structures
Few openings give rise to such strong opinions as the Dutch. After 1.f4 “Black is already slightly better,” said Ulf
Andersson half-seriously. The former World Number 4 points out that moving the f-pawn not only creates a weakness
in the middlegame (opening the a2-g8 diagonal exposes the king) but also in the endgame because of the seventh rank.
But there are players who say that there is no better way to sharpen your attitude than playing for a win from move 1
with an unbalancing move. Hikaru Nakamura and Magnus Carlsen are fighting players who use 1...f5 from time to
time.
Black wants to support his f-pawn with the e- or g-pawn, creating a pawn tandem. But White can put a spoke in the
wheel by:
Fulfilling all those steps would take the sting out of Black’s dynamic play, while opening up for an attack towards h7.
It’s no longer convenient for Black to block the diagonal with ...g7-g6, because that square lacks support from a pawn
on f7.
So the Dutch gives us mixed emotions. We are happy that Black has created a long-term weakness, but at the same
time it’s a move signalling that we won’t have a relaxing journey.
The normal routes for White’s knights are Ng1-f3-e5-d3 and Nb1-d2-f3-e5. Finally f2-f3 kicks away a black knight
from e4, the base of his hoped-for attack.
However, there is a fair chance that Black will have already created a dangerous attack during those seven moves.
Another idea is therefore to play f2-f3 straight away, never allowing the knight to land on e4. The normal piece set-up
is Nc3, Bd3, Bd2 (the bishop would be better on f4, but that’s not possible), Qc2, Nge2 and then probably short
castling. But it’s a good idea to keep the option of castling long.
The downside of this strategy is that White doesn’t control the e5-square. Black can thus push ...e6-e5, but normally he
takes on c4 first to avoid getting a weak d-pawn (another preparatory move is to shift the king away from the a2-g8
diagonal).
But even when the black e-pawn has advanced, it isn’t so dangerous while Black still has his weaknesses.
After having decided where to castle, White can consider c4xd5 when:
Another pawn lever is e3-e4 straight away, but White should be careful since it weakens the d4-pawn.
One argument in favour of White’s pawn structure is that he can add a third pawn with b2-b4, while ...g7-g5 would
weaken the black king. But the opposite argument is valid as well: being first and winning on the queenside is not
worth much if Black breaks through on the kingside a few moves later.
Normal development is Be2, 0-0 and Qb3 (defending the bishop on b2 and looking down towards Black’s king). The
queen’s knight can jump to either d2 or c3.
Black can’t play for a strong knight on e4. His advantage is that ...e6-e5 is possible, but there is usually a strong
antidote with 1...e5 2.cxd5 cxd5 3.d4 e4 4.Ne5, when White nevertheless uses the dark squares. White’s main plan is to
keep and strengthen the control over them. That is done by exchanging the dark-squared bishop (with the pawn on d2,
Ba3 doesn’t run into ...Qa5†) and preferably also both knights. Every exchange makes Black’s possible attack on our
king less dangerous.
Note that c4xd5 is an interesting exchange, just as in the Stonewall with the pawn on d4. It gives up some space (the
pawn on c4 has advanced further than the pawn Black recaptures with), but there is also a downside for Black no matter
how he recaptures.
Games
The first game shows several standard moves in the Stonewall structure where White controls the e4-square with f2-f3
and thus deprives Black of his usual counterplay.
Sokolov, April 2016: “This set-up is definitely an interesting option, not less dangerous (for Black) than the ‘main
fianchetto lines’.”
8...Qc7 9.h3
This seems preferable since it’s hard to see how Black should win on the dark squares, whereas the light squares
could be sensitive after 9.g3. In the latter case, whenever White plays e3-e4 and f3xe4, he will have to look out for
...e6-e5 plus ...Bh3 and ...Ng4.
9...dxc4
White was threatening 10.c5 since the bishop has to leave the diagonal. After 10...Be7 11.b4, the next move is 12.e4.
Black’s king has moved out of the pin and the bishop has fulfilled its job on the a2-g8 diagonal. Now it stops ...e6-e5
by hitting the pawn on f5.
12...Na6 13.a3
The standard reaction against ...Na6. 13.Bd2 was also possible on account of 13...Nb4? 14.Nxb5! winning a pawn.
13...b4 14.Na4
Again the standard reaction. The black knight is not allowed to reach b4.
14...c5
Black has to get rid of his backward pawn before White has time for Rc1.
15.Bd2 cxd4
Black is forced to open the position before he has completed development. It also helps White’s knight to reach d4,
where it keeps an eye on the e6- and f5-pawns.
19.Bb4! was a tactical way to defend the knight on a4. After 19...Bxb4 20.axb4 White’s next move is 21.Nc5, with a
clear advantage on the dark squares.
This short game was instructive, since Sokolov played almost without making any mistakes, except for accepting the
draw offer in the final position. The active knights still gave some advantage, even though Sokolov doesn’t think it’s
much.
In the next game, Kramnik kept the king in the centre one move longer than Sokolov. But it would have been even
better to postpone castling for another move.
1.d4 d5 2.c4 e6 3.Nc3 c6 4.e3 f5 5.Bd3 Nf6 6.Nge2 Bd6 7.Qc2 0-0
8.f3
8.cxd5 cxd5 9.Nb5 is too early. Black has either 9...Bb4† or 9...Nc6 10.Nxd6 Nb4!.
8...Kh8 9.0-0
9.Bd2 is a move White has to play sooner or later anyway, and it keeps the option of castling queenside.
9...a6
Defending against 10.cxd5 cxd5 11.Nb5.
10.c5!?
A decision taken on the basis that Black can no longer play ...b7-b6 followed by ...a7xb6. It was by no means forced
and White could also have played 10.Bd2.
14.bxc5
Now it’s the b-file and the b6-square that will be used. Note that Black’s cramped position makes it difficult to
manoeuvre the f8-rook to the open file.
14...a5 15.Rab1 g6
Or 15...Ba6 16.Bxa6 Rxa6 17.Rb7 and White has penetrated. After Rfb1, the rook’s range could be expanded by Be1-
g3.
21.Nb5! dxe3 22.Bxe3 Nxd3 23.Rd1 White wins back the piece with a decisive advantage.
29...Kg8
After 29...Rd7 30.Nxb7 Rxb7 31.Rfc1! Black can’t defend the eighth rank, since 31...Kg8 32.Rc8† Kf7 33.Rc7†
Rxc7 34.bxc7 Rc2 35.Rc1 queens.
The set-up with f2-f3 is designed against the Stonewall, where the e4-square is vital for Black’s counterplay. But it also
works well against the Classical Dutch, as we will see when changing from the French to the Swedish League.
12.d5!?
The simple 12.b4 was also possible.
12...Ne5 13.Be2
With f2-f4 followed by Bd3 and e3-e4, Grandelius is on the way to achieving the desirable pawn lever.
18.Nxd5 cxd5!
The control over e4 is the only positive thing Black has. However, careful play from Grandelius makes the following
phase one-sided.
19.Rfe1 Nd6 20.Bc3 Rac8 21.Qb3 Qe6 22.Bd4 b6 23.Ba6 Rc7 24.Rac1 Nc4 25.Qd3 Ne8
The alternative was 25...Rff7, which keeps the dark-squared diagonal closed.
30...Rxc1 31.Rxc1
31...Rc7
A nice try to break out, but it doesn’t work.
The engine suggests 31...Re7 32.a4 Nf7 33.Bd4 bxa4 34.bxa4 when the strategically lost position is not tactically lost
yet.
1.d4 f5 2.Nf3
Instead 2.c4 Nf6 3.e3 d6 4.Nf3 g6 5.b4, and so on, is our move order into the game; b2-b4 should not be played
before ...g7-g6.
Kempinski has also placed the knight on d2 in other games. This time he uses it on a4 to reinforce control over the c5-
square.
8...e6 9.Be2
Instead 9.Bd3?! prepares e3-e4, but on the other hand it blocks the queen and steps into ...Nbd7-c5/e5 (when the
pawn on d4 moves).
16...g5 17.Be2
Threatening to trap the proud knight on e4 by f2-f3.
31...Qb4!
White must not be allowed time to pick up e4 for free.
34...Bh3 35.Rb1 Qxa3 36.Qc4 Qd6 37.Qh4 Bf5 38.Re1 Rf8 39.Qg5
Or 39.Rxe3 Qc5 and the pin decides.
39...Bh3 40.Qh4
40.Qxe3 is not possible due to 40...Bh6 followed by a queen check on c5.
40...Qf6 41.Qxf6 Bxf6 42.Ne2 a4 43.Bb2 Rd8 44.Nc3 Be7! 45.Nxa4 Rf8 46.Bxe5 Bb4 47.Nc3 Bxc3
White resigned because of 48.Bxc3 e2 49.Rxe2 Rf1 mate.
0–1
Recap
Even though White lost the last game after a delayed kingside attack, I think it’s safe to conclude that he had quite good
control along the way. But as he managed to get rid of the knight on e4 only one move before it was too late (Bxg4
followed by f2-f3xe4), I nevertheless prefer to keep the f3-square for a pawn.
Dutch players usually keep faith with their system whatever White plays, so it’s your choice.
Theory
Leningrad
1.Nf3 f5
1...d6 is sometimes played with the idea to provoke 2.d4 before Black goes ...f7-f5. But he also has 2...Bg4, an old
favourite of Tony Miles. A Norwegian player recommended 3.Qd3, avoiding doubled f-pawns and creating the sneaky
threat of 4.Qb5†. It was soon played by Magnus Carlsen against Gilberto Milos in a rapid tournament in 2014 in Brazil.
The World Champion followed up with Nbd2, h2-h3 and e2-e4 and I don’t think we need to know more.
After 7...Ne4 8.Nbd2 White wants to take on e4. If Black exchanges on d2, he has lost time but also opened up for his
fianchettoed bishop.
7...e5!? 8.dxe5 Ng4 9.Qb3 Nxe5 10.Nxe5 dxe5 11.c5† Kh8 12.Nc3 gives a position where Black will struggle with his
development.
7...e6
The most common move. In Kempinski – Klekowski, Poland 2015, the inventor of the line put the knight on c3.
Black gets his desired pawn tandem after 12...e5 but White is quick on the queenside with Rac1, Ng5, Nb5 and c4-c5.
Classical Dutch
6...d6
Black can consider other moves:
1...f5 is a move that requires a closed position, and 6...c5 allows it to be opened. I don’t think they fit together, at least
not when the knight on b1 can still go to c3.
6...b6 7.Bb2 Bb7 is a good choice. Black’s bishop is best placed on b7, but can’t usually go there because White
fianchettoes first. 8.d4!? Not a sign that the b1-knight belongs on d2, but to gain the possibility of 8...Ne4 9.Nfd2!?.
7.d4 Ne4
Instead 7...b6 8.Bb2 Bb7 allows 9.d5! when Black would be happy with 9...e5 were it not for 10.Ng5 Bc8 11.Nc3 h6
12.Ne6 Bxe6 13.dxe6 when Black is not in time to consolidate and pick up the pawn: 13...c6 14.Qc2 g6 15.g4! with a
collapse.
8.Bb2 b6 9.Nbd2
9...Bb7
One plausible line is 9...Nxd2 10.Qxd2 Bb7 11.Ne1 Nd7 12.Nd3, with Nf4 followed by d4-d5 on the agenda.
10.Nxe4
White is happy that the bishop is on e2 and not d3.
1.d4 f5
Another move order is:
1...d5 2.Nf3 c6 3.c4 e6 4.e3 f5 This leads to a Stonewall where we cannot play:
a) Bd3 plus Nge2 and f2-f3
b) b2-b3 plus Ba3 without allowing a queen check on a5.
So if you see that your opponent plays the Stonewall, use another move order. 1.Nf3 d5 2.e3 c6 3.c4 e6 4.b3 and 1.d4
d5 2.c4 c6 3.e3 e6 4.Nc3 are two options that give the system proposed in this book. But what to do if the opponent is
clever enough to trick us? One simple option is to insist upon Ba3 after: 5.Bd3 Nf6 6.0-0 Bd6 7.b3 Qe7 8.a4
Black can take control over b4 with 8...a5 but it’s easy to exchange a knight landing on that square. After Bc1-a3,
White makes use of the dark squares, the c-file and possibly the b-file as well (with b3-b4). It’s the usual academic plus
even though I am sure Stonewall players won’t agree.
3...e6
Instead 3...d6, threatening 4...e5 in one move, is best met by 4.Nf3, transposing into Kempinski’s line. A short summary
of what could come next:
a) Leningrad: 4...g6 5.b4 followed by Bb2, Be2 and 0-0.
b) Classical: 4...e6 leads to a rather rare position and the specific order of moves 1-4 that I give to reach this position has
not been used in any of the 8,000,000 games in my database. With a bishop on e7, it’s less tempting to put a pawn on
b4, so White plays 5.Be2 Be7 6.0-0 0-0 7.b3.
4.Bd3 d5
4...Bb4† is an option, since Black is happy with 5.Nc3 Bxc3† 6.bxc3 when he can attack the c-pawns in Nimzo-
Indian style (...c7-c5, ...b7-b6, ...Nc6-a5, ...Ba6 and ...Rc8) while the f-pawn keeps control over the centre (I don’t say
that he is right to be happy, though).
White does better to play 5.Bd2 Bxd2† 6.Qxd2 and the exchange is not logical for Black. He is left with the bishop
that’s restricted by 1...f5.
White keeps the option of castling long, and even if the king usually stays on the kingside, the flexibility is annoying
for Black. White is ready against active plans.
Some rules of thumb: Always play a2-a3 against ...Na6, and answer ...dxc4 plus ...b7-b5-b4 with Na4. And seldom
go for e3-e4.
8...Kh8!
A prophylactic move, stepping out of the pin c4-g8 before playing ...d5xc4.
8...dxc4 9.Bxc4 b5 10.Bb3 and with the threat against the f5-pawn, Black has to move the king anyway. 10...Kh8 11.e4!
Now 11...b4?! was played in Lobron – Sveshnikov, Budapest 1996, but White simply enjoys the dark-squared bishop
after 12.e5!.
11...e5 is an improvement suggested by Jacob Aagaard in Stonewall II (2007). 12.Bg5!? prepares to castle long.
12...exd4 (12...b4 13.Nd5! doubles Black’s f-pawns or gives an extra pawn after 13...cxd5 14.dxe5 Bxe5 15.Bxd5 Nc6
16.Bxc6) 13.Nxd4 b4 (The knight is undefended for the moment, but Black’s dark-squared bishop has no good
discovered attack: 13...Bxh2?? 14.0-0-0 is ridiculous.) 14.Na4! With the king on the queenside, White has to stop ...a7-
a5-a4. 14...fxe4 15.0-0-0 exf3 16.gxf3 A pawn is not so relevant when the players have castled on opposite wings.
White continues with Kb1 and Rhg1 with great compensation.
8...Nbd7 was Aagaard’s favourite move (in 2007), but without any variations. I take one step forward with 9.Bd2. Will
the next move come after another ten years?
9.Bd2 Qe7
9...dxc4 10.Bxc4 e5 can be played immediately, but doesn’t make any difference after 11.0-0.
10.0-0 dxc4
10...Nbd7 can be met by 11.cxd5 Nxd5 12.Nxd5 and the knight on d7 makes it impossible to take back with the e-
pawn. White is better developed when the position opens up after 12...cxd5 13.e4!.
11.Bxc4 e5
The recommendation in Win with the Stonewall Dutch (Sverre Johnsen and Ivar Bern).
After 11...b5 we play 12.Bd3 since 12.Bb3 doesn’t create a threat, as it did with the king on g8.
12.dxe5 Qxe5 13.f4 Qe7 14.Ng3! g6 15.e4
White’s kingside majority is clearly doing better than Black’s queenside majority.
Black’s pawns look better than they are. White plays Qc2 (or a2-a3 first if Black has a knight on c6), b2-b4 and
continues with slow moves such as Rb1 and Bd2. It’s not easy to take the next step, but it’s even more difficult for
Black.
Exercise 1
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 f5 2.c4 Nf6 3.e3 e6 4.Bd3 d5 5.Nc3 c6 6.Nge2 Bd6 7.f3 0-0 8.Qc2 Kh8 9.Bd2 a6 10.c5 Bc7 11.0-0-0 e5 12.dxe5
Bxe5
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 f5 2.c4 Nf6 3.e3 e6 4.Bd3 d5 5.Nc3 c6 6.Nge2 Bd6 7.f3 0-0 8.Qc2 Kh8 9.Bd2 Qe7 10.0-0 dxc4 11.Bxc4 e5
12.dxe5 Qxe5
White to move
Show/Hide Solution
Show/Hide Solution
1.Nf3 (1.e3 d5 2.c4) 1...d5 2.e3(!) Nf6 (2...Bg4) 3.c4 e6 4.b3 250
The Anti-Queen’s Gambit can be reached from the English (1.c4), or with a delayed ...d7-d5: 1.Nf3 Nf6 2.e3(!) 2...e6
3.c4 d5 4.b3. Generally, there aren’t any tricky move orders – just make sure that Black has played the following moves
before you go b2-b3:
1) ...d7-d5
2) ...e7-e6
3) ...Nf6
The last of these moves doesn’t contribute to one of Black’s critical lines: to go for a Tarrasch with ...c7-c5 and ...d5-
d4. The difference can be seen after 1.Nf3 d5 2.c4 e6 3.b3?! c5 4.cxd5 exd5 5.e3 Nc6 6.Bb2 a6! as in Kogan – Gulko,
Nova Gorica 1997.
After 7.Be2 d4 8.exd4 cxd4 White would have wanted to develop with 0-0, Re1 and Na3-c4. But 9.0-0?? fails to
9...d3.
If Black spends time on ...Nf6, he doesn’t get this possibility.
Don’t worry, there is a simple rule of thumb: delay b2-b3 until move four.
Instead, the difficulties arise a few moves later, when White has to decide whether to transpose back to normal lines
with d2-d4. That’s discussed in the theoretical section.
Pawn Structures
The Anti-Queen’s Gambit Accepted was discussed in Kasparov – Comp Fritz 3 in Chapter 9, but only when Black
played ...e7-e5. We will now see that White benefits from delaying d2-d4 even when Black develops normally.
Tomashevsky – Ganguly is actually one of only two games in the book where White plays d2-d3 (Eljanov – Bruzon is
the other).
The Anti-Queen’s Gambit normally gives a position where Black has an isolated pawn. But throughout this book the
main idea is to play with this isolated pawn, not against it! So, as a reassurance, Mista – Tomczak shows a surprising
way to fight against the pawn. We will not reach that position in our repertoire, but I think it’s instructive and says
something about similar positions.
Structure 1
White’s standard pawn lever is d2-d4, steering towards an IQP position or a hanging pawn couple (in case of ...b7-
b6xc5). Generally speaking, Black is okay since:
a) White doesn’t have the optimal set-up to put pressure on the pawn (Bg5, Bg2, Nc3)
b) Black can take on d4, since he played ...c7-c5 before d2-d4. White doesn’t want to take back with any piece:
i) The bishop or the queen would be captured by the knight on c6.
ii) e3xd4 gives an isolated pawn and restricts the bishop on b2.
iii) Nxd4 is best, but the knight would prefer to stay on f3 to defend the kingside.
It’s still unclear, but there is a more sophisticated plan. White delays d2-d4 and starts with Nc3, Rc1 and Na4, when it’s
difficult to defend the pawn on c5. The queen has to keep an eye on the knight on f6 (due to a possible Bb2xf6) and
...b7-b6 leaves the knight on c6 undefended (d2-d4 is the possible problem).
Instead, Black should go for ...d5-d4, giving a position where the evaluation is based on concrete factors: Can Black
push ...d4-d3? Can White win the pawn?
Structure 2
A bishop on c4 would point in three directions, attacking a6 and e6 while defending the d3-pawn. In contrast a bishop
on b7 is doomed to passivity. So strategically, White is much better.
White’s piece set-up is Nbd2-c4 (the knight visits the c4-square before the bishop), Be3/f4, Rc1 and 0-0. The squares
on the queenside are probably not enough, and White often uses the f-pawn as a battering ram. Strategically, it’s a good
choice to open up on the side where Black lacks a minor piece. And f2-f4-f5 can be followed by a march with the g-
pawn.
Black has the d4-square, but can’t make a good use of it; White can always capture a knight that lands there. And ...f7-
f5, to make use of the bishop on b7, is a utopian fantasy. Black won’t get time to move the knight from f6, defend the
pawn on e6 and push the f-pawn.
Games
I met Aleksander Mista when playing for Montpellier in the Top12 event in France 2015. It was quite a different
experience from Sweden. When we arrived at the house the club had rented, there were two go-go dancers in the
Jacuzzi, and during our stay we were served by a cook and a chauffeur. Swedes are not comfortable being waited on,
but the atmosphere was enjoyable, and there was time to discuss some of the games.
