2

First published in 2013 by Gloucester Publishers Limited, Northburgh House,

10 Northburgh Street, London EC1V 0AT

Copyright © 2013 Steve Giddins

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

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

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

British Library Cataloguing-in-Publication Data

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

ISBN: 978-1-78194-034-1

Distributed in North America by The Globe Pequot Press, P.O Box 480,
246 Goose Lane, Guilford, CT 06437-0480.

All other sales enquiries should be directed to Everyman Chess, Northburgh House,
10 Northburgh Street, London EC1V 0AT
tel: 020 7253 7887 fax: 020 7490 3708
email: [email protected]; website: www.everymanchess.com

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

Everyman Chess Series

Chief advisor: Byron Jacobs
Commissioning editor: John Emms
Assistant editor: Richard Palliser

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.

3
About the Author
Steve Giddins is a FIDE Master and a former editor of British Chess Magazine. He spent a number of years
of his professional life based in Moscow, where he learnt Russian and acquired an extensive familiarity with
Russian chess literature and the training methods of the Russian/Soviet chess school. He’s the author of
several outstanding books and is well known for his clarity and no-nonsense advice. He has also translated
over 20 books, for various publishers, and has contributed regularly to chess magazines and websites.

Other Everyman Chess books by the author:

The Greatest Ever Chess Endgames
The English: Move by Move

4
Contents
About the Author
Series Foreword
Bibliography
Introduction

Game 1 Smyslov-Botvinnik (USSR Championship 1944)

Game 2 Tolush-Botvinnik (USSR Championship 1945)
Game 3 Timman-Korchnoi (Leeuwarden match 1976)
Game 4 Van Seters-Korchnoi (Skopje Olympiad 1972)
Game 5 Suetin-Uhlmann (Berlin 1967)
Game 6 Byrne-Korchnoi (London 1979)
Game 7 Byrne-Vaganian (Moscow 1975)
Game 8 Ragozin-Botvinnik (Training match 1944)
Game 9 Hartston-Uhlmann (Hastings 1972)
Game 10 Bogdanovich-Uhlmann (Sarajevo 1965)
Game 11 Fischer-Uhlmann (Buenos Aires 1960)
Game 12 Felgaer-Korchnoi (Bled Olympiad 2002)
Game 13 Vidarsson-Ward (Reykjavik 1998)
Game 14 Kanafsck-Atalik (Mar del Plata 2003)
Game 15 Inarkiev-Vitiugov (Russian Championship 2008)
Game 16 Cheparinov-Grischuk (Baku Grand Prix 2008)
Game 17 Kuipers-Stellwagen (Dutch League 2011)
Game 18 Atlas-Kindermann (Austrian League 2006)
Game 19 Hort-Petrosian (European Team Ch’ship, Kapfenberg 1970)
Game 20 Najer-Korobov (Czech League 2011)
Game 21 Karpov-Nogueiras (Rotterdam 1989)
Game 22 Chistiakov-Petrosian (Moscow Championship 1956)
Game 23 Fischer-Kovacevic (Rovinj Zagreb 1970)
Game 24 Stripunsky-Shulman (US Championship 2010)
Game 25 Bezgodov-Kornev (Russian Championship 2003)

5
Series Foreword
Move by Move is a series of opening books which uses a question-and-answer format. One of our main aims
of the series is to replicate - as much as possible - lessons between chess teachers and students.
All the way through, readers will be challenged to answer searching questions, to test their skills in chess
openings and indeed in other key aspects of the game. It’s our firm belief that practising your skills like this is
an excellent way to study chess openings, and to study chess in general.
Many thanks go to all those who have been kind enough to offer inspiration, advice and assistance in the
creation of Move by Move. We’re really excited by this series and hope that readers will share our enthusiasm.

John Emms
Everyman Chess

6
Bibliography
Dangerous Weapons: The French, J.Watson (Everyman Chess 2007)
Ein Leben lang Französische Verteidigung – richtig gespielt!, W.Uhmann (Thomas Beyer Verlags Gmbh
1991) – also available in English as Winning with the French (Batsford 1995)
Play the French, vols 1-4, J.Watson (Pergamon/Everyman Chess, 1986-2012)
The Complete French, L.Psakhis (Batsford 1992)
The Flexible French, V.Moskalenko (New in Chess 2008)
The French Defence Main Line Winawer, J.Moles (Batsford 1975)
The Winawer: Modern and Auxiliary Lines, J.Moles & K.Wicker (Batsford 1979)
The Wonderful Winawer, V.Moskalenko (New in Chess 2010)

7
Introduction
The Winawer French is one of the great black openings, and is characterized by the initial moves:
1 e4 e6 2 d4 d5 3 Nc3 Bb4

Black brings out his bishop to pin the white knight, thereby attacking the e4-pawn, and also threatening to
capture on c3, thus damaging the white pawn structure. The result is usually an unbalanced middlegame, in
which Black concedes the two bishops and takes on weakened dark squares, but where he has the superior
pawn structure.
The variation is named (at least, in English-speaking countries) after Simon Winawer (1838-1919), a
Polish player, although, as so often with chess openings, his right to be identified with the line is less than
wholly convincing. He played it against Steinitz in his first international appearance at Paris 1867, losing a
miniature, but it had already appeared before that in other games, including a Steinitz-Blackburne match
game in 1863. Despite this, it is Winawer’s name that has stuck.
These early efforts notwithstanding, the man who initially put the Winawer on the map was Aron
Nimzowitsch, and indeed, on the Continent, the variation is sometimes referred to as the Nimzowitsch
Variation. In the post-Steinitz era, the two bishops were regarded as a great strength, and the Winawer was
consequently frowned upon. However, Nimzowitsch was always a great ‘knights man’, and it is therefore not
surprising that he should have been attracted to the Winawer, especially when one considers the similarity
with his eponymous defence to the queen’s pawn 1 d4 Nf6 2 c4 e6 3 Nc3 Bb4.

8
 Here too, Black frequently surrenders the bishop-pair, in order to double the white pawns, and the affinity
between the two openings is clear for all to see.
Most of Nimzowitsch’s opponents played 4 exd5, which Tarrasch had argued was a good response against
the Winawer, on the basis that the bishop was slightly misplaced on b4. Capablanca also used this
continuation as White, most famously in the opening game of his 1927 World Championship match against
Alekhine. His defeat in that game did much to make white players turn to 4 e5 as the main line.
That leads us on to the next great Winawer pioneer, who was Mikhail Botvinnik. He used the variation as
his principal defence against 1 e4, from the early 1930s right up until the end of the 1950s, and if anyone’s
name should be associated with the opening, it should be his. After initially preferring 5 ... Bxc3+ in the main
line, he also subsequently researched 5 ... Ba5, nowadays generally referred to as the Armenian Variation, but
used by Botvinnik as far back as the 1946 radio match between the USSR and the USA, where he employed
the line to beat Reshevsky in a famous game. Botvinnik’s successes resulted in the popularity of the Winawer
growing markedly at all levels, and by the 1960s it had clearly supplanted 3 ... Nf6 as the most popular third
move.
The third of the great trio of Winawer practitioners was the East German Grandmaster, Wolfgang
Uhlmann, who emerged in the international arena in the late 1950s. He has used the Winawer almost
exclusively for over 50 years, still employing it to this day, in the veterans events and Bundesliga games that
he still plays.
Two other names without whom the pantheon of great Winawer players is not complete are perennial arch-
rivals Tigran Petrosian and Viktor Korchnoi. The former’s practice tended to concentrate more on the closed
lines with 4 ... b6, whilst Korchnoi was for much of his career the most successful player of the main lines.
Both have left us a wealth of instructive games in this opening.
In more recent years, double Soviet Champion Lev Psakhis has probably been the leading Winawer player,
with the Armenians Vaganian and Lputian also being important figures, especially in relation to the 5 ... Ba5
lines. The Winawer has lost ground in the popularity stakes at top GM level over the past 15 years, but the
Spanish-resident Ukrainian Grandmaster, Viktor Moskalenko, remains a faithful practitioner, and his books
on the opening are highly recommended (see the Bibliography).
I have myself played the Winawer for some 25 years, with great success. The unbalanced nature of the
positions, and their clear strategical outlines, make it an opening that can be learned relatively easily by the
average player, in my opinion. With one or two exceptions, it is an opening where memorisation of detailed
tactical lines is much less important than understanding the plans and ideas, which makes it ideal for the
9
amateur player. It is also a surprisingly flexible opening, where Black has many different move orders, which
enable him to avoid specific preparation by the opponent, and also to choose the set-up he likes best.
I can heartily recommend the Winawer to players at all levels, and I hope this book will contribute to helping them understand the key ideas
and typical plans behind this inexhaustibly rich and fascinating opening.

Acknowledgements
Thanks go to John Emms and Byron Jacobs of Everyman, for their usual highly professional and supportive job, and to the staff of the La
Torretta café in Rochester, whose friendly atmosphere and superb coffee provided such welcome relaxation between sessions working on this
book.
This book is dedicated to GM Neil McDonald and IM John Watson, which will doubtless come as a shock to both! It was watching Neil’s
successes with the French, whilst I was playing alongside him on the Gravesend team in the mid-1980s, that inspired me to take up the French
myself. Once I decided to do so, the first volume of John Watson’s Play the French, which appeared in 1986, provided the basis of my
repertoire and continued to do so for the next 25 years. Between them, these two gentlemen changed my chess career, and brought me both
great practical success and enormous enjoyment. I am deeply grateful to both.

Steve Giddins
Rochester, Kent
October 2012

10
Game 1
V.Smyslov-M.Botvinnik
USSR Championship, Moscow 1944

1 e4 e6 2 d4 d5 3 Nc3 Bb4

This is the move which characterises the Winawer Variation. Black brings out his bishop to an active
position, pinning the white knight, and thus threatening to capture on e4. The main point is to try to force
White to clarify the position in the centre.

Question: Isn’t it true that Fischer once condemned the Winawer as bad?

Answer: It is indeed. Annotating his game against Darga in My 60 Memorable Games, he wrote: “I may yet be forced to admit the Winawer
is sound. But I doubt it! It is anti-positional and weakens the kingside”.

Question: What exactly did he mean?

Answer: Well, there are several points. Firstly, the old adage says one should develop knights before bishops in the opening, and the
Winawer breaks that rule. However, that is a minor point. More importantly, in the French Defence, Black establishes his central pawns on
light squares. With a light-square pawn structure, Black would usually need to retain his dark-squared bishop to cover the black squares, yet in
most lines of the Winawer, he exchanges his king’s bishop for the white knight on c3. That risks leaving his dark squares very weak, and also
his kingside is vulnerable to the queen raid Qg4, attacking the undefended g7-pawn. That can be hard to defend, because the move ... g6 would
weaken the dark squares very severely.

Question: Sounds like Fischer had a point!

So why are you advocating the Winawer for Black?

Answer: Fischer was right up to a point, in that the Winawer has these drawbacks. But it also has some advantages. The exchange on c3
doubles the white pawns, and with White having been induced to close the position with e5, he frequently finds it hard to get his dark-squared
bishop into active play – after all, most of White’s pawns are on dark squares, obstructing that bishop. Black therefore argues that his own dark-
square weaknesses are not as important as they may seem, because they are hard to attack. At the end of the day, a weakness is only a weakness
if it can be effectively attacked. Black hopes that White’s own weaknesses, notably his doubled c-pawns and isolated a-pawn, will prove at
least as vulnerable as the dark squares in Black’s camp.

11
 Question: So what you are saying is that
the Winawer produces an unbalanced game?

Answer: Exactly, and that is one of its great attractions.

4 e5
Certainly the main move, although White has no fewer than five alternatives, of varying degrees of
respectability. These fourth move alternatives are examined in Games 23-25.
4 ... c5
Similarly, Black has several alternatives here, which are considered in Game 22. However, the text is the
most logical response. Black attacks the enemy pawn centre in the classic French Defence manner. He also
allows his queen to come out to a5, from where it attacks the split white queenside pawns.
5 a3
Once again, not the only move, but the main line. White’s fifth move alternatives are examined in Game
21.
5 ... Bxc3+ 6 bxc3 Ne7
Black can also play 6 ... Qc7 (Game 19) and 6 ... Qa5 (Game 13), as well as even 6 ... Nc6, but the text is
again the main line, and brings about something of a tabiya for the Winawer.
What can we say about this position? From White’s side, he has the two bishops, and, as we have already
noted, he can hope to make especially good use of his dark-squared bishop. It is clear that if this bishop can
occupy the a3-square, for example, it could cut through the black position most effectively. White also has
more space on the kingside, and if the black king castles kingside, White has ideas such as Qg4, Bd3, etc,
with threats against the king.

Question: So what has Black obtained in return?

Answer: He has damaged the white pawn structure on the queenside, and has also forced White to close the position to a significant degree,
by the advance e4-e5. This means that White’s bishop-pair may not find it so easy to become active, whereas Black’s knights can jump around
within the closed structure. Black has ideas such as ... Nbc6, ... Qa5, maybe ... Nf5, etc. He can also close the position even more, by advancing
... c5-c4. This will further block things, and also drive White’s king’s bishop from an active post on d3.

Question: What about the black king? Isn’t it vulnerable on the kingside?

12
 Answer: Maybe, although often it proves less easy to attack than one might think. In addition, Black is by no means committed to short
castling. Frequently, he castles on the other side, especially if he has closed the position with ... c5-c4.

Question: And what about Black’s traditional ‘bad bishop’ on c8?

Answer: That has a few options. One plan is ... b6 and ... Ba6, but that is only possible in relatively rare cases. More often, it comes to d7,
and then either looks to take up a post on a4, harassing White’s queenside, or else to emerge on the other wing, by means of ... Be8, ... f6 and ...
Bg6/h5.

Question: That looks rather slow!

Answer: Yes, and it often comes about only later in the middlegame, but if the bishop does manage to emerge over that side of the board, it
can be a very good piece. With the position being relatively closed, slow manoeuvring is frequently the order of the day anyway.
7 a4

Question: What is the idea of this move? Shouldn’t White be developing his pieces?

Answer: Actually, this is one of the two main approaches for White, in this position. As we noted above, a key element in White’s strategy
is to activate his unopposed dark-squared bishop, and the most obvious place to put it is a3, from where it will bear down the a3-f8 diagonal,
cutting through the heart of Black’s position. The text immediately sets about trying to do that.

Question: So is this good for White?

Answer: Not necessarily. There is one big drawback to the move, namely that the pawn on a4 can actually be quite vulnerable. Once the
bishop goes to a3, the a4-pawn will be undefended, and Black can attack it with ... Qa5, and also maybe ... Bd7. In fact, if Black is determined
to take the a-pawn, it is quite hard for White to defend it in the long run.

Question: So 7 a4 is a gambit?

Answer: Of sorts, yes. It is not a traditional gambit, seeking rapid development and a quick attack; rather, White is prepared to make a long-
term positional sacrifice of the pawn, in order to get his bishop on to a good diagonal.
White can also adopt the same positional approach with 7 Nf3 which often transposes. However, he also
has a radically different approach to the position, with the sharp move 7 Qg4, immediately attacking the g7-
pawn. This move is examined in Games 16-18.
7 ... Nbc6
A natural developing move, bringing the knight to a good square, and exerting pressure against the d4-
13
pawn.
8 Nf3 Qa5

Once again, the most natural move. Black brings his queen to an active post, attacking White’s vulnerable
queenside pawns. The c3-pawn, in particular, is threatened with check, so must be defended.
9 Bd2

Question: Wait a minute – I thought this bishop was aiming at the square a3?

Answer: Indeed it was, and Black will claim that he has scored a small success, by inducing it to settle for the more passive square d2.

Question: So why doesn’t White defend the pawn with 9 Qd2?

Is he afraid of the queen exchange after 9 ... cxd4?

Answer: As we will see later, White is not really afraid of the immediate queen exchange, but it is one problem with the move 8 Qd2. In
fact, the latter is a perfectly respectable option, that has been preferred by many white players in the Winawer (including both Fischer and
Smyslov, as we will see). We will examine it in Games 8-12.

14
 9 ... c4!?

Question: Not sure why you mark this with ‘!?’. Isn’t it almost forced,
in view of the threat of 10 c4, attacking the queen?

Answer: That was certainly what most people thought, for decades, and ... c5-c4 became almost a reflex action in such positions. However,
the modern preference is to retain the tension as long as possible in such positions, here with the move 9 ... Bd7.

Question: So you are not afraid of 10 c4 then?

Answer: No. We will discuss these issues more fully in Game 3, but the bottom line is that this advance often leaves the white centre too
weak and his position undeveloped.

Question: Going back to the position after 9 ... c4, what is going on?
I thought you said this was part of Black’s plans anyway?

Answer: It is to some extent. For years, the move was seen almost as a sine qua non of Black’s play in such positions. It was assumed that
he has to close the position, to prevent the white bishops becoming too active. Botvinnik won quite a few games in such positions, usually by
annexing the a4-pawn and eventually triumphing, but gradually white players learnt how to coordinate their forces more effectively, and the
downsides of the advance ... c5-c4 began to emerge more clearly.

Question: And these are?

Answer: By releasing the pressure on d4, Black gives White a free hand to reorganise his position and start building pressure on the
kingside. He also weakens his dark squares further, by fixing another pawn on a white square; for example, the a3-f8 diagonal is now more
inviting than ever for the white queen’s bishop, should he ever succeed in freeing himself from the need to defend c3.

Question: And what about the loss of the a4-pawn?

Answer: This is an irritant for White, certainly, but if he organises his counterplay with sufficient energy, practice suggests that he can
exploit the time taken by Black to win that pawn. Even after winning it, it is very hard for Black to exploit his material advantage on the
queenside.
10 Ng5
A standard reaction to the early closing of the centre by ... c5-c4. Now there is no pressure against d4,
White redeploys his knight, freeing his f-pawn to advance, and also looking to provoke weaknesses in the

15
black kingside. However, with the white bishop not yet having left f1, there is also a case for the move 10 g3
with the idea of Bh3, 0-0, Nh4 and f4-f5. White would have good chances of developing an initiative in such
a fashion.

10 ... h6

Question: Is this necessary? After all, you said White wanted to

provoke some kingside weaknesses, so why oblige him?

Answer: That is a good point. On g5, the knight annoys Black, perhaps, but it is not clear how much real effect it has. The attack on f7 stops
long castling, and moves such as Qh5 may be in the air, but it is still not clear whether Black could not just play a move such as 10 ... Bd7,
although then after, say, 11 g3, he still has to decide what to do about his king. It is probably hard to avoid the move ... h6 for too long.
11 Nh3

Question: So the threat now is f4-f5?

Answer: Actually, the bigger threat is probably 12 Nf4, followed by Nh5. It is quite awkward to defend g7, and Black is reluctant to play ...
g6 himself, because of the drastic further weakening of his dark squares.
11 ... Ng6
See the last note. Black stops 12 Nf4.
12 Qf3?

Question: Why is this bad?

Answer: It is just too slow. White intends to exchange knights on f4, recapturing with the queen, but, as we will see, his plan is too slow and
Black can meanwhile take the a4-pawn with relative impunity. Keres recommended 12 Be2 Bd7 13 Bh5, another typical manoeuvre in such
positions. After 13 ... Nce7 14 0-0 we transpose into the game Planinc-Timman, analysed in the notes to Game 3.
12 ... Bd7 13 Nf4 Nxf4 14 Qxf4 Ne7!

16
 An excellent move, which underlines the inadequacy of White’s approach. The knight move does two
things: firstly, it transfers the knight to the kingside, where it can fulfil a defensive role (for example, Qg4 can
now be met by ... Nf5, defending g7), and secondly, it unmasks the bishop on d7, which simply intends to
capture on a4. As we have emphasised above, taking the a-pawn is certainly something Black is very happy
to do if he can get away with it, and White needs to react energetically on the kingside in response. Here, he
has no real kingside initiative by way of compensation for the impending material deficit.
15 h4 Bxa4 16 h5
White sets up the basis for a kingside pawn storm, with g4-g5 to follow, but this proves too little, too late.
16 ... Qb5

By unpinning the bishop, Black threatens another pawn on c2. The effectiveness of a bishop on a4 in such
structures is something we will see elsewhere in this book.
17 Kd1

17
 Question: Eh? Surely this is not forced?

Answer: No, but White reasons that he needs to get his king away from the kingside anyway, so he can pursue his ambitions there with g4-
g5, and he reckons his king is relatively safe on d1.

Question: Hmm. It all looks a bit desperate to me!

Answer: Indeed it is. Things have gone seriously wrong for White, and he knows it.
17 ... Rc8 18 Bc1 Rc6
The rook is on its way to a6.
19 Be2 Ra6
Threatening 20 ... Bxc2+, hence White’s next.
20 Kd2

20 ... 0-0!
Resolutely refusing to be frightened of ghosts. Botvinnik judges that the coming White pawn storm with
g4 is nothing to be frightened of.
21 g4 f6!
The point of Black’s play. Far from going passive on the kingside, he hits back, opening the f-file and
including his rook in the defence along the sixth rank.
22 exf6 Rxf6 23 Qc7 Rf7 24 Qd8+ Kh7 25 f4 Qa5 26 Qb8 Nc6
The computer points out that the move 26 ... e5! is very strong here, with the point that 27 fxe5 (if 27 Qxe5
Nc6 28 Qe3 Re7 29 Qf2 Re4 and d4 drops) 27 ... Rf4 unexpectedly leaves White unable to defend the d4-
pawn. Botvinnik’s move is less incisive, although he has similar ideas.
27 Qe8 Re7 28 Qg6+
28 Qf8 Qd8 29 Qxd8 Nxd8 would enable White to put up tougher resistance, but it is quite understandable
that Smyslov was reluctant to trade queens into a pawn-down endgame.
28 ... Kg8 29 Ba3
29 g5 may look threatening, but in fact, after 29 ... Nxd4 30 gxh6 Nxe2 31 Kxe2 Qxc3 White has no real
threats against the black king.

18
 29 ... e5
Unmasking the rook on a6.
30 fxe5
30 dxe5 is slightly more tenacious, but even so, after 30 ... Nd4! (30 ... Nxe5 31 Qf5 is much less
convincing) 31 Bb4 Qb5 32 Bxe7 Rxg6 Black should win.
30 ... Nxd4 31 Bb4 Qd8
31 ... Qb5 is even better.
32 Qxa6
32 Bxe7 Qxe7 33 cxd4 Rxg6 34 hxg6 Qb4+ 35 Kc1 (35 Ke3 Bxc2 is no better) 35 ... c3 is decisive.
32 ... bxa6 33 cxd4 Rb7 34 Rxa4
Momentarily, White has a winning material advantage, but his scattered forces and exposed king soon
suffer ruinous losses.
34 ... Qg5+ 35 Kd1 a5
Good enough, but the computer points out that 35 ... c3 is mate in seven!
36 Bf3 Rxb4 37 Bxd5+ Kf8 38 Rf1+ Ke8 39 Bc6+ Ke7 40 Rxb4 Qxg4+ 0-1

19
Game 2
A.Tolush-M.Botvinnik
USSR Championship, Moscow 1945

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Nf3

In the previous game, White chose 7 a4. The text is the other main positional attempt, and frequently
transposes into 7 a4. However, as we will see, sometimes White can try to dispense with the move a3-a4 and
exploit the tempo thus saved for more vigorous play on the kingside.
7 ... Qa5
7 ... Nbc6 is perhaps the more common move order nowadays, although the text can transpose.
8 Bd2

Question: Is 8 Qd2 possible here?

Answer: It certainly is, and we will look at such Qd2 plans in Games 8-12. However, one could argue that with the white pawn not yet
having advanced to a4, the move Bd2 makes more sense, since he can dispense with the pawn advance altogether and play on the other wing.
8 ... c4!?
Once again, we see this thematic black idea, but also in a somewhat premature form. As we pointed out in
the notes to Game 1, the threat of an early c3-c4 by White is less serious than it looks, whilst the white bishop
has not yet committed itself to e2 or d3, so Black probably does better to remain flexible and develop with 8
... Nbc6. This would transpose into more normal lines.

20
 9 a4?!

Question: You don’t like this typical move?

Answer: I don’t greatly like it in this specific position, no. As we have discussed already, when Black closes the position with an early ...
c5-c4, White needs to react vigorously on the kingside, else he is in danger of failing to develop any initiative on that wing, and just ending up
watching while Black helps himself to the a4-pawn. That is what happened in Game 1, and we will see much the same scenario unfold in the
present game.
As we pointed out in the last note, the white bishop being committed to d2 does not fit terribly well with
the move a3-a4. I therefore think that, in this position, White should dispense with the advance of the a-pawn
and get on with purposeful play on the other flank. There are two main ways of doing so:
a) The first is 9 g3, taking advantage of the fact that the bishop is still on f1. We pointed out such a
possibility in Game 1 also. Play might continue 9 ... Nbc6 10 Bh3 Bd7 11 0-0 0-0-0 12 Ng5 (12 Nh4 is also
possible) 12 ... Rdf8 13 f4 with unclear play.
b) 9 Ng5! is probably better still. White adopts the standard idea of provoking kingside weaknesses, whilst
freeing his f-pawn to advance. After 9 ... h6 10 Nh3 Black has:

21
 b1) 10 ... Nbc6 11 Nf4 g6 (11 ... Bd7 12 Nh5 and Black is forced into contortions to protect the g7-pawn)
12 g4!? Bd7 13 Be2 0-0-0 14 0-0 and White will follow up with Ng2 and f4, eventually preparing the break
f4-f5. Once again, he must be rather better, albeit in a tough position.
b2) 10 ... Ng6 11 Be2 Nc6 12 Bh5 Nce7 13 0-0 with a similar position to that considered in Game 1, but
with White having gained a tempo by omitting a3-a4. In addition, his a-pawn is much less vulnerable to
capture on a3. The position is a tough one to play (Black is always very solid in such positions), but White
must be somewhat better here.

9 ... Nd7!?

Question: This looks a bit unusual. Doesn’t the knight usually go to c6?

Answer: Yes, but in this particular position, the text makes some sense. Black intends to win the white a4-pawn with the knight, by means
of ... Nb6xa4. In Game 1, we saw him do this with the bishop, after ... Bd7xa4, but the knight is more active on a4 than the bishop. It attacks c3,
and can be supported by ... b7-b5. Ceteris paribus, Black would rather take with the knight on a4, than the bishop.