When he saw a position with an isolated pawn, Mista said: “Don’t you know that I have found a new way of fighting the
isolated pawn? Exchanging it!”
And then he showed the following game.
1.e4
We can reach the same position via 1.e3 e5 2.d4 exd4 3.exd4 d5 4.Nf3 c5 5.Bb5† Nc6 6.Nbd2?!, but with this move
order, there would be no reason to place the knight passively.
1...e6 2.d4 d5 3.Nd2 c5 4.exd5 exd5
Accepting an isolated pawn when White can’t play Nc3.
8...Nge7
The knight is better on f6, but Black had to stop the check on e1.
11.Bd3 h6 12.h3
Avoiding 12...Bg4, as White doesn’t have 13.Bxh7† as a reply.
12...Nf5
Black wants to develop with ...Bc7, ...Qd6, ...Bd7 and ...Rae8, but starts with the knight move to avoid Be3-c5.
The threat of 16.g4 forced Black to weaken his king’s structure, but he still has a good structure in the centre. Yes –
the isolated pawn is an advantage that takes space without being weak (for the moment). White has to find something
other than just upholding the blockade.
Take it as an exercise with White to move. The solution is counter-intuitive, but if you have accepted that the d5-
pawn is a strength it may be possible to find.
16.Bd2!
16.g4 seems to win a pawn, but Black has: 16...Bb6! (threatening 17...Qg3†) 17.Kg2 Bxf2! 18.Kxf2 Qg3† 19.Ke2
Qxh3
20.gxf5!? (20.Bd2 Ng3† gives sufficient compensation, but not more since White has Ke2-d1-c1) 20...Bxf5 And the
threat against c2 forces White to sacrifice the queen with 21.Qxf5 gxf5 22.Kf2. His position would be winning if he
had time to coordinate his pieces, so Black should continue to play actively with 22...Rae8. The idea is to exchange
both rooks. The engines prefer Black, but I think they underestimate how powerful the pieces become if they work
together.
16.Be3?! invites an interesting exchange, but Black has 16...Ne5! 17.Nxe5 Qxe5 when White has to weaken his king as
well.
16.Qd2!? was Hans Tikkanen’s suggestion. The plan is to take on f5 and h6, but I don’t see a good continuation after
16...h5.
18.Qxc4
Without the shelter on d5, Black’s line-up against h2 falls apart as soon as White has played Rad1. But the
weaknesses around Black’s king remain. The first threat is to take on f5 and h6.
20...Bxf5 21.Rad1
There is no square for Black’s queen! He loses the h6-pawn by force.
21...Rfe8
Or if 21...Qd3 then 22.Qxd3 Bxd3 23.Bxh6 also wins a pawn.
In Mista’s own words: “Firstly I didn’t like my position after 15...g6 at all and couldn’t find any move. And after long
thinking I said to myself that it was not even possible to exchange the d5-pawn for my c3-pawn and equalize! Then I
came up with the idea of 16.Bd2! and realized that I was better.”
Bent Larsen said that it was an eye-opener when he understood that he could not only block the pawn, but also had to
attack it. Now there is a third possibility in that line of thought – exchanging it.
It’s of course still an exception, but the following game is not. Tomashevsky’s play can make it directly into any
textbook about restricting one of the opponent’s bishops and then playing on the wing where it cannot take part.
6...c5 7.Rd1! b5
Or 7...Nc6 8.d4! cxd4 9.exd4 Be7 10.Nc3 and White threatens 11.d5.
12.Nc4 Ng4
On its way to exchanging the knight on c4, but White’s bishop will be no less powerful on that square.
13.g3
13.e4?? Nd4 walks into mate.
16...Rc8 17.Bc4!
17.Bf4 Qf6 18.Rac1 Be7 19.d4! also makes sense, using the development advantage to force through d4-d5. White
would get a dominant position after 19...cxd4?! 20.Rxc8† Bxc8 21.e5 Qf5 22.Bc2 Qh3 23.Be4.
A strong player, who was once a junior world champion in his category, became a Christian and was eager to make
the world a better place. He stopped being a brutal attacker and developed a pacific playing style. When asked if he
really thought this was the meaning of chess, he said: “I don’t think that question ever came up. I just felt it was the
right thing to do.”
Eventually he left the attitude behind and became more of a universal player. However, even if every victory counts
equally, a strategic victory gives more pleasure than a tactical win. And White’s strategic plan in this position isn’t d2-
d4, but f2-f4-f5, opening up on the wing where Black’s light-squared bishop isn’t participating.
17...Be7 18.f4 Qc7 19.Rf1 Bf6
19...0-0 20.f5 threatens f5-f6 and after 20...exf5 21.Rxf5 Black is forced to play ...Kh8 and ...f7-f6 to defend the f-
pawn, leaving his king stalemated.
22.h4!
Black has no counterplay whatsoever.
There is no defence against 26.g4 Bxh4 27.g5, so understandably Black becomes desperate.
25...Rd4 26.Bxd4
Or 26.g4 Rxc4 27.dxc4 Bxe4 28.g5 with an extra piece.
Recap
We can’t copy Mista’s creative idea, but there is a lesson to learn from it. The advantages that the isolated pawn give
can be nullified if we open our minds and exchange the pawn.
In the Anti-Queen’s Gambit, we will see how White again benefits from delaying d2-d4. I labelled 9.Nc3 with the
novelty sign, but then found out that the position had been reached in a correspondence game in 2001. But the game
ended there, with 0–1. I think it deserves better than that.
There is no established theory, but the lines seem quite straightforward and easy to play. So here we go.
Theory
There are two parts: Anti-Queen’s Gambit and Anti-Queen’s Gambit Accepted.
Anti-Queen’s Gambit
1.Nf3
1.e3 d5 2.c4 dxc4 3.Bxc4 was analysed in Kasparov – Comp Fritz 3 in Chapter 9 (page 118).
Black can’t give up the pawn with 6...Nd7? and open the position with 7.Qxb7 c5, as would have been the case with a
pawn on d4.
So Black has to defend with 6...Qc8 and after 7.d4 White answers ...c7-c5 with d4-d5. He develops with Nc3, Be2 and
Bd2, followed by f2-f4 or Ne4 plus f3xe4.
3.c4 e6 4.b3
Black has two options: A) 4...Be7 and B) 4...c5. Against 4...b6 and 4...Bd6 I recommend transposing to the Queen’s
Indian with d2-d4 within a few moves.
Instead after 6...b6 7.cxd5 exd5 8.d4 the bishop looks stupid on b2, but it will wake up when/if Black solves his
problems on the c-file with ...c7-c5. And then we have a position with a hanging pawn couple where Black’s dark-
squared bishop would have been better on d6.
7.cxd5
7...Nxd5
Or 7...exd5 is similar to 4...c5, but with the bishop on e7 instead of d6. It has one advantage: the queen has free sight
to support ...d5-d4-d3. However, after 8.d4! White doesn’t have to worry about the unprotected h2-pawn after Nf3xd4
– Black has no bishop pointing towards h2. 8...Nc6 9.Rc1 avoids 9...cxd4 10.Nxd4. White’s next moves are likely to be
Be2 and 0-0.
8.Nxd5
A few hours after I finished a draft of this chapter, Karjakin played 8.Qc2 Nc6 9.h4 against Viswanathan Anand in
the Candidates Tournament, and won.
8...exd5
If 8...Qxd5 then 9.Bc4 and d2-d4 makes use of the development advantage.
9.d4
A move played by Mikhail Botvinnik in 1943 but only repeated a few times since. It opens up for a check on a5 or b4,
but stops Black from playing ...d5-d4.
9...Nc6
At first I thought Black’s point was to avoid the isolated pawn with: 9...Qa5†N 10.Qd2 Qxd2† 11.Kxd2 b6!
However, after 12.Rc1 White puts on the pressure with d4xc5 plus Bb2-a3 and Nf3-d4. If ...a7-a5, then Bf1-b5 keeps
control over a4.
When preparing for a game, I suddenly realized that 9...Nc6 is much stronger. It stops White from taking back on d4
with the queen and thus threatens to take there followed by a check on b4.
10.dxc5 Bxc5 11.a3
Avoiding the check, but delaying castling allows Black to play with ...d5-d4 ideas. The best way is the following:
a) 14...Bxe2 15.Nxe2 Qxd1† 16.Kxd1 Rfd8† 17.Ke1 may look like enough compensation, but it’s surprisingly difficult
for Black to create anything before White develops the rook on h1. A possible line is 17...a5 18.a4 Rd3 19.Rb1 with
20.Ng3 and 21.Ke2.
b) 14...Re8 15.Bxg4 Bxd4 16.Ra2! To protect the second rank in case Black sacrifices on e3. 16...Bc3† 17.Kf1 Again
it’s not as easy for Black as it looks. White plans g2-g3 and Kg2, but note that 17...Qf6 18.Qe2 is necessary to defend
the e3-pawn.
B) 4...c5
Black makes sure to be first to capture after d2-d4. In addition, he will soon threaten ...d5-d4.
5.Bb2 Nc6
5...dxc4 was played in Korchnoi – Short, Groningen 1996. Short won, but Korchnoi condemned the fifth move after
the game. “You played a match for the World Championship, so you should understand the position better.” 6.bxc4
White continues with Be2, 0-0, Nc3 and Qc2, with a position resembling one in the Anti-Catalan (1.Nf3 Nf6 2.c4 e6
3.g3 d5 4.Bg2 Be7 5.0-0 0-0 6.b3 c5 7.Bb2 Nc6 8.e3 dxc4 9.bxc4).
I think the bishop is better on e2 than on g2, since it stops Black’s main plan of ...b7-b5. After castling short, White
moves the knight from f3 and attacks with f2-f4, or plays d2-d4 under good circumstances.
6.cxd5 exd5
6...Nxd5 7.Nc3 is a semi-Tarrasch where Black must spend a tempo on defending g7 before he can develop the
bishop on f8.
7.Be2(!)
Not 7.d4?! which allows Black to check on b4.
7.Bb5 is the main move. After 7...Bd6 8.d4 cxd4!, Black is happy to divert the knight from f3 – compare with the
Panov. He is even happier that it’s possible to play 9.Nxd4 0-0! Taking the pawn is not recommended:
a) 10.Bxc6 bxc6 11.Nxc6 Qc7 when ...Bb4† or ...Ba6 stops White from castling.
b) 10.Nxc6 bxc6 11.Bxc6 Bg4! with ...Rc8 and a check on the a5-e1 diagonal.
When the Black players found out that they could play like this, 7.Bb5 lost much of its point.
7...Bd6
7...d4 8.Bb5! was Csom – Velimirovic, Titograd 1984. Black can’t hold on to d4 and 8...dxe3 9.Qe2 followed by
10.dxe3 gives a solid edge with well-placed pieces.
7...a6 is interesting. It’s a useful move if White goes for 8.d4, and otherwise Black can play 8...d4 without allowing the
knight on c6 to be pinned. 8.0-0 d4 9.exd4 cxd4 10.Re1 Be7 11.Na3 White plays around the d4-pawn. Useful moves
are Nc4, Rc1 and Nfe5. With less space, White is happy to exchange minor pieces, after which the d-pawn might
become weak, especially if it advances to d3. With good squares for all our minor pieces, White has a good version of
the Benoni.
9...Re8
9...Qe7 10.Rc1 Rd8 is Black’s second set-up, but 11.Nb5 Bb8 12.Ba3 b6 13.d4 gives concrete problems.
12...a6!?
Two other options are:
12...Bg4 13.h3 d3! 14.Bxd3 Bxf3 15.Qxf3 Be5! 16.Nc3 It’s unlikely that someone will find this at the board, partly
because there are a lot of tactics along the way, but also as the compensation is very slow. But Black is okay here.
12...Be5 13.Nxe5 Rxe5 14.a4 and for now, the space compensates for the pair of bishops.
5...c5
White plays in the same way after 5...a6.
6.Qe2 a6
Just as in the normal lines, Black should not delay ...a7-a6. 6...Nc6 7.Rd1! Be7 8.Nc3 0-0 9.d4 cxd4 happened in Tari
– Minguel Xu, Maribor 2012. After the game, Aryan Tari wrote: “I should have taken with the knight. It’s an advantage
to exchange them since I send a rook to c1 and keep control.” 10.Nxd4! Nxd4 11.exd4
With the knights on the board, Black would have ...Na5. Now the only possibility to stop d4-d5 is 11...Nd5 when
White can take twice and win a pawn with opposite-coloured bishops, or follow Gelfand’s 12.Qf3 Nxc3 13.bxc3 with
Bb3 and c3-c4 to come.
7.Rd1!
Keeping the d-pawn flexible.
7...b5
Weaker is: 7...Nc6?! 8.d4 cxd4 (It’s already too late for 8...b5 9.dxc5! Qc7 10.Bd3 Bxc5 when the b-pawn is more of
a weakness without its comrade on the c-file. White continues with 11.a4.) 9.exd4 Be7 10.Nc3 White threatens d4-d5
and after 10...Nb4 11.Ne5 he is happy to have omitted a2-a4.
8.Bb3 Bb7
If Black tries to keep the queenside structure intact with 8...Nc6 9.a4 Rb8, White opens the position with: 10.axb5
axb5 11.Nc3 c4 12.Bc2 Nb4 13.d3!
This is given by Alexander Delchev in The Modern Reti.
9.a4! b4
9...Nbd7 10.axb5 axb5 11.Rxa8 Qxa8 defends the pawn on b5 indirectly. After 12.Na3! b4 13.Nc4 White uses the
same set-up with d2-d3 and e3-e4.
10.d3!
See Tomashevsky’s strategic victory over Ganguly.
Exercise 1
1.Nf3 d5 2.e3 Nf6 3.c4 e6 4.b3 Be7 5.Bb2 0-0 6.Nc3! c5 7.cxd5 exd5 8.Rc1 Nc6
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.Nf3 d5 2.e3 Nf6 3.c4 e6 4.b3 c5 5.Bb2 Nc6 6.cxd5 exd5 7.Be2 Bd6 8.0-0 0-0 9.Nc3 Re8 10.Rc1 d4 11.exd4 cxd4
12.Nb5
Black to move
Show/Hide Solution
Show/Hide Solution
A) 2...Bf5 3.c4 c6 4.Qb3! Qc7 261
4...Qb6 261
4...Qc8 262
5.cxd5 cxd5 6.Nc3 e6 7.Nb5 Qb6 8.Qa4 Nc6 9.Ne5N 262
B) 2...Bg4 3.c4 c6 4.h3 Bh5? 5.cxd5 cxd5 6.Qb3 Qc7 7.Bb5† Nd7 262
7...Nc6 262
8.Bxd7† Qxd7 9.Ne5 262
Move Orders
The position above can be reached via several move orders, and most of them are valid. However, Black can also
postpone ...Nf6 and develop the light-squared bishop on move two or three.
Move two: 1.Nf3 d5 2.e3 Bf5 3.c4 c6 or 2...Bg4 3.c4 c6
Move three: 1.c4 c6 2.Nf3 d5 3.e3 Bf5 or 3...Bg4
As already mentioned, White doesn’t want to allow Black to develop the bishop for free. The b7-pawn is usually
attacked with Qd1-b3. That’s not dangerous in the Slow Slav (1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 Bf5 or 4...Bg4), but
works much better in the lines in this chapter. The reason is that the knight on c3 puts more pressure on Black’s
queenside than the pawn on d4.
Concretely, the difference can be seen in the following two lines after 1.Nf3 d5 2.e3 Nf6 3.c4 c6 4.Nc3:
a) 4...Bf5 5.cxd5 cxd5 6.Qb3 Qb6 7.Nxd5! – winning a pawn.
b) 4...Bg4 5.Qb3 Qb6 6.Ne5 Bf5 7.Qxb6 axb6 8.cxd5 Nxd5 9.Nxd5, with a small structural advantage.
Neither of the bishop moves is common (4% and 1% respectively), so our move order seems to avoid the Slav – or
give White the best possible Slav – hence my claim of Nirvana.
However, it’s a different story with the Semi-Slav (4...e6) and the ...a6 Slav (4...a6). At first, I planned to recommend
the anti-lines that Alexander Delchev gives in The Modern Reti. But I don’t have much to add, so players interested in
avoiding the ...a6 Slav and the Meran altogether can read his splendid book. There is little point in discussing the lines
only briefly, as they are too complex to do more than scratch the surface.
Instead, I recommend 5.d4, which transposes to the Meran (Chapter 13), but still gives positions that fit our repertoire
well.
Pawn Structures
Structure 1
Are the doubled pawns a weakness that can be attacked? No, after playing through a hundred games from different
move orders I didn’t find a single example where Black lost the b6-pawn (but I did see one where Black won White’s
a-pawn).
Instead, White’s advantage is the b5-square. The initiative develops with Nc3-b5 or Bb5† followed by Nf3-e5. The
bishop on f5 can easily run into a threat from one of White’s knights and Black must sometimes allow Nxf5.
(White’s initiative would actually be even stronger if the queen exchange took place on b3 instead of b6, since he
would get the a-file in addition to the b5-square. But note that it’s necessary to have a knight on c3.)
Games
The following game shows one of the ways that Black can be punished when he develops the bishop to f5 or g4. But it’s
not easy and the move order is crucial, as so often.
San Segundo Carrillo, April 2016: “I managed to fly back from Turin to Spain without ID, and got past all the airport
controls. As to the game, I believe I played reasonably well, for a change.”
7.Nf3
Two World Champions have entered this position with Black: Zukertort – Steinitz, USA (5) 1886 (1–0, a nice
attacking game) and Alekhine – Capablanca, New York (12) 1924 (½–½). And Black is in fact quite solid. He has lost
two tempos but reached an Exchange Slav with the bishop on c1 instead of f4.
White chooses between playing on the queenside straight away (Bb5, Bd2, Rfc1, Na4-c5), and reinforcing a knight
on e5 (Nf3-e5, f2-f4). The second option gives the opportunity of attacking the king with Rf3-h3, even though we
should think thrice every time we are on the way to making a rook lift.
7...e6
7...Nc6 8.Ne5 probably transposes.
The main idea behind the rook lift is to double, then play 16.Qc2 and force 16...g6 – due to 16...h6?! 17.Bh7† Kh8
18.Rxf7. It’s also possible to triple on the f-file with Be1 and Qc2-f2. The bishop could, if White is given a free hand,
continue to h6 via g3 and f4.
14...g6
This was not necessary yet, and it allows White to play something other than Qc2.
15.Raf1 Rb8
Instead 15...f5 16.exf6 Rxf6 17.Rxf6 Bxf6 defends against the first wave of the attack, but g6 will be weak as long as
White keeps the queens on the board; Ne2-f4 or Be1-g3 are two interesting manoeuvres, just like in the game.
16.Be1 b5
It’s correct to play aggressively on the queenside, even though the pawn has nothing to come into contact with.
17.Ne2?!
The knight would have been better on d1, to defend against Black’s only counterplay: ...Na5-c4xb2.
17...b4?!
After 17...Na5 18.Qd1 Nc4 White has no intuitive way of defending the b-pawn. However, he can still hope for an
attack if he defends the b-pawn with the queen, plays b2-b3, and finally manoeuvres the queen to the kingside.
20.Bxb5 Rxb5
21.b3
There was a tactical shot 21.Bh4! since 21...Bxh4? 22.Nxe6 Qe7 23.Nxf8 Qxf8 24.Rxf7! Qxf7 25.Rxf7 Kxf7
26.Qf1† wins the loose rook on b5.
21...Rb7 22.Rh3!
There is no defence against 23.Nh5.
22...Qc8
If Black evacuates the seventh rank with 22...Bg5 23.Nh5 gxh5 24.Qxh5 f6 he runs into the slow 25.exf6 Bxf6
26.Bxb4 Rff7 27.Qg4† and whatever Black plays, there follows 28.Qxe6 with a pin. White will then divert the queen
from the defence with 29.Bxa5. For example: 27...Kh8 28.Qxe6 Rbd7 29.Bxa5 Qxa5 30.Qe8† Kg7 31.Rg3† with
mate.
Recap
San Segundo Carrillo’s play was a good example of how to react after 1.Nf3 d5 2.e3 Nf6 3.c4 c6 4.Nc3 Bf5. Black has
three other ways to develop the bishop in the Slav Nirvana.
The key is to understand when White plays h2-h3 (one position), c4xd5 (one position) and Qb3 (two positions). We will
see below which one is which.
Theory
1.Nf3 d5 2.e3
We have three lines to consider: A) 2...Bf5 3.c4 c6, B) 2...Bg4 3.c4 c6 and C) 2...Nf6 3.c4 c6 4.Nc3 Bg4.
A) 2...Bf5 3.c4 c6
4.Qb3! Qc7
White doesn’t have to start with 4.cxd5 since it makes a great difference with an extra pair of knights still on the
board.