22
 Question: So why doesn’t this get played more often?

Answer: Well, as we have already discussed, Black cannot really force this position. If he wants to play the plan with ... c4 and ... Nd7, he
needs to close the centre very early (as he did in the present game), but then White will usually have the option to dispense with the move a3-a4
and play for a more vigorous set-up on the other wing.
10 Be2?!
Once again, White plays too slowly and routinely, and allows Black to reach a good position. The move 10
Ng5 was still best, although the tempo spent on a3-a4 makes it less effective here than in the previous note. A
game D.Bronstein-V.Saigin, Moscow 1945, continued 10 ... h6 11 Nh3 Nb6 12 Nf4 g6 13 h4 Bd7 14 h5 with
the initiative. However, even this is not so clear. In the game, Black played the terrible move 14 ... g5? and
after 15 Ne2 0-0-0 16 g4 Rdg8 17 Bh3 Nxa4 18 f4 gxf4 19 0-0 his kingside was collapsing. But the Irish
theoretician John Moles, in his classic 1975 treatise on the Winawer, pointed out that Black’s structure is still
very hard to crack after 14 ... 0-0-0!: for example, 15 Be2 Nxa4 16 hxg6 fxg6! 17 Bg4 Qa6! and it is not easy
for White to step up the pressure. This line is a good illustration of the solidity of Black’s position in such
lines, even when White manages to work up a decent initiative.
10 ... Nb6 11 0-0 Nxa4 12 Nh4?
Another mistake, this time more serious. Thanks to Black’s reply, the prospects of White ever getting in
the break f4-f5 are reduced virtually to zero. Once again, White should try 12 Ng5 h6 13 Nh3.
12 ... Ng6!
When the white knight cannot retreat to g2, this move is highly effective against Nh4. The exchange on g6
is virtually forced.
13 Nxg6 hxg6

Now Black has an excellent position. The transfer of the pawn to g6 leaves White with no real prospects
on the kingside, and he has no compensation at all for his pawn minus.
14 Re1 Bd7 15 Bf1 b5 16 Qf3 Rb8

Question: What is the idea of this?

Answer: We saw a similar manoeuvre in Game 1. The rook is coming to b6, from where it will patrol the sixth rank. In this case, there is
also a deeper purpose. Botvinnik sees that White is preparing to bring his bishop to a3, and he prepares a counter-measure against this.
17 Reb1 Qc7 18 Bc1 a5 19 Ba3 Rb6 20 Qg3 Qd8

23
 Preparing ... Qh4, exchanging queens.
21 Bd6

21 ... Rxd6!

Question: What’s this? An exchange sacrifice?

Answer: It is a key idea for such French positions generally, and Winawer positions in particular. This positional exchange sacrifice
eliminates White’s best minor piece. Black already has a pawn for the exchange, and he will be able to round up the one on d6 as well.
Materially, he will have enough for the exchange, but positionally, his knight will be worth a rook anyway.

Question: How come? I was always taught that rooks are better than knights!

Answer: The essential point is that the position is blocked. To be effective, rooks need open files. Here, they have none. This is why
exchange sacrifices so frequently occur in the French, and we will see some further examples in this book.
22 exd6 Bc6

24
 23 h3
White wants to be able to keep the queens on after Black’s ... Qh4, hence he ensures the h2-pawn will not
be attacked.
23 ... Kd7

Question: Now that is taking the mick, surely?

Answer: It is a nice move, but is actually perfectly logical. Once again, it is all a consequence of the blocked nature of the position. The
king is as safe as houses on d7, as White has no effective way to open the central files and get at it with his heavy pieces.
24 Re1 Qh4 25 Qe5
White desperately tries to keep the queens on, in order to complicate Black’s task. If he acquiesces in the
exchange, Black will just take on d6 and then set his central pawns in motion with ... f6 and ... e5. The latter
is ultimately his key winning plan, and the only hope of preventing it is to keep the queens on, so that the
central pawn advance will always result in some exposure of the black king.
25 ... Qf6 26 Qg3

25
 26 ... Rh4!
The start of another nice regrouping. Botvinnik intends to bring the rook to f4 and his queen to h4. Once he
achieves that, the break ... b5-b4 will undermine the support of the d4-pawn. White’s problem is that he has
no counterplay.
27 Re3 Rf4 28 Be2 Qh4
Now White cannot avoid the exchange of queens.
29 Bf3 b4
Not only is ... b3-b2 a threat, but the d4-pawn is now liable to drop off at any moment.
30 Qxh4 Rxh4 31 g3 Rh8
31 ... Rxh3?! 32 cxb4 axb4 33 Rb1 would give White much more hope. Instead, Botvinnik brings his rook
back to defend the b4-pawn.
32 cxb4 axb4 33 Rb1 Rb8 34 h4 Rb7 35 Kh2 Kxd6
White is defenceless.
36 g4 Nc3 37 Ra1 Nb5 38 Rd1 Ra7 39 h5 g5 40 Kg2 Ra2 41 Be2 0-1
Here, the game was adjourned and White resigned without resuming. After Black takes on c2, his
queenside pawns cost White a lot of material.
A classic Winawer game. It shows once again the need for White to react with vigour to Black’s plan involving ... c5-c4 and the win of the a4-
pawn. The other instructive aspect of the game is the power of Black’s positional exchange sacrifice, and such sacrifices are very typical of the
opening.

26
Game 3
J.Timman-V.Korchnoi
3rd matchgame, Leeuwarden 1976

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 a4 Nbc6 8 Nf3 Qa5 9 Bd2 Bd7

This is nowadays the main line of the Nf3/a4 treatment in this variation. Both sides develop according to
their desired scheme, and Black refrains from an early commitment to closing the position with ... c5-c4.
10 Be2

Question: This looks rather modest. Is it best?

Answer: For a long time, it was the main line. The problem is that the superficially natural-looking 10 Bd3? simply encourages Black to
play 10 ... c4 with tempo. However, White does have a major alternative in 10 Bb5 and this move has actually become the most popular in this
position, largely thanks to the variation seen in the present game, which White players have been unable to crack. The move 10 Bb5 will be
examined in Game 7.
10 ... f6
This is the key idea, which revitalised Black’s position in this variation, when it was first introduced in the
late 1970s. Black still avoids closing the centre, and instead attacks it again, creating the maximum tension in
the centre of the board.

Question: But once again, Black ignores the threat of

a discovered attack on his queen, after 11 c4?

Answer: Yes, and this game will show what happens if White carries out that threat.
Before the discovery of 10 ... f6, Black almost always played 10 ... c4 here.

27
 Question: Is that bad?

Answer: Probably not, but Black did experience a few reverses in this line during the 1970s, and these served to put black players off the
position a bit.

Question: In Games 1 and 2, Black seemed to do very well

in these blocked structures. So what changed?

Answer: Basically, white players learned how to handle them better. As we mentioned in the notes to the first two games, White needs to
play dynamically. Botvinnik’s opponents in the 1930s and 40s were still feeling their way in these lines, which were newly-developed at that
time, and even players of the class of Smyslov took some time to work out exactly how they should handle their pieces in such positions. As an
illustration of the modern understanding of the white structure, we will examine in full one of the key 1970s games, where White got things just
right. The game in question is A.Planinec-J.Timman, Wijk aan Zee 1974. After 10 ... c4 (I have changed the move order somewhat to fit the
present game), play continued:
11 Ng5!
We have pointed out this standard idea several times in the notes to Game 1 and 2.
11 ... h6 12 Nh3 Ng6
Stopping the unpleasant Nf4-h5, which ties Black down to the defence of g7.
13 0-0 0-0-0
Black almost always castles long in this variation, but John Watson suggests that 13 ... 0-0!? followed by
... f7-f6, may be perfectly playable.
14 Bh5 Nce7 15 f4
The same year, against the same opponent, Planinec varied his move order slightly with 15 Qe1. However,
his basic strategy was unchanged, and he won even more crushingly: 15 ... Kb8 16 Bc1 Rc8 17 Ba3 Qd8 18
a5 Rg8 19 Qc1 Rc6 20 a6 b6 21 f4 f5 22 exf6 gxf6

28
 23 f5 Nxf5 24 Rxf5 exf5 25 Qxh6 f4 (incredibly, the computer actually claims Black is okay after 25 ...
Qe8 but even if true, this does not detract from the instructional value of Planinec’s handling of the whole
variation) 26 Bxg6 Bxh3 27 Qxh3 Rxg6 28 Qh7 1-0 A.Planinec-J.Timman, Amsterdam 1974.
15 ... Rhg8 16 Nf2 Kb8

17 Qe1!

Question: That looks a funny move!

Answer: Maybe, but it is very logical. Planinec wants to re-route his queen’s bishop to the a3-f8 diagonal, so first he simply defends the c3-
pawn.
17 ... Rc8

Question: Why doesn’t he take the a4-pawn, as we saw in Games 1 and 2?

29
 Answer: He could, and maybe even objectively should, just so as to have something for which to suffer. But whether he takes with bishop
or knight (the latter after ... Nc8-b6xa4), he removes a guard from the key square f5, thus facilitating White’s plan of preparing f4-f5. Black
will also be tied up on the queenside, quite probably being forced into a subsequent ... b7-b5, which exposes his king. Note that in both Games
1 and 2, Botvinnik did not castle queenside!
18 Bc1 Qd8 19 Nd1!
Another nice regrouping move. The knight is coming to e3, to support the advance f4-f5.
19 ... f5
Black has to find some play from somewhere, so he tries a kingside advance of his own, but this comes
unstuck. However, it is hard to know what to recommend. If he sits tight, he is terribly passive and White can
strengthen his position on both wings, with a free hand.
20 Ne3 Nh8 21 Ba3 g5 22 a5
Another thematic idea in these lines. The despised a-pawn becomes a battering-ram against the black king
position.
22 ... Be8 23 Bxe8 Qxe8 24 Qb1 gxf4 25 Rxf4 Neg6 26 Rf1 h5 27 Kh1 f4

28 Nd1!
This knight has already got through a lot of work in this game, and there is more to come. White is looking
at playing Nb2-a4, and then something with Bd6+ and Nb6+.
28 ... Nh4 29 Bd6+ Ka8 30 Nb2
So strong is that threat that White does not even pause to defend g2.
30 ... Nf7 31 Na4 Nxd6 32 exd6 Qg6 33 Rf2 Nxg2 34 d7 Rcd8 35 Nc5 Rb8

30
 36 d8Q!
The final touch. White simply wins a piece.
36 ... Rgxd8 37 Rxg2 Qf7 38 Qb6!
A very nice way of attacking the e6-pawn!
38 ... f3 39 Rf2 Qf4 40 Nxe6 1-0
An extremely impressive performance by White. This game, and others like it, served to put Black off the
standard plan with an early ... c5-c4, and sent him in search of the more dynamic approach with 10 ... f6, to
which we now return.
11 c4
Not forced, but the critical move. The quiet 11 0-0 is seen in Game 4.
11 ... Qc7

Question: So how have the last few moves changed things?

31
 Answer: The main point is that we are heading for a much more open game than the sort of closed struggles we have seen hitherto in this
book.

Question: But shouldn’t that favour White, with his two bishops?

Answer: One might think so, but other factors are also relevant here. The main one is that White’s centre has lost its stability. It is under
attack from all sides, and cannot be maintained as a pawn unit. In particular, White cannot avoid the exchange on f6, which not only devalues
his centre, but also opens the g-file. The safety of the white king frequently becomes an issue in such positions.
12 cxd5
Another Korchnoi game, played two years later, saw White hold back his second c-pawn, to try to keep his
centre more stable, but he lost drastically: 12 exf6 gxf6 13 cxd5 Nxd5 14 dxc5 0-0-0 15 0-0 e5 16 c3 (by
comparison with the main game, White covers the d4-square, keeping the enemy knights out, but Black is
very active nonetheless) 16 ... Rhg8 17 Kh1 Nf4 18 Bc4?.

Exercise: Black can now win material. How?

Answer: 18 ... Rxg2! 19 Bxf4 Rg4! (this is the move White had missed; his position is now falling apart) 20 Qe2 Rxf4 21 Nd2 Na5 22 f3
Nxc4 23 Nxc4 Qxc5 24 Nd2 Bf5 25 Ne4 Qc6 26 Rae1 Be6 27 Rg1 f5 28 Ng5 Bd5 29 Rg3 h6 0-1 L.Lederman-V.Korchnoi, Beersheba 1978.
Returning to move 14, a line which has been seen a number of times is 14 c3 0-0-0 15 0-0 Rhg8 16 Re1 e5
17 c4 Bh3 18 Bf1 Nb6 19 d5 ...

32
 ... and now 19 ... Nxc4! with the largely forced sequence 20 dxc6 Qxc6 21 g3 Bxf1 22 Rxf1 e4 23 Qb3 (or
23 Qc2 Qd5 24 Bf4 exf3 when 25 Rac1! Ne5 26 Qxc5+ Qxc5 27 Rxc5+ Nc6 28 Rf5 is equal, but I.Mazi-
I.Farago, Bled 1992, continued 25 Rfc1?! when 25 ... Rge8! with the idea 26 Qxc4?? Re1+! favours Black)
23 ... Qd5 24 Rac1 (24 Rfc1 Nxd2 25 Nxd2 Qxd2 was E.Sharapov-K.Piorun, Grodzisk Mazowiecki 2007,
and now 26 Rxc5+ is equal) 24 ... Nxd2 25 Nxd2 Qxd2 26 Rxc5+ Kb8 27 Rb5 Rg7 28 Rb1 Rdd7 29 Qe6 e3
30 fxe3 Rge7 31 Qg8+ Rd8 32 Qb3 Rdd7 33 Qg8+ Rd8 34 Qb3 Rdd7 ½-½ B.Spassky-V.Korchnoi, 4th
matchgame, Belgrade 1977.
Returning to the immediate exchange on d5:
12 ... Nxd5 13 c4
John Watson considers this the most accurate move order. If instead 13 exf6 gxf6 14 c4 Black has the
attractive option of 14 ... Nf4!.
13 ... Nde7 14 exf6 gxf6 15 dxc5 0-0-0

33
 Question: Wow! The position has changed rather drastically
over the past few moves! What is going on?

Answer: The centre has cleared completely, and White has even won a pawn, at least temporarily. His bishops have full scope, particularly
the dark-squared bishop.

Question: So why isn’t he better?

Answer: Several things. Firstly, his queenside pawns are quite weak. Secondly, although his bishops look active, the same can be said for
the black knights. In the middlegame, knights may be inferior to bishops, when they lack secure central outposts, but if they have such outposts,
they can be a match for bishops, even in an open position. Here, Black can play ... e6-e5, securing the d4-square, and then ... Nf5, when d4
offers a great central outpost for the knights. A further factor is that White’s king will have to castle short, into the open g-file, which should
offer Black good counterplay. All in all, the position seems dynamically balanced, but Black’s practical results have been good.
16 Bc3 e5 17 Qd6
Timman decides that his king could be the more vulnerable in the middlegame, and therefore goes for a
queen exchange. However, Black still has good, active play in the resulting ending, and Black has no real
problems here.
17 ... Nf5 18 Qxc7+
18 Qxf6? is far too risky after 18 ... Rhf8 19 Qg5 e4, with a virtually winning attack.
18 ... Kxc7 19 0-0 Nfd4 20 Nxd4 Nxd4 21 Bd1 Kc6

22 Bxd4?
This is the real start of White’s problems. The exchange gives Black a powerful passed pawn, which in the
end, costs White the game. He could maintain the balance with 22 f4.
22 ... exd4 23 Bf3+ Kxc5 24 Bxb7 Bf5
With his strong passed pawn, and much the more active king, Black is simply better here.
25 Bf3 Rhe8 26 Ra2 Rb8 27 Rd2 Rb1
27 ... Rb4 and even 27 ... Rb3 are more natural, but Korchnoi wishes to prevent any chance of White
ganging up on the passed d-pawn and winning it.
28 g4 Ree1 29 Rxe1 Rxe1+ 30 Kg2 Be4 31 Bxe4 Rxe4 32 Kf3 Re5

34
 With his king cut off on the kingside, White is helpless against the enemy king and passed d-pawn.
33 h4
33 Rd1 Kxc4 34 Rc1+ Kb3 35 Rd1 Kc3 36 Rc1+ Kd2 37 Rc7 d3 is no better. The king and pawn ending
after 33 Re2! is also hopeless: for example, 33 ... Kxc4 34 Rxe5 fxe5 35 Ke2 Kc3 36 Kd1 Kd3 37 h4 Ke4 38
h5 Kf4 39 f3 e4 and White loses his kingside pawns.
33 ... Kxc4 34 Rc2+ Kb3 35 Rc7 d3 36 Rxh7 Rd5
Now the d-pawn is unstoppable.
37 Rb7+ Kc2 38 Rc7+ Kb1 39 Rb7+ Ka1 40 Rb5 Rd8 0-1

This game, and those quoted in the notes, is a good illustration of how Black’s dynamic play enables him at least to hold the balance in the
open positions resulting from 10 ... f6 11 c4 Qc7. Although White’s bishops gain more scope, his crumbling pawn centre and potentially
exposed king give Black enough compensation. The realisation that Black does not need to close the position with ... c5-c4 was a major
development in the understanding of this variation.

35
Game 4
F.Van Seters-V.Korchnoi
Skopje Olympiad 1972

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Nf3 Bd7 8 a4 Qa5 9 Bd2 Nbc6 10 Be2 f6
For the sake of clarity in presenting the lines, I have changed the move order here. In the game, play in fact
proceeded 10 ... c4 11 0-0 f6, transposing.
11 0-0

White declines to take up the gauntlet with the move 11 c4, instead preferring to complete his development
and maintain his pawn centre. He reasons that Black still has to concern himself with the possibility of a
subsequent c3-c4, which White will look to achieve in more favourable circumstances.
11 ... c4

Question: So Black plays ... c5-c4 after all. Isn’t this a failure on his part?
After all, he was trying to avoid this move, wasn’t he?

Answer: No, not really. In fact, Black’s argument is that he has gained from the transaction. What he wanted to do was play ... c5-c4, but
without allowing the strong reply Ng5, followed by the plan we saw in Planinec-Timman, in the notes to Game 3. With his pawn already on f6,
Black can now close the centre without fear of this counter-plan. Having said that, however, the move 11 ... 0-0-0 is also very playable, and
would transpose into Game 3 if White then took up the challenge with 12 c4.
12 Re1 0-0
12 ... 0-0-0 is also playable here, but the most common move is 12 ... fxe5: for example, 13 dxe5 (13 Nxe5
Nxe5 14 dxe5 0-0 is also comfortable for Black) 13 ... 0-0 14 Bf1. It is instructive to follow this line a few
more moves, as we see some interesting play by Black.

36
 14 ... Rf5!

Question: Why is this good?

Answer: Black prepares to double rooks along the f-file, and also exerts pressure against the enemy e5-pawn. The latter can potentially be
augmented by moves such as ... Ng6 and ... Qc7, and the e5-pawn can possibly become a serious target. This is perhaps one argument in favour
of the simplifying knight capture at move 13, when White always has the possibility of f2-f4, defending the e-pawn. After 15 g3 Raf8 16 Re3
R5f7 Black was comfortable in V.Ciocaltea-R.Vaganian, Budapest 1973.
13 exf6!?
13 Bf1 was perhaps more consistent. The text helps Black to double rooks on the f-file.
13 ... Rxf6 14 Bg5 Rf7 15 Qd2 Raf8

Question: So how do you assess this position?

37
 Answer: I think Black is quite comfortable here. His pieces are almost all well placed, his rooks exert pressure down the f-file, and he has
pressure against the weak white queenside pawns.

Question: But what about his bad bishop on d7? And also his dark squares –
the squares e5, c5, d6, etc, look like so many open wounds!

Answer: True, but one just has to get used to this in the Winawer. As we have remarked earlier, the positions are frequently unbalanced;
that is to say, equal, but containing plusses and minuses on each side. Black has these weaknesses, sure, but he also has other trumps, such as
the weak white queenside pawns. The bishop on d7 should not be despised, as it fulfils some useful functions – it defends the weak e6-pawn,
and it also points at the a4-pawn. On a really good day, this bishop can find a new life on the kingside, after ... Be8-g6/h5.
16 Bh4

Question: What is this about?

Answer: It is a typical manoeuvre in such positions, the bishop aiming for g3, from where it controls e5 and attacks d6. However, Black is
well placed to meet this here, since he can eliminate the bishop after ... Nf5.
16 ... Nf5 17 Bg3 h6

A typical move in such positions.

Question: It is just defensive, guarding against Ng5, right?

Answer: Not entirely! As we will see, it also has the idea of a later ... g7-g5 advance.
18 Rad1?
This looks pretty pointless, and is a sign that White was struggling to find something constructive to do.
Maybe he should have taken advantage of Black’s failure to take on g3 last move, with 18 Be5. Black can
then still eliminate the bishop, of course, but a subsequent ... Nxe5 will be met by the knight recapture on e5,
which causes some problems for Black.
18 ... Qd8

Question: What is the point of this? Wasn’t the queen well-placed on a5?

Answer: Up to a point, but with White having regrouped and brought his queen to d2, the pressure against c3 is not going anywhere. In
addition, of course, the a4-pawn cannot be taken in view of 18 ... Qxa4?? 19 Ra1 Qb5 20 Reb1, trapping the queen. Black therefore decides that
his queen has done her job on a5, at least for the time being, and so she switches to a more central position.
19 Rb1 Bc8
38
 Black prefers this to 19 ... b6, so as to preserve the possibility of the queen returning to a5 at a later stage
(see move 23).
20 Qc1

20 ... g5

Question: Crikey! He really did it!

Answer: Yes. This is an idea that is always in the air in these structures. Black threatens to take on g3 and push ... g5-g4, breaking through
to the f2-pawn.
21 Ne5?
This leads to exchanges which leave Black clearly better. White should probably just return the rook to f1,
defending f2. Although this looks passive, it is not easy for Black to improve his position.
21 ... Nxg3! 22 fxg3
The tactical justification of Black’s last is 22 Nxf7? Nxe2+, winning two minor pieces for the rook.
22 ... Nxe5 23 dxe5 Qa5!

Question: What is the upshot of the last few moves?

Answer: Black has greatly improved his position. The weakness on e6 is now shielded along the e-file, the e5-pawn is isolated and weak,
and the white queenside pawns are vulnerable. Black is clearly better.
24 Qe3 b6?!
Korchnoi seeks to maintain control, but it is not clear why he did not simply take on a4, also attacking c2.
25 Bh5 Rf2?!

39
 26 Rf1?
This allows Black a winning simplification. White should play 26 Rec1, when the threat of 26 Bf3 forces
the rook to retreat. After 26 ... R2f5 27 Bg6 Black cannot really avoid a repetition.
26 ... Rxf1+ 27 Rxf1 Qc5!
27 ... Rxf1+ 28 Kxf1 Qxa4 is certainly much better for Black, but Korchnoi realizes that the bishop ending
resulting from the text is winning for Black, thanks to the weakness of the white queenside.
28 Qxc5
28 Rxf8+ Kxf8 29 Kf2 is no better after 29 ... Qxe3+ 30 Kxe3 Bd7, when the a4-pawn drops.
28 ... Rxf1+ 29 Kxf1 bxc5 30 a5
This is the point of the exchange on c5, but the a-pawn is only prolonging its life, not curing its terminal
illness. The black king will march over and round it up in due course.
30 ... Bd7 31 Ke2 Kf8 32 h4 Be8 33 Bg4 Ke7 34 hxg5 hxg5 35 Kd2 Kd7 36 Kc1 Bf7
Now it is clear that the a-pawn is doomed, and the rest is silence.
37 Kb2 Kc6 38 Ka3 Kb5 39 Kb2 Kxa5 40 Ka3 Kb5 41 Kb2 a5 0-1
Although the a-pawn cannot itself be forced home, Black can create a second passed pawn on the d-file, which will be decisive.

40
Game 5
A.Suetin-W.Uhlmann
Berlin 1967

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 a4 Qa5 8 Bd2 Nbc6 9 Nf3 Bd7 10 Be2 c4 11
0-0 f6 12 exf6
Another different way of handling the position. In the previous game, White delayed this move until Black
was able to recapture on f6 with the rook. Here, he takes at once, forcing the reply ... gxf6 which leads to
another slightly different structure.
12 ... gxf6 13 Re1 0-0-0

With the pawn on f6, castling short is clearly not very attractive, so this move is the most natural. The
position before us could also be reached by other move orders, of course.

Question: So what are the differences between this set-up

and that which we saw in the previous game?

Answer: The pawn being on f6, rather than g7, strengthens Black’s centre (especially the e5-square), and gives him the open g-file, along
which he can attack the white king. White himself perhaps has slightly more chances of attacking the black king here, using the open b-file, but
it is still not easy to do so, as the doubled c-pawns hamper White’s communications. Very often in these structures, with the black pawn having
gone to c4, White finds his position rather cut in two along the c-file, and it is rather cumbersome to shift pieces between the king’s and
queen’s flanks. Overall, this structure with ... gxf6 strikes me as very comfortable for Black.
14 Bf1 Nf5
Note that one idea for Black in such positions is to bring this knight to e4, via d6. If it ever gets there, it is
likely to be very powerful, so White has to fight against this plan.
15 Qc1 h5 16 Qa3 Rdg8 17 Rab1

41
 17 ... Nd8!
Another typical Winawer manoeuvre, worth noting.

Question: What is the idea?

Answer: The knight moves away to meet the threat of 18 Rb5. At the same time, it defends the e6-pawn, thus potentially freeing the bishop
on d7 for more fruitful work, possibly taking on a4, perhaps coming to c6, to defend b7. The knight on d8 also defends b7, of course, so Black
is already well prepared for the possible white attack down the b-file.
18 Rb4 Rh7
The rook gets ready to double on the g-file, as well as adding yet another potential defender to the b7-
square.
19 h3

Question: What is the point of this?

Answer: White wants to get his king off the g-file by Kh2.
19 ... Rhg7 20 Bf4 Bc6 21 Kh2

42
 21 ... Kd7!

Question: What’s this? Surely after all the defensive moves to

protect b7, the king does not need to flee the queenside?

Answer: No, that is not the idea. Black is actually using his king to defend the weakness at e6, so as to free his knight on d8 to come via f7
to d6 and ultimately, e4!

Question: Hardly an everyday manoeuvre, is it?

Answer: Indeed not, but such deep manoeuvring is characteristic of these blocked Winawer positions.
22 g3?!
Uhlmann criticises this passive and weakening move, and recommends instead 22 Reb1, when Black must
always reckon with a potential exchange sacrifice on b7.
22 ... Nf7 23 Qc1
White’s last intended 23 h4 and Bh3, but now runs into 23 ... N7h6! and a check on g4, so White changes
his mind.
23 ... N7d6 24 Bxd6?!
This is also a huge concession. White’s reluctance to let the knight into e4 is understandable, but losing his
dark-squared bishop is a high price to pay. Now not only does any hope of dark-square counterplay disappear,
but the g3-square loses a crucial defender.
24 ... Nxd6 25 Qe3 Ne4 26 Nd2
This was White’s idea, challenging the knight and seemingly forcing it away, but Uhlmann now exploits
the hidden dynamism of his position.

43
 26 ... Qc7!
This very strong move threatens a rook sacrifice on g3, and virtually forces White to capture on e4, thus
opening the diagonal of the oft-despised ‘bad’ bishop on c6.
27 Nxe4 dxe4

Question: But isn’t Black just losing a pawn here?