4...Qb6 5.cxd5 Qxb3 6.axb3 cxd5 (here’s the difference mentioned above: Black would have taken with the knight if it
was on f6) 7.Nc3 e6
The initiative leads to something concrete after 8.Nb5 Na6 9.Rxa6! bxa6 10.Nc7† Kd7 11.Nxa8 Bd6 12.Bxa6. The
doubled pawns will never queen, but they control more squares than Black’s a-pawn. White develops with Nd4, d2-d3,
Ke2, Bd2 and Rc1 and is better since there is Bb5† as a response to ...Rb8.
9.Ne5N
The idea is 10.Nd4 followed by 11.Bb5.
It’s good to know why 4.Qb3?! is inexact. The problem is not 4...Qb6?! 5.Qxb6 axb6 6.cxd5 which still gives an
advantage for White.
But 4...Qc7! is better. Just as with the bishop on f5, Black’s X-ray threat against the bishop on c1 makes it impossible
to capture twice on d5. 5.Ne5! Be6 6.d4 Nd7 7.Nxd7
Black can play 7...Bxd7 because Nc3 and ...Nf6 have not been included, and he also has 7...dxc4 8.Bxc4 Bxc4
9.Qxc4 Qxd7. With two minor pieces exchanged, Black has no problems despite having less space.
After 4.h3, Black’s best is to capture on f3 and transpose to other lines. Let’s see what happens if he tries to avoid that:
7...Nd7
7...Nc6 8.Qxd5 wins a pawn.
4...Bg4
Black has a range of other options:
d) 4...Bf5 5.cxd5 (5.Qb3 allows 5...Qb6! without winning a pawn) 5...cxd5 (5...Nxd5 gives up the centre) 6.Qb3 Bc8
(6...Qb6 7.Nxd5 is simply a pawn up) 7.d4 was seen in San Segundo Carrillo – Reefat.
5.Qb3
5.h3 is inaccurate in our repertoire, due to 5...Bxf3 6.Qxf3 e6 (6...e5?! weakens the light squares) 7.d4 with a
transposition to the Slow Slav with 4...Bg4, where we prefer to postpone Nc3.
5.cxd5 is met by 5...Bxf3! 6.Qxf3 cxd5, even though White can fight for an advantage here.
7.d4 Nbd7
The only way to avoid a kind of Schlechter Slav (...g7-g6).
If he wants, White can take the bishop pair with either Nc5 or Nb6.
Exercise 1
1.Nf3 d5 2.e3 Nf6 3.c4 c6 4.Nc3 Bf5 5.Qb3 Qb6 6.cxd5 Qxb3 7.axb3
Black to move
Show/Hide Solution
Show/Hide Solution
A) 4...Bg4 5.Bxc4 e6 6.h3 Bh5 7.Nc3 275
A1) 7...Nbd7 8.0-0 Bd6 (8...Bb4) 9.e4 e5 10.g4 Bg6 11.dxe5 Nxe5 12.Nxe5 Bxe5 13.f4! Qd4†(13...Bxc3)
14.Qxd4 Bxd4† 15.Kh2 Bxc3 (15...h5) 16.bxc3 Bxe4 17.g5 Bd5! (17...Nd7) 18.Bd3 Nd7 19.Re1† Kf8
(19...Kd8) 20.Ba3† Kg8 21.Re7 275
A2) 7...Nc6 8.Bb5 Bd6 9.e4 Nd7 10.Be3 0-0 11.0-0 Nb6 12.Be2 f5 13.Ng5!N 276
A3) 7...a6 8.g4! Bg6 9.Ne5 Nbd7 10.Nxg6 hxg6 11.Qf3 c6 (11...Rb8; 11...c5?!; 11...b5) 12.Bd2(!) Bb4!
(12...g5; 12...Bd6) 13.Bb3 276
B) 4...b5 5.a4 b4 6.Bxc4 e6 7.0-0 Be7 8.Nbd2 0-0 277
C) 4...e6 5.Bxc4 c5 (5...a6) 6.0-0 a6 (6...Nc6?!) 7.a4 Nc6 (7...cxd4) 8.Qe2 cxd4 278
C1) 8...Qc7 9.Rd1(!) Bd6 (9...Be7) 10.dxc5 Bxc5 11.b3 0-0 12.Bb2 e5 13.Nc3 e4 (13...Nb4) 14.Ng5 Bg4
15.Nd5! Bxe2 16.Nxc7 Bxd1 (16...Bxc4?) 17.Nxa8 279
C2) 8...cxd4 9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nb4 (11...Nd5) 12.Bg5 Bd7 281
a) 12...Re8 282
b) 12...Nfd5 282
c) 12...h6 282
13.d5 exd5 14.Nxd5 Nbxd5 15.Bxd5 Nxd5 (15...h6?!) 16.Rxd5 Bxg5 17.Nxg5 h6 18.Qd2! hxg5 19.Rxd7
Qf6 (19...Qb6) 20.Rxb7 Rad8 (20...Rfd8; 20...Rab8) 21.Qc2N 283
Move Orders
The concept of the e3 poison is to stay away from the main lines and choose viable systems that keep a lot of play. It’s
hardest after 1.d4 d5. With a pawn in the centre, Black can “punish” a semi-passive move like e2-e3 in other ways than
are possible in the King’s Indian or the Grünfeld.
If our opponent normally plays the Queen’s Gambit Accepted, a good option is 1.Nf3 d5 2.e3 Nf6 3.c4 (delaying d2-
d4), as covered in Chapter 16. But that’s not a choice after 1.d4 d5.
1.d4 d5 2.Nf3 Nf6 3.e3 is possible, but it doesn’t avoid anything. Black can still play the Queen’s Gambit Accepted
with 3...e6 4.c4 dxc4 or 3...Bg4 4.c4 dxc4.
So instead we reach a critical position after six moves: 1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 a6
a) 7.a4 gives a standard IQP-position
b) 7.a3 transposes to Schön – Gerola, a game given in the chapter about the Tarrasch (page 177).
c) 7.b3, as played by Grischuk, tries to reach the same line without losing time on a2-a3-a4, but Black has 7...cxd4! and
then:
i) 8.exd4, but b2-b3 is not a move that should be combined with the isolated pawn.
ii) 8.Qxd4 is not dangerous before Black has weakened the queenside with ...b7-b5.
7.a4 is rewarding to play and study, since the type of IQP positions that arise are common in the e3 poison. The
downside is that it’s mainstream theory, and I was on the verge of cutting the lines from the book.
When in addition there is a long line starting with 13.d5, it felt like I had betrayed the idea behind this book: to avoid
theory in favour of playable positions. But an extra pawn is an extra pawn, and we will see that White gets the usual
academic plus.
Take your time to decide which 7th move you will play, and remember that what we avoid defines us as humans at least
as much as what we choose. If a stranger asks who I am, I would tell her what I don’t do, what I don’t eat and what I
don’t appreciate.
Pawn Structures
Structure 1
We have already discussed IQP positions in the chapters about the Queen’s Indian and the Panov, but with the a-
pawns on their initial squares. So here are only some differences:
a) White’s bishop didn’t get the chance to choose between c4 and d3 – it had to take back a pawn on c4. Nevertheless,
d3 is not an attractive square with a2-a4 played, since a knight on b4 would harass the bishop. It changes White’s
strategy completely. Instead of attacking h7, he plays for d4-d5 or tricks with Ne5xf7.
b) Since Black has secure access to the b4-square, he rejects ...Nbd7 in favour of ...Nc6(-b4). As soon as the knight
leaves c6, White usually goes Ne5.
c) Black seldom develops his bishop to b7.
Also note that Black hasn’t been forced to play ...Nf6xd5. His king is safer, and even though d4-d5 is more common,
it’s still seldom achievable against an opponent who plays well.
Structure 2
A pawn structure that’s hard to play – both sides have advantages. White has chased away the knight from f6 and has
many aggressive moves: Nf3-g5, Nd2-e4, Qg4(-h3), Bd3, Bh6 and h2-h4-h5. The attack would almost play itself if
Black wasn’t quick to challenge the centre. But he is.
The d4-pawn is not that important in itself; we can sometimes allow Black to win a pawn with ...c5xd4. But the
problem is that Black’s knights get good squares: c5, c6 and d5. At least it’s a relief that he hasn’t three knights!
The knights make it hard for White to manoeuvre his pieces, and also threaten to exchange them. That would make
the attack less dangerous and allow Black to make use of the squares White has weakened with e3-e4-e5.
However, the tempo-sensitive nature of the position makes it difficult to use general advice during a game. The
evaluation between two positions that look very similar could be fundamentally different only because of a tempo or a
small detail.
Structure 3
This is a structure where the evaluation depends a lot on the placement of the kings. White wants his on b1, but Black
could play against this by quickly challenging the knight on c3 with ...Bb4 and ...Nd5. For the other pieces, a normal
set-up is Bd2, Bg2 (but the bishop normally has to take back on c4), Rd1 and Qf3.
Black also wants to castle long, but he has to take care of the f7-pawn first (looking out for g4-g5 followed by Qxf7).
A good move is ...g6-g5, which also fixes White’s backward pawn on h3.
Since White has the pair of bishops, he is happy with a slow game, opening the position only when he has coordinated
all his pieces. Black should be more aggressive, look out for opportunities to play ...c6-c5 or ...e6-e5, and put pressure
with ...Bb4 and ...Nd5, as mentioned earlier.
If he doesn’t succeed, White has an advantage when he ultimately plays e3-e4.
The same structure can be reached via the Slow Slav, with 1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 Bg4.
Games
The first game is a standard IQP position with the thematic d4-d5. Then comes one game with 4...Bg4 and one with the
unusual 3...b5.
Oleksienko on grandcoach.com: “It is basically my first game with the isolated pawn.”
A flexible move, waiting for ...Bd7 or ...Nb4 in order to place the queen on e2 (without losing the d4-pawn).
11...Bd7
To choose another diagonal with 11...b6 provokes d4-d5. The best version is 12.Bxf6 Bxf6 13.d5 when one
illustrative line is: 13...exd5 14.Nxd5 Bxb2 15.Rb1 Bf6? 16.Rxb6 and Black is actually dead lost since the rook looks
towards f6.
If White is given a free hand, he would put the knight on e5 and the bishop on b3 before pushing d4-d5. Black’s last
and next moves are a standard manoeuvre to exchange a minor piece, but it doesn’t work well here.
14...Nh5?
Oleksienko gives: 14...Nb4 15.Ne5 Bc6? (15...Nfd5 16.Bxd5 wins a pawn, so Black has to defend d5 with another
piece)
16.Ng6! But it’s too late. 16...fxg6 (16...Re8 17.Nxe7† Qxe7 18.d5 also wins material) 17.Bxe6† Rf7 18.Bxf6 gxf6 The
bishop has to stay on e7 to defend the knight on b4. 19.d5 Bd7 20.Qg4 with a decisive double threat against d7 and g6.
After the text move, it looks like White has to do something about the hanging bishop on h4. But Oleksienko didn’t miss
the chance to create momentum with the isolated pawn.
15.d5!
15.Bxe7 Nxe7 keeps Black prepared against d4-d5.
15...Nf4
15...exd5 16.Bxe7 Nxe7 17.Qxe7 Rxc4 18.Nxd5 doesn’t win any material, but Black is unable to defend against the
pin along the d-file. For example 18...Rc8 19.Qe5 with the double threat of Qxh5 and Ne7xc8.
16.Qe3 exd5
17.Qxf4!
17.Bxe7 Nxe7 18.Qxe7 Rxc4 saves Black: d5 is defended and White doesn’t have time to chase the knight: 19.Qxd8
Rxd8 20.g3 Bg4!
17...dxc4?
After 17...Bxh4 18.Nxd5 Bg5 19.Nxg5 hxg5 White’s superior rooks give a clear advantage wherever the queen
moves. Nevertheless, it was Black’s best chance.
Jankovic, April 2016: “My opponent wasn’t aware of the danger in that line.”
5.Bxc4 e6
6.Nc3
It’s safer to start with 6.h3 Bh5, just in case Black decides to take the knight on f3 later on.
6...Nbd7
6...a6 has the idea of placing the knight on c6, when the pressure against the d4-pawn stops White from playing e3-e4.
8...Bh5 9.e4 e5
10.g4
This is not as brave as it looks – the queens are normally exchanged within a few moves.
a) 15...hxg4 16.fxg6 fxg6 17.Kg2 and Black’s compensation should not be enough. White has a lot of pieces that are
ready to defend the king; Bc1-f4 could be the next free move.
b) 15...Nxg4† 16.Kg2! delays capturing one of the pieces until the next move.
c) 15...Bh7 16.g5 Bxc3 17.Qa4† Qd7 18.Qxd7† Kxd7 19.bxc3 Nxe4 20.Ba3 with a strong initiative for White.
17.g5!
White can’t win a piece: 17.Re1 0-0-0 18.g5 Bd5! and Black saves himself with a lifeline.
17...Bd5
17...Nd7 is analysed in the theoretical section.
18.Re1† Kd7?!
Black had to hide the king on g8, but his position would have been terrible anyway.
19.Bd3
Jankovic: “After 18...Kd7 I was out of book, but it wasn’t that hard. I figured out that I should restrict his pieces and
that his bishop on d5 is holding his position. That’s how I found 19.Bd3 and 20.c4.”
19...Ne8
Or 19...Rhe8 20.Rd1! and White wins a piece.
20.c4
The bishop is trapped if it goes to e6.
20...Bc6 21.Bf5†
But from c6, it can’t block this check.
21...Kd8 22.Bb2
White threatens 23.Rad1† Nd6 and then 24.c5 or 24.Bxg7, both winning. Black tries his best defence, but it’s
insufficient.
22...Ba4 23.c5!
Jankovic: “Another restriction, this time against his knight.”
Romanov and others have also tried 1.d4 d5 2.c4 dxc4 3.e3 b5, but a “refutation” was found by the Norwegian Gunnar
Bue: 4.a4 b4 5.Qf3! c6, when it’s harder for Black to achieve ...c6-c5.
17...Qa7 18.e4!
Finally.
18...Rac8
19.e5
“If you have said A...” Ellinor Frisk wrote in her annotations to the game. But moving the pawn doesn’t only open the
diagonal for the bishop on b1, but for the one on b7 as well.
19...Nd5 20.Qh5?!
20.Qg4 is better.
22.Nc4
If it was White’s move, she would play 23.h4 h5 24.Qg3 with the threat to take on g6. Black’s next move creates
another opportunity.
22...Nd7?
22...Nc3 was still best, even though it’s a pawn sacrifice this time.
23.Nxe6! fxe6 24.Qxe6† Kh8
Going onto the long diagonal shortens the end, but 24...Kf8 25.Bxg6 hxg6 26.Qxg6 is also ‘game over’ due to e5-e6
with Qf7 mate or Bb2-g7†.
Recap
Do you like the isolated pawn, but are frightened by the long lines in the following section? (The essence of the question
is not whether it’s possible to remember them, but whether it’s a bad practical choice.)
Don’t worry, it’s also possible to play on your own – but Black is always a small step closer to so-called equality than
in lines where he doesn’t put a pawn in the centre on Move 1.
And never panic just because White has a static weakness. You don’t have to be Tal to play with an isolated pawn.
Theory
We need a further split. Black can play A1) 7...Nbd7, A2) 7...Nc6 and A3) 7...a6.
A1) 7...Nbd7
Stopping g2-g4 plus Ne5, but d7 is a passive square for the knight.
8.0-0 Bd6
After 8...Bb4 then 9.Be2 is a normal quiet move. White takes good care of the extra bishop.
9.e4
9.Be2 is still okay.
17...Bd5!
17...Nd7 18.Re1 f5 19.gxf6 Nxf6 20.Bd3 Kf7 21.Rxe4! and White’s two bishops dominate the board.
18.Bd3
This is a slight improvement compared to Jankovic – Reutsky. The point of delaying the check is that Black’s knight
doesn’t get access to e8.
12.Be2 f5
When I called Stellan Brynell to ask for advice, he opened Delchev’s book and saw Kinsmann – Brynell from 1998.
“So I am the leading theoretician in the variation,” he joked, without remembering what he concluded after the game.
13.Ng5!N 13...Bxe2
Now, Delchev gives 14.Nxe2 and after 14 more moves, he concludes that Black can hold a rook ending. I am not sure
I agree, but it’s not in the spirit of the book (or of any practical player) to enter such a theoretical discussion. I have
another suggestion.
14.Qxe2 Qd7 15.Nxe6 Qxe6 16.d5 Qd7 17.dxc6 Qxc6 18.exf5 Rxf5
White’s pawn structure is better, since Black always has to worry a little about his king. The engines don’t consider
this to be much and it certainly isn’t, but it is something. Fine.
11...b5 12.Bb3 c5 13.g5 leaves the black knight isolated on h5, or gives White an initiative after 13...Ng8 14.d5 as in
Matlakov – Hasangatin, Loo 2013.
12.Bd2(!)
In the long run, White wants to push the e-pawn one step and open the position for the bishops, but he should not
hurry. Advancing too quickly plays into the hands of Black’s knights. Long castling is the first priority, but Black will
try to hinder that with aggressive moves. Let’s see some illustrative lines:
12...Bb4!
12...g5 is a decent positional move, but the first priority should be to fight against White’s castling long.
12...Bd6 13.0-0-0 Qe7 With f7 defended, Black prepares long castling, but if he is too quick he is in for a surprise:
14.Kb1 0-0-0? 15.Bxa6! bxa6?! 16.Qxc6† Kb8 17.Qxa6 White’s attack is not so fast, but the problem is that Black is
unable to untangle. He is completely lost. 17...Nf8 tries to defend with the queen, but after 18.Qb6† she has to stay on
e7 to defend the rook on d8.
So Black has to start with 17...Rhe8 but 18.g5 Nh7 19.Rc1 Ndf8 20.Nb5 Qb7 21.Qa4 leaves him defenceless. The
next move is Ba5.
After 12...Bb4! then 13.0-0-0 Nd5 either weakens White’s king position or exchanges the bishop pair. Black has
counterplay after 14.Ne4 N7b6 15.Bb3 Bxd2† 16.Rxd2 a5. So White has to start with a prophylactic move.
13.Bb3
If Black does nothing, White can play 14.0-0-0 Nd5 15.Ne4 Bxd2† 16.Rxd2 without the tempo-gaining ...N7b6. A
nice trick appears after 16...Qe7 17.h4 0-0-0? 18.Qxf7! Qxf7 19.Nd6† winning.
B) 4...b5
When this move first appeared in practice, some players thought that it was a fingerfehler. Black spends a tempo and
weakens squares on the queenside without the intention of keeping the extra pawn. The idea is to control the c3-square,
but White’s knight is not unhappy with Nb1-d2-b3 or Nb1-d2-c4.
I think Black is doing okay, but he’s still the one who has to equalize. It’s not important to remember concrete moves.
9.b3 Nbd7 10.Bb2 c5 11.Qe2 a6 transposes to Frisk – Percivaldi.
C) 4...e6 5.Bxc4 c5
5...a6 6.a4 b6 is another way to play. If White develops his knight to c3, the game could transpose to a Vienna, but there
is also an independent option: 7.0-0 Bb7 8.Qe2!? Nbd7 9.Rd1 By delaying Nc3, White tries to avoid ...Bb4. The point
is 9...Bb4 10.Bd2 as in Caruana – Nakamura, Moscow 2012. Nakamura retreated with 10...Bd6 but the value of getting
the bishop to d2 is questionable. In the game it ended up on e3 anyway.
6.0-0
6...a6
6...Nc6?! is inexact. The advantage of keeping the pawn on a7 is that ...b7-b6 doesn’t create a weakness, but that
move makes more sense with the knight on d7. By delaying ...a7-a6, White gets time to react in a way other than a2-a4.
7.Qe2 cxd4 (7...a6 8.Rd1! cxd4 9.exd4 b5 10.d5 and White wins) 8.Rd1 Be7 9.exd4 0-0 10.Nc3 a6 11.d5 with a huge
initiative.
Quite a few games have seen 10...Na5 11.Bd3 b6 and both Boris Avrukh and Alexei Kornev give the still uncommon
12.Qe5!? in order to transfer the queen to g3 and slowly build up an attack.
7.a4 Nc6
7...cxd4 stops Qe2 but allows Bg5 straight away. 8.exd4 Nc6 9.Nc3 Be7 10.Bg5 0-0 11.Re1
a) 11...b6 12.d5! is a standard break. Take your time to understand why it’s strong. The payback will come sooner
than you suspect. 12...Nxd5 (12...exd5 transposes) 13.Nxd5 and now:
a1) 13...Bxg5 14.Nxg5 Qxg5 (14...exd5? 15.Nxf7 Rxf7 16.Qxd5! and Black can’t defend c6, f7 and e8) 15.Nxb6 Ra7
Defending the pawn on a6, but White wins another with 16.Nxc8 Rxc8 17.Rxe6!.
a2) 13...exd5 14.Qxd5! Black has to take on g5 to defend against both threats (Qxc6 and Bxe7). 14...Bxg5 15.Nxg5
Qxd5 16.Bxd5 Bd7 For the moment, Black has everything defended and if he gets the extra move ...h7-h6, he would be
fine. However, White increases the pressure with 17.Rac1 Rac8 18.Red1 Be8 and Black is more than a move away
from equality.