Answer: He is losing the c4-pawn, yes, but this is irrelevant. He has terrible threats against the white king, such as ... h4 and ... f5-f4.
28 Bxc4
It is hard to know what to recommend for White. The queen exchange 28 Qf4? loses at once after 28 ...
Qxf4 29 gxf4 e3!, and Uhlmann himself points out that 28 Kh1 runs into 28 ... Rxg3 29 fxg3 Rxg3 followed
by a lethal discovered check with ... e3+. The best chance was 28 h4, but after 28 ... Bd5, Black prepares ...
f5-f4, and White’s position remains very unpleasant.
28 ... h4?
This natural move looks crushing, but the computer points out an unlikely tactical resource, which allows
White to snatch a draw. Correct is the preliminary 28 ... a5, driving the white rook off the fourth rank, and
only after 29 Rbb1 h4, with a dangerous initiative.

44
 29 Rg1?
Amazingly, White can force a perpetual check with 29 Bxe6+! Kxe6 30 d5+! Bxd5 31 Rxe4+ Bxe4 32
Qxe4+ Kf7 33 Qe6+ Kg6 34 Qg4+ etc.
29 ... f5 30 Rbb1 hxg3+ 31 fxg3 Rxg3!?
Uhlmann awards this an exclamation mark, but once again, the computer points out a superior defence for
White. Objectively, Black should therefore have played 31 ... Bxa4 32 Bb5+ Bxb5 33 Rxb5 b6 with a solid
advantage.
32 Rxg3 f4 33 Rxg8?
Here, the remarkable 33 Rf1! Rxg3 34 Qxf4 Qxf4 35 Rxf4 Rxc3 36 Bb3 allows White to put up tough
resistance.
33 ... fxe3+ 34 Kg2
No doubt missing Black’s next. 34 Rg3 is a more tenacious defence.
34 ... Kd6!
This nice move sets up ... Qf7, with decisive effect.
35 Be2?
Collapsing at once, but other moves also lose: for instance, 35 Rf1 Bxa4, or 35 Rg3 Be8 36 Be2 Qxc3.
35 ... Qf7 0-1

Despite the tactical opportunities missed by both players over the last few moves, this is a highly instructive game for Uhlmann’s middlegame
manoeuvring. Wolfgang Uhlmann was of course the world’s greatest Winawer exponent for over 40 years, using it as virtually his exclusive
defence to 1 e4, and winning many fine games with it. We will see several others later in this book.

45
Game 6
R.Byrne-V.Korchnoi
London 1979

This game was played in the BBC televised tournament, The Master Game, in which the players afterwards
recorded an abbreviated version of their thoughts during the game. I have made use of these comments below,
as they shed some interesting light on the whole Winawer variation.
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Nf3 Qa5 8 Bd2 Nbc6 9 Bd3!?

Rather an unusual move, but one with which Byrne had experimented previously. White is prepared to
sacrifice a tempo, in order to provoke the advance ... c5-c4, clarifying the position.
9 ... c4 10 Bf1

Question: Uh? What’s this?

Answer: White’s idea is to play g3 and Bh3. We have seen in previous games how White often does this anyway, having first played Be2,
0-0 and Re1; here, he hopes to gain time on that.
10 ... f6!?
This is a distinctly risky move in this position, as Black’s king is rather exposed after the exchange on f6.
After 10 ... Bd7, Korchnoi was worried about the standard idea 11 Ng5, provoking weaknesses. He therefore
decided to play the provocative text, partly motivated also by the tournament situation, which required him to
play for a win. It is a typical example of the kind of ‘heroic defence’, often practised by Lasker (Korchnoi’s
hero) and Nimzowitsch – Black invites difficulties, in order to achieve a fighting game, and create an
imbalance, where winning chances are also available.
11 exf6 gxf6 12 Nh4!
This is the drawback to Black’s plan. Now Qh5+ is a highly unpleasant threat. Against quiet development,
Black could play ... Bd7 and ... 0-0-0, reaching a position similar to the previous game.
12 ... 0-0
Practically forced, but the king is, of course, not terribly secure on the kingside.
13 g3 Rf7 14 Bg2 Bd7 15 0-0 Ng6 16 Nxg6 hxg6 17 h4
46
 Question: I don’t like the look of Black’s game here!

Answer: Indeed, White’s opening plan has worked out quite nicely, and he has a useful initiative. In his TV commentary, Korchnoi
admitted that White’s position “is clearly better – his king is better protected and he has the two bishops. They don’t play right now, but sooner
or later, the position will get open”.

Question: So why have you included this game?

It hardly looks like model play by Black!

Answer: No, but that is in many ways why I have included it. It is a really instructive example of how hard these positions can be to break
down, even when they go a bit wrong. The fact of the matter is that, without doing anything obviously wrong, Byrne’s position seems to tread
water from now on, and he is gradually outplayed. Korchnoi’s handling of the position is masterly, and shows how to play such structures.
17 ... Rh7!
The first step. Black must prevent the advance h4-h5.
18 Re1!?
This can hardly be bad, but it is the start of this ‘running on the spot’ approach that White finds himself
unwittingly following. Perhaps he should have played 18 Bf3, enforcing h4-h5.
18 ... Re8 19 Qf3 Kg7
White attacks things, Black defends them! At each step, White is invited to come up with something else
to do.
20 Bf4

47
 20 ... Nd8!
A really excellent move, and a manoeuvre which we saw to good effect in the last game also. Black needs
to protect his dark squares, so the knight comes to f7. From there, it rather ‘dominates’ the white bishop on
f4, covering the squares d6, e5, g5 and h6, which the bishop attacks.
21 Re2 Nf7

Question: Isn’t White still better here?

Answer: Objectively, perhaps he should be, but he has real trouble finding something to do. In fact, Byrne made a memorable comment
here: “It’s always a problem; every time I found the same thing. It looks at times as though in this opening both sides get stuck in a sticky
pudding after about 15 moves”. This is rather a good description of what often happens – the heavyweight manoeuvring struggles that often
arise in these blocked Winawer structures resemble a bit of a sticky pudding, with both sides finding it hard to improve their positions. But in
this example, Black continues to do so, whilst White struggles.
22 Re3

Question: That looks rather strange! The rook has only just gone to e2.

Answer: Yes, and it is a sign that Byrne was really struggling for a plan. His last move had intended 22 Rae1, but now he backed out,
because he was afraid Korchnoi would simply take the a-pawn, when White has no follow-up. With the white queen having to guard c3, we see
that two of White’s most powerful pieces are tied down to defending the weak queenside pawns.
22 ... Rhh8

48
 Question: Idea?

Answer: Just slightly improving his position. Korchnoi looks forward to the day when he will be able to unite his rooks behind the e-pawn.
At the moment, he still has to worry about h4-h5, but the day may come, and meanwhile, he continues to invite White to think of something to
do.
23 Qe2

Question: Now what is he up to?

Answer: White intends to bring his queen to d2, defending c3 from a different square. He is still rather fumbling around for a constructive
idea.
23 ... Bc6!

Question: Why the exclamation mark?

Answer: It is as much for psychological effect as anything else. Korchnoi reminds White that his c3-pawn is still under attack, so Black can
dangle the e6-pawn under White’s nose with impunity.
24 Qd2
24 Rxe6? Rxe6 25 Qxe6 Qxc3 would leave half the white position en prise. The text defends c3 again, but
now h4-h5 is no longer supported, and Korchnoi immediately takes advantage of this.
24 ... Re7! 25 Re2
White simply does not know what to do, whereas Black steadily improves his position.
25 ... Rhe8

49
 Suddenly, Black is ready to break in the centre with ... e6-e5. White has made no progress at all over the
past 7-8 moves, whereas Black has regrouped most of his pieces – the knight has come to f7, the bishop to c6
and the rooks to e7 and e8.
26 g4
Rather a desperate lunge, but Byrne was by now feeling enormously frustrated by the position, and felt that
he had to do something before the black pawns started to roll.
26 ... e5 27 Bg3 e4 28 g5 fxg5 29 hxg5 Bd7!
With the change in central structure, Black’s bishop no longer fulfils any function on c6, and so it seeks
pastures new.
30 Bf4
The g5-pawn is now starting to be vulnerable, and so White defends it, in anticipation of his next move.
30 ... Bg4 31 Re3 Qa6!

50
 Question: What is the idea of this?

Answer: The queen has done a sterling job on a5, tying the white pieces down to the defence of the c3- and a3-pawns. Now at last she
transfers to the other wing, where there are new fish to fry, namely the g5-pawn.
32 Rg3 Qe6 33 Qe3 Qf5 34 Kf1 Rh8
The rook is looking at h5, attacking the g5-pawn again, or at a penetration down the h-file.
35 Ke1 Bf3!
A very nice move. With the removal of the bishop from g2, the black rook will be able to penetrate along
the h-file.
36 Kd2
36 Bxf3 exf3 37 Qxe7 Rh1+ 38 Kd2 Qxf4+, winning a piece, is the tactical point of Black’s last move.
36 ... Rh4
The white army is being harried at every turn, while the poor rook on a1 has still not moved.
37 Bb8 Bxg2 38 Rxg2 Rh3
Decisive. Now the g5-pawn drops off.
39 Rg3 Rxg3 40 Qxg3 Nxg5

The game is basically over. White is a pawn down, and still has weaknesses on f2, c3 and a3. In addition,
the black knight now penetrates very effectively on the light squares, which White’s once-proud ‘anti-
Winawer’ bishop cannot protect.
41 Kc1 Nf3 42 Bxa7
White recovers his pawn, but his pieces are hopelessly scattered and inactive, whereas Black’s compact
army is poised to overrun the kingside.
42 ... Qg5+ 43 Qxg5 Nxg5 44 Bb8 Rf7 45 Rb1?
45 Bg3? leads to the instructive variation 45 ... e3 46 fxe3 Rf1+ 47 Kb2 Rxa1 48 Kxa1 Ne4 with a
textbook knight against bad bishop ending.

51
 This is rather ironic, considering that the dark-squared bishop is supposed to be White’s pride and joy in
the Winawer, whereas it is Black whose bishop is traditionally bad! The pawn on e3 is useless, whilst the g5-
pawn will cost White a piece.
However, 45 Be5+ was a little more tenacious, although 45 ... Kg8 (45 ... Kh6 46 Kb2 Rf3 is probably also
good enough) 46 Bg3 e3 47 fxe3 Rf1+ 48 Kb2 Rxa1 49 Kxa1 Ne4 is still winning.
45 ... e3
45 ... Rxf2 also wins, but Korchnoi prefers to stick with the plan outlined in the previous note.
46 f4 Ne4 47 Kb2

47 ... g5!
Forcing White to open the f-file, after which the rook supports the passed e-pawn.
48 fxg5 e2 49 Re1
Korchnoi points out that Black also wins after 49 Be5+ Kg6 50 Bf6 Nxg5!: for example, 51 Bxg5 Kxg5 52
Re1 Re7! 53 Kc1 Kf4 54 Kd2 Kf3 55 a4 Kf2 56 a5 Re4 and White is in zugzwang.
52
 49 ... Rf2 0-1
There is no defence to the knight transfer to f3.

A really great Winawer game, which is enormously instructive. It shows the kind of patience Black must be prepared to show in such positions,
especially when things have gone a little wrong for him. Korchnoi does not panic, but carefully covers his weaknesses, meets White’s threats,
and then gradually improves his position. White, on the other hand, finds it extremely hard to come up with a constructive plan, and is
gradually outplayed. One should also bear in mind that Robert Byrne was no fool – a very strong and experienced grandmaster, he had
appeared in the Candidates Tournament just five years before this game was played. When a player of his class finds the white side of the
Winawer so hard to play, all white players should pause for thought.

53
Game 7
R.Byrne-R.Vaganian
Moscow 1975

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 a4 Nbc6 8 Nf3 Qa5 9 Bd2 Bd7 10 Bb5

Question: Instead of 10 Be2. What is the idea?

Answer: The move was first played by the American Grandmaster, Robert Byrne, but most of the credit for popularising it goes to
England’s John Nunn. The bishop takes up a more active post, exerting greater pressure against the centre (notably supporting e5, by
threatening to take on c6 at some moment).

Question: But can’t Black just chase it away with ... a6?

Answer: Indeed; White hopes that this will prove a weakening, and/or a loss of tempo.
10 ... Qc7

Question: So what is the point of the text?

Answer: Black gets the queen out of the way of the bishop on d2, and prepares to adopt a set-up which is very similar to one we will see in
the lines with an early ... Qc7 by Black – see Game 14.
The text is one of four main alternatives:
a) 10 ... a6 is the most principled response, simply putting the question to the bishop at once. Byrne’s
original idea was 11 Be2, hoping that the pawn being on a6 will represent a weakness for Black, but practice
has not really borne this out. After 11 ... f6, play proceeds much as Games 3-5. The b7-pawn may sometimes
be a little weaker than usual, especially if Black castles short (a later Rb1 by White can no longer be met by
... b6), but it is hard to believe that Black need lose a great deal of sleep over this nuance. White has also tried
11 Bxc6, but then both 11 ... Nxc6 and 11 ... Bxc6 seem perfectly adequate for Black.

Question: So 10 ... a6 looks like a very easy and

54
 straightforward response to 10 Bb5 then?

Answer: Yes, and for that reason, it is John Watson’s recommendation for Black.
b) 10 ... c4 is another principled response, cutting off the bishop from its home base and ensuring it must
exchange itself. However, as we have already noted in this book, the pendulum of opinion has swung against
the commitment ... c5-c4 in many Winawer lines, and this is no exception. An excellent practical example of
the merits of White’s play was the game M.Chandler-V.Ivanchuk, Manila Interzonal 1990. Chandler’s
opening play was frequently influenced by his friend John Nunn, and this was one successful example: 11 0-0
0-0-0

12 Bc1! (exploiting the tactic 12 ... Qxc3?? 13 Bd2 to re-route the bishop to the a3-f8 diagonal, without
having to waste a tempo defending the c3-pawn) 12 ... f6 (another bad advert for Black’s set-up was the
following: 12 ... h6 13 Ba3 f5 14 exf6 gxf6 15 Re1 Rde8 16 Nh4 Qxc3 17 Bxe7 Nxe7 18 Bxd7+ Kxd7 19
Rb1 b6 20 Re3 Qa5 21 Qg4 Nc6 22 Qg7+ Kc8 23 Ng6 Rhg8 24 Qxf6 and Black did not last much longer: 24
... Qd2 25 Rxe6 Nxd4 26 Ne7+ Kb7 27 Rexb6+ Ka8 28 Rb7 1-0 J.Nunn-S.Kindermann, Munich 1991) 13
Qe1 Rhe8 (Psakhis gives as “a probable improvement” 13 ... Rde8 14 Ba3 Nf5 “with quite good chances of
equalising”) 14 Ba3 Kb8 15 Bxc6! (this exchange was also seen in the previous note, and is a standard feature
of the variation – White times the move so as to cause Black maximum embarrassment) 15 ... Nxc6 (15 ...
Bxc6? 16 exf6 gxf6 17 Qxe6 simply loses a pawn) 16 Qe3 Ka8 17 Rfb1 Bc8 18 Bd6 Rd7 19 Rb5 Qd8 20 a5
and White had obviously achieved much of what he wants in such structures.

55
 Chandler went on to win a fine game, which is worth seeing further: 20 ... g5 21 Bc5 g4 22 Bb6! Rc7
(already desperation, but 22 ... axb6 23 axb6+ Kb8 24 Qc1 mates quickly, whilst 22 ... Qe7 23 exf6 Qxf6 24
Ne5 is also very good for White) 23 Ne1 fxe5 24 dxe5 Bd7 25 Qc5 Re7 26 a6! bxa6 27 Rxa6 Be8 28 Bxc7
Rxc7 29 Qd6 Qxd6 30 exd6 Rb7 31 Rxb7 Kxb7 32 Ra1 and Chandler won the ending.
c) 10 ... f6 is another critical try, apeing Black’s best line against 10 Be2.

However, as already pointed out, here the bishop on b5 exerts counter-pressure against Black’s pressure on
e5, so it is arguably less logical to try to attack the e5 square in this line. After 11 Qe2 Qc7 (11 ... fxe5? fails
tactically; after 12 Nxe5! Nxe5 13 Qxe5 Bxb5 14 c4! Qd8 15 cxb5 White was clearly better in J.Nunn-
J.Brenninkmeijer, Groningen 1988) 12 0-0 a6 13 Bxc6 Nxc6 14 Bc1 cxd4 15 exf6 gxf6 16 Nxd4 Black’s
position is a little rickety, J.Nunn-S.Kindermann, Vienna 1991.
Returning to Vaganian’s 10 ... Qc7:
11 0-0 0-0
The game J.Nunn-I.Farago, Dortmund 1987, saw Black transpose into another c4-structure, with 11 ... Na5
56
12 Bd3 c4 13 Be2 0-0-0, but this one looks a little odd, since his knight is rather misplaced on a5 and the
queen passive on c7. White was considerably better after 14 Bc1 Kb8 15 Qd2 h6 16 Qf4 Be8 17 Qg4 Rg8 18
Nh4 Ka8 19 f4 Bd7 20 Ba3 Nf5 21 Nxf5 exf5 22 Qf3 Be6 23 Rfb1, although the game was eventually drawn.
12 Re1

Question: How is this position for Black? He looks more passive than
in Byrne-Korchnoi, for example, with his queen on c7.

Answer: Black is very solid here. We will discuss this set-up in more detail in Game 14, but it is safe to say that Black has no particular
problems. He intends play on the c-file, and also retains the chance of a kingside break with ... f6.
12 ... h6 13 Bf4?!
13 Bc1 is more usual here, but Black is still fine. Psakhis then gives 13 ... Na5 14 Ba3 b6 15 dxc5 bxc5 16
Bxd7 Qxd7 17 Bxc5 Rfc8 “with adequate compensation”. As we will see later, such pawn sacrifices are a
standard feature of this variation.
13 ... Ng6 14 Bg3 Nce7!
The white bishop is not well placed on g3, and Vaganian starts to exploit it at once. This knight usually
goes to a5 in this variation, but here, it has its eye on the enemy bishop.
15 Bd3
White does not want to exchange light-squared bishops.
15 ... c4 16 Bxg6
A concession, but as Psakhis points out, 16 Bf1 Qa5! leaves White really missing his dark-squared bishop
on the queenside.
16 ... fxg6!

57
 Question: That’s a surprise!

Answer: It is an excellent move, after which Black takes over the advantage. He opens the f-file for his rooks, and prepares a later ... g6-g5,
followed by ... Be8-h5, when his so-called ‘bad’ bishop will become a very strong piece. White stands worse on both sides of the board.
17 h4

Question: Goodness me! That looks a bit much!

Answer: It is a sign of desperation. Byrne can find no other way to stop the plan of ... g5 and ... Be8-h5, but of course, the move is seriously
weakening. It is remarkable how rapidly the position has turned against White, since his ill-considered transfer of the bishop to g3.
17 ... Rf7 18 Qb1 Kh7

Question: What is the point of this move?

Answer: A good question! I am not sure it is necessary at all, but it does not spoil anything. Black improves his king, and emphasizes that
he is in no great hurry. White does not really have a constructive plan.
19 Qb4 Nf5 20 Bh2 Raf8 21 Re2 Bc6 22 g4?
As Psakhis puts it, “Byrne loses his nerve”. On the other hand, he does not suggest what else White can do
about the threat of 22 ... Qd8, winning the h-pawn.

58
 22 ... Ng3! 23 Ng5+ hxg5 24 Bxg3
White hopes to have avoided the worst, but now the black queen switches to the kingside, after which the
game does not last long.
24 ... Qd8! 25 h5 gxh5 26 gxh5 Rf3
Threatening ... Qe8.
27 Kg2 Qe8 28 Rh1 Qf7

Now 29 ... Rxg3+ is the threat.

29 Kh2 Qf5 30 Qe7 Rf7 31 Qd8 Bxa4 0-1
Byrne had seen enough. One pawn has gone, and the moves ... Rxc3 and ... Qg4 will soon account for a couple more.

59
Game 8
V.Ragozin-M.Botvinnik
Training match, Moscow 1944

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Nf3 Qa5 8 Qd2

In contrast to the previous games, here White defends the c3-pawn with his queen, reserving the option of
bringing the bishop to a3.

Question: Isn’t this more logical if White is going to play a4?

Answer: Arguably, yes, although as we have seen already, the bishop can sometimes redeploy from d2 to a3 later in the middlegame. But
Qd2 saves time, and certainly appears more logical on the surface.

Question: What does “on the surface” mean?

I sense a ‘but’ somewhere in the background!

Answer: Well, yes, I’m afraid so. As we will see, the bishop cannot come to a3 so easily in these lines, without allowing Black a
comfortable ending. Because of this, the Qd2 lines are not generally regarded nowadays as being so dangerous for Black.
8 ... Nbc6 9 Be2

60
 9 ... Bd7

Question: So why doesn’t Black take the ending here with 9 ... cxd4?

Answer: Well, he has to be careful. For a start, trading on d4 forces off the queens, but only at the cost of undoubling White’s c-pawns. As
we already know, the Winawer involves trading positional advantages – by taking on c3 with his bishop in the opening, Black surrenders the
bishop-pair and weakens his dark squares, but in return, he mangles the white queenside pawns. Undoubling the pawns by an exchange on d4
risks throwing away his compensating advantages, and leaving White with a free bishop-pair.

Question: So are you saying the endings are bad for Black?

Answer: Not necessarily! In fact, here we come to a major issue. General rules of thumb must always be treated with some caution, as chess
has numerous exceptions, but we can state a useful such rule here: the endings arising after the exchanges on d4 and d2 are usually pretty
comfortable for Black, if White cannot recapture on d2 with the bishop. However, where he can play Bxd2, Black must be rather more careful
about entering the ending.

Question: Why does it make such a difference?

Answer: Basically because Black needs to be able to play a rapid ... Na5-c4 in such endings. If he is unable to do so, the white bishop-pair
can often assume considerable strength. If White recaptures with the bishop on d2, then ... Na5 is prevented.
The classic illustration of this rule of thumb is the game V.Smyslov-R.Letelier, Venice 1950, which we
will present in full. White had actually played a4 instead of Be2, and play then continued:
9 ... cxd4 10 cxd4 Qxd2+ 11 Bxd2!
The key point. Smyslov immediately stops the manoeuvre ... Na5-c4.
11 ... Nf5

61
 12 Bc3!
Another excellent move. Black was hoping his opponent would defend the d4-pawn with the move 12 c3?!
which would allow ... Na5, but Smyslov is wise to that trick.
12 ... Bd7 13 Bd3 Rc8 14 Kd2!
Again accurate. Black was threatening the cheapo 14 ... Nxe5, which Smyslov stops by defending his
bishop. Meanwhile, Black is still prevented from playing his knight into c4.
14 ... 0-0 15 a5!
Now Na5 is stopped once and for all, whilst the black b-pawn is fixed on b7, and can be attacked along the
b-file.
15 ... Rc7 16 Rhe1!

A subtle prophylactic move, directed against the ... f6-break.

Question: What is your assessment of this position?

62
 Answer: White has some advantage. He has stopped Black’s main source of counterplay, with ... Na5-c4 and play along the c-file, and he
has a potential target on b7. His also has the bishop-pair, and control of the dark squares, his traditional Winawer birthright.
16 ... f6?
Playing into White’s hands. Smyslov suggested instead 16 ... a6 with the idea of ... Na7 and ... Bb5, trying
to exchange the bad bishop and increase Black’s control of the light squares. White would have only a small
advantage after that.
17 Bxf5!
A nice positional idea. White surrenders the bishop-pair, in the cause of keeping the f-file closed, thus
preventing Black obtaining counterplay.

Question: But now we have opposite-coloured bishops.

Answer: Yes, but with plenty of other pieces on the board, this is not such a significant drawing factor.
17 ... exf5 18 exf6 Rxf6 19 Rab1
White now enjoys an indisputable advantage. Black has pawn weaknesses on b7 and d5, whilst the dark
squares c5 and e5 are very vulnerable. Smyslov exploits these plusses in exemplary fashion.
19 ... h6 20 Rb5! Be6 21 Reb1 Rff7 22 Ne1!
An excellent manoeuvre. The knights heads for d3, from where it attacks the two weak dark squares on c5
and e5, and also has the option to come to f4, attacking the d5-pawn again.
22 ... f4
Black stops the latter idea and tries to work up some counterplay on the kingside, but his pieces are rather
too passive for it to be effective.
23 f3 g5 24 Nd3 Kh7 25 Re1 Rf6

26 Rc5!
Now specific tactical threats start to appear. The first is 27 Nb4, winning a pawn.
26 ... Rc8
26 ... b6 is met by 27 axb6 axb6 28 Rb5 when the b-pawn is lost, since 28 ... Rb7? runs into 29 Nc5.
27 Nb4! Nxb4
After 27 ... Ne7 Smyslov points out the nice combination 28 Nxd5 Nxd5 29 Rxe6! Rxc5 (29 ... Rxe6 30
Rxc8) 30 Rxf6 Rxc3 31 Rd6 and White regains the piece, with an extra pawn.
28 Rxe6! Rxe6 29 Rxc8 Nc6 30 a6!
63
 Black was hoping to hang on after 30 Rc7+ Re7 but a fresh tactical blow destroys the last line of defence.
30 ... bxa6 31 Rc7+ Kg6 32 Rd7 Ne7 33 Bb4

Decisive. With the loss of the d5-pawn, Black cannot stop the two connected white passed pawns.
33 ... Nf5 34 Rxd5 Ne3 35 Rd8 Nxg2 36 d5 Rb6 37 Bc5 Rb7 38 Rc8! Nh4 39 Ke2 Nf5 40 Rc6+ Kh5 41
d6 Rd7 42 Rc7 1-0
A textbook game by Smyslov, which has taught generations of Winawer players not to exchange queens in
these endings, if White can take with the bishop.
Returning to Botvinnik’s 9 ... Bd7:
10 a4

10 ... Rc8!?

Question: So what is the point of 10 ... Rc8?

64
 Answer: It is a kind of high-class waiting move. We know from above that taking the ending is not so good here, because of 10 ... cxd4 11
cxd4 Qxd2+ 12 Bxd2! à la Smyslov-Letelier.

Question: But wait a minute – can’t Black play 12 ... Rc8 in this position?
Now he threatens 13 ... Nf5, when 14 Bc3 is impossible because of
14 ... Nxe5! so White would have to play 14 c3.

Answer: A nice try, but White will answer that with 13 Ra3! covering c3, and meeting 13 ... Nf5 with 14 Bc3. Botvinnik’s 10 ... Rc8 is in
fact a more subtle version of your idea here.

Question: Okay, I see. So does Black have any other moves?