Instead of 18...Be8, if 18...Ne5 then 19.Rxc8 wins a pawn by challenging the knight on e5 next move.
When Black has developed the queenside, it’s time for him to move a knight. But there is no good move. 13...Nb4
14.d5! exd5 15.Qxe7 Rxc4 16.Bxf6 gxf6 17.Qxd8 Rxd8 18.Nxd5 Nxd5 19.Rxd5 gives White some advantage, as does
13...Nd5 14.Bxd5! Bxg5 15.Be4 followed by d4-d5.
8.Qe2
When Bc1-g5 isn’t possible, White puts a rook on d1 to force a reaction in the centre.
C1) 8...Qc7
Insisting on avoiding ...c5xd4.
9.Rd1(!)
Keeping the option of either Nc3 or Nbd2.
9...Bd6
9...Be7 10.dxc5 0-0 has the idea of pushing ...e5 before taking back on c5, which forces White to spend a tempo on
h2-h3 to avoid ...Bg4. The line goes: 11.b3 e5! 12.h3 e4 13.Nd4 Ne5 14.Qc2N An old novelty of mine, but I have not
been close to having the opportunity to use it, and now I doubt that it’s better than other moves. The point is to force
Black to take on c4 immediately. 14...Nxc4 15.Qxc4 Qxc5 16.Qxc5 Bxc5 17.Ba3 and White plays for the b6-square.
We get a position where White has a better queen and a better bishop, but he can’t play e3-e4 since the rook has left
the defence of the pawn on f2. Normally White plays h2-h3 to stop ...Ng4, but Vladimir Kramnik found out that it’s not
necessary.
12...e5
Delchev doesn’t mention 9.Rd1, but in a similar position he gives 12...Na5 13.Nbd2 and feels that Black’s position is
unpleasant.
Instead, again in the similar position, he suggests 12...Bd7, but in our position I see no reason to avoid the more active
diagonal on b7.
However, 12...b6 also accepts a slightly worse position. White develops with 13.Nbd2 and considers taking on f6.
13.Nc3
The problem in omitting h2-h3 is that Black can play ...e5-e4 followed by ...Bg4, but as we will see, this is not a big
problem.
13...e4
A variation that only the engine can come up with is: 13...Nb4 14.Na2 e4 (14...Nxa2? 15.Bxe5! wins a pawn)
15.Nxb4 exf3 (15...Bxb4 16.Ne5! with an initiative) 16.Qd2 Ng4 (16...fxg2 17.Nd5 Nxd5 18.Qxd5 with strong
bishops and a deadly threat in Qg5) 17.g3
It seems that Black can’t get access to g2 or h2. The straightforward try is 17...Qb6 18.Nd3 Qh6 19.h4 Ba7 20.Nf4
but it’s impossible to challenge the pawn on h4.
14.Ng5 Bg4
15.Nd5!
The tactical point of Kramnik’s play.
17.Nxa8 Bc2!
Defending the pawn.
18.Rc1 Nb4 19.Ba3 Nd3
20.Bxf7†!
This move is not mentioned in Ftacnik’s and Huzman’s analysis.
I prefer White even though it’s not easy to advance with the b-pawn. The first goal is to exchange the knight.
Normal is 13.Ne5 Nxc3 14.bxc3 Nxe5 15.dxe5 Qc7 16.Rd3 but such rook moves are often either good or dubious,
not neutral, and here I am sceptical.
Instead Kramnik played 13.h4 against Kasparov in their London 2000 match. He came close to winning and Alexander
Khalifman recommended the move in Opening for White according to Kramnik.
a) Accepting the sacrifice is not recommended: 13...Nxc3 14.bxc3 Bxh4 15.Nxh4 Qxh4 16.Rd3! with a dangerous
initiative.
b) Kasparov played 13...Ncb4 14.h5 b6 15.Ne5 Bb7 16.a5 b5 17.h6 g6. The engines recommend 18.Bd2N to be able to
keep control over c2 with Rac1, while the bishop keeps an X-ray gaze on the a5-pawn. This move deviates from
Starting Out: Queen’s Gambit Accepted (2007).
c) Delchev recommends 13...h6!, a set-up that’s only been tested in correspondence games. 14.Bc2 Qc7 (Delchev’s
move) 15.Ne5 leads to a position with several attacking possibilities: Ng4 followed by a sacrifice on h6, Qd3, taking
on f7.
12.Bg5
The same position exists in the Bogoljubow Slav (that’s the trendy Slav line with 4.Nc3 dxc4 5.a4 e6 6.e3 c5) but
with the a-pawn on a7. We will see that it makes a few differences.
12...Bd7
There are other options:
a) 12...Re8 13.Ne5 Nfd5 14.Bxe7 Rxe7 15.Qf3 is a standard position that is slightly better for White. The exchange
Bc4xd5 ...exd5 gives Black a structure with a pawn fixed on the same colour as his bishop.
b) 12...Nfd5 gives the same type of positions as 12...Re8, but with Black losing a tempo since he has to take on e7 with
the knight (while returning to d5 anyway).
c) 12...h6 avoids the sequence White plays against 12...Bd7, but it gives away a tempo. 13.Bxf6 Bxf6 14.Ne4 b6
(14...Be7 15.Ne5 and with two strong knights, White will create an attack with Ra3-g3) 15.Ne5 and:
c1) 15...Bb7 loses an exchange, but White has to know just a few more moves: 16.Nxf6† Qxf6 17.Nd7
And now 17...Qg6 18.d5! with a double threat: if Black moves the rook from f8, White plays 19.Ne5 Qg5 20.Nf3
followed by dxe6.
But after 17...Qg5, 18.d5? exd5 19.Nxf8 doesn’t threaten the queen, so Black would win with 19...dxc4!. Instead,
White has 18.f4! Qxf4 19.Nxf8 Rxf8 20.Rac1 and Black will have a hard time holding a draw.
c2) 15...Ra7 is an extra possibility with the pawn on a6. The rook goes to c7 to threaten the bishop and stop White from
mobilizing an attack. But after 16.a5 b5 17.Nxf6† Qxf6 18.Qd2 Qe7 19.Bb3 White will take control over the c-file.
c3) 15...Be7 16.Ra3 Bb7 17.Qg4 is possible when the bishop no longer defends c8; h6 is a weakness when the rook is
swung to the kingside.
17...Nd5 This move is Black’s best resource with the pawn on a7. But now 18.Qg4 Bg5 19.Nxg5 Qxg5 20.Qxg5 hxg5
21.Bxd5 Bxd5 22.Nd7 wins a pawn.
17...Qe7 is not played in the position with the pawn on a7, but it seems to be a move that holds everything together after
18.Qg4 Bxe4!.
12...Bd7 is the start of a long line that gives an extra pawn for White.
19...Qf6
19...Qb6 defends the pawn for the moment, but not for long after 20.a5!.
20.Rxb7
Black is happy to have the a-pawn on a6 instead of a7, but he is still a pawn down.
In the World Championship game Botvinnik – Petrosian, Moscow (10) 1963, White got a rook ending with 3 versus 2
that was easy to hold. In two other top games, White got a 3–2 rook ending with both rooks on the board, with some
chances. But both were drawn.
20...Rad8
20...Rfd8 is probably an improvement, but White still keeps the pawn.
20...Rab8 was played by Gashimov versus Topalov, with the pawn on a7. It doesn’t make any difference and White is
slightly better after 21.Rxb8 Rxb8 22.Rb1 Qe6 23.h3.
21.Qc2N
I suggest this as an improvement over 21.Qc3, to avoid getting the theoretical rook endings.
21...Rc8 22.Qb3
White plans Rb7-b6. The passive Rf1 is also a useful move.
Exercise 1
1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4 Nc6 8.Qe2 cxd4 9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nb4
12.Bg5 Bd7
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4 cxd4 8.exd4 Nc6 9.Nc3 Be7
White to move
Show/Hide Solution
Show/Hide Solution
1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 Nc6 7.Qe2 cxd4 8.Rd1 Be7 9.exd4 0-0 10.Nc3
Exercise 3
Black to move
Show/Hide Solution
Show/Hide Solution
A) 6...b6 7.cxd5(!) exd5 8.Nc3 295
The tricky thing with the move orders in the Queen’s Gambit is not how to reach the position above, but that it easily
transposes to other lines:
Pawn Structures
Our system with an early e2-e3 isn’t showing its best face against the Queen’s Gambit. Black develops with ...Be7 and
goes for a kind of delayed Tarrasch with ...c7-c5.
d4xc5 can often be met by ...d5xc4!, with a symmetrical position that should be okay for Black. But if he starts with
...Nbd7 (as many do), we get the kind of slow but rich positions we like.
White’s alternative is to go for a hanging pawn couple where Black has arranged his pieces in the ideal way (the
knight on c6!). The position is unclear, but Black’s pressure against the centre makes it difficult for White to get the
set-up he desires.
Structure 1
The bishop would be passive on b2 in a position with an isolated pawn. But why does it help that White has kept a
pawn on b3?
a) If Black exchanges on c4, it’s easier for White to push d4-d5 and open the diagonal.
b) If Black doesn’t take, White can go for a queenside majority with c4-c5, when the bishop controls the e5-square and
keeps Black’s majority at bay.
Normal development is Qe2, Bd3, Nc3, Re1 and Rc1. A common plan is then Nf3-e5 and f2-f4, making use of the
semi-open file. Black can’t play for the e4-square in the same way. He develops with ...Nf6, ...Bb7, ...Be7 (keeping the
queen’s sight open), ...0-0 and ...Rc8.
The placement of the queenside knight is often a factor that determines whether he gets pressure against White’s
hanging pawn couple. The knight wants to go to c6, attacking d4 and being able to disturb with ...Nb4 or ...Na5.
Normally, ...Nc6 can be punished with c4xd5, but that’s not the case after the slow move b2-b3, as we will see.
However, Black’s pressure doesn’t win anything, but merely stops White from getting the activity he would have
wanted.
Structure 2
In a symmetrical position without pawn levers, the pieces have to do the job. With queens on the board, a normal plan
is Qe2 and Ba6 to exchange Black’s bishop (in case he has fianchettoed). But it’s also possible to restrict the bishop
with e3-e4 and f2-f3
Without queens, White hopes that the extra tempo makes his knights faster with one of the following manoeuvres:
Games
Just before the following game, I visited Pal Benko’s place in Budapest. As mentioned in Chapter 7, the Benko Gambit
was a reaction to the boredom he felt when studying theory. His score with Black was, however, not that great.
Avoiding theory is easier when we have an extra tempo.
In the Candidates tournament in Curaçao 1962 he opened his first eleven games with 1.g3, beating Bobby Fischer and
Mikhail Tal along the way. And inspired by his arguments for such an attitude, I went to the playing hall and opened
with 1.e3 – following the e3 poison for the first time.
9.Nc3
9...Ne4
9...c5 10.cxd5! The standard reply to ...c7-c5. Black has three options:
a) 10...cxd4?! helps White’s knight into the centre.
b) 10...exd5 11.Rc1 with a position from the Queen’s Indian, but there Black’s bishop was on d6. It costs him a tempo,
while White develops with Qe2, Rfd1, Bf5 and Ne5.
c) 10...Nxd5 11.Nxd5 and Black would have wanted to take back with the queen – one of the reasons to go ...c7-c5
before developing the knight to d7. 11...Bxd5 White has two good bishops and many moves, like Rc1 or Qe2 (with the
intention e3-e4). It looks pleasant, although Solozhenkin (2001) thought that Black was okay.
11.d5 exd5 12.cxd5 Nxd5 13.Nxd5 Bxd5 14.Bxh7† Kxh7 15.Qxd5 White’s safer king and the slight lead in
development is enough to disturb Black. He can neutralize the bishop with 15...Bf6, but White has 16.Qf5† g6 17.Qg4!
Bxb2 18.Rad1 with a strong rook on d7 that’s supported with Rfd1 and Ng5(-e6). After 18...Bd4 19.exd4 c4 20.Ne5!,
White gets a strong initiative.
10.Qc2 f5
10...Ndf6 11.Ne2 and the knight on e4 gets kicked away with Ne5 and f2-f3.
11.Ne2
White normally exchanges first with 11.cxd5 exd5 and only then plays 12.Ne2, but that allows Black’s heavy pieces
to fight for the e5-square from e7 and e8.
11...Bd6 12.Ne5
There are no other constructive moves.
12...Bxe5
Black acts before White plays Nf4 or f2-f3.
12...c5!? 13.Nf4 Qe7 14.cxd5 exd5 15.Nxd7 Qxd7 16.dxc5 and despite the passive bishop on b7, Black should accept
an isolated pawn. Instead 16...bxc5?! 17.f3 Ng5 18.h4! wins a pawn.
13.dxe5 Nec5
13...Ndc5 keeps the strong knight on e4 for the moment, but the d5- and e6-pawns nevertheless are under pressure
after 14.Nf4.
14.Rfd1!
I wanted to play 14.cxd5 until I saw 14...Nxe5! 15.Bxe5? Qxd5 winning a pawn.
14...Qe7?
Black lost attention for a moment. Better was 14...Nxd3 followed by 15...c6, taking back on d5 with the c-pawn.
White gets the c-file, but it’s difficult to penetrate.
15.cxd5 Bxd5?!
15...exd5 16.Bxf5 was what Mindlin had forgotten.
15...Nxd3! 16.Rxd3 Bxd5 and the c7-pawn is tactically defended due to ...Rc8-c2. So Black gets time to defend with
...c7-c6, when he’s only slightly worse.
16.Nf4 Bb7?!
A set-up with 16...c6 was still the best of the bad options.
17.Bc4
Threatening 18.b4, winning the e6-pawn.
17...a5
17...g5 also defends against the threat, but 18.Nh5 and 19.Nf6† gives Black a strategically lost position on the dark
squares.
18.Ba3!
The e6-pawn is lost.
20...Kh8 21.Bd5?!
A decent practical decision, considering that White is still much better, with a position that’s easy to play.
21.Bxf5 was my first intention, but it seemed unnecessary to allow 21...Bxg2! 22.Nxg2 Nf3† 23.Kf1 Nxh2†, even
though it’s probably good for White.
However, 21.Rac1 was more straightforward – White doesn’t have to worry about ...g7-g5 since the pin along the long
diagonal saves the piece.
21...c6 22.Bc4 a4
Many players sacrifice such flank pawns without understanding that the extra pawn is valuable. But Mindlin lacked
good alternatives.
23.bxa4 Ba6 24.Bxa6 Rxa6
Black threatens 25...c4 with a semi-blockade.
29.g3!
Stopping the last hope of counterplay. It turns out that ...f7-f5 is a weakness even in the ending – remember Ulf
Andersson’s bold claim about the Dutch.
29...g6 30.Rxc5 Rb2 31.Qc4 Rc8 32.a5 Kg7 33.a4 Kh6 34.Rxc6 Rxc6 35.Qxc6 Qxc6 36.Rxc6 Rb4 37.a6 Rxa4 38.Kg2
Kh5 39.Rc7 h6 40.a7 g5 41.h3
1–0
A strong player who read the manuscript concluded that if White opens with 1.d4, the Queen’s Gambit may be the e3
poison’s weak spot. One example is the following game, where Black is allowed to arrange his pieces in the way he
wants. But it’s still unclear.
As discussed in the chapter about the Queen’s Indian, 6.b3 avoids 6...dxc4 7.Bxc4, but allows 6...Bb4†.
6.0-0 Be7
7.b3?!
Our repertoire continuation is 7.Nc3 cxd4 (7...d5 8.cxd5! with a check on b5) 8.exd4 d5 and the bishop can develop
to a better square than b2.
10...Nc6!
With pressure against d4, it’s harder for White to arrange his pieces in the ideal way.
11.Rc1
Normally Black isn’t in time for ...Bb7 and ...Nc6 without being punished by 11.cxd5 Nxd5 12.Nxd5 Qxd5 and Be4.
But even if White replaces Bb2 with Re1, he isn’t in time to disturb Black. That’s a downside with a slow move like
b2-b3.
11...Rc8 12.Re1
White’s best set-up is Qe2 plus Rfd1, directing the rook against Black’s queen and supporting Ne5 without having to
worry about the d4-pawn. 12.Qe2? dxc4 13.bxc4 Nxd4?? 14.Nxd4 Qxd4 loses a piece after 15.Nd5 Qc5 16.Bxf6!
Bxf6 17.Qe4!.
However, Black’s knight on c6 shows its strength with 12...Nb4! 13.Bb1 dxc4 14.bxc4 Bxf3 when White has to play
15.gxf3 to avoid losing the pawn on c4.
12.Re1 vacates the f1-square, where the bishop keeps defending the pawn on c4.
12...Nb4 13.Bf1
I don’t think White is better here, but this game shows that it’s not an innocent set-up.
13...Re8
Kalle Kiik’s move 13...Nc6 is not bad, even though one of the plans with ...Nb4 was to force a2-a3, weakening the
b3-square.
13...Ne4 is recommended in Opening for Black according to Karpov. 14.a3 Nxc3 15.Rxc3 Nc6 16.c5! deviates from
that book and creates the queenside majority that’s well-known in the Panov.
14.Ne5 Bf8
14...Nc6 is again interesting. The point is 15.cxd5 Nxe5 16.dxe5 Nxd5 when White can’t play the desired 17.Ne4?!
due to 17...Bb4!, a move available only before White has played a2-a3.
Black wants to play 16...Nxd5 17.Nxd5 Qxd5 18.Bc4 but the tactical point is that 18...Qd8 (18...Qd6 doesn’t save
Black: 19.Nxf7! Kxf7 20.d5 Na5 21.dxe6† Kg8 22.e7† Nxc4 23.exf8=Q† with an extra pawn) 19.Nxf7! wins.
19...Kxf7 20.Qh5†! Kg8 21.Rxe6 Rxe6 22.Bxe6† Kh8 23.Bf5! and the dark-squared bishop triumphs after 23...g6
24.Bxg6 Rc7 25.d5†.
16...Nxe5 17.dxe5 Nxd5 18.Ne4 is promising for White. Black has to watch out for both Nd6 and Ng5.
17.g3
Best is 17.Bb5! when Black’s rook is harassed after 17...Re6 18.Ne2!.
17...Nxe5 18.dxe5?!
18.Rxe5 would have kept the position even.
18...d4!
White is not in time to defend the long diagonal properly.
Black has pressure against c3 and f3. It seems like White can keep his position together with exact moves; but he will
always be worse.
24.Be2 Re8 25.Qd2 Qe5 26.Rc2 Qf5
Black threatens to take on e2, f3 and c2, in that order.
28...Qe6 29.Bd3
A double threat, winning due to 29.Bc4 Bxf3†! 30.Kxf3 Qe4 mate.
29...Qxb3
0–1
If the hanging pawn position doesn’t suit your appetite, White also has the option of going for symmetrical positions
where he still has the extra tempo. One example is 1.d4 Nf6 2.Nf3 e6 3.c4 b6 4.e3 Bb7 5.Bd3 d5 6.b3 Be7 7.0-0 0-0
8.Bb2 c5 where:
a) 9.Nc3 cxd4 10.exd4 transposes to the game
b) 9.dxc5!? dxc4! 10.Bxc4 Bxc5 is the alternative
Recap
None of the games contained the dull positions that White risks getting in the Queen’s Gambit, but it’s not a good idea
to keep the eyes closed forever. So here they are.
Theory
An idea that’s stupid, but that I may give a try one day, is 5.cxd5 exd5 6.Bd3
I didn’t take a similar set-up seriously when I faced it in 2007, neither during preparation, nor in the game. And it
looks like Black’s set-up with ...Re8, ...Bd6, ...Nf6-e4 gives a promising attacking position.
That set-up is not possible in the Exchange variation of the Queen’s Gambit, with a bishop on g5. It’s worse on c1,
but White can and should play for e3-e4 with Nc3 and Re1. It gives an IQP position with a pair of knights exchanged,
which favours Black. But still I think that White’s chances are underestimated. He is not worse, and against an
unsuspecting opponent...
However, I am sure that Black has a move order, probably something with ...Bg4, that puts an end to the optimism.
5...0-0
White develops with Bd3, 0-0 and Bb2, but in which order? The question is how to answer ...c7-c5. There are
basically two alternatives:
1) Continue to develop and reach a position with a hanging pawn couple
2) Take on c5
6.Bb2
The only move that allows Plan 2.
A) 6...b6 7.cxd5(!)
7.Bd3 Bb7 8.0-0 c5! The bishop on d3 is the only thing that breaks the symmetry. It’s on a better diagonal than Black’s,
but nevertheless it allows Black to equalize. (8...Nbd7 was Smith – Mindlin.)
Now 9.Nc3 transposes to 6...c5 7.Bd3 and keeps the tension. There is a problem with 9.dxc5:
a) 9...bxc5?! 10.Nc3 and it’s too late for Black to take on c4, and with hanging pawns, the bishop would have been
better on d6. 10...Nc6 threatens 11...d4, but White simply plays 11.cxd5 exd5 12.Rc1 with an advantage.
b) 9...dxc4! Hitting the bishop on d3. 10.Bxc4 Bxc5 White has lost the “advantage” with the more active bishop, and
has to play a symmetrical position with 11.Qe2. But he is still White!
Note that in the Queen’s Indian, with the bishop on d6, it is White who hits the bishop with 9.dxc5. That’s an argument
in favour of the passive e7-square!
7...exd5 8.Nc3
With Bd3, Rc1, 0-0 and Ne5 – the same position was reached from the Queen’s Indian.
B) 6...c5 7.dxc5
7.Bd3 with Plan 1, as in Franco Ocampos – Goldin, avoids the dull position.
7...dxc4!
Again the same problem. It’s only been played a handful times, but one of them was Topalov – Anand, Stavanger
(blitz) 2015.
After 7...Bxc5 8.Nbd2, we reach a position where it’s usually White’s move since Black has moved the c-pawn twice.
8...b6 9.cxd5 gives Black an isolated pawn, which doesn’t fit well with ...b7-b6, or a symmetrical position with slightly
worse pieces. Better is 8...Nc6 9.a3. Now we have:
a) 9...a5 10.Bd3 With the b4-square under control, White enjoys a position with two good bishops.
Better is 10...Nxd4 11.b4 Nxf3† 12.Nxf3 Qxd1† 13.Rxd1 Be7 and Black seems to get a playable position with ...a7-a5
and ...Nf6-d7/e4-c5.
8.Bxc4
Taking on d8 helps Black to develop his rook.
8...Qxd1†
Otherwise White avoids the exchange.
Exercise 1
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 Be7 5.b3 0-0 6.Bb2 c5 7.Bd3 cxd4 8.exd4 b6 9.0-0 Bb7 10.Nc3 Nbd7
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 e6 5.b3 c5 6.Bb2 cxd4 7.exd4 Nc6 8.Bd3 Be7 9.0-0 0-0 10.Nc3 b6 11.Re1 Bb7
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 d5 2.Nf3 Nf6 3.c4 e6 4.e3 Be7 5.b3 0-0 6.Bb2 b6 7.Bd3 Bb7 8.0-0 c5 9.dxc5
Black to move
Show/Hide Solution
Show/Hide Solution
A) 4...Bg4 5.h3 Bh5 6.g4!? Bg6 7.Ne5 e6 8.Nd2! Nbd7 9.Nxg6 hxg6 10.Bg2 g5 11.0-0 303
B) 4...Bg4 5.h3 Bxf3 6.Qxf3 e6 7.Bd3 Nbd7 (7...Bb4†) 8.0-0 Bd6 9.Nc3 0-0 10.e4!? dxc4 304
a) 10...dxe4 304
b) 10...e5 304
11.Bxc4 e5 12.d5 Nb6 13.Bd3 cxd5 14.exd5 h6 304
14...Bb4 304
15.Rd1N Rc8 16.Be3 305
Move Orders
1.c4 or 1.Nf3 are good moves against Slav players. Black gets semi-punished if he develops the light-squared bishop –
hence the name Slav Nirvana.
However, after 1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3, both 4...Bf5 and 4...Bg4 are good options. White has to take another
step into the moorland of theory, with more trampled ground than we would have preferred.
Pawn Structures
Structure 1
Black is satisfied to develop his bishop outside the pawn structure, but he would have been even happier if he had
time to save it with ...h7-h6. In the diagram position, White has taken the bishop pair with Nh4xg6. The open h-file is
in Black’s favour and White usually spends a tempo with g2-g3 or h2-h3 to defend the h-pawn.
a) c4-c5 (with tempo) plus f2-f4, creating a bind. It’s important that it’s possible to answer ...b7-b6 with b2-b4, without
getting undermined by ...a7-a5.
b) Play a long ending with the bishop pair – usually after castling in the same direction as Black.
c) e3-e4, maybe with f2-f3 first.
In order to fight against all of these plans, Black often takes on c4 and tries to open the position with ...e6-e5, ...c6-c5 or
even...g6-g5.
Games
For a while, the Slow Slav was so popular with the world’s elite that it was almost as if a consensus arose that the
bishop pair and an academic advantage was a fair deal. Neither White nor Black could expect more of the opening, so
both players got some rest when preparing, because concrete variations are less important in the resulting positions.
Or are they? Of course, preparations went ever deeper – also analysing deeply in positions where the engines claim
many moves to be of equal value.
In the theoretical section, I will keep things simple. For now, Carlsen’s win is a good example of the way White wants
to play.
Carlsen in an interview with Sjakkfantomet, March 2016: “The number one opponent I prefer to meet, quite clearly, is
Boris Gelfand. He plays a kind of chess that I really like. With no nonsense: straightforward plans, very dynamic and
positionally well-founded. When we play it always becomes interesting positionally-heavyweight duels, and mostly I
manage to out-calculate him in the end.”
1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.e3 Bf5 5.Nc3 e6 6.Nh4 Bg6 7.Nxg6 hxg6
8.Bd3!
Veselin Topalov beat Viswanathan Anand in their 2006 World Championship match with 8.a3. He wanted to follow
up with 9.f4 and 10.c5, creating a strong bind. And the b4-pawn must be defended in advance after 10...b6 11.b4 a5, so
White gets time to do something about the undefended rook on a1.
Later on, we will see the same idea in a more efficient manner.
10...dxc4
A move ago, it was too early for 10.f4 dxc4! but after 10.h3 if 10...Qe7 then 11.c5 Bc7 12.f4 would create the bind.
It’s no longer obligatory to put the pawn on a3, since 12...b6 13.b4 a5 14.b5! works with the knight on d7 instead of b8.
14...bxc5 15.bxc6 Nb8 16.Ba3 gives White an extra pawn.
11.Bxc4
11...0-0
A year earlier, Gelfand played 11...Nb6 12.Bb3 e5 against Carlsen and castled long. However, 13.d5! gave White an
advantage in a correspondence game.
A lot of set-ups were possible for both players and I only understand two moves: Black’s rook against the queen and
Carlsen’s decision to keep the bishops in the garage until the endgame. It’s not too different from what Peter Heine
Nielsen did against Vugar Gashimov. White’s task is to advance the pawns only when it doesn’t give up squares to
Black’s knights.
15...e5 16.dxe5 Bxe5 17.Rac1 Rcd8 18.Be1 Rxd1 19.Rxd1 Rd8 20.Rxd8† Qxd8 21.g3 Qe7 22.Bg2 Bd6 23.Bd2
On the way to a new parking place.
23...Qe6 24.b3 Nbd5 25.Ne2 Nb4 26.Qb1 Qf5 27.e4 Qc5 28.Nf4 g5
Fixing White’s pawn majority; a decent decision even though the pawn will need constant protection when it gets to
g4.
31.Be3 Qc2
Maybe Gelfand missed 31...Qa5 32.Qd1! with two pieces hanging on the d-file. Black is forced to retreat with
32...Qc7 33.a3 Na6 34.b4 when Carlsen would play on the kingside, far away from Black’s knight on a6.
With Be2xg4 as an irresistible threat, Gelfand starts to chase pawns. But even if he gets one on the queenside,
White’s 4 versus 2 kingside majority is decisive.
36...Nb4 37.a4 Nf6 38.e5 Nd7 39.Bd4 Nc2 40.Bc3 Nc5 41.Be2 Ne4 42.Bxg4 Nxc3 43.Nxc3 Bb2 44.e6!
Game over – Black is two pawns down and Carlsen will not exchange knights.
44...fxe6
Or 44...Bxc3 45.e7 with a new queen.
45.Bxe6† Kf8 46.Ne4 Nd4 47.Ng5 Ke7 48.Bg8 Kf8 49.Bc4 Ke7 50.Kg2 b5 51.Bg8 Kf8 52.a5!
1–0
In the same interview as the quote above, Carlsen was asked about Rotlewi – Rubinstein 1907. He answered: “Everyone
got that combination with the mother’s milk; everyone that has a little chess in their blood.”
Do you have it?
25...Rh3!!
There is no defence against the mate on h2.
Recap
White needs patience. It’s not possible to make a clear-cut plan on how to use the bishop pair, but even so it’s a long-
term advantage. Carlsen placed them on f1 and e1 and only after 41.Be2 did Black finally pay for the absence of his
light-squared bishop: it was not possible to defend the pawn on g4.
White’s bishops did a lot of work – in total 18 out of Carlsen’s 52 moves. And the game was decided when the light-
squared bishop sacrificed itself on the last move.
Theory
A) 4...Bg4
9...dxc4!?
9...e5? 10.cxd5 Nxd5 11.Nc3 gave White a pleasant edge in Smith – Mads Hansen, Vaxjo 28th June 2017, played
while the book was being proofread.
10.Nd2 e5
a) 10...Qa5 11.0-0 Nb6 defends the pawn, but only for a while. White plays 12.Qc2 Qa6 13.Rd1 and plans a4-a5. Bf1
is also a possibility if necessary.
b) 10...Nb6 11.Qc2 e6 12.Nxc4 Nxc4N 13.Qxc4 Nd5! is logical because White would have wanted to keep more
minor pieces at the board. White should castle long to avoid an attack starting with ...f7-f5. 14.Bd2 Bd6 15.0-0-0 g5!
16.Kb1 White will prepare f2-f4 with Rhf1.
14.Na5N 14...Rd7
Or 14...Nd5 15.Bd2² with Ke2 and Rac1 next.
15.Nxb7!?
White gets a rook, two pawns and some temporary threats for his two pieces. Since Black has knights and not
bishops, I prefer White.
9...0-0 10.e4!?
White loses control over e5, but he is doing okay anyway.
10...dxc4
Black’s standard reaction to e3-e4. The knight on d7 gains access to the dark squares in the centre.
a) 10...dxe4 11.Nxe4 Nxe4 12.Qxe4 Nf6 13.Qh4 is a Fort Knox with White’s pieces actively placed. It’s not a great
chance to win with a Bxh6-sacrifice, but the queen is nevertheless strong on h4. After developing everything it’s time
to consider advancing the kingside pawns (even if the queens are exchanged).
b) 10...e5 11.exd5 exd4 12.Ne4 Nxe4 (the hanging bishop on d6 stops 12...cxd5) 13.Qxe4 g6 14.Qxd4 and Black will
probably get back the pawn, when the pair of bishops gives some advantage.
13...cxd5 14.exd5 h6
14...Bb4 wins the pawn on d5 but 15.Rd1 Bxc3 16.bxc3 Nbxd5 17.Bc4! forces Black to give up an exchange with
17...e4 (only move) 18.Qg3 Rc8 (only move) 19.Bxd5 Nxd5 20.Ba3 Rxc3 (only move) 21.Qd6!. He has two pawns,
but they are split and don’t give enough compensation.
15.Rd1N
Not giving Black a second chance to play ...Bb4.
15...Rc8 16.Be3
White’s next move is either 17.Bf5 or 17.Ne4. If Black goes 16...Nc4 it’s possible to snatch the a7-pawn with
17.Bxc4 Rxc4 18.Qe2 Rc8 19.Bxa7 just like Ivanchuk did.
6...Be4
6...Bg4 is a rare line (4%) that deserves to be even rarer. After 7.Qb3 Qb6 8.h3 Bh5 we get a position when it’s
usually Black to move (4...Bg4 saves a tempo). White continues with 9.g4. The same goes for 7...Qc7. White continues
with 9.g4.
Instead after 6...Bg4 7.Qb3, Black can try to complicate matters with 7...b6 8.h3 Bh5 9.g4 Ne4, but it works out badly
after 10.Nxe4 Qxh4 11.Ng3 Bg6 12.cxd5 exd5 (12...Bd6 13.e4! exd5 14.e5 was Ellinor Frisk – Stephen Jablon,
Reykjavik 2013) 13.e4! dxe4 14.Bg2 with a great initiative.
7.f3 Bg6
This is almost as popular as 6...Bg6. With the pawn on f3, White is less inclined to castle short. Instead it’s logical to
search for variations where he can:
a) Exchange the queens
b) Use f2-f3 to play e3-e4
8.Qb3
8...Qc7
We will consider a couple of alternatives:
a) 8...Qb6 9.Nxg6 hxg6 10.Bd2 Nbd7 11.0-0-0 and White continues in slow fashion with Kb1, Bd3 and h2-h3. One
line that is good to know is: 11...Qxb3 12.axb3 Bd6 13.h3 Nh5
Glenn Flear evaluated this position as equal in Starting Out: The Slav and Semi-Slav (2005), but since Markus Ragger
allowed this ten years later, I am sure it’s an interesting option for White. 14.Bd3 Ng3 15.Rhe1 Black’s idea is that a
knight on f5 is good prophylaxis against e3-e4 – but it could be captured.
b) 8...b5!? was an innovation Tiger Hillarp found at the board against Anish Giri in Wijk aan Zee 2009. Even back then,
Giri was already extremely solid, but he played far too cautiously and lost his first game with White for a long time.
White can take the pawn, or go for: 9.c5!? a5 (9...Nbd7 10.a4 with good use of the a-file after 10...a6 11.Qa3)
10.Nxb5! a4
11.Nd6†N 11...Bxd6 12.Qb7 Even though Tiger Hillarp has argued for Black’s compensation after 12...Nbd7 13.cxd6.
9.Bd2
9...Be7
9...Bd6?! doesn’t create a threat: 10...Bxh2 11.Rxh2 Qxh2 12.Nxg6 hxg6 13.Qxb7 wins for White, even though the
queen gets semi-trapped on a8. And White is allowed to play a 10th move as well.
9...Nbd7 10.cxd5! is a justified exchange when Black’s knight can’t go to c6. Now:
a) 10...exd5 gives White an advantage after 11.Nxg6 hxg6 12.e4!? dxe4 13.fxe4. The pawn on h2 is still immune and
e4-e5 will come soon.
b) 10...cxd5?? 11.Rc1 wins.
c) 10...Nxd5 is the main line. White usually plays 11.e4 Nxc3 12.bxc3 but 12...Bh5 seems pretty okay for Black. And
12.Nxg6 hxg6 13.bxc3 allows 13...Bd6 when 14.g3 is impossible with the open h-file.
So I recommend starting with 11.g3. White has the same plan, but with greater flexibility. He could take back on c3
with the queen, or take on d5 before playing e3-e4. Two supporting lines:
c1) 11...Bd6 12.e4 Nxc3 13.bxc3 Bh5 is not a problem when there’s time for 14.Be2 as White answers 14...Be7 with
15.Ng2.
10.cxd5
Only played in 5% of the games, and slightly illogical with the black knight on b8. But I like the move, since it avoids
10.0-0-0 dxc4.
White continues with Kb1, Rc1 and g2-g4, when his king is safe. If Black castles short, there is an attack coming
right at him.
10...Qe7 11.c5 Bc7 12.f4 was given in the comments to the game.
12...e5 13.b5
It’s clearly a race. Black can choose whether to play for the e4-square (...e5xd4) or a kingside attack (...e5-e4).
13...Re8
13...e4 14.Be2 Bb8 15.bxc6 Qc7 16.g3 was played in Parligras – Bakre, Fujairah 2012, but White’s king never faced
real problems after Kg2 and Rh1 (when necessary).
Exercise 1
1.d4 d5 2.Nf3 Nf6 3.c4 c6 4.e3 Bf5 5.Nc3 e6 6.Nh4 Bg6 7.Nxg6 hxg6 8.Bd3 Nbd7 9.0-0 Bd6 10.h3 0-0 11.c5 Bc7
White to move
Show/Hide Solution
Show/Hide Solution
The Baltic Defence
1.d4 d5 2.Nf3 Bf5 3.c4 e6 4.cxd5 exd5 5.Qb3 Nc6 6.Nc3 Nf6 (6...Nb4; 6...Bb4) 7.e4!? Nxe4N 8.Qxb7 Bd7 9.Qb3
Rb8! 10.Qxd5 Nf6! 11.Qc4 Nb4! 12.Qe2† Qe7! 314
The Rausis
A) 4...b5 5.a4 e6 6.axb5! cxb5 7.b3 Bb4† 8.Bd2 Bxd2† 9.Nbxd2 a5 (9...c3) 10.bxc4 b4 11.Ne5 Nf6 12.Be2! 0-0
(12...Bb7?)13.Bf3 Ra7 14.0-0 Qc7 15.Qa4 Bd7 16.Nxd7 Nbxd7 17.Qc6! 315
B) 4...Be6 5.Nc3 b5 (5...Nf6) 6.Bd2(!) Nf6 (6...a6; 6...a5) 7.a4 b4 8.Ne2 316
2...a6
1.d4 d5 2.Nf3
Move Orders
This chapter examines the extra options Black gets after 2.Nf3 (instead of 2.c4):
None of the lines are common, and since they don’t have many similarities, the pawn structures can’t be organized into
a few typical ones. We will only see one typical structure, followed by two short games with Black’s other choices. But
the lines will, of course, often give structures from other chapters.
Pawn Structures
Structure 1
The London System (1.d4, 2.Nf3 and 3.Bf4 followed by e2-e3 and c2-c3) is infamous because White uses the same
set-up almost whatever Black does. There’s no reason to be judgmental, but I have seen many players get comfortable
(lazy) and continue to play the same moves throughout their lives.
I guess they were happy when our hero Vladimir Kramnik started using it. But he has added a little bit of finesse.
Kramnik doesn’t play the London when Black can restrict the bishop on f4 with ...d7-d6 and ...e7-e5.
We are even more selective and go for the London set-up only when Black has played:
a) ...d7-d5
b) ...e7-e6
c) ...a7-a6
The advance e3-e4 can be good in some positions, opening the board for every piece except the bishop on c8. But
usually White plays with the pieces: Bf4, Bd3, Nbd2, Nf3-e5 (maybe supported by Ndf3, or Bf4-g3 and f2-f4). White
prepares an attack on the kingside (where the c8-bishop isn’t taking part), but probably White castles short anyway.
Black plays for ...e6-e5, to release his bishop and gain space in the centre. That’s normally stopped with Nf3-e5, but
if Black has spent too much time, it’s possible to allow him to open the position and then add fuel to the fire with e3-e4.
It’s not only Kramnik who has toyed with the London System lately. Alexander Grischuk is another top player who has
tried the London, and in Wijk aan Zee 2016 it was used by Magnus Carlsen, Sergey Karjakin and Anish Giri.
Games
Playing the Rausis, Zoltan Varga shows that even though Black defends the pawn on c4, his idea isn’t really to keep it.
He grabs space on the queenside with ...b7-b5-b4 and above all controls the c3-square. The price he has to pay is a
backward c-pawn and the game circles around two questions: Can White exploit the backward c-pawn? Can Black free
himself with ...c6-c5?
5.Nc3
5.Ng5?? Qa5† is embarrassing, but now White is ready for 6.Ng5 Bd5 7.e4.
5...b5
A justified move with the knight on c3, but now Black would have preferred a better square for his light-squared
bishop.
6.a4
In the theoretical section, we will propose Hans Tikkanen’s invention 6.Bd2. White plans 7.a4 b4 8.Ne2 a5 9.Rc1
followed by 10.Nf4 and the c4-pawn is doomed. Winning back the pawn would give White nice squares on the c-file –
a better version of what we have seen in the Queen’s Gambit Accepted.
6.a4 is what White usually plays and in the game it transposes. However, Black had another option along the way.
8.Bd2(!)
Often a good move in the e3 repertoire.
12...exd5?!
12...cxd5 gets rid of the weak c-pawn, but not the weak light squares (b5 and c6). However, while the light-squared
bishop is the best minor piece on the board, White’s other bishop is the worst. It’s possible to play only on the c-file
with Bb5, Qe2, 0-0, Rc2/c6, Rfc1 and Ne5-c6 or Ne1-d3-c5, but with an extra piece it should be possible for Black to
defend patiently. Therefore White also needs to play f2-f3 and Be1-g3/h4. That would weaken the dark squares and is
therefore best done after a queen exchange.
14...Qe7 15.g3
Not obligatory for the moment. Avoiding unnecessary prophylactic moves is often the difference between an
advantage and an equal position. At the same time, contradictory advice such as “Loose Pieces Drop Off” also has its
points. It’s about finding a balance between being practical (lazy?) and objective (obstinate?).
With 15.g3, Graf prepares Nf3-h4 followed by f2-f3 and e3-e4, a plan that’s logical with the knight stuck on b8. But I
think it’s more logical to retreat to d2 in order to exchange Black’s knight on e4, so my suggestion is 15.0-0 0-0 16.Be1
followed by 17.Nd2. The knight continues to b3 if Black doesn’t take it.