Answer: Yes, he can also play 10 ... f6 which we will examine in Game 10.
11 0-0
This allows Black to show his idea. White has two more testing tries: 11 dxc5 (see Game 9), and 10 Bd3!?
(Game 11). Note that 11 Ba3 would allow Black to reach the same comfortable version of the ending as
occurs in the game: 11 ... cxd4 12 cxd4 Qxd2+ 13 Kxd2 Nf5 14 c3 Na5, etc.
While we are talking about this position, however, I should point out one other way Black can go wrong in
these endings. Even when the white bishop is on a3, rather than d2, it is important to play the move ... Nf5
early. It is often a mistake to allow White to trade both his bishops for Black’s knights, and leave an ending
with the white knight against Black’s light-squared bishop. An instructive recent example of this was the
game Y.Vovk-S.Williams, Gatwick 2012. In the position after 13 Kxd2 in the above variation, but with the
white bishop on f1 and the black rook still on a8 ...

... Williams continued 12 ... Na5!? (12 ... Nf5! 13 c3 Na5 is equal) 13 Bd3 Nc4+? (13 ... Nf5!), after
which the Ukrainian GM seized his chance: 14 Bxc4 dxc4 15 Bxe7! Kxe7 16 Kc3.

65
 This ending is not nice for Black. Although his bishop has increased scope now the pawn has left d5, it is
still inferior to the agile white knight, whilst Black has weak dark squares, a weak pawn on c4, and (most
crucially) lacks effective counterplay. Vovk went on to win instructively, as follows: 16 ... Rac8 17 Nd2 Bc6
18 f3 f5 19 a5 g5 20 Rhb1 h5 21 Rb4 (the c-pawn is already dropping off) 21 ... g4 22 Rg1 gxf3 23 gxf3
Rhg8 24 Rxg8 Rxg8 25 Rxc4 and White proceeded to the full point.
11 ... cxd4!
Now Black reaches his desired endgame.

Question: But isn’t White ready to recapture on d2 with the bishop?

Answer: Yes, but his king is not in the centre and your tactical idea above prevents White from establishing the Bc3/Kd2 formation which
Smyslov used.
12 cxd4 Qxd2 13 Bxd2 Nf5 14 c3
Now forced, since 14 Bc3? is refuted by 14 ... Nxe5, which reveals the point of Botvinnik’s 10th move.
14 ... Na5

66
 Question: So Black has got his knight to a5. Why is this such a big deal?

Answer: The point is that with the pawn on c3, the white dark-squared bishop has much less scope, whilst Black has much more active play
on the queenside. His knight will come to c4, he can double rooks on the c-file, etc. He has also fixed the white a-pawn on a4, at least
temporarily, and he has no issues with pressure down the b-file, such as we saw in Smyslov-Letelier. A comparison between this position and
Black’s position in that game will clearly reveal the difference.

Question: So is Black better here?

Answer: I am not sure that he is objectively better, but he is certainly comfortable. In practice, Black quite often wins such endings,
although that may be because more often than not, he is the stronger player.
15 Rfb1

15 ... Rc7

67
 Question: Why didn’t Black block the b-file with 15 ... b6 here?

Answer: He could, and that would be a perfectly valid move. Timman suggests that Botvinnik probably wanted to avoid the risk of White
later breaking with a4-a5, after the black knight goes to c4.
16 Bc1 f6

Question: This looks a bit odd. Will Black really take on e5?

Answer: Probably not. The point of the move is simply to give his king a square on f7, so he can bring the king’s rook over to the
queenside.
17 Ba3 h5!
Nice prophylaxis. 17 ... Kf7?! runs into 18 Bb4 Nc4 19 g4 Ne7 20 Bd6 winning the b7-pawn, as pointed
out by Timman.
18 Bb4 Nc4 19 a5 Bc6 20 Bc5 a6

White has played well so far and held the balance comfortably.
21 Bd3 Kf7 22 Re1 Re8 23 Rab1 Bb5 24 h3 Ne7!
Heading for c6. White has to be careful about his a-pawn.
25 Bxc4
As Timman points out, this is hard to avoid in the long run, if only to save the a-pawn, so White tries to stir
up some queenside play while he has the chance.
25 ... Bxc4 26 Bd6 Rd7 27 Nd2 Rc8 28 Rb6 Bb5

68
 29 exf6?!
Timman is critical of this, and recommends 29 Nb3, after which a drawish position arises following the
practically forced variation 29 ... Rxc3 30 Nc5 Rxc5 31 Bxc5 Nc8 32 Rxb5 axb5 33 exf6 gxf6 34 Rb1.
29 ... gxf6 30 Bb4 Rc6 31 Rxc6 Nxc6 32 f4
He wants to prevent the advance ... e6-e5.
32 ... Nxb4 33 cxb4 Rc7
Suddenly, Black is better and White must fight to hold the balance.
34 Nb3 Rc4 35 Nc5 Rxd4 36 Rxe6 Rxb4 37 f5
Timman applauds this as the best chance, although the computer disagrees, preferring 37 Rb6 Rxf4 38
Rxb7+ Kg6 39 Rb6. Black is certainly better after 39 ... Rf1+ 40 Kh2 Ra1 41 Nxa6 Rxa5, but with the
position that much ‘smaller’ (i.e. the passed pawn not so far distant from the kingside), White would have
drawing chances.
37 ... Rf4?!
This justifies White’s last. The computer’s 37 ... Be8 looks better.
38 Rb6?
38 Nxb7 should draw: for example, 38 ... Rxf5 39 Re1 Rf4 40 Nd6+ Kg6 41 Nxb5 axb5 42 Rb1 and White
can hold the rook ending.
38 ... Rxf5 39 Rxb7+ Kg6 40 Rb6
40 Ne6 is slightly more tenacious, but after 40 ... Re5 Black should still win.
40 ... d4 41 Nxa6 d3
Now the d-pawn is too strong.
42 Nc7 d2 43 Rd6 Rf1+ 44 Kh2 d1Q 45 Rxd1 Rxd1 46 Nxb5 Rd5 47 Nc3 Rxa5

69
 48 Ne2 Kf5 49 Kg3 Ra3+ 50 Kh4 Ra2 51 g4+ Ke5 52 Nc1 Rc2 53 Nd3+ Ke4 54 Nb4 Rc4 0-1
The knight will be hunted to extinction on the a-file.

On the basis of this game and the notes, we now know most of the essentials of these Winawer endings arising after the queen exchange on d2.
Basically, Black needs to avoid two main things – the white set-up with Bc3/Kd2, preventing Black’s ... Na5-c4 manoeuvre (Smyslov-Letelier),
and also the position where White exchanges both bishops for knights (on c4 and e7), as in Vovk-Williams. Providing he avoids both of these,
Black should be comfortable in most such endings.

70
Game 9
W.Hartston-W.Uhlmann
Hastings 1972/73

1 e4 e6 2 Nc3 d5 3 d4 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 Nf3 Bd7 8 a4 Qa5 9 Qd2 Nbc6 10 Be2 Rc8 11
dxc5!

Question: Why is this good?

Answer: Introduced by Vasily Smyslov, it is the most critical reply to 10 ... Rc8. White puts paid once and for all to any endgames after an
exchange of pawns on d4.

Question: Okay, but surely he mangles his pawn structure in doing so?
And, if the idea is so good, why didn’t he play it last move?

Answer: Well, to take the second question first, he is reacting to Black’s last. This has deprived Black of any chance to castle queenside. In
addition, as we will see, White hopes to bring his knight to d6, in which case he will gain a tempo on the rook. As for the first question, White
does mangle his pawns, yes, but he gets various compensations. Firstly, he frees the square d4, as a base for his knight. From there, the knight
can potentially jump via b5 to d6, or maybe to b3, from where it would defend the extra pawn on c5. A second point is that White opens the
position – he rules out the move ... c5-c4, and if Black recaptures on c5, then the a3-f8 diagonal is opened for White’s dark-squared bishop. He
hopes that his dark-square play will more than compensate for the weak queenside pawns.
11 ... Ng6

Question: Why doesn’t Black just recapture on c5?

Answer: That would be bad after 11 ... Qxc5? 12 Ba3 Qa5 13 Bb5 when Black will have trouble castling, as 13 ... 0-0?? 14 Bxc6 wins
material.
12 0-0
The immediate 12 Qe3 is also possible, hoping to transpose into Smyslov-Uhlmann below. Old analysis
from the 1970s claims that Black can equalize with the pawn sacrifice 12 ... d4!? 13 Nxd4 Nxd4 14 Qxd4
Qxc5 15 Be3! (the best way to pose Black problems; he was fine after 15 Qxc5 Rxc5 16 f4 Rxc3 17 Bd3 0-0

71
in W.Wittmann-W.Uhlmann, telex 1990) 15 ... Qxc3+ 16 Qxc3 Rxc3 17 Kd2 Rc8 18 Rhb1 Bc6 19 Bb5 0-0
20 Bxc6 Rfd8+ (allegedly a key improvement over 20 ... bxc6 21 Bxa7 Rfd8+ 22 Kc3 with a clear advantage
to White, W.Hartson-S.Webb, England 1973) 21 Kc1 Rxc6 22 Rxb7 a6 (Eales).

However, running this line on further reveals that after 23 f4 Rdc8 24 Ra2 Nh4 25 g3 Nf3 26 h4 Ne1 27
Rb6 Rxc2+ 28 Rxc2 Nxc2 29 Kd2 Nxe3 30 Kxe3 Ra8 White is vastly more active in the rook ending. I
would not touch this ending with a barge-pole as Black.
Instead after 12 Qe3, Black should probably prefer 12 ... Nf4, when it is hard for White to avoid the
repetition after 13 Bf1 Ng6 14 Be2 Nf4.
12 ... Ngxe5
This is an attempt to avoid the continuation of the stem game, which went 12 ... 0-0 13 Qe3 Qc7 14 Nd4!
Qxe5 15 Nb5 Qxe3 16 Bxe3 a6 17 Nd6 Rc7 18 a5 and White had a strong bind in V.Smyslov-W.Uhlmann,
Mar del Plata 1966. Smyslov went on to win a memorable game.
13 Nxe5 Nxe5

72
 14 Qe3
14 Qg5 is more active, and conceals a fiendish trap: 14 ... Ng6 15 c4!. Now the trap is 15 ... dxc4?? 16 c6!
winning a piece. This has caught several players, sadly the most notable being 1970s French Defence
authority and Irish international John Moles, who fell for it against Gligoric in a clock simul. In his seminal
volume on the Winawer, he not only quoted the game, but added the intriguing comment “in vino veritas!”,
presumably implying that he had consumed a drink or two when playing the Gligoric game!
Instead, 15 ... 0-0 is best, when Moles assesses the position as equal, saying “White’s extra pawn is
worthless and his two bishops are not particularly dangerous.” Watson analyses further: 16 cxd5 exd5 (Fritz’s
16 ... f6 17 Qd2 Qxd2 18 Bxd2 exd5 19 Be3 Rfe8 is also fine) 17 Qxd5 Bc6 18 Qd2 and now the nice tactic
18 ... Qxc5! 19 Ba3 Qf5 20 Bxf8 Nf4 when both ... Qg5 and the neat ... Qh3 are threats. Fritz confirms that
White has nothing after 21 Rfd1 Rxf8.
14 ... Ng6

Question: So what is the verdict on this position?

Answer: I think Black is fine. He has regained his pawn, and has much the sounder pawn structure. White still has the bishop-pair, and the
dark squares in Black’s camp look like their usual bad accident, but as Winawer players, we can live with that, can’t we? In fact, Uhlmann even
assesses the position as better for Black, and although, like many of his assessments in this opening, that may be a touch optimistic, it is not
long before Black really does stand better.
15 Ba3 Bc6 16 f4
This is necessary, as otherwise Black will mobilize his central pawn majority with ... e6-e5.
16 ... 0-0 17 Qh3?!
This somewhat pointless queen manoeuvre looks like the start of White’s difficulties. In view of what
happens in a few moves’ time, White should probably follow the computer’s advice and secure his a-pawn,
with 17 Bb4 Qc7 18 a5.
17 ... Rcd8 18 Bb4 Qc7 19 Bd3 f5

Question: I am not sure I would have played that! Doesn’t this make
Black’s bishop even worse and leave weaknesses on e6 and e5?

Answer: It is a case of attackable and non-attackable weaknesses. Black wants to stop the advance f4-f5. After the text, this is prevented
forever, and the white bishop on d3 is severely curtailed in its mobility. Meanwhile, the e5- and e6-squares are easily defended, and far from
easy to attack effectively.

73
 20 Qg3 Rf6 21 Qe3
Uhlmann points out that this was the last chance to play 21 a5.
21 ... a5

Question: Doesn’t this just win a pawn?

Answer: Yes, but not entirely free of charge. As Uhlmann points out, White is able to bring his dark-squared bishop to the long diagonal,
and work up some initiative in compensation.
22 Ba3 Bxa4 23 Bb2 Bc6 24 c4 Rf7 25 g3
25 Qxe6? Nxf4 is not good for White, but the e-pawn is now threatened.
25 ... Re7

26 Rae1?
Uhlmann does not comment on this move (perhaps because he thought Black was just better all the way
through!), but the computer suggests that it is a turning point, after which White is always struggling. It does
not consider Black to have any advantage after 26 cxd5 Bxd5 27 Bc3 Ra8 28 Ra3. Indeed, such are the chains
in which Black is held on the dark squares, that it is hard to see him making serious progress.
26 ... Qd7
Now, by contrast, White can never exchange on d5, because of the queen recapture.
27 Qd4 a4 28 h4 dxc4! 29 Qxd7
After 29 Qxc4 Uhlmann intended 29 ... h5, but 29 ... Qd5 looks like a safer way to secure the advantage.
29 ... Rdxd7 30 Bxc4 Rd2 31 Rf2
Forced, because 31 Rxe6 loses after 31 ... Rg2+ 32 Kh1 Rxe6 33 Bxe6+ Kf8, and if 34 Bd7 then 34 ...
Rf2+.
31 ... Rxf2 32 Kxf2 Kf7

74
 Question: Black is still a pawn up. Is he winning this ending?

Answer: Well, my old Russian trainer Igor Belov always used to deplore the use of the word “winning” in such positions, saying: “It’s only
winning when you attack something big and he can’t defend it!”. But certainly, Black has good practical winning chances. The remainder is of
peripheral interest to our main theme, which is the opening, so I will skip through the rest of the moves relatively briefly.
33 h5 Nf8 34 Be5 Re8 35 Bd6 Nd7 36 g4 fxg4
Uhlmann gives 36 ... b5! as best: for example, 37 cxb6 Nxb6 38 Ba2 Bd5 39 Rb1 Nc8! 40 Bxd5 Nxd6
“with a clear endgame advantage”.
37 f5 Nf8 38 Kg3 Kf6
This time, Uhlmann’s recommendation is less convincing: 38 ... Bf3 (“!”) 39 fxe6+ Nxe6 “and Black
dominates”, but in fact, it is hard to see how he can untangle after 40 Bb5 Bc6 41 Rf1+ Kg8 42 Bc4.
39 Be5+?
The losing move. White should hold after 39 fxe6 Nxe6 40 Kxg4, since Uhlmann’s 40 ... Bf3+ (“!!”) is
met adequately by 41 Kg3! Bxh5 42 Bb5 Ra8 43 Bc4: for instance, 43 ... Ng5 44 Be7+ Kg6 45 Kf4 h6 46
Bd3+ Kf7 47 Bc4+ etc.
39 ... Kxf5 40 Bxg7 Nd7 41 Bd3+ Kg5 42 Bxh7
42 h6 is slightly more tenacious.
42 ... Kxh5 43 Re2 Kg5

75
 Now it is all over.
44 Bd4 Bf3 45 Rd2 Nf8 46 Bd3 Ng6 47 Rh2 e5 48 Be3+ Nf4 49 Rh7 a3 50 Bc4 Kg6 51 Rh2 Ra8 52 c3
Bd5 0-1

There is a slightly sad postscript to this game. At the start of it, Hartston needed half a point from his two remaining games, for a Grandmaster
norm. After losing here, he was outplayed in the final round by Larsen, missed out on his norm, and never did become a Grandmaster.

76
Game 10
R.Bogdanovic-W.Uhlmann
Sarajevo 1965

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 a4 Nbc6 8 Nf3 Qa5 9 Qd2 Bd7 10 Be2

10 ... f6
Instead of 10 ... Rc8, as seen in the previous game, Black adopts a more dynamic approach, immediately
attacking the white centre.

Question: Is this better than 10 ... Rc8?

Answer: Objectively, I am not sure there is much to choose between the two moves, but 10 ... f6 is more dynamic, and also more risky.

Question: What about the standard idea 10 ... c4 here?

Answer: In this position, closing the position is definitely to be frowned upon. By comparison with the examples we have already
considered in this book, it is obvious that White is better off with his queen on d2 and his bishop on c1 – he is a couple of tempi up on the best
he can hope for in the Bd2 lines.

Question: Yes, but how important are a couple of tempi in a blocked position?

Answer: Less important than in an open one, certainly, but still not to be despised. Examples such as Planinec-Timman showed how nice
White’s position can be in these structures, if he gets himself organised well; to give him some extra tempi is really asking for it.

Question: So do you have an example from grandmaster practice?

Answer: Well, perhaps because 10 ... c4 is so illogical, there have not been many practical examples. Moles quotes a game A.Lutikov-
V.Lyavdansky, USSR 1965, which went 11 0-0 0-0-0 12 Ba3 f6 13 Rfb1 and White was simply better. This is one case where I must just
appeal to you to trust me!
11 exf6
This is the most critical. 11 0-0 fxe5 12 dxe5 (or 12 Nxe5 Nxe5 13 dxe5 0-0) 12 ... 0-0 is equal.

77
 11 ... gxf6

12 dxc5

Question: Why is this necessary? Can’t White just play 12 0-0 here?

Answer: He can, of course, but Uhlmann claims easy equality in the ending after 12 ... cxd4 13 cxd4 Qxd2 14 Bxd2 Nf5.

Question: Is that necessarily so? From the previous game,

I thought White was doing well if he can play 15 Bc3 in such
a position? After all, 15 ... Rc8 can be met by 16 Ra3.

Answer: Yes, but even so, without his king in the centre on d2, White does not usually have a lot in such positions. Furthermore, the
exchange on f6 has changed things radically, because Black has the possibility of ... Nd6-e4. He is comfortable here.
12 ... 0-0-0 13 0-0 e5

78
 Question: Well now, what is happening here? It looks rather sharp!

Answer: It is, indeed. White has an extra pawn, at least for the moment, and as always, the opening of the position potentially makes his
bishop-pair stronger. On the other hand, his pawn structure on the queenside is shot away, and Black has a strong pawn centre. Chances are
probably about equal, but in a very dynamic way, and any inaccurate play by either side could have serious consequences.
14 c4
14 Ba3 is given as best by Moles. The game W.Hartston-S.Hutchings, Swansea 1972, then continued 14 ...
Rhg8 15 Rfb1 Ng6 16 g3 and now 16 ... h5 gives counterplay, in an unclear position.
14 ... d4!?

Question: Is this best?

Answer: Possibly not. Black seizes central space and keeps his powerful pawn centre, but remains a pawn down. In a later game, L.Stein-
V.Doroshkevich, USSR Championship, Riga 1970, Black tried 14 ... dxc4 15 Bxc4 Qxc5 16 Bb3. Now 16 ... Bg4?! 17 Qh6! Nd5?! (17 ... Bxf3
18 Qh3+ with a plus) worked out badly: 18 Nd2! Rhg8 19 Ba3 Qa5 20 Ne4 Nd4 21 Kh1! Be6 22 Rad1 Qc7 23 Nxf6! Rg6 24 Nxd5 Rxh6 25
Nxc7 Kxc7 26 Bxe6 Rxe6 27 c3 with a clear advantage. Moles suggests no fewer than three improvements at move 16: 16 ... Nf5, 16 ... Ng6,
and 16 ... e4, all of which may offer reasonable chances.
However, the computer’s suggestion 14 ... Qxd2! 15 Bxd2 e4 may be best of all, when Black really does
have a good initiative for his pawn.
15 Qh6
A critical try is 15 Qxa5! Nxa5 16 Nd2. Uhlmann now gives 16 ... f5 as slightly better for Black, but this
assessment seems questionable after 17 Nb3 which seems to favour White. Fritz prefers 16 ... Rhg8 17 g3 and
only now 17 ... f5 18 Nb3 Nac6; the point being that 19 Bg5 is no longer possible. The position then remains
unclear.
15 ... Ng6 16 Nd2 f5 17 Nb3 Qc7

18 f4
Uhlmann hails this as the best for White, but Fritz likes 18 a5 a6 19 Rd1, considering the position clearly
in his favour.
18 ... Nb4 19 Bd1?!
Uhlmann gives the line 19 fxe5 Nxc2 20 Rb1 Bc6 21 Rxf5 Be4 “with a strong attack for Black”, but once
again, the silicon beast is far from convinced, reckoning White is clearly better after 22 Rf1.
19 ... d3 20 cxd3 Nxd3 21 fxe5 Qxe5 22 Bf3!

79
 The best. 22 Bg5 is another important option, mentioned by Uhlmann. He analyses a long combination
beginning 22 ... Bc6, but the computer shows his analysis to be full of holes. Instead, 22 ... f4! 23 Bf3 (23
Bxd8?? Qe3+ mates) 23 ... Rde8 is equal, according to Fritz.
22 ... f4 23 Bd2 Bc6 24 Bxc6?
White can win by force here with 24 Qh3+! Kb8 (if 24 ... Rd7 25 Bg4) 25 Bxc6 bxc6 26 Ba5 when the
rook has no retreat on the d-file.
24 ... bxc6 25 Rab1 Rd7 26 Qh3 Re8 27 Qf3 Qe4 28 Na5?!
28 Qxe4 Rxe4 29 Na5 holds the balance. Now Black starts to take over.
28 ... Qxf3 29 Rxf3 Re2

30 Bc1?
Losing at once. 30 Bc3 was the only move, although Black is now clearly better after 30 ... Nh4.
30 ... Nh4 31 Rxf4
A final amusing example of the flaws in Uhlmann’s pre-computer analysis of this game is his note to 31
Rh3. He gives 31 ... Rxg2+ 32 Kf1 Rf2+ 33 Kg1 Rg7+ 34 Kh1 Rf1#, overlooking the mate in one with 31 ...
Re1#.
31 ... Rxg2+
Good enough, but there is a forced mate with 31 ... Re1+ 32 Rf1 Nf3+ 33 gxf3 Rg7+ etc.
32 Kf1 Nxf4 33 Bxf4 Rd3 0-1

Question: Hmmm. I am not too convinced by all this! It looks as

though White was just better for much of the game.

Answer:
It is interesting that Moles, writing back in 1975, described this game as “difficult to analyse”; Uhlmann’s efforts rather bear this out! Of
course, it is easy for me to carp, sitting here, armed as I am with a 21st century computer, and I mean no disrespect to the East German GM, but
it does look as though his whole assessment of Black’s chances in this game was extremely optimistic. Indeed, the fact that so strong a player
as Leonid Stein should have been prepared to go down the white side of the line five years later, against Doroshkevich, suggests that he also
thought so. Nonetheless, I think Black is doing fine after 10 ... f6, if he heeds the improvements at move 14. However, there seems little doubt
that 10 ... Rc8 is a safer way to handle the line for Black.

80
Game 11
R.Fischer-W.Uhlmann
Buenos Aires 1960

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Nf3 Bd7 8 a4 Qa5 9 Qd2 Nbc6

10 Bd3

Question: Is this better than 10 Be2? Doesn’t it invite the reply 10 ... c4 with tempo?

Answer: It does invite that reply, but therein lies the point. White argues that he wants to see the pawn committed to c4, and is willing to
sacrifice a tempo to provoke the move. 10 Bd3 was a favourite of Fischer’s.

Question: What about 10 Bb5 in this position?

Answer: Actually, by comparison with the similar position in the Bd2 variation, here the move 10 Bb5?! is dubious, for tactical reasons.
Black can reply 10 ... Nxe5! 11 Nxe5 Bxb5. White can regain the pawn with 12 Ba3 Ba6 (Fritz’s 12 ... f6 may be even better) 13 Bxc5, but
after 13 ... f6 14 Nd3 Nc6 15 0-0 Kf7 Black was equal in E.Mnatsakanian-V.Korchnoi, Yerevan 1965, and went on to win.

81
 10 ... c4?!
Uhlmann takes the bait, but this move is no longer regarded as best, and Black usually prefers 10 ... f6.

Question: I think I am detecting a pattern here – in most

Winawer lines, Black seems reluctant to play ... c5-c4 these days!

Answer: I think that is definitely true. The understanding of the closed structures with ... c5-c4 has moved on over the years, and the initial
impression made by Botvinnik’s victories with the ... c5-c4 plan in the 1930s has now worn off. White players have a much better idea of how
to handle the position and coordinate their forces, and as a result black players have increasingly shifted attention to more fluid approaches,
based around ... f6 and leaving the pawn on c5. Indeed, despite his success in the present game, Uhlmann himself switched to 10 ... f6 in
another game against Fischer, played two years after this one. Play continued 11 0-0 fxe5 12 Nxe5 (12 dxe5 0-0 is also comfortable for Black)
12 ... Nxe5 13 dxe5 0-0 14 c4 Qxd2 (14 ... Qc7 is more ambitious, but the text is fine for Black) 15 Bxd2 Bc6 16 a5 Rad8 17 Be3 d4 18 Bd2
Ng6 19 f4 Nh4 and Black was fine.

The game ended in a draw after 20 Rf2 Rd7 21 Re2 Rdf7 22 Rb1 g5! 23 fxg5 Bxg2! 24 Rxg2 Nxg2 25
Kxg2 Rf2+ 26 Kg3 Rxd2 27 Rxb7 Rf7 28 Rb8+ Kg7 29 Re8! Rdf2 30 Rxe6 R7f3+ 31 Kg4 Rf4+ 32 Kg3 ½-
82
½, R.Fischer-W.Uhlmann, Stockholm Interzonal 1962.
11 Be2 f6 12 Ba3! Ng6?!
This early attack on e5 proves premature, because it is too dangerous to take the pawn anyway. Uhlmann
recommends 12 ... 0-0-0 13 0-0 and now Black has several options, although White remains a little better in
all cases.

Question: Why?

Answer: Basically, he has a good set-up against the ... c5-c4 structure. His bishop is strong on a3 and his pieces are well placed. 13 ... Rdf8,
a suggestion of Botvinnik, is perhaps the most interesting of Black’s tries here.

Question: What is the idea?

Answer: It is quite a typical manoeuvre for such blocked structures. The rook intends to come to f7. This will protect the e7-knight, threaten
activity along the f-file after the exchange ... fxe5, and also frees d8 for the black knight. We saw in Game 5 that the manoeuvre ... Nc6-d8 can
be useful in such positions, to defend b7 and potentially allow the light-squared bishop to take on a4.
13 0-0!

83
 13 ... 0-0-0

Question: So why can’t Black take on e5?

Answer: The line 13 ... fxe5?! 14 Nxe5 Ngxe5 15 dxe5 Nxe5 16 Qg5! Ng6 17 Bh5 is too dangerous for Black.
14 Bd6

Question: How do you assess this position?

Answer: I think White is certainly somewhat better. He has defended his e5-pawn conveniently, and obtained a relatively favourable
version of the blocked structure. However, once again, we see the resources of the black position being brilliantly exploited by Uhlmann, which
is why I have included the game in this book.
14 ... Nce7!
The knight is heading for f5, to attack the bishop.
15 Nh4!

84
 Question: Huh? That’s a surprise!

Answer: Fischer neatly exploits the tactical opportunities, to achieve his strategic objective of keeping his bishop on a strong diagonal.
15 ... Rde8 16 Nxg6! hxg6 17 exf6!

Question: What is going on?

Answer: Fischer is consistently pursuing the plan begun by his 15th move. He clears the h2-b8 diagonal, and intends that his bishop should
remain on that diagonal when it is driven away from d6.
17 ... gxf6 18 h3!

This is the point. The bishop now has a retreat on h2, and the diagonal h2-b8 is very strong.
18 ... Nf5 19 Bh2 g5 20 f4?
Up to now, Fischer has played brilliantly, and has achieved a great position, but this premature attempt to

85
shatter the black structure runs into a brilliant refutation. Instead, 20 Rfe1! leaves White with a clear
advantage. His bishops are both great, and the black king is distinctly uncomfortable.
20 ... Nd6! 21 Bf3?!
This was no doubt part of Fischer’s plan, but he misses the reply. He should have tried 21 fxg5!? Ne4 22
Qe3 (22 Qf4 e5!) 22 ... Qxc3 23 Qxc3 Nxc3 24 Bf3 fxg5 25 a5 (Uhlmann), when White is still slightly better,
although he has lost a fair bit of his previous advantage.

21 ... g4!!

Question: What is this all about?

Answer: It is a brilliant positional idea from Uhlmann, which transforms the position completely. As so often in the Winawer, the strategic
battle is over the activity of the white bishop pair; if White can get the bishops working, he will usually be doing well, whereas if Black can
tame them, his own chances will usually be good. Here, Uhlmann sacrifices a pawn to bury the bishop on h2.
22 hxg4
22 Bxg4? Ne4 23 Qe3 f5 24 Bf3 Qxc3 25 Qxc3 Nxc3 is now just better for Black.
22 ... f5!
The second step of the plan. Having sacrificed a pawn, Black now grants his opponent a protected passed
pawn!

Question: Sounds like madness!

Answer: Yes, but there be method in’t! It is all about killing the white bishop-pair, especially the bishop on h2. A few moves ago, it was the
pride of White’s position, glaring menacingly down the open h2-b8 diagonal, towards the black king. Now it will spend the rest of the game
staring at its own pawns on f4 and d4, neither of which can move and let the poor prelate back into the daylight.
23 g5
23 gxf5 exf5 just gives Black the open g- and e-files to play with.
23 ... Re7
There has been a total transformation of the position. The extra pawn is hardly felt at all, whilst Black has
a clear plan to strengthen his position: his bishop can come to h5, his rooks can double on the h-file, and the
knight constantly threatens to jump into e4. In addition, as usual in these structures, the a4-pawn is on death
row, and can only await Black’s announcement of a date for its execution.
24 Bg3 Be8 25 Qe3 Ne4 26 Bxe4
The knight clearly cannot be tolerated on e4.
86
 26 ... dxe4

27 Kf2
Fischer was always a great materialist, and duly clings on to his extra pawn. It was worth considering 27
d5!? to bring his bishop back to life.
27 ... Reh7
27 ... Qd5 at once was probably more accurate.
28 Rfb1
Passing up the last chance to liberate his bishop with 28 d5.
28 ... Qd5!
Slamming the door on the unfortunate prelate, which will spend the remainder of the game in cloistered
inactivity. The only issue here is whether Black can actually break through and win. Given the blocked
position, the game should probably end in a draw, but Black has what chances there are.
29 Qe1?
Fischer completely overestimates his position and misses Black’s 30th. Correct was 29 Qd2!? or 29 a5,
maintaining the blockade.
29 ... Rh1! 30 Qxh1??
This just loses. Uhlmann gives the line 30 Qe3 Rxb1 31 Rxb1 Bxa4 32 Ra1 Bxc2 33 Rxa7 Kb8 34 Ra1
Bd3 as offering better defensive chances.

87
 30 ... e3+!
This zwischenzug is key. Now the opening of the a8-h1 diagonal means that the oft-despised ‘bad’ bishop
will become extremely powerful, whereas its counterpart on g3 continues to resemble a big pawn.
31 Kg1
Instead, 31 Kxe3? Qe4+ 32 Kf2 (32 Kd2? Rxh1 33 Rxh1 Qxg2+ wins) 32 ... Rxh1 33 Rxh1 Bc6! 34 Rh2
Qxc2+ 35 Kg1 Qxc3 (Uhlmann) wins a piece, and 31 Ke2? also loses after 31 ... Rxh1 32 Rxh1 Qxg2+.
31 ... Rxh1+ 32 Kxh1 e2! 33 Rb5!?
A desperate attempt to drum up some counterplay, but it merely makes Black’s task easier. Even so, after
33 Rg1 Qe4 34 Rac1 Kd8 Uhlmann considers White’s position to be hopeless. The bishop will come to c6
and the c3-pawn will drop, after which the c2- and d4-pawns will follow.
33 ... Bxb5 34 axb5

34 ... Qxb5!
Trading the e-pawn for a passed a-pawn.
88
 35 Re1 a5 36 Rxe2 a4 37 Rxe6 a3 38 g6
Both 38 Re5 Qxe5, and 38 Re1 a2 39 g6 Qa6 40 g7 Qh6+ (Uhlmann) also lose.
38 ... Qd7!
But not 38 ... a2?? 39 g7 a1Q+ 40 Kh2

when it is White who wins!

39 Re5 b6!
Depriving White of the move Ra5, and with it, his last chance.
40 Bh4 a2 41 Re1 Qg7 42 Ra1 Qxg6 0-1
43 Rxa2 Qh5 44 g3 Qf3+ 45 Kh2 Qxc3 46 Bf6 b5 (Uhlmann) leaves White helpless.

This is one of the great Winawer games, thanks to Uhlmann’s brilliant strategic idea at move 21. The battle to tame the white bishop-pair is
central to the Winawer, and this is a memorable example, which every black player should bear in mind.

89
Game 12
R.Felgaer-V.Korchnoi
Bled Olympiad 2002

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 a4

White’s other main alternative here is 7 h4.

Question: Another non-developing move! What is the idea?

Answer: White logically aims to weaken further the kingside dark squares, which are always likely to be the Achilles’ heel of Black’s
position in the Winawer. If permitted, the pawn will come all the way to h6. Positionally, the idea is well founded, but as you will note, the
problem is that White neglects his development. Black has a number of approaches: 7 ... Qa5 (7 ... Qc7!? 8 h5 h6 9 Nf3 b6 is also playable) 8
Bd2 Nbc6 9 h5.

90
 This is the parting of the ways. Essentially, Black has three options here.
a) 9 ... cxd4 10 cxd4 Qa4 is one line. Black attacks the d4-pawn and intends to follow up with ... b6 and ...
Ba6, exchanging White’s key attacking bishop. The critical line now is Kasparov’s dynamic pawn sacrifice
11 Nf3 (11 c3 Qxd1+ 12 Rxd1 h6 is no problem for Black, nor is 11 Bc3 b6) 11 ... Nxd4 12 Bd3.

Question: So what does White have for his pawn?

Answer: He has caught up in development and hopes that his dark-squared bishop will develop active play. Typical ideas include Rh4
(usually prefaced by Kf1, to avoid the tactic ... Nxf3 with check), h6, etc. G.Kasparov-V.Anand, Linares 1992, continued 12 ... Nec6 (12 ... h6
13 Kf1 Nxf3 14 Qxf3 is another approach, when the analysis runs deep in the sharp lines after 14 ... b6 15 Qg3 Ba6! 16 Qxg7 Bxd3+ 17 cxd3
Rg8 18 Qxh6 Qd4, etc) 13 Kf1 Nxf3 14 Qxf3 b6 15 h6. White has good compensation, although the position is extremely sharp. This is not a
line Black can afford to enter without deep preparation.
b) The quiet approach is 9 ... h6!?. White retains a kingside space advantage, with the so-called Quartz
Grip on the kingside potentially offering long-term chances in the endgame, but on the other hand, Black

91
secures his dark squares against further significant weakening and can develop calmly with ... Bd7 and ... 0-0-
0. Psakhis then gives 10 Nf3 Bd7 11 a4 0-0-0 12 Qc1 f6, assessing the position as equal.
c) 9 ... Bd7 is the course preferred for many years by Uhlmann, as well as other strong players, such as
Psakhis, Lputian and Grischuk.

Black simply continues developing, placing his faith in that aspect of his position. After 10 h6 gxh6 White
has:
c1) 11 Nf3 0-0-0!? (despite being so comprehensively outplayed in the present game, Short retained his
faith in the h4-plan, and later used it to beat Lputian, who preferred the immediate 11 ... Ng8!? 12 c4 –
playing to open the game to the maximum – 12 ... Qc7 13 cxd5 exd5 14 dxc5!? 0-0-0 15 Be2 Bg4 16 Kf1 h5
and Black was doing perfectly well N.Short-S.Lputian, China 2004) 12 Bd3 (12 a4 Rdg8!? 13 Rxh6 Nf5 14
Rf6 Qd8 15 Be3 Rg6 16 Rxf7 Be8 17 Rxf5 exf5 18 g3 Bd7 19 Qd2 Rhg8 was reasonable for Black in
V.Bologan-W.Uhlmann, Dresden 1997, although White eventually won; 12 Rxh6 Ng8 13 Rh4 f6 also gives
Black reasonable counterplay) 12 ... c4 13 Be2 Ng8!.

92
 Question: That looks funny!

Answer: It is a typical idea in this line. Black is not so much concerned with trying to hang on to the h6-pawn, as preparing the break ... f6.
After 14 Kf1 f6 15 Qe1 fxe5 16 Nxe5 Nxe5 17 dxe5 Ne7 18 Bxh6 Rhg8 Black stood well in N.Short-L.Psakhis, Port Erin 1999. He went on to
outplay his powerful opponent totally, before collapsing in time-trouble and losing.
c2) 11 Rb1 is a newish and fairly critical idea here. White activates the rook on the b-file, attacking b7, and
setting up ideas of Rb5. He has scored quite well with this, and the best line for Black is not entirely clear at
present. A reasonably solid approach is 11 ... Qc7 (11 ... 0-0-0 12 Rb5 Qxa3 13 Rxc5 a6 14 Qb1 Kc7 15 c4
b6 16 cxd5! bxc5 17 d6+ Kc8 18 Rh3 Qa4 19 Rb3, D.Vocaturo-T.Hillarp Persson, Reykjavik 2009, with an
unclear position, is an example of the sort of sharp play than can ensue) 12 Nf3 Ng8 13 Rb5 c4 14 Qb1 0-0-0,
K.Nemcova-D.Petrosian, Dubai 2011, with a position that is not entirely clear, but where Black’s dark
squares look weaker than usual in such structures. As I say, 11 Rb1 is a critical test for this whole 9 ... Bd7
rapid development approach.
In summary, the early h4 plan remains an interesting and unclear way for White to proceed. The line with
9 ... h6 is probably the way for Black to play, if he wants a quiet position, without needing to prepare sharp,
forcing lines.
Returning to Felgaer’s 7 a4:
7 ... Qa5 8 Qd2 Nbc6 9 Nf3 f6

Question: This is another new move order, isn’t it?

Answer: Yes. Instead of 9 ... Bd7, Black immediately attacks the centre.

Question: So what is the difference?

Answer: I am not sure that it will always make a huge difference from the line with 9 ... Bd7 10 Bd3 f6. Assuming that Black ends up
playing his bishop to d7 fairly soon, then a transposition is likely, but, as the present game shows, there can be important differences.
10 Bd3
Korchnoi had used the 9 ... f6 move order before. In particular, he demonstrated one of its advantages after
10 Ba3. Rather than enter the standard Winawer ending with 10 ... cxd4 (comfortable though that should be
for Black), he can do even better: 10 ... fxe5! 11 dxe5 Qxa4! (once again, Korchnoi exploits the pin on the a-
file to annex a pawn; here, the fact that White is one move further from castling than would be the case if the

93
moves ... Bd7 and Be2 were already played, means that Black has time to extract his queen, before the threat
of Bxe7 becomes real) 12 Be2 b6 (simply defending the extra pawn) 13 c4 (13 0-0 Qe4 was presumably the
idea; White can gain a tempo or two on the queen, but he has no way to achieve anything concrete, and a
pawn is a pawn – all in all, a typical Korchnoi pawn-grab) 13 ... Nd8 14 0-0 Qd7.

Black has had to resort to one or two contortions to escape with his booty intact, but White was unable to
take advantage of his small development lead. After 15 Ng5 h6 16 Nh3 0-0 17 Rfd1 Ndc6 18 Bb2 Bb7 19
Nf4 d4 in F.Kuijpers-V.Korchnoi, Wijk aan Zee 1971, Black just had a solid extra pawn and a clear
advantage.
However, in view of what happens in the main game, perhaps White should prefer the safer 10 Be2.
10 ... fxe5 11 dxe5 0-0 12 0-0
It is hard to suggest any other move for White, but now there follows a typical and strong Winawer
exchange sacrifice.

12 ... Rxf3!
94
 This cries out to be played.
13 gxf3 c4 14 Be2 Nxe5

Question: So what is going on?

Answer: I think Black has a very attractive position. He has one pawn for the exchange, and has exposed the opponent’s king. Even more
importantly, as we pointed out in Game 2, in such Winawer structures the lack of open files means that rooks often struggle to become fully
effective. That is the case here, with Black’s knights proving very adept in the blocked structure. The computer may assess the position as
slightly in White’s favour, but I think that, in practice at least, Black has the easier position to play. The further course of the game bears this
out.
15 Ba3 Nf5 16 f4 Ng6 17 Bg4 Ngh4 18 Bb4 Qc7 19 a5 Qf7

White has secured his queenside from the typical pressure exerted by the enemy queen from a5, but
instead, Black switches his attention to the other wing.
20 f3 Bd7 21 Rae1 Qg6 22 Kh1 Bc6 23 Qf2 d4!
Once again, this positional pawn sacrifice just begs to be played. The bishop on c6 now becomes a
powerful piece, bearing down the long diagonal. There is also the incidental threat of ... Qxg4, hence White’s
next.
24 Bxf5 Nxf5 25 cxd4 Bd5

95
 Question: White has a whole extra exchange, so shouldn’t he be better?

Answer: In most positions, he would be of course, but here the rooks just lack effective scope. It is hard to argue that the rook on e1 is
really a better piece than the knight on f5. That is the secret of such positional exchange sacrifices in the Winawer.
26 c3 Qh5 27 Re5 Qh6 28 Re4

Question: White offers the exchange back?

Answer: Indeed! This is the acknowledgement that the rook is no stronger than the black minor piece.
28 ... Qh3
Black blithely ignores the rook.
29 Rg1 b6 30 Qg2 Qh5 31 a6 Rc8

32 Bd6??

96
 White tries to exploit the tactics to transfer his bishop to e5, but in fact, this move is just a losing blunder.
After a move such as 32 Rf1, the balance would be maintained. White cannot really do anything, but it is also
hard for Black to strengthen his position.
32 ... g6
Now that the mate threat on g7 is disposed of, Black threatens both ... Nxd6 and ... Nh4. White cannot
defend against both.
33 Be5 Nh4 0-1
White’s position caves in on the long diagonal.

This game is another nice illustration of the positional exchange sacrifice in these Winawer structures. The relatively blocked structure, with its
lack of open files, means that knights are frequently as effective as rooks. White was holding the balance until his blunder near the end, but
Black always had full compensation for his material investment.

97
Game 13
J.Vidarsson-C.Ward
Reykjavik 1998

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Qa5

Question: Yet another move order!

Answer: Yes! This is a line which has become very popular in recent years, as a way of avoiding the problems of 6 ... Ne7 7 Qg4. This
move order actually goes back to the 1950s, when it was used regularly by Lajos Portisch. In the 1960s, it attracted notice in the games of the
colourful US Virgin Islands player, Bill Hook, who played it in a famous game against Bobby Fischer. For these reasons, John Watson calls it
the Portisch-Hook Variation, a nomenclature which I have no difficulty following.

Question: So what is the point?

Answer: With his early queen manoeuvre, Black aims to bring quick pressure against the white pawns, hoping thereby to restrict White’s
options. Firstly, Black attacks c3, which needs to be defended. Black then intends to follow up with ... Qa4, which attacks d4 and c2, and also
blockades the white a-pawn, thus preventing the standard plan of a4 and Ba3.

Question: So do you think this line is an improvement on the usual 6 ... Ne7 line?

Answer: I am not sure I would be that categorical, but the early ... Qa5 is certainly a very logical way to play as Black, and I am surprised
that it has taken so long to become popular.
7 Bd2

Question: Is 7 Qd2 also possible here?

Answer: Of course. Now 7 ... Qa4?! makes less sense, since d4 and c2 are both defended, and 8 dxc5 looks an effective reply. 7 ... Ne7 is
likely to transpose into lines we have already examined, but Black can also pursue other plans, one of the main ones being 7 ... Nc6 8 Nf3 b6.

98
 Black hopes to trade off his bad bishop by Ba6. One oft-quoted example then is J.Poulton-N.Pert, British
League 2002, which went 9 dxc5 bxc5 10 a4 Ba6 11 Bxa6 Qxa6 12 Ba3 Rc8 13 Qd3 and now Watson gives
13 ... Qa5 with a good game for Black. Overall, 7 Qd2 looks no threat at all to Black.
7 ... Qa4 8 Qg4
This is the most aggressive reply. White insists on forcing his opponent to reveal how he plans to defend
the g7-square.
The other major option for White here is 8 Qb1.

Question: That looks odd! What is the point?

Answer: Basically, White wants to use the threat of 9 Bb5+ to force Black to close the position with 8 ... c4 (8 ... a6?! is not very attractive
after 9 Qb3).

Question: But this is just a standard structure, isn’t it?

99
 Answer: Yes, but White has clarified things. In addition, he has relieved the pressure against d4, which is often a feature of the Portisch-
Hook variation, and he has stopped the standard plan of ... b6 and Ba6, exchanging off the bad bishop. On the other hand, Black has his queen
blockading the white queenside on a4, and he is relieved of any worries about Qg4, probing the kingside.

Question: So how should we assess the position?

Answer: I think chances are about equal. A typical recent example, featuring a top player as White, was the game P.Svidler-D.Lima,
Khanty-Mansiysk 2011: 9 h4 (White has many possible set-ups, of course, but Svidler’s choice here is the most popular) 9 ... Nc6 10 h5 h6 11
Nf3 Bd7 12 g3 0-0-0 13 Bh3 (the most active formation; White has gained space on the kingside with the h-pawn march, and now he takes aim
at the e6-pawn, hoping to discourage Black from breaking with ... f6) 13 ... Nge7 14 0-0 Rdf8 15 Re1

15 ... g5 (Watson suggests 15 ... g6 as “more active”) 16 Nh2 f5 17 exf6 and now Watson’s suggestion is
17 ... Rxf6 18 f4 gxf4 19 Bxf4 Rg8 with counterplay. White has made progress in activating his dark-squared
bishop, but at the cost of weakening his king position. 8 Qb1 is a critical try against the Portisch-Hook, and
forces a closed, heavyweight struggle, but the average Winawer player should be used to such battles.
Objectively, the game is close to equal.
Now back to 8 Qg4, after which Black has a major choice to make.

100
 8 ... g6

Question: That looks incredibly ugly!

Answer: I agree! Weakening the dark squares even further in this way is grossly unaesthetic and creates permanent, long-term weaknesses.
However, it does defend g7, without giving up castling rights. Black intends to castle long, securing his king, and hopes that his positional
gains on the other wing will give him an adequate game.

Question: But surely White will infiltrate on f6 or h6, won’t he?

Answer: Black intends to leave his knight on g8 for the time being, to cover those squares. Meanwhile, he will pressurize the white centre
and queenside pawns, in order not to give White a free hand on the kingside.
The other option is 8 ... Kf8.

Question: But now Black cannot castle!

101
 Answer: True, but he has avoided a permanent weakening of his kingside dark squares. He hopes that his king will prove safe on f8.
Meanwhile, White must decide what to do about his hanging c2-pawn, hence 9 Qd1.

Question: This is all pretty confusing! Why retreat the queen again?

Answer: White’s argument is that his queen has done her job on g4, by provoking the enemy king into moving. Now the queen returns to
guard the centre. This is the most popular move here, and was Fischer’s choice in the game against Hook.
Alternatively, White can defend everything with the clumsy-looking 9 Ra2, after which a famous Hook
scalp continued 9 ... Nc6 10 Nf3 b6 11 Bd3 Ba6 12 0-0 Bxd3 13 cxd3 c4 14 dxc4 Qxc4 15 Raa1 Na5,
V.Liberzon-W.Hook, Nice Olympiad 1974, with a comfortable position for Black, who eventually scored a
famous victory.
Fischer-Hook continued 9 Qd1 b6. This is the standard plan in such positions. Black proceeds to exchange
light-squared bishops. Another logical treatment is 9 ... Nc6, increasing the pressure against d4. A high-class
example from a few years ago then continued 10 Nf3 Nge7 11 Qb1 c4 12 h4 Ke8 13 h5 h6 14 g4 Kd8 15 g5
hxg5 16 Nxg5 Ke8 (to and fro; the blocked position makes such manoeuvring viable) 17 Be2 b6 18 Bg4 Bd7
19 Qd1 Ng8 20 Be3 Nh6 and Black had good chances in S.Karjakin-P.Nikolic, Wijk aan Zee 2005.

He went on to stand clearly better, before falling for a tactic and losing.
Note that White’s last move, 9 Qd1, defended not only c2, but indirectly d4 as well: 9 ... cxd4 10 cxd4
Qxd4?? 11 Bb4+.
After 9 ... b6, Fischer went 10 h4.

102
 Question: Still not developing any pieces!

Answer: Once again, the relatively closed nature of the position makes such extravagances possible. As we have already seen, h4-advances
are another common theme in these Winawer structures. White hopes to push the pawn all the way to h6, weakening the enemy dark squares
further, whilst the rook can often develop via h4.
10 c4 is a critical alternative.

Question: What is the point?

Answer: White seeks to open the position, to activate his bishops and get at the uncastled black king. However, one adequate solution is 10
... Ne7 11 cxd5 Qxd4 12 dxe6 Qxe5+ 13 Be2 Bb7 14 Bf4 Qc3+ 15 Bd2 Qe5 16 Bf4 Qc3+ 17 Bd2, as in A.Grischuk-Y.Shulman, Khanty-
Mansiysk 2005.
After 10 h4, 10 ... Ne7 was Hook’s choice against Fischer. 10 ... Ba6 is also fine for Black: 11 Bxa6 Nxa6
(equally good is 11 ... Qxa6: for example, 12 h5 h6 13 Rh4 Ne7 14 Rf4 Nbc6 as given by Moskalenko, when
103
Black is doing well) 12 h5 h6 13 Rh4 Rc8 14 Rf4 cxd4 15 cxd4 Rxc2 (or 15 ... Ne7 16 Qf3 Ke8 17 c3 Rf8 18
Nh3 Nb8 19 Qg4 Rg8 20 Qf3 Rf8 21 Qg3 Rg8 22 Qf3 and ½-½ in I.Cheparinov-P.Nikolic, Wijk aan Zee
2005) 16 Bc3 Qb3 17 Rf3 Ne7 18 Rb1 Qa2 19 Ra1 Qb3 20 Rb1 was agreed drawn in P.Konguvel-
V.Moskalenko, Badalona 2006.
R.Fischer-W.Hook, Siegen Olympiad 1970, proceeded 11 h5 h6 (almost always necessary; the pawn
cannot be allowed to come to h6) 12 Rh4 Ba6 13 Bxa6.

The game went 13 ... Nxa6 with unclear play, but Watson suggests that 13 ... Qxa6 followed by ... Qc4
would be even better.
After that long digression, we now return to Ward’s 8 ... g6:

9 Ra2
White has many options here, just as after 8 ... Kf8. However, Black’s plan is largely the same – if he can
take on c2, he usually does so; if not, he generally plays ... b6, ... Ba6 and prepares queenside castling. Let us
look at some of the options:
104
 a) 9 Rc1 defends c2 more naturally than the game move, but of course, it involves sacrificing the a3-pawn.
Now 9 ... Nc6 10 Nf3 h6!?.

Question: What is this for?

Answer: Black just covers the dark squares a little more, taking g5 from White. 10 ... b6 is also perfectly reasonable.
11 Bd3 c4 12 Be2 Bd7 13 Qf4 Rh7

Now we see the other point of Black’s 10th. The rook defends f7, so Black can castle long.

Question: It all looks very clumsy!

Answer: It does, but such manoeuvres are often possible in blocked positions. Don’t forget that Black is effectively a pawn up already,
since he will be able to take on a3 whenever he feels like it. “A pawn ahead is worth a little trouble”, as they say.
B.Parma-L.Portisch, Bled 1961, proceeded 14 h4 0-0-0 15 0-0 Qxa3 16 Ra1 Qf8 17 Nh2 f5 18 exf6 Nxf6

105
with an unclear position.
b) 9 h4 is a natural way to attack the kingside, but now Black sets up a light-square blockade: 9 ... h5 10
Qg3 Ne7 11 Nf3 b6 12 Bd3 Ba6 13 Qg5 Nd7.

The other knight covers f6 instead! M.Jimale Abdulle-W.Hook, Bled Olympiad 2002, continued 14 g4
Bxd3 15 cxd3 c4 (a typical idea in such structures; once the exchange of bishops has taken place on d3, this
move secures an outpost for Black on c4) 16 gxh5 Rxh5 17 Qe3 Qa6 18 dxc4 Qxc4 19 Qe2 Rc8 20 Qxc4
Rxc4, and Black was better in the ending (weak pawns on a3, c3 and h4, plus a bad white bishop) and went
on to win a long game.
c) 9 Qf4.

Question: What is this – a pawn sacrifice?