White’s structural advantage is gone and he was happy to get a draw a few moves later.
As long as the pawn remained on c6, Black stayed with a knight on b8 to keep it defended. It sounds like a dubious
strategy, but it worked when White didn’t manage to create enough on the other wing before Black achieved ...c6-c5.
The next game is not strategical at all.
Sosonko, April 2016: “I remember the game but, unfortunately, nothing special about it. Wasn’t it 100 years ago?”
1.d4 d5 2.c4
Of course our move order in this chapter is 2.Nf3.
2...Bf5
3.Nc3
The Baltic Defence is considered to be worse when Black can’t take back with the e-pawn after 3.cxd5. To avoid
losing a tempo after 3...Qxd5 4.Nc3, he parts with his bishop: 3...Bxb1
5...Nc6!
Black’s standard defence, creating a counterattack against c2 if White captures on b7. But the knight will be
misplaced on c6 if Black doesn’t get anything concrete.
6.Nf3 Nb4?!
6...Bb4 and 6...Nf6 are the main moves, for a good reason.
7.e4! dxe4 8.Ne5
White has two threats: 9.Qxf7 mate, and 9.a3 followed by 10.Qxb7.
10...Nc2†?
10...Qe7 11.0-0 is simply lost for Black. Without the check on c2, he has nothing to make up for White’s
development.
The only move was 10...Nd3† 11.Nxd3 exd3 but after 12.Qb5†!, the queen takes on b7 first and picks up d3 in the
future.
11.Kf1
Black can’t defend f7 without stepping into Nc3-d5.
11...Qf6 12.Nd5
1–0
Recap
The last game was clearly just an aperitif, but it held one important message that can’t be repeated too often: Black
should not be allowed Black to develop with ...Bf5 without being punished. The first move to consider is Qd1-b3,
attacking b7.
As we will see in the next section, that’s not always possible.
Theory
4.cxd5
After 4.Qb3 Nc6! 5.Qxb7 Nb4 6.Na3 Black can force a repetition with 6...Rb8 7.Qxa7 Ra8.
4...exd5 5.Qb3
It was important to start by exchanging on d5, as otherwise Black would have taken back with the queen.
5...Nc6
Still Black’s idea. He keeps the pawn, but also has to prove that the knight is not just badly placed in front of the c-
pawn. Things are not so easy for Black after the next move.
6.Nc3
6...Nf6
Instead 6...Nb4 lost quickly in Sosonko – Ree after 7.e4!.
6...Bb4 7.Bg5! and Black probably gives up his bishop on c3, and one way to get compensation is to make use of the
e4-square. That’s not possible after 7...Nge7 8.e3 or 7...f6 8.Bd2. Nor does 7...Nf6 work – d5 is too weak after the
simple 8.e3.
7.e4!?
This time the e4-break is only interesting, instead of devastating.
A calmer alternative is 7.Bg5 when Black has to double his f-pawns to keep material even: 7...Na5 8.Qa4† c6 9.Bxf6
gxf6 This is similar to the Panov game Aronian – Grischuk at the 2015 Sinquefield Cup. Black’s bishops make up for
the pawn structure.
7...Be7 is given by Alexey Bezgodov in The Liberated Bishop Defence (2014). After 8.Bxf6 Bxf6 9.Qxb7 Nxd4
10.Nxd4 Bxd4 11.Qc6† Bd7 12.Qxd5, White has an extra pawn, but Black has good chances to hold with 12...Bxc3†
13.bxc3 0-0 14.e3 Be6. Okay, let him try.
7...Nxe4N 8.Qxb7 Bd7
The only move.
12.Qe2† Qe7!
Exchanging on e7 speeds up Black’s development, so the engine gives 13.Be3 Bf5 14.Ne5 as balanced. Black has
compensation, but he can’t win the exchange due to Ne5-c6.
The Rausis
1.d4 d5 2.Nf3 c6 3.c4 dxc4 4.e3
Since Black had problems reinforcing his queenside after 2...Bf5 3.c4, he starts with the prophylactic 2...c6. 3.e3 Bf5
would have been safe, and after 3.c4 dxc4 he has time to defend the pawn, either with A) 4...b5 or B) 4...Be6.
A) 4...b5
5.a4
The ...b5-advance makes more sense with a knight on c3. Black could then push with ...b5-b4.
5...e6
Now, Boris Avrukh’s line is strong.
5.Nc3 b5
5...Nf6 6.Ng5! Bd5 7.e4 h6 A standard move, and the only one. 8.exd5 hxg5 9.dxc6 Nxc6 10.d5 Ne5 11.Qd4! White
has enough control to claim that the bishop pair is an advantage in itself. The next move will be 12.Bxg5.
6.Bd2(!)
Hans Tikkanen’s move, which is not mentioned in the Slav books.
6...Nf6
By delaying a2-a4, Black gets time for 6...a6 7.a4 Nd7 but he is in for a surprise: 8.Ng5 Bf5
9.Nxf7! Kxf7 10.Qf3 e6 11.g4 Ne7 The weak king and the possibility to open the centre gives fair compensation after
any move: 12.gxf5, 12.Ne4-g5, 12.e4 or 12.Bg2.
Another way to interfere with White’s plans is 6...a5 7.a4 b4 8.Ne2 Nd7 9.Rc1 Nb6 10.Nf4 Bd5 but Black can’t keep
the pawn after 11.Ne5.
We can note that the double threat 11...Nf6 12.Nxc4 Bxc4 13.Bxc4 g5!? 14.Ne2 Nxc4 15.Rxc4 Qd5 leaves Black far
too underdeveloped. After 16.Qc1 Qxg2 17.Rg1 Qxh2 18.Rxc6 his extra pawn can only be used as payment for a
worthy funeral.
7.a4 b4 8.Ne2
We have reached Graf – Varga.
2...a6
1.d4 d5 2.Nf3 a6
The intention is 3.c4 dxc4 4.e3 b5, a line known to be very solid. It forces us out of our routine set-up and the
question is what to play without having to learn a whole new concept, since unfortunately it’s not possible to punish
Black after 3.e3 Bf5 4.c4 e6.
3.Bf4
White has not yet decided whether to continue with c2-c4 or play an improved version of the London System.
3...Nf6 4.e3 e6
Black can catch the bishop with 4...Nh5 5.Bg5 h6 6.Bh4 g5 7.Ne5! and ...Ng7-f5 or ...Nf6-e4, but the weakened king
is a fair price.
4...Bf5 is, as always, critical. But 5.c4 c6 6.Nc3 is not really an ...a6 Slav. In the Slav, 4...a6 threatens to take on c4 and
forces White to save the pawn with c4-c5, a2-a4 or e2-e3. With this move order, White has managed to develop the
dark-squared bishop outside the pawn chain without any concession. 6...h6 (6...e6 allows the same move as in the Slow
Slav: 7.Nh4 grabbing the pair of bishops) 7.Qb3
Now Black has a couple of ways to try to save the b-pawn:
a) If 7...b5 then 8.cxd5 cxd5 9.a4 is standard and strong. With the bishop on f4, we can play: 9...b4 10.Qxb4! Nc6
(10...e5?? 11.Qb7 hits the rook) 11.Qb7 Na5 12.Qc7 The extra pawn isn’t beautiful, but it’s still there.
b) 7...Ra7 8.cxd5 cxd5 9.Bxb8! A difference with the bishop on f4 – White wins a pawn.
5.Nbd2
Instead 5.c4 c5 gives Black a good Tarrasch, so White goes for a London – but only when Black’s bishop stays on c8.
Even though the opening could be played for a lifetime, it’s enough to learn a few standard moves and a well-known
trap.
8...0-0 9.Ne5
Not 9.0-0?! Bxg3!.
9...Qe7?
Black defends the bishop on d6 and thus prepares 10...Nd7.
But it’s better to ask a question of the e5-knight with 9...Qc7; then White plays 10.f4 and probably castles short.
11...Bxd7?
Right into the trap.
2...c5
1.d4 d5 2.Nf3 c5
Against this rare move, White can choose between the following options:
a1) 4.e3 gives Black the opportunity to make use of the move order with 4...cxd4 5.exd4 5...Be6!?, or 5.Bxc4 Qc7!?,
hitting the bishop.
a2) The critical move is 4.d5! when theory considers White to be better, although opponents who play this line will have
something up their sleeve.
b) I recommend being practical and staying in the repertoire with 3.e3(!), as covered after 1.Nf3 d5 2.e3 c5 3.d4 on page
135.
Exercise 1
White to move
Show/Hide Solution
Show/Hide Solution
Rossolimo
1.c4 e5 2.e3 Nf6 3.Nc3 Bb4! 4.Nge2 0-0 335 (4...c6) 5.a3 Be7! (5...Bxc3) 6.d4 exd4 7.Qxd4 Nc6 (7...c6!) 8.Qd1 336
Alapin
1.c4 e5 2.e3 Nf6 3.Nc3 c6 4.d4 e4 (4...exd4) 5.d5 Bd6 6.g4! 337
Grand Prix
1.c4 e5 2.e3 Nc6 3.Nc3 f5 4.d4 e4 5.Nh3 338
Old Indian
1.c4 e5 2.e3 Nf6 3.Nc3 d6 4.d4 Nbd7 338
Delayed Rossolimo
1.c4 e5 2.e3 Nf6 3.Nc3 Nc6 4.Nf3 Bb4 5.Qc2 0-0 339 (5...Bxc3; 5...d6) 6.Nd5 Re8 340 (6...Be7; 6...a5) 7.Qf5 d6!
(7...Nxd5) 8.Nxf6† gxf6 (8...Qxf6) 9.Qh5 e4! (9...d5) 10.a3! exf3 11.gxf3!? Re5! (11...Bc5) 12.Rg1† Rg5 13.Rxg5†
fxg5 14.Bd3 Qf6 (14...Bc5) 15.Qxh7† Kf8 16.axb4 Nxb4 17.Be4 341
Miscellaneous
1.c4 e5 2.e3(!) Nf6 (2...g6) 3.Nc3 Nc6 4.Nf3 d6 (4...Be7) 5.d4 g6 6.Be2(!) Bg7 7.dxe5 Nxe5 8.Nxe5 dxe5 9.Qxd8†
Kxd8 10.b3 343
Move Orders
Positions with ...e7-e5 are covered in this chapter. And since 1.d4 and 1.Nf3 rule out 1...e5, there are not many possible
move orders. 1.Nf3 d6 2.c4 e5 is one, but there’s nothing wrong with 2.d4.
To make things clear: 1.c4 e5 2.e3 is not theoretically threatening. It gives Black greater freedom than the traditional
2.Nc3 and 3.Nf3 (so does 2.d3, as Alexander Grischuk played against Viswanathan Anand in London 2015).
So Black has greater freedom – in some lines. But Black is also deprived of some possibilities.
The important thing is, as usual, to know what you’re up to. We will see what happens if Black plays like in a
Sicilian. Does the extra tempo count in the Alapin, the Grand Prix or the Open Sicilian?
At the same time the coverage has to be kept rather brief. The e3 English could fill a whole book, even though such a
book doesn’t exist yet, as far as I know. Here, it’s one of 25 chapters.
If the game starts as a Symmetrical English with 1.c4 c5, the most common move is 2.Nf3. But 2...Nc6 3.e3 allows
3...e5 4.Nc3 f5 5.d4 e4. After 1.c4 c5 2.e3(!) it is worse to play 2...e5 3.d4! since 3...e4 doesn’t hit a knight on f3.
So after 1.c4, White plays e2-e3 already on the second move.
Pawn Structures
White’s plan is to play e2-e3, d2-d4 and e3xd4, but sometimes he needs some patience to prevent Black from getting a
favourable Exchange French with ...exd4, ...Bb4 and ...d7-d5.
Structure 1
With a pawn on c4, it’s favourable for White to exchange the d-pawn for Black’s e-pawn. The strategy is to make use
of the d5-square by Nc3-d5 (after completing development).
1) Kick away the knight with ...c7-c6, which weakens the pawn on d6.
2) Play ...Nxd5, when White recaptures with the c-pawn, kicks away a knight from c6 (is there is one) and puts pressure
along the c-file.
3) Live with the d5-knight as a cuckoo in his nest, and survive with less food (freedom).
If Black tries to use his semi-open file in the same way and plays ...f7-f5 to control the e4-square, he weakens his king.
And without a pawn on f5, ...Ne4 isn’t dangerous – we just take it. Black is generally happy with the exchange (no
Nc3-d5!) but it costs time.
The pawn on c4 gives White’s queen a good square (c2), while Black’s queen is not happy either on e7 (running into
Nd5), nor on d7 (opposing a rook on d1).
However, the pawn structure is better for White in theory than in practice. There are several reasons:
1) The bishop would have been better on g2, keeping control over d5 and threatening the pawn on b7.
2) If Black has developed quickly, he can try to release his position with ...d7-d5 (maybe with ...c7-c6 first) – especially
when a knight on d4 blocks the queen on d1.
3) Black can be disruptive with an early ...Bb4.
So, White may not be better when the structure appears, but he has great potential. One slow move by Black may be
enough to consolidate the d5-square and claim a long-term advantage.
Structure 2
Black has not taken on c3 voluntarily (I hope), but instead been forced to in order to castle. The centre always
threatens to expand, but is not weak as in the Grünfeld (when the e-pawn is on e4).
White arranges his pieces with Rb1, 0-0, Nf3 and Be2/d3. The dark-squared bishop can bide its time on c1, keeping the
options to go to a3 but also e3 or g5, if White plays e3-e4.
White should keep the queens on, since in theory Black is striving for a pawn ending with a distant pawn majority
(after ...c7-c5xd4). Another plan for Black is to play ...e5-e4, with an attack on the kingside, but it seldom works. White
can play c3-c4, control important squares and soon be ready to push further.
Structure 3
Everything in this position circles around the pawn on e4, which has given White and Black two advantages each:
1) White has a nice pawn tandem on c4-d4.
2) White has a square on f4.
3) Without a knight on f3, White’s kingside is poorly defended. But since ...Bd6 isn’t really possible, it will take time
for Black to build up an attack.
4) The bishop on c1 is restricted. White should avoid b2xc3 (after ...Bb4xc3), when the b2-g7 diagonal would close as
well.
White’s centre pawns have no potential to improve. Instead, he usually plays f2-f3, which gives active pieces but also a
weak pawn on e3. Whether this is good or bad depends on the placement of the pieces. And compared to the normal
move order (1.c4 e5 2.Nc3 Nc6 3.Nf3 f5 4.d4 e4) White has saved one to two tempos by not jumping around with the
knight (1.c4 e5 2.e3 Nc6 3.Nc3 f5 4.d4 e4). The odds are thus great that f2-f3 will leave White on top.
Structure 4
Without its colleague on the d-file, the c-pawn would have preferred to stay on c3. The reasons are:
1) If White plays e3-e4, he weakens the d4-square.
2) The pawn on e5 controls two squares close to the centre, while the c-pawn only controls one.
3) The e-pawn could be supported with ...f7-f5.
For this reason White only enters this pawn structure when he can play Qxd8† (forcing ...Kxd8), 0-0-0† and Ba3, and
when it gives an ongoing initiative.
Structure 5
Black is generally happy with this structure. The doubled pawns are not weak, but they restrict the bishop on c1. The
pawn on c3 would have been better on c2 to give the bishop free sight from b2.
The downside with a doubled-pawn complex is shown when it tries to advance: d2-d4 would make the pawn on c4
significantly weaker, and another step (d4-d5) gives up the c5-square. That would not have been the case without
b2xc3.
Instead, White prefers e3-e4 and d2-d3. Black can, on the other hand, support the pawn on e5 with either ...f7-f5 or
...c7-c6 plus ...d7-d5, even though the latter exchanges a doubled pawn. Also a common move is to disturb a knight on
f3 with ...e5-e4.
All in all, it doesn’t sound too good for White, but he only enters this pawn structure in one position in the repertoire,
and that one is a story all itself.
Games
The first game is a reversed Open Sicilian and already after six moves it’s clear that the extra move makes a big
difference.
Gajewski, April 2016: “The sacrifice was very simple, even automatic – but I am bit loose with sacrifices!”
1.c4 Nf6
Our move order is 1...e5 2.e3 Nc6 3.Nc3 Nf6 4.Nf3 d5.
5.cxd5 Nxd5
In the normal Sicilian, with Black (who would actually be White!) to move, theory recommends 6...Ndb4 or 6...Nxc3
7.bxc3 e4.
6.Bb5!
The threat on e5 forces Black to capture on c3.
9...Qg5!
The only light in the tunnel. The double threat forces the bishop to retreat to f1.
13.0-0 Bg4
The point of this manoeuvre is to avoid f2-f3.
13...b6 is Black’s preferred set-up, but 14.f3! exf3 15.Bxf3! causes tactical problems.
a) 15...Rxe3 16.Ne4! Rxe4 17.Bh5! wins the queen. But not 16.Bxc6 Rxg3† 17.hxg3 Qxg3† 18.Bg2 Qh2† 19.Kf2
Qg3† with a perpetual.
b) 15...Bb7 and White is much better after 16.e4, or winning after 16.Bh5.
16...a6?!
It was better to give up an exchange in the complications starting with 16...Bd7 17.Bxe4 Rxe4 18.Qxe4 Nxd4
19.Qxg6 hxg6 20.Ra5 Nc6. The rook escapes, but to a passive square.
White can also play 17.Rb2, winning half a tempo since Black has to return with the bishop to f5.
Practically everything got worse for Black once the e4-pawn disappeared. With both strategical (pawn centre) and
dynamic compensation, there is no risk that the compensation will evaporate in a few moves. Finally, there’s another
important factor that’s worth at least half a pawn: the sacrifice is incredibly practical as White’s play is childish from
now on – just push the pawns! ”
18...Qh5 19.Bg2
19...Na5 20.e4 c5!
Tomczak opens lines for the rook.
No draw. Black has to start worrying about Kg2, h2-h3 and g3-g4.
27.h3 f5?
27...Kh8 was relatively best.
28.Bd5† Kh8 29.Ng5
Moving from the Open Sicilian to the Rossolimo, we will again see that the extra tempo makes a substantial difference.
Grischuk at the press conference: “I was lucky because usually Anish knows everything, but today he didn’t know the
theory.”
5.Qc2
Allowing the pawns to be doubled would restrict the bishop even further and 5.Nd5? is not good as long as Black has
5...e4!.
7.Qf5!
7.Nxb4 loses too much time; Black would open the position with ...e5-e4 and/or ...d7-d5.
With reversed colours, the e-pawn would still have been on e2 (e7) and the queen move could then have been strongly
answered with 7...Nxd5 8.cxd5 Nd4. But now 7...Nxd5 8.cxd5 Ne7 allows 9.Qxe5.
7...d6
7...Be7 8.Nxe5 loses a pawn, so Black can’t avoid the f-pawns being doubled.
For the moment, it’s difficult to attack the doubled pawns, but they will always be there. At the press conference
Grischuk said: “Of course White is slightly better” and claimed that 8...gxf6 was better, keeping on the queens with
very concrete play. That option is examined later.
14.d4
14...Ng6
I think Black should have played 14...a4 since:
1) White can’t play a3-a4xb5 if Black tries to open the b- or c-file with ...c7-c6.
2) The b-file can be opened for the rook with ...Ba5 (a move in the right direction).
3) The b3-square can become useful.
None of these arguments is strong enough on its own, but the main thing is to direct focus to the queenside, where Black
has a sound pawn structure.
Given this, Grischuk could have played 14.a4, but it is, as always, also a question of time.
15.a4 Bg4
16.0-0-0!
A surprising decision, since it’s still possible for Black to open up with ...c7-c6. The reason behind castling long is
that the normal 16.Be2 runs into 16...Bxf3! 17.Bxf3 exd4 18.Bxb7 Rab8 19.Bc6 dxe3! sacrificing the exchange. After
20.Bxe8 exf2† 21.Kf1 Rxe8 Black has nice compensation since White can’t attack the pawn on f2.
16...exd4?!
Exchanging pieces increases the impact of the doubled pawns.
Grischuk said that Black needs a miracle to survive, and Giri explained that he completely underestimated the danger.
The only counterplay is ...Nf8-d7-c5xa4, but the pawn is easy to defend and White breaks through on the kingside
before the knight has time to return. Black’s king can’t simultaneously defend f6 and h7.
In another Rossolimo, Black takes early on c3 to avoid the quick attack with Nd5 and Qc2.
Eljanov, September 2016: “I think it was a nice strategical game. But the opening was harmless...”
5...Bxc3 6.bxc3(!)
The main move is 6.Qxc3 when (for example) 6...Qe7 7.a3 d5 8.cxd5 Nxd5 9.Qb3 Nb6 10.d3 gives a position where
White plays the Hedgehog with the pair of bishops.
It’s in line with the main strategy in this book: having a long-term academic plus that needs some work before it is
felt.