Answer: Indeed. The strong Cuban GM Dominguez has played this twice, so it deserves respect: 9 ... Qxc2 (the critical move, else White

106
has just improved his queen position free of charge) 10 Nf3 b6 (10 ... c4 11 Be2 h6 12 0-0 Nc6 was unclear in L.Dominguez Perez-
W.Arencibia Rodriguez, Havana 2008) 11 dxc5 bxc5 12 Rc1 Qe4+ 13 Qxe4 dxe4 14 Ng5 f5 15 exf6 Nxf6 16 Bc4 Ba6 17 Ba2 Nbd7 18 Nxe6
Rc8 was unclear and eventually drawn, L.Dominguez Perez-R.Vaganian, Barcelona 2007.
d) 9 Qd1 is probably soundest, but less threatening than the Qf4 plan. After 9 ... b6 (9 ... Nc6 10 Nf3 b6 11
h4 h5 – 11 ... cxd4!? is more critical – 12 dxc5 bxc5 13 Be3 c4 14 Be2 Nge7 15 0-0 was V.Bologan-
L.Psakhis, Internet (blitz) 2004, and now 15 ... Nf5 is unclear) White has:
d1) 10 h4 was once Magnus Carlsen’s choice, but after 10 ... Ba6 11 Bxa6 Nxa6 12 h5 cxd4 13 hxg6 fxg6
14 Rh4 Ne7 15 cxd4 Rc8 16 c3 Nc5! 17 Rh3 Qxd1+ 18 Rxd1 Black had no problems in A.David-L.Psakhis,
Bad Wiessee 1999.

Black’s treatment here, allowing h5xg6 and recapturing ... fxg6, is quite notable. Psakhis is one of the
world’s premier French Winawer experts, and his handling of the positions is always worth a close look.
d2) After 10 Nf3 Ba6 11 Be2 Bxe2 12 Qxe2 c4!? (12 ... Nc6 also looks fine for Black) 13 0-0 Qxc2 14 a4
Nc6 15 Rfc1 Qf5 16 Be3 h6 17 Nh4?! Qe4 18 g4? g5 19 Ng2 Nge7 Black was clearly better and went on to
win in A.Hunt-N.Pert, Witley 1999.

Question: So which do you prefer, 8 ... g6 or 8 ... Kf8?

Answer: I am honestly not sure. 8 ... g6 certainly offends my eye, but, interestingly, it was the choice of Portisch himself, and I have great
respect for his positional judgement. Black certainly has a safer king after 8 ... g6, and, as Nigel Short likes to point out, checkmate ends the
game – it is no good accumulating positional advantages, if your king gets mated! That is one argument for 8 ... g6, rather than 8 ... Kf8. In
truth, I think both lines are equally possible, and the choice is largely one of taste.
Finally, we return to 9 Ra2:
9 ... c4
9 ... b6 à la Hook, is more natural. Ward’s treatment is always an option in such positions, but strikes me
as rather passive.
10 Qd1 Nc6 11 Be2 Bd7

107
 12 g4?

Question: What is this?

Answer: I think this is just rather a rash lunge. White should just develop normally with 12 Nf3 0-0-0 13 0-0 when he must be somewhat
better, although the usual problem of breaking down the solid black structure is still there, of course. The queen’s blockading post on a4 is
useful here, but the fact that the only kingside break with ... f6 involves such weakening of the dark squares and the e6-pawn means that it is
hard to create any active counterplay.
12 ... f6
Now that White has so weakened himself with 12 g4, this is justified.
13 Nf3 0-0-0 14 h4 h5!
White’s play has been extremely optimistic. Where is he going to put his king? The answer is “Nowhere” –
it just has to remain in the centre, and Black will open lines there and attack it.
15 gxh5 Rxh5 16 exf6 Rf5 17 Rg1
Missing his chance. There was the tricky tactic 17 Ne5! Nxf6 (if 17 ... Rxf6 18 Bg5) 18 Nf7, after which
Black must shed an exchange with 18 ... Ne4 19 Nxd8 Nxf2 20 Qb1 Nxd8, with an unclear position.
17 ... Nxf6
Now Black simply stands better. The white king is a sitting duck in the centre of the board.
18 Ng5 Rf8 19 Be3 Qa5!
Nicely luring the white queen on to the fatal square d2.
20 Qd2

108
 20 ... Rxg5!
The decisive blow. The knight jump to e4 is devastating.
21 hxg5 Ne4
The rest is rather gory, so I would suggest you remove any children from the room, before playing through
the remaining moves!
22 Qc1 Qxc3+ 23 Bd2 Qxd4 24 Be3 Qc3+ 25 Bd2 Qh3 26 c3 Qh2 27 Rf1 Rxf2 28 Rxf2 Qxf2+ 29 Kd1
Na5 30 Qc2 Qg1+ 0-1
31 Be1 Nf2+ 32 Kd2 Qxg5 is mate.

This main game was rather one-sided, after some over-optimistic play by White in the early middlegame, but nonetheless I think the Portisch-
Hook system is a very viable way to play for Black. If the latter is looking to avoid lots of forcing lines, such as arise after 6 ... Ne7 7 Qg4, it is
ideal, as it relies much more on general plans.

109
Game 14
G.Kanefsck-S.Atalik
Mar del Plata 2003

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7

Question: Is this move of independent significance?

Answer: It can be, but usually, as here, it just amounts to a transposition.

5 a3 Bxc3+ 6 bxc3 c5 7 Nf3 Qc7

Question: So, yet again something new! The queen looks more passive here.

Answer: This is the start of another system entirely. Rather than place the queen on a5, Black puts it on c7. From there, it still exerts
pressure down the c-file ( ... cxd4 is sometimes a threat), but also attacks e5. It is less threatening than 7 ... Qa5, but is a very solid and
respectable system. It enjoyed great popularity in the late 1960s and early 1970s, and was later revived with success in the 1980s/90s, by the
then leading GMs Yusupov and Predrag Nikolic. It remains a reliable and, in my view, underrated line for Black.
The real significance of this system with ... Qc7 is in terms of the move order 4 ... c5 5 a3 Bxc3+ 6 bxc3
Qc7. This is a weapon against an early Qg4. Black intends to meet 7 Qg4 by moving his f-pawn, when the
queen defends g7 along the rank. This line is examined in Game 19. However, if White meets this line with 7
Nf3, then after 7 ... Ne7, we reach the game position.
8 a4
8 Bd3 is a tame move, which allows Black to implement his main idea: 8 ... b6!, revealing a key point of
these ... Qc7 systems. Black prepares to exchange his traditionally bad light-squared bishop, by means of ...
Ba6. In general, if he can achieve this without suffering any immediate consequences, then he stands
comfortably. Not only has he offloaded his own potentially bad bishop, he has also deprived White of one of
his key attacking pieces, after which castling short becomes much safer for Black.
8 ... b6 9 Bb5+

110
 Question: What’s this – “patzer sees a check, patzer gives a check”?

Answer: Not at all. This move actually embodies quite a subtle idea. White is trying to prevent Black from implementing his ... Ba6 plan, to
exchange bishops.

Question: But how can he do that? Surely ... Ba6 will be

possible, after the check has been blocked?

Answer: Well, as John Nunn always used to say to me, when I tried to spout general principles, “What move are you going to play?”

Question: Well, how about just 9 ... Nbc6 then?

Answer: Okay, but now you can’t play ... Ba6, because it will be en prise!

Question: Oh, I see. Okay, I will play 9 ... Nec6 instead,

and then I can follow up with ... Ba6.

Answer: Yes, but now you have removed the only minor piece that defends your kingside. You will get to make the bishop exchange, sure,
but your kingside is rather bereft. The old game A.Lein-A.Bastrikov, USSR 1959, continued 10 0-0 Ba6 11 Ng5! h6 12 Nh3 and moves such as
Qg4 and/or Nf4-h5 are in the air. This is not terribly comfortable for Black.

Question: You have an answer to everything, don’t you! All right,

let’s play simply 9 ... Bd7 and now we will exchange bishops on b5 instead!

Answer: “Cherez moi troop!”, as they say in Russian ! (“Over my dead body!”). The point of White’s idea is now revealed:
9 ... Bd7

111
 10 Bd3!
This is the key – now the black bishop has been lured away from the a6-square, White can retreat his
bishop to d3, and the exchange of bishops has been avoided, albeit at the cost of a tempo.
10 0-0 is less logical, and does not offer any advantage, although it is not entirely without venom and is
worth a more extended look: 10 ... Bxb5 11 axb5.

Question: Doesn’t Black now have a backward a-pawn?

Answer: He does at this moment, but after 11 ... a5! he does not any more! Play might continue 12 Ng5 (White again adopts this standard
way of generating kingside play; it is his only hope, as Black has solved his positional issues on the queenside, and will stand very comfortably
if he is allowed simply to castle short and play ... Nd7, ... Rc8, etc) 12 ... h6 13 Nh3 Nd7 14 Nf4.

Now 14 ... 0-0 is good enough for equality: for example, 15 Nh5 Kh8! 16 Qg4 Rg8 17 Qh3 Nf8 and Black
was extremely solid in B.Ivkov-R.Byrne, Sousse Interzonal 1967, but 14 ... cxd4 is more critical, and possibly
objectively stronger: 15 cxd4 Qc4 16 Ba3. Here is the rub. White will generate play on the dark squares. A
112
famous game L.Stein-R.Byrne, Sousse Interzonal 1967 (played a few rounds before Ivkov-Byrne above – 14
... 0-0 was Byrne’s attempted improvement) ended in drastic defeat for Black, but improvements are possible:
16 ... Nf5 17 c3 g6 (17 ... h5!? 18 Nxh5 Qxc3 19 g4 Qh3! – Moles) 18 Qf3 Qxb5 (and here Moles suggests 18
... h5!?) 19 g4 Ne7 (19 ... Nh4!? is Moles’s final suggestion) 20 Rfc1 Rc8 21 Rab1 Qc4? (21 ... Qa6) 22 Ng2!
Rc6 23 Ne3 Qa4 24 Ra1 Qb5

25 c4! 1-0. Black actually resigned, chiefly in view of the line 25 ... dxc4 26 Nxc4 Rxc4 27 Qa8+ Nc8 28
Qxc8+ Rxc8 29 Rxc8#. However, although this game gave Black’s line, starting at move 14, a bad reputation,
it remains far from clear, as the various suggested improvements show.
Now back to the main line.
10 ... Nbc6 11 0-0

Question: So what is happening in this position?

113
 Answer: I think chances are about equal. Black has developed his pieces, and has the standard queenside structural advantage. He will
castle short, play ... Na5 and possibly ... Rc8, with counterplay on the queenside. The kingside thrust ... f7-f5 is another key idea. For his part,
White has his bishop-pair, chances of kingside play, and hopes to activate his queen’s bishop on the dark squares. It is a typical Winawer fight.
12 ... h6

Question: Is this necessary?

Answer: Indeed it is, at least if Black wants to castle short. The immediate 11 ... 0-0? allows the old Greek Gift sacrifice: 12 Bxh7+ Kxh7
13 Ng5+ Kg6 14 Qd3+ Nf5 (or 14 ... f5 15 Bf4 Qb7 16 Qg3) 15 g4 with a dangerous initiative.
12 Re1
The older line, which has similar ideas, starts 12 Ba3 Na5 13 Nd2 0-0 14 dxc5 bxc5 15 Nb3! (this forces
the exchange on b3, straightening out White’s pawns) 15 ... Nxb3 (15 ... Nb7? leaves the knight misplaced on
b7) 16 cxb3

16 ... f6! (this active move was an improvement on 16 ... Rfc8?! which landed Black in trouble after 17
Re1 Ng6 18 Bc1 Nxe5 19 Bf4 f6 20 Re3 Be8 21 Qe2 Rab8 22 Bc2 c4 23 b4 Rd8 24 Rxe5 fxe5 25 Bxe5 Qd7
26 Bxb8 Rxb8 27 Qe5 and Black never escaped his chains on the dark squares in A.Williams-R.Keene,
British Championship, Eastbourne 1973) 17 exf6 Rxf6 18 c4?! (18 b4 c4 19 Bc2 Be8 is unclear) 18 ... d4 19
b4 cxb4 20 Bxb4 Nc6 21 Ba3 e5 22 Rb1 Na5 23 Be4 Bc6 24 Bd5+ Kh8 25 f4 Rd8, C.Pritchett-G.Botterill,
British Championship, Brighton 1977. Chances are about equal, although White soon went wrong and lost.
The reader will note the many references in this section to the English Oxbridge generation of the late
1960s and early 1970s – Keene, Whiteley, Hartston, Botterill, Williams, Moles etc. These players contributed
greatly to the theory of this ... Qc7 set-up.
12 ... 0-0

114
 13 Ba3
As in the previous note, White looks to activate his bishop on its traditional a3-f8 diagonal. The alternative
is 13 Bf4, which is similar to Game 7. White wants to use his bishop on the kingside with Qd2 and a possible
sacrifice on h6. White’s kingside chances should not be underestimated, but it seems that Black defends
adequately after 13 ... Ng6 14 Bg3 cxd4.

15 Nxd4 Na5 (a standard idea in this line; the knight heads to c4, whilst the queen’s pressure down the c-
file is unmasked) 16 Qg4 (16 Re3 Nc4 17 Bxc4 Qxc4 18 h4 Bxa4 19 h5 Ne7 20 Bf4 was unclear, and
eventually drawn in J.Nunn-P.Nikolic, Amsterdam 1988) 16 ... Qxc3 17 Rad1 (17 Ne2 Qc7 18 Bxg6 fxg6 19
Qxg6 Bc8! is equal, as given by Atalik) 17 ... Nc6 18 Nb5 Qb4 gives White compensation for the pawn, but
no more than that.

Question: Why not straighten out the pawns with the other recapture at move 15?

115
 Answer: The problem is that 15 cxd4 Nb4! eliminates White’s key light-squared bishop.

Question: Isn’t the immediate bishop sacrifice on h6 possible at move 13?

Answer: A good question. It took some years for the English theorists to work out that it was unsound, but it appears that White has
insufficient compensation for the piece after 13 Bxh6 gxh6! 14 Qd2 Kg7 15 Qf4 Ng8!.

Following 16 Re3 Nce7 17 Nh4 Ng6 18 Rg3 Be8 19 Re1 c4 20 Bxg6 fxg6 21 Qg4 Ne7 22 Qxe6 Qd7
Black soon won in R.Sutton-J.Moles, Skopje Olympiad 1972.
Yet another impressive demonstration of the English players’ mastery of this variation was the game
C.Zuidema-A.Whiteley, Cala Galdana 1974, which saw the white player massacred: 13 Qd2 f5 14 exf6 Rxf6
15 dxc5 bxc5 16 Qe3 c4 17 Bf1 Raf8 18 Ba3 R8f7 19 Rad1?! Qa5 and White was already dropping a pawn,
with no counterplay on the kingside.

He tried to save himself tactically, but without success: 20 Qc5 Qxc3 21 Bxc4 dxc4 22 Re3 (22 Rxd7?
Rxf3) 22 ... Qxc2 23 Rxd7 and now the exchange sacrifice finished him off: 23 ... Rxf3! 24 gxf3 Qb1+ 25
116
Kg2 Rf5 0-1.
13 ... Na5
This is basically forced, to defend the c-pawn.
14 dxc5 bxc5 15 Nd2
Again, a standard manoeuvre. White frees the way for his queen to come to g4 and his rook via e3 to g3.
15 ... Bxa4
Critical, although French expert Igor Naumkin has played 15 ... Ng6 here, which also seems adequate.
16 Qg4!

This move is the most critical. J.Nunn-A.Yusupov, Linares 1988, was equal/unclear after 16 Bxc5 Qxc5 17
Rxa4 Qxc3 18 Re3 Kh8!. Black eventually won, and although this was not directly the result of the opening,
it is notable that Nunn later switched to 16 Qg4 here.

Question: This is all a bit obscure. What is going on?

Answer: Basically, White has sacrificed a pawn, for a kingside attack. He will follow up with Nf3 and Bc1, and try to blast Black off the
board on the kingside.

Question: So is he better?

Answer: I don’t believe so, but such a position is very concrete, and careful analysis is needed. On the basis of the game under
consideration, Black seems to be holding at least, but he needs to play accurately. White has actually scored very well in practice from this
position, but Black’s 16th move may change that.
16 ... Be8!
Atalik attributes this to the Russian GM, Sergey Ivanov (another Winawer expert, whose games are always
worth looking at). Previously, the usual move had been 16 ... Bd7, when 17 Nf3 Rab8 (17 ... Nc4 18 Bc1 f5
19 exf6 Rxf6 20 Bxc4 dxc4 21 Qxc4 was better for White in J.Nunn-P.Nikolic, Wijk aan Zee 1992, whilst the
point of Black’s 16th move in the game is shown if Black here tries: 17 ... Nac6? aping Atalik’s play in the
main game; this now fails, because after 18 Bxc5 Nxe5 19 Nxe5 Qxc5, the bishop on d7 hangs) 18 Bc1 Kh8
19 Qh3! Ng8 20 g4! Bb5 21 g5 Bxd3 22 cxd3 was A.Colovic-D.Komarov, Barletta 1999, where White’s
attack was too strong.
17 Nf3 Nac6

117
 18 Bc1?!
The natural follow-up, but Atalik, annotating the game in ChessBase Magazine, suggests it may be the
cause of White’s troubles. Instead, he analyses 18 c4 Nb4 19 cxd5 Nexd5 20 Be4 Bc6 21 Bc1 and now both
21 ... Rae8 and 21 ... Kh8 are enough for equality, but no more. Indeed, after the latter, 22 Qh5 Kg8 23 Qg4
Kh8 could produce an immediate draw.
18 ... f5 19 exf6
19 Qg3 Kh8 is adequate for Black.
19 ... Rxf6

20 c4
20 Rxe6? Bd7 21 Bxh6? Rxh6 (Atalik) is the tactical point of Black’s play.
20 ... Bg6
Atalik mentions 20 ... Bd7 as an alternative, whilst the computer likes 20 ... Rb8. In all cases, White has
enough kingside play to compensate for his pawn minus, but no more than that.
118
 21 Bb2!?
Atalik records that his opponent was unable fully to calculate the consequences of 21 Rxe6. The point is
that after 21 ... Qd7 22 cxd5 the reply 22 ... Bxd3 is forced, because 22 ... Nxd5? allows the winning queen
sacrifice 23 Rxf6!! Qxg4 24 Rxg6 Qc8 (if 24 ... Qh5 25 Rxc6) 25 Bxh6 with a decisive advantage. After 22 ...
Bxd3, 23 cxd3 Nxd5 24 Re4 Qxg4 25 Rxg4 is equal.
21 ... Bf5
Atalik says he assumes White had missed this, but in fact, Black has no advantage anyway.
22 Qh5?
Now Black assumes the advantage. As Atalik points out, White had to exchange queens with 22 Qg3 Qxg3
23 hxg3 when the position is equal.
22 ... d4
Now the white dark-squared bishop is shut out of the game.
23 Ba3

23 ... Ng6
Atalik awards this two exclamation marks, and claims that after 23 ... Qd6 24 Nd2, “I will not get much”,
but the computer disagrees: 24 ... Ne5 25 Bxf5 exf5 26 f4 Nd7 just seems to leave Black a pawn up, with a
clear advantage.
24 Bxf5 exf5 25 Bxc5 Nf4 26 Re8+
After 26 Qh4 Rg6, the white queen is done for after ... Rg4.
26 ... Rxe8 27 Qxe8+ Kh7 28 Re1?
Here, the computer agrees with Atalik, that White is still holding after 28 Qe1!. The text loses.
28 ... Qa5!

119
 This extremely strong move leaves the white bishop without a safe square anywhere on the board!
29 Bxd4
29 Bf8 Ng6! traps the bishop, whilst 29 Be7 Rg6 30 g3 Qc3 is “Goodnight Vienna”.
29 ... Nxd4 30 Nxd4 Re6! 0-1
This nice interference/back rank motif decides the game.

Rather a cataclysmic finish, but the analysis shows that White has enough for his pawn in this line, if no more than that. Overall, the ... Qc7
lines are a very solid and respectable way for Black to play.

120
Game 15
E.Inarkiev-N.Vitiugov
Russian Championship, Moscow 2008

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Nf3 b6

Question: Yet another different set-up!

Answer: Yes! One of the great advantages of the Winawer is that Black’s position is extremely flexible and allows numerous plausible
ways of handling the line.

Question: This time, I think I understand the idea! Black

intends ... Ba6, exchanging off his bad bishop?

Answer: Correct. We saw this plan in the previous game, but with Black’s queen already on c7.

Question: Isn’t this more logical? After all, if Black intends to play ... b6 and
... Ba6, why not do so at once, and keep flexibility with the queen?

Answer: That is a very fair point. However, as we pointed out in the notes to the previous game, the real significance of the ... Qc7 line is
the move order 6 ... Qc7, intending 7 Qg4 f5. With the move order in the present game, Black saves a tempo on ... Qc7, but he has to have
another way to meet 7 Qg4.
8 Bb5+
We also saw this idea in the previous game. White intends to avoid the exchange of bishops. Other moves
offer White nothing.
8 ... Bd7 9 Bd3 Ba4

121
 Question: Aha! This did not happen in the previous game.

Answer: Well, this is the advantage of the tempo saved on not playing ... Qc7. The white pawn has not had time to advance to a4, so this
blockading move becomes possible.
Black can also flick in the preliminary 9 ... c4, blockading the queenside once and for all, and then after 10
Bf1 play 10 ... Ba4. Indeed, this was originally the main line, but in recent times, Black has started to delay ...
c4.

Question: Why is that?

Answer: Mainly to preserve flexibility. Black reasons that,,,c5-c4 is a move he will always be able to play anyway, so it makes some sense
to hold it back, so as to complicate White’s task. The latter has to select a development plan, taking into account the possibility of both ... c5-c4,
and also some other line, such as ... cxd4 or just development by Black. At some point, if White has committed himself, Black can then play ...
c5-c4, hopefully in more favourable circumstances.
But while we are talking about the blocked positions after ... c5-c4, we may as well introduce an idea,
which we will see repeatedly in such positions. It is a manoeuvre first played, I believe, by Yusupov. After 9
... c4 10 Bf1 Ba4: 11 h4 (11 g3 is met in similar fashion: 11 ... h6 12 Bh3 Kd7 13 0-0 Qg8 14 Ne1 Kc7 15
Ng2 Nd7 16 f4 g6 17 Ne3 h5 18 Bg2 b5 19 h3 f5 20 exf6 Nxf6 21 Qe2 Kd7 22 Re1 Nf5 23 Nf1 Re8 was
agreed drawn in F.Jenni-A.Yusupov, German League 2002) 11 ... h6 12 h5 Kd7!.

122
 Question: Good grief! What an extraordinary move. What on earth is Black up to?

Answer: This plan has become almost a reflex action for Black in such positions. Of course, it looks strange to move the king so early, and
forego castling, but with the position so blocked, the king is as safe as houses on d7.

Question: Okay, that may be so, but I still don’t see

why Black does not just prepare ... 0-0-0

Answer: You will see why in a moment. Black has plans for his queen, namely 13 Bf4 Qg8!. Here it is! The queen is heading for the square
h7, from where it will join in with its bishop on a4, in attacking the c2-pawn!

Question: It looks mad!

Answer: Maybe, but in such a blocked position, it is not clear what else the queen can do. The pawn on b6 prevents her reaching a5. The

123
double attack on c2 is quite a nuisance for White, who has to devote a rook and maybe his queen to defending the pawn. In addition, from h7,
the black queen always has the possibility of jumping into the game, via f5 or e4. This plan with ... Ba4, ... Kd7 and ... Qg8-h7 has now been
played numerous times in this lines, and is known as a sound idea.

Question: But even if it is sound, I still don’t see

how Black is going to break through?

Answer: Ah, well that is another matter. The truth is that these blocked positions are quite drawish. That is another reason why Black tends
to delay ... c5-c4 these days, in the hope of preserving more dynamism in the game. I cannot pretend that this 7 ... b6 system is a great winning
weapon for Black, but it is excellent if he wants a solid position.
A.Shirov-A.Yusupov, Moscow Olympiad 1994, continued 14 g3 Qh7 15 Rc1 Nbc6 16 Bg2 a5 17 Nh2 b5
18 Qd2 Nf5 19 0-0 Nfe7 20 Ng4 Rag8 21 Bf3 Kc8 22 Kg2 Kb7 23 Rfe1 Rd8 24 Re2 Ka6 25 Rh1 Nc8 26
Ne3 Nb6 27 Ree1 Qg8 28 Ng4 ½-½. This was a typical game for the line – if White does not do anything
silly, it is hard for either side to make progress.
Returning to Vitiugov’s 9 ... Ba4:
10 h4 h6

11 Bf4
After the logical follow-up 11 h5, Black has a choice. He can close the position with 11 ... c4 (as he can do
at almost any move, of course), and then play reverts to very similar patterns to those in the last note: for
example, 12 Be2 Kd7 13 Bf4 (13 Nh4 Qg8 14 Bg4 Qh7 15 Ra2 Na6 16 Rh3 Nc7 17 Kf1 Raf8 was yet
another Yusupov example, also won by Black after a long battle, K.Georgiev-A.Yusupov, Las Palmas 1993)
13 ... Qg8 14 Nd2 Qh7 15 Ra2 Na6 16 g4 Nc7 led to a long manoeuvring battle and an eventual draw in the
game V.Bologan-N.Vitiugov, Russian Team Championship 2010. 11 ... Nbc6 12 Rh4 c4 13 Be2 Kd7 14 Be3
Qg8 15 Qd2 Qh7 16 Rc1 Kc7 17 Rf4 Raf8 was similar in A.Sokolov-A.Yusupov, 1st matchgame, Riga 1986,
one of the first games with the ... Kd7 and ... Qg8 plan. Black eventually won.
Alternatively, he can keep things fluid with 11 ... Qc7 12 0-0 Nd7 13 Re1 a6.

124
 Question: What is the idea of this move?

Answer: It is another idea worth noting. Black intends ... Bb5, forcing the exchange of bishops after all. It is frequently seen in these lines
where Black does not close the game with ... c5-c4. After 14 Nh4 Bb5 15 a4 Bxd3 16 cxd3 0-0 Black was doing fine, and the game was
eventually drawn, Z.Lanka-N.Sedlak, Lienz 2011.
11 ... Nbc6
Once again, the standard idea 11 ... c4!? 12 Be2 Kd7 is perfectly possible, whilst the game Z.Lanka-
I.Farago, Austrian League 2011, proceeded in similar fashion to Lanka-Sedlak in the previous note: 11 ... Qc7
12 Qd2 Nd7 13 h5 a6 14 g4 (this time White adopts a much more aggressive plan, ruling out short castling
for both sides; however, Black is still okay) 14 ... Bb5 15 Kf1 Bxd3+ 16 cxd3 Rc8 17 Kg2

17 ... c4 (an idea we have seen before after cxd3 – Black secures an outpost on c4 and pressure down the c-
file) 18 Rhc1 cxd3 19 Qxd3 Qc4 20 Qe3 b5 and another long battle was eventually drawn.

125
 Question: The words “long battle” seem synonymous with this variation!

Answer: That is true. The positions reached are real heavyweight stuff.
12 h5 a6!?
Now we know what the point is – Black wants to exchange bishops on b5.
13 Qb1
Stopping the bishop exchange. The other way of trying to do so, 13 Qe2?? c4, happens in a lot of online
blitz games!
13 ... Qc7 14 0-0 Na5 15 Ra2!

Question: What is the point?

Answer: White has several ideas – he overprotects c2, and prepares Rb2, attacking b6.
15 ... Rb8?!
This proves a waste of time. In ChessBase Magazine, Vitiugov recommended 15 ... 0-0, when chances are
balanced.
16 Re1

16 ... Kd7?!

Question: Cripes! This looks a bit optimistic here, with the centre still not closed.

Answer: I totally agree. It really looks a bit much to me. 16 ... 0-0 is still perfectly satisfactory.
17 Rb2
Now 18 Bxa6 is a threat, since Black cannot cut the bishop off with 18 ... b5.
17 ... c4
Because 17 ... Nc4?! fails to 18 Bxc4 dxc4 19 dxc5 bxc5 20 Rd1+ Kc8 (20 ... Nd5?? 21 Rxd5+ wins) 21
Nd2 with a strong attack, the text is forced.
18 Be2

126
 Question: OK, we have finally reached the standard
blocked structure. Isn’t Black fine here?

Answer: Well, the problem is that he has made a lot of moves which do not really fit. His queen is much less active on c7 than h7, for
example, and his knight on a5 is doing nothing. White certainly has a better version of the structure than he normally gets. Indeed, Vitiugov
went so far as to assess the position as clearly better for White. Having said that, the further course of the game goes to show how hard it can be
for White to make progress in such positions, even when he starts with a favourable version.
18 ... Nac6 19 Qc1 b5!?
This is a typical part of the plan of pushing the queenside pawns, but it also shuts the bishop on a4 out of
the game.
20 Nh2!
The knight is on its way to e3, which is frequently a good post for it in these lines.

Question: Why is that?

Answer: From e3, the knight frees up the kingside pawns to advance, with g4, f4 and f5. The knight also attacks d5, f5, etc, and defends c2.
20 ... a5

127
 21 Ra2!
I like this move.

Question: Why?

Answer: It is a neat piece of prophylaxis against the advance ... b5-b4. Black cannot play this, until he has ensured his bishop will not hang
after axb4.
21 ... Qd8 22 Bg4 Qg8!
The black queen heads for her traditional post on h7, albeit rather late.
23 Nf1 Kc7 24 Ne3 Kb7 25 Bh3 Nc8 26 Bg3 Nb6

27 f4?
f4-f5 is White’s main plan, but he mistimes it. Vitiugov recommended 27 Bh4!?.
27 ... f5! 28 Qd1?!

128
 The second mistake. White loses the thread. 28 Bh4 was Vitiugov’s recommendation.

Question: But surely White should be opening

the position with 28 exf6, shouldn’t he?

Answer: That is the move he would like to play, but here, after 28 ... gxf6 29 Bh4 f5, Black is fine. The e6-pawn is easily defended, White’s
f4-f5 advance has been killed forever, and his dark-squared bishop is now obstructed by his own pawn on f4.
28 ... Ne7 29 Rf1
Perhaps White should try 29 Kh2!? g5 30 hxg6 Nxg6 31 Qh5 (Vitiugov).
29 ... g5 30 hxg6 Nxg6

31 Bxf5!?
Inarkiev realises that he has lost much of his advantage, and tries to force matters. Instead, 31 Kh2 h5 is
equal.
31 ... exf5 32 Nxf5 Ka6 33 Qf3?!
The game now starts to be affected by time-trouble. This allows Black to advance the h-pawn and start
counterplay. 33 Qh5 maintains the balance.
33 ... h5! 34 Ne3 h4 35 Bh2 h3 36 g4 Nh4 37 Qg3 Rb7 38 f5 Ng2

129
 39 g5?
Losing at once. Correct was 39 Nxg2 hxg2 40 Rf2! Rbh7 41 Rxg2 Rh3 42 Qf4 Rxc3 and Black retains
some advantage, but the position is still very unclear.
39 ... Rh5 40 g6 Nxe3 41 Qxe3 Rf7!
Now the white kingside pawn chain collapses.
42 f6 Qxg6+ 43 Bg3 h2+ 44 Kh1 Rh3 45 Rf3 Bxc2 0-1

This heavyweight battle is absolutely typical of the 7 ... b6 line. It is a very respectable and solid way for Black to play, and White has a very
tough time breaking his opponent down, even if the latter commits a few inaccuracies. The main drawback of 7 ... b6, from Black’s standpoint,
is that it is quite hard to play for a win, unless White goes crazy. To win such blocked positions, the black player needs the patience of Job and
the positional understanding of Petrosian – rather like Artur Yusupov, in fact!

130
Game 16
I.Cheparinov-A.Grischuk
FIDE Grand Prix, Baku 2008

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Qg4

This move initiates White’s sharpest response to the main line Winawer. His queen comes out to attack the
key square g7, which has been weakened by the early development of Black’s dark-squared bishop.
7 ... cxd4
And this, in turn, starts Black’s sharpest and most uncompromising reply. He sacrifices his kingside
pawns, to destroy the white central pawn structure, and set up an extremely double-edged position. Quieter
alternatives, principally 7 ... 0-0, are examined in Game 18.
8 Qxg7 Rg8 9 Qxh7 Qc7 10 Ne2 Nbc6 11 f4 Bd7
11 ... dxc3 12 Qd3 d4 is a trendy line at present, and is examined in the next game.
12 Qd3 dxc3

131
 This is the principal tabiya of the whole variation.

Question: What on earth is going on? Black seems to be on

the verge of losing a pawn, with his kingside destroyed!

Answer: Both of those things are true, but Black has various forms of compensation.

Question: Such as?

Answer: Firstly, he is ahead in development, which is certainly not to be sniffed at, in such a sharp and open position. Secondly, his king
will be pretty safe after ... 0-0-0, whereas White’s king may struggle to find a safe haven. He cannot easily castle kingside, because moving his
king’s bishop is likely to leave the g2-pawn hanging, and even if he does manage it, the open h- and g-files mean that Black is likely to have
serious counterplay on that wing.

Question: But what about the white kingside pawns?

Isn’t the passed h-pawn just going to overwhelm Black?

Answer: It may look that way at first, but in practice, it is often very hard to make effective use of these pawns, at least in the middlegame.
White is likely to be too busy trying to hold back the black counterplay, and shield his own king.

Question: Okay, but isn’t Black going to lose his

c-pawn and thus be a pawn down as well?

Answer: Yes, but he will then have counterplay down the c-file as well.

Question: Hmm. I am still a bit sceptical!

Answer: That is understandable. The position is hard to judge at first sight, and it is only after seeing some of the concrete problems both
sides face, that one can start to appreciate the strength of the black position, in particular. Indeed, many GMs still do not believe in the black
position, and the strong American Grandmaster, Yasser Seirawan, has even gone so far as to claim that Black is objectively lost! That is an
extreme claim (Seirawan is notorious for his materialism), but it is fair to say that most GMs would prefer the white side of the position.
Computers have made people more sceptical of such long-term material sacrifices, and 7 Qg4 is undoubtedly the critical response to the
Winawer. It is because of many black players’ reluctance to play this sharp variation that move orders such as the Portisch-Hook line have
become popular in recent years.
13 Nxc3
This is a major parting of the ways, with White having many alternatives here. Of course, it is impossible
132
to give a detailed theoretical coverage of the whole line in a book such as this, but we will look at a few
sample lines, with a view to illustrating some of the typical ideas for both sides.
a) With 13 Rg1 White declines to take on c3, instead leaving the pawn alone and concentrating on trying to
advance his kingside pawns.

Question: But why not take the pawn?

Answer: Well, as we will see, taking on c3 often allows Black counterplay on the c-file, after a later ... Rc8. It also takes time, which further
delays White’s development. By leaving the c3-pawn alone, White instead attends to his own plans.
Black responds 13 ... d4.

Question: Huh? Isn’t that en prise?

Answer: Yes, but it is a typical pawn sacrifice in such positions. 14 g4 (after 14 Nxd4 Nxd4 15 Qxd4 Bc6 Black has excellent
compensation, in the shape of splendidly active pieces, and the white king stuck in the centre; moves such as ... Nf5, ... Rd8, etc. are in the air)
14 ... 0-0-0 15 h4 Nd5 (15 ... Qb6 is also a good move here, defending d4 and preparing ... Nd5) 16 Nxd4 Nxd4 17 Qxd4 Bc6 18 Be2 (18 Qxa7
f6 starts central counterplay) 18 ... f6! (now we see the drawbacks of White’s plan with g4 and h4, his king is denuded and very exposed, if the
centre opens up; Black is willing to invest several pawns for that purpose) 19 h5 (19 exf6 e5! is Black’s idea; after 20 Qxe5 Qb6 followed by ...
Rge8, he has fierce counterplay) 19 ... Qh7 20 Ra2 (20 exf6!) 20 ... Qe7 21 exf6 Nxf6 22 Qxc3 Ne4 23 Qh3.

133
 Thus far we are following A.Rasmussen-N.Zhukova, Athens 2008. Now 23 ... e5 would have resulted in a
typically chaotic and unclear position for this variation. White has two extra pawns, but his king is a
permanent problem and Black has a strong initiative.
b) 13 h4 is another variation on the plan seen in the last game. White again ignores the c3-bait and starts
pushing his passed h-pawn, hoping to tie down the black forces to stopping the pawn, and thus hamper
Black’s counterplay in that way. Play goes 13 ... Nf5 (another typical theme in this variation; advancing the
h-pawn weakens g3, to which Black immediately directs his knight) 14 Rb1 d4 15 h5 0-0-0 16 Rg1.

Question: What a mess!

Answer: After a while spent studying this variation, such positions start to look perfectly normal! Both sides have played logically enough –
Black has finished his development, whilst White has advanced his h-pawn. Now he wants to join up with g2-g4, also attacking the active black
knight.
Black can now go 16 ... Kb8.
134
 Question: That looks incredibly slow, if not to say,
complacent in such a sharp position.

Answer: In fact, it is quite a standard idea. Black gets his king into a more secure spot, and also envisages a possible ... Bc8, ... b6, and ...
Bb7 manoeuvre. However, he has alternatives:
b1) 16 ... Nce7?!.

Question: What is the point?

Answer: It is another thematic idea in such lines. The knight frees c6 for the bishop, whilst itself preparing to come to d5, when the black
knights would be ready to jump into e3. White cannot really allow this, so his next is virtually forced. However, Black’s 16th has the drawback
of depriving his knight of the e7-square, and is probably less good in this position.
So 17 g4 Nh6.

Question: Isn’t the knight badly placed here?

Answer: It is quite precarious, yes. It attacks g4, and if the g-pawn advances, then the knight returns to f5. The problem comes if White
manages to secure the g4-pawn and then play f4-f5. O.Korneev-A.Kveinys, Nova Gorica 2004, continued 18 Bh3 Bc6 19 Rg3.

Question: What’s wrong with 19 f5 here?

Answer: Black has prepared the sacrifice 19 ... exf5! 20 Bxh6 Be4 21 Qg3 (21 Qd1 d3 is crushing) 21 ... Bxc2 22 Rc1 d3 with excellent
compensation for the piece.
After 19 ... Bd5 20 Nxd4 Bc4 21 Qxc3 ...

135
 ... Kveinys found 21 ... Rxd4, a typical tactic: 22 Qxd4 Nef5 23 Qc3 Nxg3 24 Qxg3 Be2 (ingenious, but it
should not be sufficient) 25 g5 Bxh5 26 Rb3 Kb8 27 Bg2?? (27 Qc3 would be better for White) 27 ... Qxc2
(suddenly, the game is over) 28 Rxb7+ Kc8 29 Kf1 Be2+ 30 Kg1 Qd1+ 31 Kh2 Ng4+ 32 Qxg4 Bxg4 33 Be3
Qe1 0-1. In my experience, this is a typical scenario for this variation – examining the game afterwards, with
the aid of a computer, suggests that White is objectively better, but one slip proves fatal. Over the board,
White’s position is extremely hard to handle.
b2) 16 ... f6 would be thematic, when the position is unclear after 17 g4 Nh6 18 exf6 Rxg4 19 Rxg4 Nxg4
20 Nxd4 Nxd4 21 Qxd4 Be8, but let’s go back to 16 ... Kb8:

V.Vehi Bach-F.Vallejo Pons, Palma de Mallorca 2009, proceeded 17 g4 Nh6 18 g5?! (see the previous
note; 18 Bh3 looks stronger, when Black must constantly worry about the position of the knight on h6) 18 ...
Nf5 19 Bg2 Na5 (another thematic idea – the knight looks at c4-e3; however, 19 ... b6 followed by ... Bc8, is
safer here, reaching the position seen after move 21) 20 Rb4 (20 Be4 is stronger) 20 ... Nc6 21 Rb1 b6 22 Be4
Nce7 23 Kf2 Qc5 24 Ng3 Ba4! (yet another standard idea in this line; now there is a strong threat of 25 ...
136
Bxc2, and White finds himself struggling) 25 Kg2? (25 Bxf5 was best, but after 25 ... exf5 White is in trouble
anyway, as his light squares are a disaster) 25 ... Ne3+ (decisive) 26 Bxe3 dxe3 27 Qa6 Bc6 28 Kh3 Rd2 29
Rbe1 Nd5 30 Kg4 Bb5 0-1. Another excellent example of how easily things can go wrong for White.
c) 13 Ng3.

Question: This looks like a sensible move. By blocking the g-file, White
wants to follow up with Be2 and 0-0. How does Black meet this?

Answer: The key idea is to remove the knight from g3: 13 ... 0-0-0 14 Qxc3 Nf5! and this is how. The exchange on f5 damages the black
pawn structure, but removes the key knight from g3. A sample game, involving French aficionado John Watson on the black side, continued 15
Nxf5 exf5 16 Bd2 Be6 17 Bd3? (17 Kf2 is objectively better, although human players are naturally averse to such king moves) 17 ... Rxg2 18
0-0-0 d4 19 Qb2 Na5 when Black was clearly better, and went on to win.
d) 13 Be3 d4!? (once again, this thematic pawn sacrifice is interesting, although there is nothing wrong with the natural 13 ... Nf5) 14 Bxd4 (the
top-level stem game with 13 ... d4 continued 14 Bf2 0-0-0 15 Nxd4 Nxd4 16 Qxd4 b6 17 Bh4 Bb5 18 Qe4 Bxf1 19 Rxf1 Rd5 20 Bxe7 Qxe7
and Black had compensation for the pawn, and even went on to win the heavy piece ending B.Spassky-V.Korchnoi, 2nd matchgame, Belgrade
1977) 14 ... Nxd4 15 Nxd4 0-0-0.

137
 Once again, Black has excellent compensation for the pawn – the open lines and exposed white king,
together with Black’s significant lead in development, give him an excellent game. After 16 Qc4? (returning
the pawn to bail out into an ending, but now White is just worse) 16 ... Qxc4 17 Bxc4 Rxg2 18 0-0-0 Ba4
Black won in D.Papakonstantinou-E.Rozentalis, Vrachati 2011.
e) 13 Rb1 is probably the most respected of the non-captures on c3. The rook takes up a useful post, aiming
down the b-file, preventing ... Qb6, and also setting a nice trap. 13 ... 0-0-0 (13 ... Nf5?! is well met by 14 h3!
followed by g4; by contrast, simple development with 14 g3 gave Black good, and thematic, counterplay in
the old game L.Shamkovich-W.Uhlmann, Sarajevo 1963, which continued 14 ... d4 15 Bg2 0-0-0 16 Kf2 Na5
17 Rd1 Bc6 18 Bxc6 Qxc6 19 Qf3 Qc5 20 Rd3 Rd7 21 g4 Nh4 22 Qg3 Rh8 23 h3 Nc6 24 Rb3 Qc4 25 Ng1
a5! with a nice advantage to Black) 14 Nxc3 Na5 (the trap set by White’s 13th move is 14 ... a6?? 15 Qxa6!).

Now White has several options:

e1) 15 Nb5 leads to little after 15 ... Bxb5 16 Rxb5 (16 Qxb5?! Qxc2 is a dubious pawn sacrifice) when 16
... Kb8 transposes to variation ‘e2’ below.
138
e2) 15 g3 Kb8 16 Nb5 Bxb5 17 Rxb5 Rc8 (or 17 ... Nc4 18 Be2 Rc8 19 0-0 Nf5 which was equal, and
quickly turned into advantage for Black after 20 Rf3 Rg6 21 Bb2?! Qc6 22 Rb4 Qc5+ 23 Kh1 Rh8 24 Bf1
Nxg3+ 25 Rxg3 Qf2 26 Rg2 Rxg2 27 Bxg2 Ne3 28 Rxb7+ Kc8 29 Rb8+ Kc7 30 Rb7+ Kd8 and 0-1 in
M.Sergeeva-S.Ganguly, Abu Dhabi 2007) 18 Be3 b6 with equality, L.Dominguez Perez-Y.Shulman, Buenos
Aires 2005.
e3) 15 Rg1 Kb8 16 g4 d4 17 Ne4 Ba4 18 Rb2 a6 19 Nd6 Nf5! 20 Nxf5 exf5 21 Bd2 Bb5 22 Qxf5 Nc4 was
unclear in L.Dominguez Perez-Zhang Pengxiang, Linares 2002, eventually won by Black.
e4) 15 h3! is the critical move: 15 ... Kb8 16 g4 Rc8 17 Nb5 Bxb5 18 Rxb5, as in L.Dominguez Perez-
D.Stellwagen, Wijk aan Zee 2009.

Now Postny has suggested 18 ... Nc4!? 19 Be2 Qc6 20 Rb1 Qc5 followed by ... Nc6-d4 with
compensation.
f) 13 Qxc3!? is the other way to take the pawn. On c3, the queen pins the enemy knight on c6, thus
restricting possible ... Na5 ideas. White also retains control of the crucial square d4, unlike in the line
beginning 13 Nxc3. Black has:
f1) 13 ... Nf5 14 Rb1.

139
 Question: Why this?

Answer: We have seen the idea many times already. The rook takes up an active post and restricts the activity of the black queen. The text
has been regarded as best here, ever since the old game R.Bogdanovic-W.Uhlmann, Sarajevo 1963, which continued 14 Bd2 Qb6! 15 a4 Rc8
16 a5 Qd8 (the queen has been driven back, but now threatens to emerge on the other flank, via h4; Black had good play, and went on to win in
powerful style) 17 Qd3 a6 18 Ra3 Na7 19 Rc3 Bb5 20 Rxc8 Qxc8 21 Qc3 Bc4 22 g3 Nb5 23 Qb2 Qc6 24 Rg1 Rh8 25 Rg2 d4! 26 Rf2 Qh1 27
g4? (27 Qb4) 27 ... Rxh2 28 gxf5 Rxf2 29 Kxf2 Qh2+ 30 Ke1 Qh4+ 31 Kd1 Qf2 32 Be1 Qxf1 33 Ng3 Qf3+ 34 Kc1 Qe3+ 35 Kd1 Nc3+ 36
Bxc3 dxc3 37 Qb6 Qd2# (0-1).
After 14 Rb1 Black can go 14 ... Rc8.

Question: Why not castle long?

Answer: As we have just seen, that is a viable option too, but Black reasons that, with the c-file open, he does better to bring his rook there
straightaway, with the idea of counterplay against c2.
S.Karjakin-P.Harikrishna, Bilbao 2007, continued 15 Rg1 d4 16 Qd3 Nce7 17 g4.

140
 Now Black should have tried 17 ... Ba4 18 Rb2 Bxc2 19 Qb5+ Kf8 with an unclear position.
f2) 13 ... 0-0-0 is another option: for example, 14 Rb1 Nf5 15 Rg1 d4 16 Qd3 Na5 (the sharp counterattack
16 ... f6 17 g4 Nh4 18 exf6 e5 must always be considered in such positions, but it came up short in
I.Kurnosov-N.Vitiugov, Ramenskoe 2006, after 19 Kf2 Be6 20 Bh3 Rh8 21 c3 Qf7 22 cxd4 exd4 23 g5;
however, 16 ... Be8!? is also worth considering, followed by ... f6) 17 g4 Ba4 18 c3

18 ... Bc2! (a nice tactical idea, and yet another which crops up in various lines) 19 Qxc2 d3 20 Qa2 Qc5!
21 Rg2 Ne3 22 Ng3 Rxg4 23 Rf2 Nac4 with an unclear position, where chances are about equal, A.Volokitin-
S.Ganguly, Moscow 2007.
After this extremely long digression, we return to the main game and 13 Nxc3:
13 ... a6

141
 14 Rb1
A major alternative here is Spassky’s 14 Ne2.

Question: What is the point? It looks an odd move.

Answer: White reasons that the knight will be threatened along the c-file after ... Rc8, so he moves it away in advance. In addition, he takes
control of the d4-square. This move caused a few problems for Black when first introduced in the 1980s, but it does look rather odd to spend a
tempo, returning the knight to a clumsy square, where it shuts in the king’s bishop. Black should be able to find adequate counterplay against
such extravagance:
a) 14 ... 0-0-0 15 g3 d4 16 Qc4 Nf5 17 Bg2 Na5 18 Qxc7+ Kxc7 19 Rb1 Bc6 20 Bxc6 Kxc6 was equal in
V.Anand-V.Ivanchuk, Nice (rapid) 2009.
b) A good alternative is 14 ... Nf5 15 h3 which was tried in A.Volokitin-Hou Yifan, Wijk aan Zee 2009, and
now 15 ... Na5! 16 g4 Bb5 17 Qc3 Qxc3+ 18 Nxc3 Bxf1 19 Rxf1 Nd4 gave Black equal chances in the old
game J.Timman-N.Short, Rotterdam 1988.
14 ... Na5
14 ... Rc8 is also a very logical continuation. After 15 h4 Nf5 (immediately targeting g3, which White’s
next defends) 16 Rh3 Nce7 (16 ... Ncd4?! 17 h5 and now Black lost at once after the blunder 17 ... Qc5? 18
Rxb7 Nb5?? 19 Ne4 1-0 in F.Nijboer-J.Timman, Hilversum 2006; Black clearly needs an improvement at
move 17, if this set-up with 16 ... Ncd4 is to remain playable, but I have to admit, it is not immediately
obvious what this improvement is) 17 Bd2 (one point of Black’s last is revealed after 17 h5 Nh6! followed by
... Nef5, when the knights blockade the kingside pawns effectively, D.Velimirovic-J.Levitt, Pinerolo 1987) 17
... Nh6 the position is typically unclear.
15 h4 Nf5
15 ... Rc8 was tried in J.Zawadzka-E.Berg, Warsaw 2008. After 16 Rh3 Nf5 17 h5 Black played the
interesting idea 17 ... Kf8!?.

142
 Following 18 h6?! (maybe this should be held back, so as to prevent the following manoeuvre) 18 ... Rg6
19 a4 Kg8 (a most interesting set-up, as the black king manages to play a role in holding up the white
kingside pawns, whilst the pawn on h6 itself prevents the white pieces attacking the enemy monarch;
objectively, it is rather hard to believe, but in the game. Black soon got on top) 20 h7+ Kh8 21 Kd1 Nc4 22
Qe2? Ng3 23 Qf2 Nxf1 24 Qxf1 d4 and Black was winning.
16 Rh3

16 ... 0-0-0
A similar example was 16 ... Qc5 17 h5 0-0-0 18 Bd2 Nc4 19 h6 Rg6 20 h7 Rh8 and now the blunder 21
Qf3? (21 Ne4 dxe4 22 Qxc4 Qxc4 23 Bxc4 Rxg2 is equal) walked into the sucker punch 21 ... Nxd2 22 Kxd2
Rxh7!. White then decided to go down in a blaze of glory, but resigned after 23 Bxa6 (23 Rxh7 Rg3 wins) 23
... bxa6 24 Nxd5 Rxh3 25 Nb6+ Kd8 26 gxh3 Qd4+ 27 Ke2 Rg3 in D.Petrosian-S.Lputian, Yerevan 2006.
17 h5 Nc4 18 Rb4
18 h6 Rg6 19 h7 Rh8 was played in a blitz game also involving Grischuk.
143
 Question: I still don’t get what is happening here! It looks
as though White is about to promote his h-pawn.

Answer: He has marched almost all the way, but the final step will be the hardest. White hopes that his advanced h-pawn will tie Black
down and disrupt his coordination, thereby depriving him of counterplay, but objectively this is not the case. Black can cope with the pawn and
still keep adequate activity. Objectively, chances are about equal. 20 Ne2 (20 Qf3 is a critical move, trying to get in the advance g2-g4, but
Black is fine after 20 ... Rg7 21 g4 Ne7) 20 ... Bb5 21 Nd4

21 ... Nxe5! 22 Qd1?? (22 Qc3 Nxd4 23 Qxd4 Bxf1 24 Kxf1 Qc4+ 25 Qxc4+ Nxc4 is only slightly better
for Black) 22 ... Bxf1 23 Kxf1 Qc4+ 24 Ne2 Qe4 and Black was winning in S.Karjakin-A.Grischuk, Moscow
2008.
Meanwhile, 18 Qf3 again hopes to achieve g2-g4, but 18 ... Bc6 objectively leaves White with nothing
better than repeating moves. Instead, White tried to do so with 19 Qe2? against a youthful Caruana, but after
19 ... Ng3 20 Qd3 Nxf1 21 Kxf1 (21 Qxf1 Nxe5 22 fxe5 d4 is even worse) 21 ... d4 22 Ne4 Rg4 Black was

144
already better, and after 23 Ng3 in B.Bruned-F.Caruana, Collado Villalba 2005, he could have won at once
with 23 ... Ne3+ 24 Bxe3 dxe3 25 Qxe3 Bxg2+!.
18 ... Bc6 19 Ne2 Bb5 20 a4

20 ... Qc5

Question: Just a moment – isn’t 20 ... Nxe5 on here?