However, the more I tried to find a move order I was happy with, the less I liked the position. Black always had
different things in mind and the risk is that we never reach the “long term”.
6.bxc3 is an old favourite of Mihai Suba. InKhalifman’s Opening for Black According to Karpov (2001), it’s written off
as “illogical” – and in general it’s less attractive to double the pawns when e2-e3 has been played. In Pawn Structure 5,
it was said that Black is happy in this kind of position, and here we have even spent a tempo on the queen. But that
actually makes the difference!
With the queen on c2, Black has only one chance of playing ...e5-e4 (now!) and that’s in a position where he can’t
support the pawn with ...Re8. If he delays the move, White releases the dark-squared bishop with 7.e4.
1) With the f6-knight on e7, he could have played actively with ...f7-f5.
2) With the c6-knight on b8, it would have been possible to play ...c7-c6 and ...d7-d5.
For example, 1.c4 e5 2.Nc3 Nf6 3.g3 Bb4 4.Bg2 0-0 5.e4 Bxc3 6.bxc3 is a popular line, where Black’s best is to break
with ...d7-d5 after 6...c6.
6...0-0
Black collapses after 6...e4 7.Ng5 d5? 8.f3 h6 9.cxd5!, but he can play 6...e4 7.Ng5 Qe7. After 8.f3, we can note that
the extra tempo by being White (e2-e3) is important.
1) With the rook on e8 (instead of ...Qe7), f2-f3 would have been bad due to a queen check on h4.
2) 8...e3 is illegal since there is a pawn on e3.
7.e4! d6 8.g3
Eljanov postpones the decision whether to play for d2-d4 or f2-f4 until he has completed development. 8.g3 would
not have been possible if Black was ready with ...f7-f5.
8...Ne7
8...Ng4 is given in Khalifman’s book. It seems like a good idea to push the f-pawn, but the line given is 9.h3 Nh6
10.d3 Qf6 and the evaluation “with equality” seems optimistic. To me it simply looks like Black has misplaced his
pieces.
In Attacking the English/Reti (2016), Alexander Delchev points out that White’s fianchetto allows Black to play for
...a7-a6 and ...b7-b5. He writes: “Of course, it should be combined with ...f5” and gives 8...Kh8 as in Suba – Delchev,
Albacete 2005, which continued with 9.d3 Ng8 planning ...Nge7.
9.Bg2 Ng6
The knight does nothing in itself, but is posted on g6 to stop Nf3-h4-f5.
13...b6
14.a4!
The knight on g6 would have liked to have a quick journey back to the queenside.
16...Rc8 17.a5
17...bxa5
17...b5 was also interesting. 18.cxb5? Qxa5 would be a failure, but there are several promising options. One is the
nice-looking: 18.d5 cxd5 19.exd5 Bd7 20.cxb5 Qxa5 21.c4 Bxb5 Necessary to avoid a very passive position. 22.Rxb5!
Qxb5 23.Qxg6! Rxc4! (a pawn better than 23...fxg6) 24.Qd3 Rb4 25.Qxb5 Rxb5
18.Ba3 c5
Bruzon had to play this move sooner or later to avoid d4-d5 followed by c4-c5.
19.dxc5! dxc5
The knight will reach d5 via f1 and e3, and it’s difficult to believe that Black can avoid taking it, even though Bruzon
tries in the game.
In the final game, we come to the normal reversed Sicilians. 4...Be7 may look passive, but it’s a sound developing move
and Black doesn’t need higher aims than that.
It’s correspondence chess, but in contrast to Schön – Gerola (Chapter 12), I have dared to attach a question mark to a
move. Maybe it’s because this game was played in 1986...
1.c4 e5 2.Nc3 Nf6 3.Nf3 Nc6 4.e3 Be7 5.d4 exd4 6.Nxd4
6.exd4 d5 is also an option and gives the same IQP-position as in the Exchange French. It’s important that Black can’t
play ...Bb4 in one move.
6...0-0 7.Be2
White can try to stop ...d7-d5 by 7.Nc2, but Black will manage anyway after 7...Nb4!.
7...d5!
Black can’t delay this pawn lever if he wants to achieve it.
8.Nxc6
Instead 8.Nxd5! Nxd5 9.cxd5 Qxd5 10.Bf3 and 11.Nxc6 gives White a better structure.
10...cxd5 11.b3
The engines prefer Black’s well-placed pieces over White’s small structural advantage. I don’t agree. In the long
term, White gets pressure on the c- and d-files, and Black can’t achieve ...a5-a4 easily to exchange one of the weak
pawns. And since White doesn’t have any weaknesses on the kingside, I think it’s possible to contain Black’s piece
activity.
However, don’t ask me to play this position against a computer.
11...Qe7 12.Bb2
12.Nxd5!? was also possible. Black is a rook up after 12...Qe5 13.Nxf6† gxf6 14.f4 Qxa1, but 15.Qc2 traps the
queen.
The material will be balanced, but Black’s open king is an important factor. If the queen escapes with 15...Bxf4
16.Rxf4 Qe5, she will not be able to defend the dark squares, and there is no other piece helping her.
12...Rd8?
By allowing the exchange of his dark-squared bishop, Black loses his activity. 12...c6 was practically forced.
White simply doubles on the c-file and Black collapses. No more comments are required; the game speaks for itself.
16.Qd4 Qb6 17.Rfc1 Ne8 18.Rc5 c6 19.Rac1 Rd6 20.Qa4 Rc8 21.Bd4 Qb7 22.Ra5 a6 23.Rxa6 c5 24.Rxd6 Nxd6
25.Rxc5
White has two pawns more and Black could have resigned. In the game he delayed it until move 46, so the game
lasted six years. What a painful period it must have been!
...1–0
Recap
The games had few similarities with each other, which illustrates the wide range of positions that can appear in the
English. There are more in the theoretical section: the Rossolimo, the Alapin, the Open Sicilian, the Grand Prix and the
King’s Indian Attack.
Theory
Rossolimo
This is a Hedgehog position where the knight is happier on g3 than it would have been on f3. It’s about equal.
After fifteen moves, Aronian started to attack and Grischuk commented at the press conference: “If I had more time I
wanted to go to this confession room and confess that I’m a complete idiot.”
But he played well and the game ended in a draw.
5.a3 Be7!
5...Bxc3 6.Nxc3 d5 7.cxd5 Nxd5 is also possible. Compared to a line we avoid (1.c4 e5 2.e3 Nf6 3.Nc3 Nc6 4.Nf3
Bb4 5.Qc2 Bxc3 6.Qxc3 Qe7 7.a3 d5 8.cxd5 Nxd5), White’s queen doesn’t lose additional time. He usually plays
8.Qc2 and decides between Bd3 and Be2 on the next move.
7...Nc6
7...c6! was played in Grachev – Nyzhnyk, Turkey 2013. I suggest putting the bishop on its best diagonal with
8.g3!?N. This move is not mentioned by Mikhalevski in Beating Minor Openings. A fitting evaluation is the usual
“potential, but not an advantage”.
8.Qd1
It’s difficult for Black to open the position before White has developed with Nf4, Be2 and 0-0. In practice, White has
scored 70% over 15 games.
Alapin
8.d4 e4
Nimzowitsch – Spielmann, Berlin 1928, saw 8...Bd7 but Black can’t be satisfied with a passive move like that. I
propose 9.0-0 0-0 10.Bd3N with the same play as in similar positions.
9.Nd2 Qg5!
10.Bf1 Qg6 11.h4
11.g3 was Gajewski – Tomczak.
11...h5
11...0-0 12.h5 makes life harder for the pawn on e4. White has Rh4 as an option, and there’s no good square for the
queen. On f5, she blocks both ...Bf5 and ...f7-f5, and on e6 she has to watch out for the pin with Bc4.
12.g3N
The idea behind including the moves with the h-pawns is shown if Black castles.
12...0-0 13.Be2!
Black has only one way of keeping the pawn.
Grand Prix
1.c4 e5 2.e3
2.Nc3 Nf6 3.Nf3 Nc6 4.e3 is a good move order if we know what our opponent plays. The main point is to avoid
3...Bb4.
2...Nc6 3.Nc3
3.d4?! seems clever, as d4-d5 hits the knight. But 3...exd4 4.exd4 Bb4† 5.Nc3 Nf6 6.d5 is too much after 6...Qe7†
7.Be2 Ne5 when White has problems with development.
3...f5
This advance is premature when the knight is still on g1.
4.d4 e4 5.Nh3
Black has given up the f4-square and White develops with Be2, 0-0 and plays f2-f3. Against ...Bb4, he plays Bd2 and
a2-a3.
Old Indian
Delayed Rossolimo
5...0-0
The most common move is 5...Bxc3 when Eljanov played 6.bxc3(!) against Bruzon instead of the more usual move
6.Qxc3.
5...d6 is seldom played, but holds a clever idea: to avoid 6.Nd5 and 7.Qf5 (5...a5 is another move with the same idea).
6.a3 Bxc3 7.Qxc3 is not that bad since, in the 5...Bxc3 6.Qxc3 line, White usually plays a2-a3 anyway, and now Black
will lose a tempo if he opens the position with ...d6-d5. But there is also a short-term matter of tempos, and 7...Bg4 is
annoying.
A better plan for White is 6.Be2 0-0 7.0-0 when the next move could be d2-d4 or Nc3-d5. 7...Bxc3 is met by 8.Qxc3
e4 9.Ne1 followed by exchanging the pawn on e4, and after 7...Re8 8.Nd5 we have two sample variations:
a) 8...Bc5 9.Ng5! with the second knight landing on e4. White doesn’t take on c5, but simply develops with a2-a3, b2-
b3/b4 and d2-d3.
When I played Jon Ludvig Hammer in the 2016 Nordic Championship, I had prepared in the same way as in all my
other White games in the tournament: by reading the draft of this book. (And varying the move order: 1.Nf3, 1.c4 and
1.d4 in the first three games and the plan was to start with 1.e3 in the fourth. But when I woke up the next morning I
had somehow changed my mind.)
When Hammer replied to 1.c4 with 1...e5 2.e3 Nf6 3.Nc3 Nc6 4.Nf3 Bb4 5.Qc2 d6 6.Be2 0-0 7.0-0 Re8 8.Nd5 e4
9.Ne1 a5, I was out of book, but it was not that hard to find a downside to delaying ...Bc5. Black doesn’t threaten
...Nxd5 plus ...Nb4 and there is thus time to fianchetto the bishop:
10.b3! Bc5 11.Nxf6† Qxf6 12.Bb2 Qg6
Now I wrongly dismissed 13.d4 (in favour of 13.f3) since I thought that it’s not kind to my knight on e1 to leave the
e4-pawn on the board. But it could escape with g2-g3 and Ne1-g2. Black’s pieces are not too good either, except for
the light-squared bishop potentially on h3 – but that one could also be attacked with Ng2-f4.
Nevertheless, after 13.d4 I think Black should keep the structure the same with 13...Ba7. As instead after 13...exd3
14.Bxd3 White has all the pluses: better bishops and a better structure.
My game against Hammer was my last one before a six-month break, and in order to avoid being a loser for such a long
time, I decided I had better win... Now, one day later, I am starting to question if the decision to take a break was wise.
6.Nd5
6...Re8
Black has a few other moves, but nothing to be afraid of.
6...Be7 7.a3 stops Black from playing 7...Nxd5 8.cxd5 Nb4 9.Qb3 c6. After 7...d6, White can either go 8.Bd3 followed
by Nxf6† and Be4, or 8.Be2, to take the bishop pair. I tried the former against Francesco Rambaldi in the Top 12 (the
French Team Championship) in 2016, and will go for the other move next time.
6...a5 7.a3 Bc5 8.Bd3 is a standard move; the bishop will be strong on e4.
7.Qf5 d6!
7...Nxd5 8.cxd5 loses a pawn, or the game after 8...d6 9.Qh5! Ne7 10.Ng5.
8.Nxf6† gxf6
See Grischuk – Giri for 8...Qxf6.
9.Qh5
After the game against Grischuk, Giri chose the other side of the board. In Qatar 2015, he played 9.Qc2 e4 10.Ng1,
winning the doubled f-pawn on move 27 and the game on move 30.
Black’s king is open, but he is the one who must play actively to punish White before he catches up in development.
But White is up for the challenge.
The following lines are more computer-based than in other parts of the book, but at least using Informant-symbols will
give you the confidence you can’t get by looking at the position.
9...e4!
After 9...d5 10.Bd3! e4 11.cxd5 Black has three options:
a) 11...Na5N 12.Bc2 exf3 13.gxf3!± with 14.Qh6 being the main threat.
b) 11...Nd4!? 12.Nxd4 exd3 with a positional advantage when the d5-pawn drops off.
10.a3!
Black is not allowed to retreat the bishop to f8 after ...Re5 plus ...d6-d5.
5.d4 cxd4
5...e4 transposes.
6.exd4 e4 7.Nd2
White tries to exchange the d4-pawn against Black’s e-pawn, to get a favourable structure.
7...Bb4 8.Be2(!)
A new concept that’s only been tried in a correspondence game.
8...0-0
Instead 8...Nxd4?! 9.Ndxe4 gives White an initiative since 9...Nxe2 10.Qxe2 Nxe4 11.Qxe4† comes with check and
10...0-0 11.Bg5 Be7 12.Bxf6 Bxf6 13.Nd5 is an illustrative example of two tremendous knights.
9.0-0
Black must decide if he should take on d4 or defend the e-pawn.
a) 9...Bxc3 10.bxc3 is dubious. White can play c4-c5 plus Nc4, f2-f3 or Re1 and Nf1. Black should at least improve
with 9...Re8 10.a3 Bxc3.
White has a positional edge due to the d5-square since the tactics work after: 12...d5 13.Rd1! Re8 14.Rxd5 Qe7
15.Bd2! Black’s back rank makes it impossible to cash in on e4.
c) Tsenkov – Braun, corr. 2010, went 9...d5! 10.a3 Ba5 11.cxd5 Nxd4 12.Ndxe4 Nxe2† 13.Qxe2 with a draw in this
position. The reason is that Black can play 13...Nxd5 14.Rd1 Nxc3 15.Nxc3 with an equal position. White’s
development and the strong knight are balanced by the pair of bishops.
After 1.c4 e5 2.e3 it is too early to play 2...c5?! without the knight on c3 or f3 because:
After 3.d4! exd4 4.exd4 cxd4 5.Nf3, it’s easy to handle Black’s potentially disruptive moves ...Bb4† and ...Qe7. A
plausible line is: 5...Bb4† 6.Bd2 Qe7† 7.Qe2 Bxd2† 8.Nbxd2 Nc6 9.Nb3 The d4-pawn will soon fall, leaving Black to
suffer with his remaining isolated d-pawn.
Miscellaneous
5.d4 g6
There are several move orders with the same idea: allowing d4xe5 followed by Qxd8†.
6.Be2(!)
6.d5 is a bad King’s Indian for White.
Two world champions have lost as Black after: 6.dxe5 Nxe5 7.Nxe5 dxe5 8.Qxd8† Kxd8 9.b3 The g-pawn would have
been better on g7 and the f-pawn on f5, but his pawn structure is nevertheless better. And both Hübner – Anand,
Munich 1994, and Illescas Cordoba – Karpov, Villarrobledo 1997, were just rapid chess.
6.Be2(!) is a clever move with the idea that White gets a better version of 6.dxe5 after the following moves.
6...Bg7 7.dxe5 Nxe5 8.Nxe5 dxe5 9.Qxd8† Kxd8 10.b3
Exercise 1
1.c4 e5 2.Nc3 Nf6 3.d3 Bb4 4.e4 Bxc3† 5.bxc3 0-0 6.g3
Black to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.e3 e5 2.c4 Nf6 3.Nc3 Nc6 4.Nf3 d5 5.cxd5 Nxd5 6.Bb5 Nxc3 7.bxc3 Bd6 8.d4 Bd7 9.0-0 0-0
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.c4 e5 2.e3 Nf6 3.Nc3 Nc6 4.Nf3 Bb4 5.Qc2 0-0 6.Nd5 Re8 7.Qf5 d6 8.Nxf6† gxf6 9.Qh5 e4!
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 4
1.e3 e5 2.c4 Nf6 3.Nc3 Bb4 4.Nge2 0-0 5.a3 Be7 6.d4 exd4 7.Qxd4 Nc6 8.Qd1 d6 9.Ng3 Be6 10.Nd5
Black to move
Show/Hide Solution
Show/Hide Solution
A) 4...Bg4 5.h3 Bh5 6.Qe2†! Qe7 (6...Be7) 7.Be3 Nc6 8.Nc3 0-0-0 (8...Bxf3) 9.g4! Bg6 10.0-0-0 Nb4 (10...Nf6)
11.Rd2 f6 12.Bg2! Bf7 13.Ne1N 356
B) 4...Nc6 5.Bb5 Bd6 (5...Nge7) 6.c4 dxc4 7.d5 a6 8.Ba4 b5 9.dxc6 bxa4 10.0-0 Ne7 11.Nbd2 357
11...0-0 358
11...c3 358
Move Orders
It’s possible to find several creative move orders leading to the Exchange French. Here are three:
We will see some of them in the illustrative games, but 1.e3 e5 is the only way our repertoire reaches the Exchange
French.
1.c4 e6 2.e4 d5 3.exd5 exd5 4.d4 is worse, since White should delay c2-c4 until Black has played a passive move or
developed his dark-squared bishop. Otherwise he gets the possibility of 4...Bb4†.
Pawn Structures
The Exchange French has a dull reputation, and quite rightly. White’s aim is often a better bishop, after exchanging the
dark-squared bishops and keeping the light-squared one on d3. He might even get the e-file and could worry Black for a
few moves before agreeing to a draw.
However, things changed when Garry Kasparov used the Exchange French a handful of times in the 1990s. Instead of
limiting himself to symmetrical positions, he pushed c2-c4 and played with an isolated pawn.
However, it would be silly to claim that White can seriously play for an advantage. Instead, I think this is mainly a
good choice against players who don’t have any experience with this opening or this structure. And they are not so few!
Structure 1
White gets an isolated pawn in many chapters of this book, and there is an introduction given with the first structure
in the Queen’s Gambit Accepted (page 267). This diagram differs from the normal positions since Black has a pawn on
c6 instead of e6.
Pluses
a) After ...c7-c6, Black can’t put a knight on c6 to put pressure on the isolated pawn, so White has greater freedom in
how to arrange his pieces.
b) If Black plays ...Nc6 instead of ...c7-c6, he will have difficulties controlling the d5-square. But even so, that is often
what he does.
c) The open e-file makes it easier for White to exchange the heavy pieces. This is not a threatening plan, but a way to
avoid a bad endgame.
The conclusion is that the position is easier to play for both sides! Whether this is good news may depend on your state
of mind, but I would say that it favours Black – the point of having an isolated pawn is to create trouble.
But as usual, it is about making the pluses more important than the minuses. With the list above, it seems
contradictory that Nf3-e5 is a standard move. Wasn’t Black okay exchanging there? The problem is that if White hasn’t
wasted tempos, it’s difficult for Black to find the ideal situation he’s looking for.
a) He may get forked with d4xe5, threatening a d6-bishop and f6-knight.
b) Parting with the bishop pair is always a concession.
c) It’s not easy to find a good square for the f6-knight.
So usually Black leaves the knight on e5 untouched. White will then follow up with Qf3, a move that is possible only
since Black has no pressure on d4. And then we have the answer to how to make the pluses more important than the
minuses.
Structure 2
The chance to play a successful c3-c4 (or ...c6-c5) diminishes with every move. Once development is completed, it’s
easier to handle the activity that the isolated pawn generates. Instead, White will try to create something with the pieces
in this symmetrical position. An ideal set-up looks like this:
Normally, White castles short, but players who want to create more unbalance can consider the other way. The risk is
however that Black does the same! And then, he is actually safer since it’s less threatening to advance the kingside
pawns.
Since the position is symmetrical, Black has in general the same objectives as White. But hopefully it will be possible to
keep using the tempo until it suddenly forces Black to make a passive move.
If Black is playing for equality, he has another plan though: exchanging pieces. And if he manages to exchange two
pairs of minor pieces without conceding anything else, the position becomes as dull as its old reputation.
Games
Lysyj: “I think the Exchange French is really underrated nowadays. But in the 90s it was one of the favourite lines of
Kasparov!”
4...Bd6 5.c4
Only now, when the bishop has already been developed.
7.cxd5! Re8†?!
This rook move is not the first priority, and White’s bishop will develop to e2 anyway.
7...Nbd7 is discussed in the theoretical section.
Lysyj: “I remember that Boris spent a lot of time in the opening (the Exchange Variation is not so researched!) but he
didn’t achieve an equal position.”
12...Bxf6 13.Qb3 a6
13...Nxd5?! works tactically, since e2 is hanging, but after 14.Bc4 the pawn on f7 is attacked once again and Black
has to give up a pawn with 14...c6.
Black threatens 23...Bd5 again, but White is happy to push the d-pawn.
26...Bxf3 27.gxf3
27...Qg5†
After 27...Qxf3 White should avoid 28.d7 Rc6 29.d8=Q† Rxd8 30.Rxd8† Kh7 when he should be happy to avoid
mate and get a draw after: 31.Rd3 (31.Qxc6?! Qxc6 can never be better for White with the open king. Rooks are not
good defenders and Black would push the pawns on the kingside.) 31...Qg4† 32.Rg3 Qxg3† 33.hxg3 Rxa6
Instead better is 28.Qd3! Qxd3 29.Rxd3 Rxb2 30.d7 Rd8 31.Re1 Kf8 (31...Rbb8 32.Re7 Kf8 33.Rde3 transposes)
32.Rde3! Rbb8 33.Re7 when Black is doomed to eternal passivity and White can try to penetrate with the king. But it’s
probably a draw.
28.Rg4?!
A step in the wrong direction. White should keep the rook behind the d-pawn.
34.Qh5
The threat is 35.Rh4.
The next game starts as a Queen’s Gambit Accepted, but the position after six moves can also be reached from the
Exchange French. The e3 poison move order is 1.e3 e5 2.d4 exd4 3.exd4 d5 4.Nf3 Bd6 5.c4 dxc4 6.Bxc4.
Fier, March 2016: “When I played 17.f4 I had around 1:35 on the clock, five minutes more than at the start.”
Semcesen, March 2016: “A funny quote? The game was not funny at all! ”
1.d4 d5 2.c4 dxc4 3.e3 e5 4.Bxc4 exd4 5.exd4 Bd6 6.Nf3 Nf6 7.0-0 0-0 8.h3 Nc6 9.Nc3
9.Qc2? stops ...Bf5, but it’s shown to be premature after 9...Nb4 when the knight continues to d5.
9...h6
There is a book about the Queen’s Gambit Accepted that recommends Black imitates White. When White moves the
bishop, Black does the same, when White creates luft, Black also needs to breathe.
9...Bf5 solves the problem with the bishop, but 10.Bg5 is annoying. 10...h6 11.Bh4 g5 12.Bg3 Bxg3 13.fxg3!
10.Qc2 Nb4
Black’s bishop has problems developing and 10...Be6 11.Bxe6 fxe6 is not nice, as shown in Smith – Lindgren in
Chapter 11 (page 191). Both sides have an isolated pawn, but White has more space and will find better squares for his
pieces.
10...a6 is given as an alternative in Vitiugov’s The French Defence Reloaded (2012). White plays 11.a3 to be able to
retreat with Bc4-d3 (after ...b7-b5) without stepping into ...Nb4.
10...Na5 11.Bd3 Be6 gives Black a knight on the rim, but also a happy bishop.
11.Qb1 c6
12.Bxh6!
Not mentioned in The French Defence Reloaded, but it is in A Practical White Repertoire with 1.d4 and 2.c4 by
Alexei Kornev (2013).
17.f4!
White prepares 18.Ng5.
17...Bf5 18.Ng5
18...Kg8?!
The best chance was 18...Bxg5 19.fxg5 Bg6. White defends the pawn on g5 with 20.h4 and Black collapses if the
rook on a1 enters the game. Black’s only chance is to immediately go for the d4-pawn, but 20...Nb6 21.Nxf7† Bxf7
22.Bxf7 Qxd4† 23.Kh1 Qg7 24.Qh5 would see White getting the piece back with g5-g6. He is winning due to the
exposed black king.
19.g4?
Better was 19.Rf3!. The rook is on the way to g3 with decisive effect.
19...Qb6?
Giving back the piece is hopeless.
It was possible to play 19...Bxg5 20.fxg5 Bg6 21.h4, since compared to 18...Bxg5, White’s king is open. That tells after
21...Qb6 22.Rad1 Qxb2! 23.h5 Be4 24.Rf2 Qc3 although White would still make a draw.
23.f6!
Who is afraid of the wolf? 23.f6 is the fastest way to mate – Black has only one check.
While the first two games were theoretical lines, the next gives the kind of position that has given the Exchange French
its dull reputation. White’s play is slow, but the game shows that it’s not so easy to handle even though it looks
innocent.
Nielsen, April 2016: “It’s a board full of pieces, and despite the symmetrical pawns, there are chances for the better
player I would say.”
Short, April 2016: “I thought it was a good choice against a player who was highly booked up. Such players often can’t
take such innocuous lines seriously. (He drifted into a poor position quite rapidly, in fact, and was then suffering for
the rest of the game.)”
1.e4
1.e3 e5 2.d4 exd4 3.exd4 d5 4.Nf3 Nf6 gives the same position, but with two moves fewer played.
1...e5 2.Nf3 Nf6 3.Nxe5 d6 4.Nf3 Nxe4 5.d3 Nf6 6.d4 d5 7.Bd3 Be7?!
The pin with Bg5 is not so dangerous, so 7...Bd6 is preferable.
8.h3
This position was reached by Garry Kasparov against the Argentinean IM Jorge Alberto Rubinetti in a clock
simultaneous exhibition in Buenos Aires in 1992. Kasparov drew the game, but won the match 9–3.
8...Nc6
Nielsen breaks the rule of avoiding placing the knight in front of the c-pawn (in positions with a pawn on d4), but this
should not be taken seriously. The idea behind the knight move is to discourage White from playing c2-c4 by putting
pressure on the d-pawn. The downside is that it stops Black from getting an isolated pawn with ...c7-c5.
13.Bf4 Rxe1† 14.Qxe1 Bd6 15.Qd2 Be6 16.Re1 Bxf4 17.Qxf4 Nh5 18.Qe3 Nf6
I don’t think Peter Heine Nielsen hoped to get a draw by repetition. The reason behind losing the tempo was to get the
d6-square for the queen.
23.c3
23...Re8
This time, White’s pawn sacrifice involves a deep tactical shot: 23...Qxb2 24.f5! Bd7 25.Nxd7 Nxd7 26.f6! Nxf6
27.Nh5!
23...Qxb2! deserves an exclamation mark, and the engine destroyed the illusion that Black played the whole game
without getting any counterplay. Thanks for that.
24.b4
Who said that the position is drawish? White has taken space on both wings and will continue to advance the pawns.
28...hxg5?!
Black should have defended the rook with 28...Bd7, even though both 29.f6 and 29.gxh6 look scary.
29.Qxg5 f6 30.Qh4?
Instead 30.Qg4 eyes g7. I guess Short wanted to avoid 30...g5, possibly followed by the exchange of queens on f4.
But that would just allow White to pick up the pawn.
30...Bd7 31.c4 Nc8?!
The engines claim that Black has time to go for a pawn with 31...Qa6! and best play should lead to a repetition. Let’s
continue to the critical position in the game without going into the details.
32.Rxe8† Bxe8 33.c5 Qe7 34.Kf2! a6 35.Ne2 Bb5 36.Bxb5 axb5 37.Qf4 Qd7 38.g4 Ne7 39.Qf3 Kf7 40.Nf4 c6?
There are only two pieces each on the board, but the squares on e6 and g6 are dangerous enough.
41.Qh1
Playing for the spectators. 41.Qh3 gives the same.
Black is in zugzwang and loses after: 44...Nxg6 45.fxg6 Qe6 46.Qh7† Kf8 47.Qh8† Qg8 48.Qh2! Qe6 49.Kf2!!
Another zugzwang. Black’s queen can’t keep control over both e8 and g8.
46.fxg6
The queen delivers mate on b8.
1–0
Recap
Black has several ways to get a decent position, but he must know what he’s doing. As a consequence, don’t play 1.e3
e5 2.d4 against players who have the French in their repertoire. It’s the same logic as behind 1.g3 e5 2.e4, tricking
Black into a Glek.
But if the trick is to do its job, we must know how to make the most out of the position. Otherwise it’s pointless.
White often reaches an isolated pawn position, but with a pawn on c6 instead of e6. This makes it easier for Black to
develop or exchange the light-squared bishop, but also stops him from putting pressure on the pawn. White uses the
freedom to play Nf3-e5 followed by Qd1-f3, leaving d4 undefended.
But c2-c4 is not so strong as long as Black can develop the bishop to b4 in one move, so we have to play other
positions as well. And that’s the subject of the next section.
Theory
3.exd4 d5
In the chapter about the King’s Indian (Chapter 5), Black gained an acceptable position in the structure without the e-
pawns by fianchettoing the bishop. There are two differences. 3...Nf6 The first is the annoying check after 4.Nf3 g6
5.Bd3 Bg7 6.Qe2† and the second is that White isn’t obliged to play c2-c4.
4.Nf3(!)
Black has four main options: A) 4...Bg4, B) 4...Nc6, C) 4...Bd6 and D) 4...Nf6.
A) 4...Bg4
Recommended by John Watson in Play the French.
5.h3 Bh5
6.Qe2†!
Kasparov’s invention. The idea is that Black must put his queen in between, to avoid the check on b5. And then
White can fianchetto his king’s bishop, while Black has a bishop on g6 in the way.
6...Qe7
GM Simon Williams mentions 6...Be7 7.Qb5† Nd7 but doesn’t believe in the sacrifice after 8.Qxd5! Ngf6 9.Qb3,
defending against ...Bxf3.
8...0-0-0
When I was a guest player in the ‘Fight of the Titans’ in October 2015, my opponent took a pawn with 8...Bxf3
9.Qxf3 Nxd4 but after 10.Qd1 his position collapsed in a non-titanic way. 10...Nf5? 11.Qxd5 Nxe3 12.Bb5† winning
the exchange.
9.g4!
Accurate, as it avoids 9...f6 and 10...Bf7.
9...Bg6 10.0-0-0
We can now see why it was useful for White to include the check on e2. Black has to move his queen to develop the
bishop on f8. White also has a simple plan with Ne5 and f2-f4.
10...Nb4
Or 10...Nf6 11.Ne5! and if Black takes the pawn, he gets his bishop trapped with f4-f5.
11.Rd2
Watson gave 11.a3 as the only critical move, but I don’t understand why.
B) 4...Nc6
5.Bb5
5...Bd6
5...Nge7 is recommended in The Flexible French (2008). After 6.0-0 a6 7.Bd3 Bf5 Moskalenko thinks that White has
played dubiously and cites an internet blitz game. Black plans to castle long, but to me the hook on a6 seems to give
White good attacking chances. It’s possible to prepare the attack right away with 8.Re1 Qd7 9.b4!?.
Kasparov took on a4 immediately against Bareev and won due to his better pawn structure, but the main difference of
starting by picking up the c-pawn is that 11...0-0 12.Nxc4 creates a threat on d6. After 12...Nxc6 13.Qxa4, White gets
another threat and continues with quick development. In comparison, 11.Qxa4 would have allowed Black to be
disruptive with ...Rb8-b4.
11...c3 is Black’s other try. After 12.Ne4N 12...0-0 13.Qxa4 Rb8 White takes with the pawn on c3 and avoids ...Rb4.
And 13...cxb2!? 14.Bxb2 Rb8 15.Nxd6! Qxd6 16.Ba3 is annoying. Better is 15...Rxb2 16.Nxc8 when White has a
slight edge due to the pawn on c6.
C) 4...Bd6
Simon Williams recommends 4...c6 in Attacking Chess: The French (2011), but I don’t see any independent value in
that move. After 5.Bd3 Bd6 6.0-0 Black has to put the knight on e7 and 6...Ne7 7.c4 transposes to 4...Bd6.
5.c4 Nf6
I will mention a couple of alternatives:
5...dxc4 6.Bxc4 Nf6 reaches a Queen’s Gambit Accepted, like Fier – Semcesen.
5...c6 6.Nc3 Ne7 7.Bd3 If Black doesn’t take on c4, White soon plays c4-c5 followed by b2-b4. 7...dxc4 This is one of
Watson’s recommendations. We reach one of the exercises after 8.Bxc4 0-0 9.0-0 Nd7 10.Re1 Nb6 11.Bb3 Nbd5 (see
page 361).
6.Nc3
6...0-0
A pawn sacrifice recommended in some French books.
6...dxc4 7.Bxc4 0-0 is a clever move order, having waited for 6.Nc3 before taking on c4. White is not in time to stop
8...Bg4. 8.0-0
b) Black could also try to place the bishop on f5, but after 8...c6 9.Re1 Bf5 10.Ne5 Nbd7 White has 11.Qf3!, which
wins the bishop pair; Qd1-f3 is a standard move in this structure.
c) 8...Bg4 9.h3 Bh5 10.g4 Bg6 If White’s initiative evaporates, he will regret his weakening moves. See the second
exercise for further analysis (page 361).
7.cxd5 Nbd7
10.Be2 Be7!
Black wants to take back the pawn without playing the weakening 10...g5.
Black could transpose to Lysyj’s game, having avoided Bb5. In any case, it’s up to Black to prove that he has
compensation for the pawn.
D) 4...Nf6 5.Bd3
Since Black still has the possibility of giving a check on b4, we limit ourselves to a solid set-up. But at least we get a
version where we might pin Black’s knight with Bg5.
5.Bg5?! should not be played while Black can still place the bishop on e7. 5...Be7 6.Bd3 h6 7.Bh4 Nh5 even gives him
the initiative.
5...Bd6
5...Be7 was Short – Nielsen.
Black can play for an isolated pawn with 5...c5 but White has 6.0-0! when the check on e1 forces Black to play: 6...c4
7.Re1† Be7 8.Bf1 0-0 9.b3 cxb3 10.axb3
White can get a nice centre and create a passed pawn on the d-file with c2-c4, but he has to take care to make sure that
Black doesn’t get too much pressure with ...Nc6, ...Bg4 and ...Bb4. The following two lines are relevant.
a) 10...Nc6 11.Bd3! We start with c2-c3 and complete development before we consider c3-c4. 11...Nb4 12.Ba3! is not a
problem.
b) 10...Bf5 11.c4! White is exactly in time to fortify the centre: 11...Nc6 12.Nc3 Bb4 13.Bb2 Bg4 14.Re3 And Black
can’t win the pawn. 14...dxc4 15.bxc4 Bxf3 16.Rxf3 Qxd4 17.Qc2 The massive threats against the knight on f6 decide.
6.Qe2†!
After 6.0-0 0-0, the drawing ratio between strong players is 95%, with White winning 3% and Black 2%. It sounds
alarming, but they don’t even try to win! 60% of the games have been 13 moves or shorter.
Playing well is a difficult task and with only one pawn exchanged, there is every possibility to win or lose. Alexander
Grischuk has tried 7.Bg5 Bg4 8.Nbd2 Nbd7 9.c4!? to break the symmetry. White plans c4-c5, and the position should
be about equal if Black captures on c4.
6...Qe7
6...Be6 7.Ng5 picks up the bishop pair.
6...Be7! is not as common. It’s solid, but at least White has gained a little something.
7.Qxe7†
Not counting quick draws, White has 3/4 here. The following moves are Bf4 (in case of ...Bxe7), 0-0 and Re1 and the
initiative is bigger than one might think. A probable line is 7...Bxe7 8.Bf4 c6 9.0-0 h6 10.Re1 Be6 11.Ne5 Nbd7
12.Ng6 with the bishop pair.
Exercise 1
1.e3 e5 2.d4 exd4 3.exd4 d5 4.Nf3 Bd6 5.c4 c6 6.Nc3 Ne7 7.Bd3 dxc4 8.Bxc4 0-0 9.0-0 Nd7 10.Re1 Nb6 11.Bb3
Nbd5
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 2
1.e3 e5 2.d4 exd4 3.exd4 d5 4.Nf3 Bd6 5.c4 Nf6 6.Nc3 dxc4 7.Bxc4 0-0 8.0-0 Bg4 9.h3 Bh5 10.g4 Bg6
White to move
Show/Hide Solution
Show/Hide Solution
Exercise 3
1.d4 d5 2.c4 dxc4 3.e3 e5 4.Bxc4 exd4 5.exd4 Bd6 6.Nf3 Nf6 7.Qe2† Be7 8.0-0 0-0 9.Nc3 Bf5 10.Bf4 Bd6 11.Ne5
Nc6
White to move
Show/Hide Solution
Show/Hide Solution
A common mistake when studying openings is to focus on details deep into the lines, and forget the first moves. One
kind of exercise that has worked well in my coaching experience is to pick a random diagram in the book (or one of the
exercises) and try reconstructing the moves that lead to the position, while explaining them. It forces active learning,
thus cementing the knowledge in a way that pure repetition never can.
Try it! The only objection I know is that it’s easier with a coach, to discuss if an alternative move order is logical or
not.
2) Compare (4 exercises)
Which of the two diagrams should White prefer?
Compare 1
Compare the following positions and select the one you think White should prefer.
1) 1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 c5 6.d5 e6 7.Nc3 exd5 8.cxd5 d6 9.0-0 Bg4 10.Nd2 Bxe2 11.Qxe2
1. Black to move
2) 1.d4 Nf6 2.Nf3 g6 3.c4 Bg7 4.e3 0-0 5.Be2 c5 6.d5 e6 7.Nc3 exd5 8.cxd5 d6 9.0-0 Bg4 10.h3 Bxf3 11.Bxf3
2. Black to move
Show/Hide Solution
Show/Hide Solution
Black or White 1
Is it White’s or Black’s move?
Show/Hide Solution
Show/Hide Solution
Black or White 2
Black or White 3
Is it White’s or Black’s move?
Show/Hide Solution
Show/Hide Solution
Black or White 4
Compare 2
Compare the following positions and select the one you think White should prefer.
1) 1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 e6 6.Nf3 Be7 7.cxd5 Nxd5 8.Bc4 0-0 9.0-0 Nc6 10.Re1 Nf6
11.Bd3(?)
1. Black to move
2) 1.c4 c6 2.e4 d5 3.exd5 cxd5 4.d4 Nf6 5.Nc3 e6 6.Nf3 Be7 7.cxd5 Nxd5 8.Bd3 0-0 9.0-0 Nc6 10.Re1
2. Black to move
Show/Hide Solution
Show/Hide Solution
Black or White 5
Is it White’s or Black’s move?
Show/Hide Solution
Show/Hide Solution
Compare 3
Compare the following positions and select the one you think White should prefer.
1) 1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4 Nc6 8.Qe2 cxd4 9.Rd1 Be7 10.exd4 0-0 11.Nc3 Nb4
12.Bg5 h6 13.Bh4 Bd7
White to move
2) 1.d4 d5 2.Nf3 Nf6 3.c4 dxc4 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.a4 Nc6 8.Nc3 cxd4 9.exd4 Be7 10.Bg5 0-0 11.Re1 Bd7
12.Qe2 Rc8 13.Rad1 Nb4
White to move
Show/Hide Solution
Show/Hide Solution
Black or White 6
Black or White 7
Is it White’s or Black’s move?
Show/Hide Solution
Show/Hide Solution
Black or White 8
Black or White 9
Is it White’s or Black’s move?
Show/Hide Solution
Show/Hide Solution
Compare 4
Compare the following positions and select the one you think White should prefer.
1) 1.c4 e5 2.e3 Nf6 3.Nc3 Nc6 4.Nf3 Bb4 5.Qc2 Bxc3 6.bxc3
Black to move
2) 1.c4 e5 2.Nc3 Bb4 3.e3 Bxc3 4.bxc3 d6 5.Nf3
Black to move
Show/Hide Solution
Show/Hide Solution
Black or White 10
Chapter 11 – Panov
Andrey Shariyazdanov – Valery Petukhov, Sochi 2004 146
Luc Winants – Gata Kamsky, Tilburg 1992 148
Axel Smith – Philip Lindgren, Malmö, 16th Jan 2016 149
Viswanathan Anand – Anthony Miles, Wijk aan Zee 1989 152
Veselin Topalov – Jordi Magem Badals, Pamplona 1994 154
Judit Polgar – Spyridon Skembris, Corfu 1990 156
Chapter 14 – Chigorin
Olaf Teschke – Ulli Reyer, Nakenstorf 2008 218
Wilhelm Steinitz – Mikhail Chigorin, World Championship, Havana (10) 1889 219
Chapter 15 – Dutch
Ivan Sokolov – Predrag Nikolic, Bled 1991 230
Vladimir Kramnik – Pavel Tregubov, France 2002 231
Nils Grandelius – Lars Karlsson, Västerås 2012 233
Robert Kempinski – David Anton Guijarro, Essen 2014 235
Chapter 21 – Miscellaneous
Alexander Graf – Zoltan Varga, Hungary – Germany, Budapest 2004 311
Gennadi Sosonko – Hans Ree, Nijmegen 1977 313
Chapter 22 – e3 English
Grzegorz Gajewski – Jacek Tomczak, Polish Championship 2014 325
Alexander Grischuk – Anish Giri, FIDE Grand Prix, Paris 2013 327
Pavel Eljanov – Lazaro Bruzon Batista, 50th Capablanca Memorial, Havana 2015 330
Bill Richards – Gerd Heidemann, World Championship M377, corr. 1986 333