Answer: It is possible, and was played in the stem game from this position, V.Hort-W.Uhlmann, Hastings 1970/71. However, after 21 Qc3
Bxe2 22 fxe5 Bxf1 23 Qxc7+ Kxc7 24 Kxf1, Moles’ assessment is that “White has anything that’s going”, and Hort did in fact go on to win.
Moles himself suggests 20 ... Bc6, with the threat of 21 ... a5, followed by 22 ... Bxa4, but Fritz is
unimpressed: 21 h6 a5 22 Rb1 Bxa4 23 h7 Rh8 24 g4 with advantage to White. Grischuk’s choice looks best,
as he takes control of the key square d4, with tempo.
21 Qc3?
After this mistake, Black seizes the advantage. 21 Ba3 had been tried in J.Becerra Rivero-Y.Shulman,
Tulsa 2008, which continued 21 ... Bc6 22 Qc3 Rg4 (it is a recurring pattern in these lines that Black makes
good use of the squares g3 and g4, weakened by the early advance of the white h-pawn) 23 h6 Rh8 24 h7?
(this just exposes the h-pawn) 24 ... Rg7 25 Rb3 Qa7 26 Bc1 Bxa4? (26 ... Rgxh7 is just better for Black) 27
Qb4?? (27 Ng3 is unclear) 27 ... b5 28 Rh5 a5 29 Qc3 Rgxh7 and Black won.
Finally, 21 Bd2 is the computer’s preference, with equal chances after 21 ... Bc6.
21 ... d4 22 Qb3 Na5 23 Rxb5
Desperation, but White has nothing better. His position collapses after both 23 Qb1 Bxe2 24 Kxe2 (if 24
Bxe2 Rxg2) 24 ... Rg3, and 23 Qa3 Bxe2 24 Kxe2 Nc4 25 Qb3 d3+.
23 ... Nxb3 24 Rxc5+ Nxc5

145
 Black has won the exchange for a pawn. White has two bishops and his kingside pawns, but the latter are
well controlled and Black’s pieces are very active. He has a clear advantage, which Grischuk has no
particular trouble converting.
25 Ng1 Rg3 26 Ba3 d3
26 ... Nxa4 is also good.
27 Rxg3
27 cxd3 Nxd3+ 28 Bxd3 Rdxd3 is slightly more tenacious, but should also lose.
27 ... Nxg3
The computer points out the typically incisive 27 ... d2+ 28 Kd1 Ne4! 29 Rf3 Nf2+! 30 Rxf2 Ne3+ which
is objectively even better, but I don’t think we can really criticise Grischuk for missing this!
28 cxd3 Nxa4 29 Ne2 Nxh5 30 g4 Rxd3 31 gxh5 Rxa3 32 h6 Ra1+ 33 Kf2 Rd1 34 h7 Rd8
Just in time! The remainder requires no comment.
35 Nd4 Nc5 36 Nf3 Kd7 37 f5 exf5 38 Bh3 Ke7 39 Bxf5 a5 40 Bb1 a4 41 Nd4 Rh8 42 Nf5+ Kf8 0-1

146
Game 17
S.Kuipers-D.Stellwagen
Dutch League 2011

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Qg4 Qc7 8 Qxg7

8 Bd3 is a major alternative here.

Question: Is that entirely logical? I thought the

whole point of 7 Qg4 was to take on g7?

Answer: It is, but the point is that White does not necessarily have to do so at once. The g7-pawn is not easy to defend, so White reasons
that he can afford to bring the bishop out first. The idea is to avoid having to shut this bishop in with Ne2, as happens in the main line.
8 Bd3 is a very sharp and dangerous line, for which Black needs to be well prepared. He has two main
approaches:
a) 8 ... cxd4 9 Ne2 dxc3 10 Qxg7 Rg8 11 Qxh7 (11 Qh6 intending to take on h7 with the bishop, is another
sharp line; theory considers that Black is okay after 11 ... Nbc6 12 Bf4 Bd7 13 Bxh7 Rxg2 14 Bg3 0-0-0
when he will sacrifice the exchange on g3 to win the e5-pawn, with typical Winawer compensation) 11 ...
Qxe5 12 Bf4 Qf6.

147
 Question: What is going on here?

Answer: The position is very sharp. White has temporarily sacrificed a pawn, but he has activated his dark-squared bishop much more than
he usually manages in these lines, and he has the typical kingside pawn phalanx. Objectively, Black is probably fine, but the position is
extremely sharp and preparation can run very deep in such lines. Watson gives extensive coverage in Play the French 4, and any black player
intending to go down this line should study his analysis carefully.
b) 8 ... c4 is the quieter solution, avoiding forcing lines, in favour of a typical closed Winawer struggle.
Strangely, Watson makes no mention of this in his latest French volume, although he covered it in the first
edition. Objectively, Black may be slightly worse in these lines, but it is a typical structure, where the black
player should feel at home, and he does avoid a lot of sharp preparation. 9 Be2 Nf5 10 Nf3 Nc6 and now 11
Nh4 Nce7 12 Bg5 is a suggestion of Psakhis, but I don’t see that Black has any real problems after 12 ... 0-0
or 12 ... Nxh4 13 Qxh4 h6 14 Bd2 Nf5 15 Qh3 Bd7.
In this line, Black pays his money and takes his choice. In purely objective terms, 8 ... cxd4 is probably the
best move, but the positions are very sharp and require detailed preparation. Given that 8 Bd3 is only a fairly
obscure sideline, many Black players may be reluctant to spend much time on such preparation, in which case
8 ... c4 is the pragmatic choice.
8 ... Rg8 9 Qxh7 cxd4 10 Ne2
Believe it or not, White does have an alternative here, in 10 Kd1.

148
 Question: That looks insane!

Answer: Maybe, but there is some logic to it. White sidesteps the threat of a lethal check on c3, without committing his knight to e2. He
wants to play the knight to the more active square f3, from where it will have the idea of jumping to g5. The main line goes 10 ... Nbc6 11 Nf3
dxc3 12 Ng5 Nxe5 13 f4 (13 Bf4 Qb6! is a neat tactical solution to Black’s problems) and now 13 ... f6! is fully adequate: for example, 14 fxe5
fxg5 15 Qh5+ Kd8 16 Bxg5 Qc5! and Black will untangle with ... Bd7 and ... Kc7. He has material equality, and the white king is likely to end
up the more exposed of the two monarchs.
10 ... Nbc6 11 f4 dxc3 12 Qd3

12 ... d4!?
This is the move that gives Black’s move order its independent significance.

Question: So what is the point? Doesn’t 12 ... d4 just give away a pawn?

Answer: Yes! We saw in the previous game that this type of pawn sacrifice is a thematic idea in the variation. Black opens lines, especially

149
the d-file and the long diagonal h1-a8, for his pieces. With the move order in the game, Black does this at once, delaying the developing move
... Bd7. The bishop will usually come to d7 later anyway, although occasionally, it develops by means of ... b6 and ... Bb7. This system with 12
... d4 has become quite popular in recent years, and is covered in detail by John Watson in the latest edition of his seminal work Play the
French. Watson uses 12 ... d4 as his main line recommendation. Instead, 12 ... Bd7 would transpose to the previous game.
13 Nxd4
Taking the pawn is the most natural response, but others are also possible:
a) 13 Rb1 is a thematic move in such positions, as we know from Game 16, but in this line, it allows Black
to exploit one of the advantages of not having played ... Bd7. 13 ... b6! logically neutralizes the pressure
along the b-file, as well as preparing ... Bb7 and ... 0-0-0. With the long diagonal already opened by ... d5-d4,
the bishop clearly belongs on that line. 14 Rg1 Bb7 15 g4 0-0-0 16 Qc4 Rd5 (Watson gives 16 ... Na5 17
Qxc7+ Kxc7 as better for Black, which it is, but there is nothing wrong with Grover’s move either) 17 Rg3
Rgd8 18 h4?! Ra5!.

Question: That looks a bit funny. Why is it so strong?

Answer: It is rather counter-intuitive to put the rook on a5, but the point is that the threat of ... Ba6 is extremely hard to meet. 19 Qd3 Ba6
20 Qh7 was E.Barbosa-S.Grover, Chennai 2011, and now simply 20 ... Nd5 gives Black a clear advantage.
b) 13 h4 is another typical white plan in such positions, as we know. In response 13 ... b6 is again good. As
Watson emphasises, this is usually the move Black wants to play, unless White does something specific to
prevent it. Now:
b1) 14 Nxd4 Nxd4 15 Qxd4.

150
 Question: So White is a pawn up?

Answer: Yes, but just as in the main game, Black has excellent compensation in the form of the open lines and exposed white king: 15 ...
Nf5 16 Bb5+ Bd7 17 Qxd7+ Qxd7 18 Bxd7+ Kxd7. Despite the pawn minus, Black has positional compensation (weakness on g2, active
knight, etc) and chances are balanced. The high-level game V.Bologan-G.Kamsky, Reggio Emilia 2009/10 confirmed this: 19 Kf2 Rg4 20 Be3
Rag8 21 Rag1 Nxh4 22 g3 Nf5 23 Rh3 Ke7 24 Kf3 Nxe3 25 Kxe3 Rd8 26 Rh2 Rgg8 27 g4 Rd2 28 Rxd2 cxd2 29 Kxd2 Rh8 30 f5 exf5 31
gxf5 Rh5 32 Rf1 Rh4 33 Ke3 Ra4 34 Rd1 Rxa3+ 35 Ke4 Ra4+ 36 Rd4 Rxd4+ 37 Kxd4 f6 38 e6 Kd6 39 Kc4 a6 40 Kb4 Kc6 ½-½.
b2) 14 h5 Bb7 15 h6 0-0-0 16 h7 Rh8 is perhaps a more logical follow-up to White’s 13th.

Watson regards it as critical and devotes a lot of analysis to it, assessing the position as “dynamically
equal”. After 17 Rh3 he suggests Fritz’s choice 17 ... Nd5, when 18 Nxd4 Nde7! is an amusing follow-up.
White has problems on the d-file, and 19 Nb5 Rxd3 20 Bxd3 (20 Nxc7?? Rxh3 wins) 20 ... Qb8 is roughly
equal.
c) 13 Ng3 is a critical attempt to exploit Black’s move order.

151
 Question: What is the point?

Answer: White takes advantage of the fact that Black’s last move gave up control of e4. The knight threatens Ne4-d6+. After 13 ... Bd7 (13
... b6 is now too slow, since after 14 Ne4 checks on d6 and f6 are already threatened) White has:
c1) 14 Ne4 0-0-0 15 Nd6+.
15 Rb1 has also been tried.

Question: With the idea that the threat of Nd6+ is stronger than its execution?

Answer: Indeed, but Black can strike with the thematic piece sacrifice 15 ... Nxe5 (15 ... Nf5 is also playable) 16 fxe5 Qxe5.

Question: What is going on here?

152
 Answer: Black has two pawns for the piece, the white king is very exposed, and the black central pawns threaten in the near future to sweep
down the board. Black is just better here, I believe, and especially in an OTB game. White’s position is hard to defend. After 17 Qe2 Bc6 18
Ng3 (18 Nf2 is mentioned by Watson, but simply 18 ... Qd5 keeps a strong initiative) 18 ... Qd5 (18 ... Qxe2+ is also fine for Black, but the text
is more fun!) 19 Qf2 (19 Rb4!? Nf5 20 Kf2 is the computer’s unlikely suggestion; then 20 ... Nd6 21 Bf4 f6 22 Bxd6 Qxd6 still looks better for
Black) 19 ... f5 20 Rb4 a5 21 Rc4 e5 (21 ... b5 was also good) it was “wagons roll!” in the game T.Calistri-D.Bunzmann, French League 2007.

Black went on to win.

Returning to the immediate check, and then 15 ... Kb8 16 Rb1.

Question: Why not simply 16 Nxf7 with an extra pawn?

Answer: This is similar to the main line, after 16 ... Rdf8 17 Nd6 Nf5.
After 16 Rb1 Black has:
c11) 16 ... Bc8 was played in D.Neelotpal-S.Sengupta, Mumbai 2003, but Fritz is unconvinced about the
black position after 17 Nxf7 (17 g3 Rxd6!? 18 exd6 Qxd6 19 Bg2 b6 20 0-0 Nf5 21 a4 Bb7 gave Black
compensation for the exchange in V.Artemiev-D.Petrosian, Izhevsk 2011) 17 ... Rdf8 18 Nd6 Ng6 (18 ... Nf5
19 Nxf5 Rxf5 is like the main line, but worse; after 20 g3 b6 21 Bg2 Bb7 22 0-0 it is questionable whether
Black has enough for the pawn) 19 Qe4 Nh4 20 g3 Nf5 21 Bg2 Ka8 22 Nxf5 exf5 23 Qd3 Be6 and although
Black eventually drew, the silicon monster wants nothing to do with his position after 24 0-0.
c12) 16 ... b6 17 Nxf7 Rdf8 18 Nd6 Nf5 19 Nxf5 Rxf5.

153
 Question: So why is this any better than the analogous line with 16 ... Bc8?

Answer: The difference is tactical; after 20 g3 Black has the strong reply 20 ... Nxe5! 21 fxe5 Bc6!. This is the point – with the bishop on c8
and pawn on b7, the bishop could not come to the long diagonal at once. After 22 Rg1 Be4! 23 Qc4 (23 Qxd4 Bxc2 24 Rb4 Rd8 is decisive) 23
... Qxc4 24 Bxc4 Bxc2 and Black’s extra pawn (soon to be two) plus initiative adds up to the advantage.
c2) Another logical move is simply 14 Be2, taking advantage of the fact that 13 Ng3 defended the g2-
pawn. White simply wants to castle short, solving the traditional problem of his king’s exposure in the centre.
This has been played by Karjakin, and so must be taken seriously: 14 ... Qb6 (a useful move order point;
Black stops the move Rb1) 15 0-0 0-0-0 16 Ne4 Nd5!? (16 ... Kb8 17 Nd6 Be8 is Watson’s recommendation,
but I prefer Fritz’s 16 ... Nf5, with an unclear position) 17 Nd6+ (17 Qb5 Qxb5 18 Bxb5 Kc7 was equal in
M.Khachiyan-Y.Shulman, Saint Louis 2009) 17 ... Kb8 18 Nxf7 was Karjakin’s choice against Kamsky.
After 18 ... Rdf8 19 Nd6 Nce7 20 Bf3 Bc6 ...

... and now, instead of Karjakin’s 21 a4?! Nb4 22 a5 Qc5 23 Qh7 (S.Karjakin-G.Kamsky, Nalchik 2009),
154
when 23 ... Nf5 is slightly better for Black, 21 g3! is critical. Watson then analyses 21 ... Nf5 (both captures
on f4 favour White: 21 ... Nxf4 22 Bxf4 Rxf4 23 Bxc6 Rxf1+ 24 Rxf1 Nxc6 25 Qh7, and 21 ... Rxf4 22 Bxf4
Nxf4 23 Qd1 d3+ 24 Kh1 dxc2 25 Qxc2) 22 Nxf5 Rxf5 as equal; Black threatens ... Bb5, and after 23 Bxd5
Bxd5 his monster bishop guarantees at least equality.
After that important digression, we can return to the annexing of the d-pawn:
13 ... Nxd4 14 Qxd4 Bd7
The immediate 14 ... Nf5?! allows the annoying 15 Bb5+.

This is the main tabiya for the 12 ... d4 variation.

Question: So what is going on?

Answer: As we have already seen, Black has sacrificed a pawn, in order to open central lines for his pieces. He has the long diagonal a8-h1,
along which lies the g2-pawn, already attacked once by the rook on g8. The open d-file is also available for a black rook, although the fact that
a7 would hang prevents an early ... 0-0-0. The knight will jump to f5, gaining a tempo on the white queen. White’s issue is how to develop his
pieces and get his king into shelter. The pressure against g2 makes it hard to develop the bishop from f1, and the move g3 always runs into an
attack along the long diagonal, displacing the rook on h1.
15 Rg1
15 Rb1 is the other main move here. We are familiar with the general idea in such positions – the rook
takes up an active post, pressurizing b7 in some lines, and also tames some of the black queen’s activity, by
taking away the b6-square. A high-class practical example then continued 15 ... Nf5 16 Qf2 Qc6.

155
 Question: That is a bit of a surprise. I was expecting the bishop to go to this square.

Answer: That is also a perfectly reasonable alternative, but quite often in this variation, it is the queen which uses the c6-square. From there,
as well as pressurizing g2, the queen can jump effectively into d5, e4, and sometimes a4.
Play may go 17 Rb4 (stopping the check on e4, but exposing the rook; 17 Rg1 could well end in an
immediate draw after 17 ... Qe4+ 18 Qe2 Qd4 19 Qf2 Qe4+ – indeed, from a winning viewpoint, this whole
variation has the drawback that, in many lines, Black has little alternative but to take a draw by repetition, if
White offers it) 17 ... Qd5.

Question: What is the idea of this?

Answer: It is another thematic idea in these lines. Black prepares ... Bc6, establishing a battery along the long diagonal, and securely
defending b7. Another point is to add protection to the d4-square, which allows the black knight to jump there in many lines.
18 Rg1 Bc6 19 Rc4 (19 Bd3 amounts to a transposition after 19 ... 0-0-0,since then 20 Qxa7? Rxg2 21
Rxg2 Qxg2 favours Black; it is noteworthy how well the bishop on c6 defends the black king in such
positions) 19 ... 0-0-0 20 Bd3 Kb8 21 Rxc3.

156
 Question: Two pawns up!

Answer: Yes, but the pawns are hard to make effective use of at present, and the white king is condemned forever to remain in the centre.
After 21 ... Qa2, practice shows that Black has enough counterplay. As Watson points out, this position has been known from at least 1965,
since it can arise by transposition from the 12 ... Bd7 lines. After 22 Be3 (otherwise, the threat of 22 ... Qa1 is awkward to meet) 22 ... Nxe3 23
Qxe3 Rxg2 24 Rxg2 Bxg2 the position is balanced. The game we are following ended in a draw after 25 Kf2 Bc6 26 Be4 Bxe4 27 Qxe4 Qb1
28 Re3 Rc8 29 Re2 Rh8 30 Qg2 Qc1 31 Qg3 Qh1 32 Ke3 a6 33 Rd2 Qc1 34 Ke2 Qh1 35 Ke3 Qc1 36 Ke2 Qh1 ½-½ S.Mamedyarov-
E.Alekseev, Ohrid 2009. Wild though the position looks, this was a pretty correct game by both sides, and a draw is highly likely in this
variation.
15 Be3 is also a natural move, intending 15 ... Nf5 (15 ... b6 is a Watson suggestion, which perhaps keeps
more tension) 16 Qc5. Watson then suggests 16 ... Qc6!.

Question: That is a surprise! I would not have thought Black wanted

to exchange queens, given that he is a pawn down and much of

157
 his counterplay is directed against the exposed white king.

Answer: That is a fair assumption, but White’s development difficulties are the other major aspect of Black’s compensation, and those
persist into many of the endgames. 17 Qxc6 (17 0-0-0 has also been tried, when Fritz likes the clever 17 ... Qb6!, which practically forces the
follow-up 18 Qxb6 axb6 19 Bxb6 Rxa3 20 Kb1 Ra4 when the double threats of ... Rb4+ and ... Rxf4 regain the pawn, again with easy equality)
17 ... Bxc6 18 Kf2 Nxe3 19 Kxe3 0-0-0 20 Rg1 Rd2 21 Rc1 Ba4, gives enough play to hold the balance easily.
Returning to 15 Rg1:

15 ... Nf5
Obviously the most natural, but Black has also tried 15 ... Rh8 here.

Question: With what idea?

Answer: It looks a bit strange, but Black wants to provoke the advance of the white h-pawn, which weakens g3. I am not sure how good it is
objectively, but it has had a couple of successful GM outings, beginning 16 h3 Nf5 17 Qf2 and now:
a) 17 ... Qc6 18 Bd3 Qd5 19 Rb1 Bc6 with a typical set-up for the variation. Once again, it is worth
following the game to the end: 20 Bxf5?! (this exchange removes the active black knight, but at the cost of a
total surrender of the light squares; it is well-known that in middlegame positions with opposite-coloured
bishops, the holder of the initiative has effectively an extra piece, and this is soon seen here) 20 ... exf5 21
Rb3 0-0-0 22 Qe2 Rhg8 23 g4 fxg4 24 hxg4 Rh8 25 Rxc3 Qd4

158
 26 Qe3? (26 Rxc6+ is a better try) 26 ... Qd1+ 27 Kf2 Rh2+ and once again, White suddenly finds himself
busted. It is a familiar scenario in this line. 28 Kg3 Rdh8 29 f5 Qd5 30 Kf4 f6 31 Rc5 fxe5+ 32 Qxe5 Rf2+ 33
Ke3 Qf3+ 34 Kd4 Rd8+ 35 Kc4 Rxc2+ 36 Kb4 Rxc5 37 Qxc5 a5+ 38 Qxa5 Rd4+ 0-1 T.Michalczak-
T.Hillarp Persson, Reykjavik 2011.
b) As usual, Black’s alternative plan is 17 ... Bc6, which has been played by the veteran Cuban GM and
French Winawer expert, Nogueiras. After 18 g4 Nh4 19 Rg3 Qa5 20 Rd3 (Fritz likes White after 20 Be2 0-0-
0 21 Rb1, but it is not clear how much better he is after, say, 21 ... Kb8 or 21 ... Qd5) 20 ... Rd8 21 Qg3 Rxd3
22 Qxd3 Kf8 23 Kf2 Kg7 White should be somewhat better, but as usual, things can easily go wrong in such
positions, which is what happened: 24 Qg3?! (24 Be3) 24 ... Nf3 25 Bd3 Rd8 26 Be3?.

Now Black struck with 26 ... Rxd3! 27 cxd3 Qb5! and White is lost, thanks to the comically helpless
position of his queen. 28 Ra2 Qb2+! 29 Rxb2 cxb2 30 f5 b1Q 31 Qf4 Qg1+ 0-1 I.Ortiz Suarez-J.Nogueiras
Santiago, Havana 2010. Rather a typical game for such lines – White looks as though he should be better, but
one slip and it is all over for him.
159
 16 Qf2 Qc6 17 Bd3
17 g4 Qe4+ 18 Qe2 (18 Be2? Nd4) 18 ... Qd5 is equal.
17 ... Qd5

18 Rb1
18 a4 has also been tried.

Question: With what points?

Answer: White has two ideas: Ra3xc3, similar to the game, and also Ba3.
18 ... Bc6 (Watson advocates 18 ... 0-0-0 with the idea of 19 Ra3 Kb8 20 Rxc3 Bxa4) 19 Ra3 Rh8 (19 ...
Qa5!? is suggested by Fritz) 20 h3 Qa5 21 Bxf5 exf5 22 Qd4 Rd8 23 Qxc3 Qd5 is another opposite-coloured
bishop structure.

White has two extra pawns, but Black is not without play: 24 Bd2 (24 Kf2 is Fritz’s preference) 24 ... Rg8

160
25 g3 Qe4+ (25 ... Rxg3! forces an immediate draw, with 26 Rxg3 Qh1+ 27 Kf2 Qh2+ 28 Ke1 Qh1+, since
26 Qxg3? Qxd2+ 27 Kf1 Qc1+ wins) 26 Kd1 Rd4 27 Re1 Qd5 28 Kc1 and White was objectively better,
although once again, that did not stop him losing in D.Svetushkin-S.Grover, Kavala 2010.
Instead, 18 Be3 allows Black to regain the pawn at once with 18 ... Nxe3 19 Qxe3 Rxg2 20 Rxg2 Qxg2 21
Be4 Qxh2 22 0-0-0, as in R.Robson-S.Shankland, Milwaukee 2009. Now the computer’s 22 ... Bc6 holds the
balance: for example, 23 Bxc6+ bxc6 24 Qd4 a5! (a real computer move; the point is to stop Qb4 in various
lines) 25 Rg1 Qd2+ 26 Qxd2 cxd2+ 27 Kxd2 Kd7 and the rook ending is equal.
18 ... Bc6 19 Rb3 0-0-0 20 Rxc3 Kb8

Question: Once again, Black is two pawns down! Is he really okay in this position?

Answer: Practice confirms that he is. His active pieces, and the insecure enemy king, give Black enough compensation to hold the balance.
21 Rc4
Lots of moves are possible here, but the scenario is basically the same – Black has enough to hold the
balance. Let’s look at a few examples from practice.
a) 21 g3 Nd4 22 Be3 Nf3+ 23 Kd1 Nxg1 (the computer’s 23 ... Qa2! looks like a more convincing way to
hold the balance) 24 Bxa7+ Ka8 25 Qxg1.

161
 With three pawns for the exchange, plus the strong dark-squared bishop, White has more than enough
compensation, and Black is struggling to maintain the balance here. 25 ... Bb5 26 Bb6 Bxd3 27 cxd3 (27
Rxd3 Qb5 28 Rd6 Rc8 29 Kc1 Qb3 was equal in M.Pacher-T.Petrik, Banska Stiavnica 2011) 27 ... Rc8 28
Qd4 Qb5 29 Qb4 Qxb4 30 Rxc8+ Rxc8 31 axb4 and White is better, although he ended up only drawing in
T.Michalczak-L.Johannessen, Rogaska Slatina 2011.
b) 21 Rc5!? Qa2 22 Rxc6!? bxc6 23 Qc5 Rxd3! 24 cxd3 Rxg2 25 Rxg2 Qxg2 is basically just a draw,
S.Ganguly-Y.Shulman, Ningbo 2011.
c) 21 Qc5 Rxg2 22 Qxd5 Rxg1+ 23 Kf2 Rdg8!!

Question: Wow! That is quite a move!

Answer: Yes, although it is only equal!

24 Qd8+! Rxd8 25 Kxg1 is equal, although Black later managed to mate her opponent in A.Muzychuk-

162
K.Lahno, Moscow 2010.
21 ... b6 22 g4 Bb5 23 Rc3 Nd4 24 Bxb5 Qxb5

Black is holding the balance, but it should be no more than that.

25 Qf1 Qa5 26 Bd2 Qa4
Watson claims that 26 ... Qd5 27 Qg2 Qa2 28 Qe4 Rd5 “gives Black excellent chances”, but in fact, 29
Rgg3 is just equal, according to Fritz.
27 Be3??
27 Qd3 Nb5 28 Rc4! Qa6 29 Qe2 Nxa3 30 Rc3 maintains equality, as shown by the computer. But this line
is enough to show how carefully White has to play, even with two extra pawns. The text is a losing blunder.
27 ... Nxc2+ 28 Kf2 Nxe3 29 Rxe3
White must lose the f4-pawn, since 29 Kxe3 Qd4+ drops a whole rook.
29 ... Qxf4+ 30 Ke2 Qxh2+ 31 Rg2
If 31 Qf2 Rd2+.
31 ... Qh7

163
 In heavy-piece endings, king safety is paramount, and here, White has a hopelessly exposed king, as well
as being material down.
32 Kf3 Rd4 33 a4 Rgxg4 0-1

Question: So what are we to make of these crazy 7 Qg4 Qc7 lines?

It looks as though everything hangs by a thread?

Answer: I think that is absolutely right. I myself played these lines for some 20 years. In all honesty, I never entirely believed my position,
yet I scored an incredible 86% as Black! I lost quite a few post-mortems, where opponents eventually managed to show they had stood better at
some point (often quite a lot better), but in the games themselves, I nearly always managed to come out on top. We have seen quite a few
examples in the foregoing as well, even with GMs on the white side. I think the big lesson of this is that Black’s position is much easier to play
than White’s, in a practical OTB game.

Question: Why?

Answer: I think it is because almost all players find it much easier to spot their own tactical opportunities than to anticipate their
opponent’s. To play the white side of this variation successfully, White needs to be able to spot all of his opponent’s resources, and very few
players are able to do this, even at GM level. One slip is all it takes to bring the white position crashing down, because his king is always so
draughty in these positions. The black player just needs to be alert to his own chances, which most players (myself included!) find much easier.

Question: What about the theoretical issue of whether

to play the old 12 ... Bd7 lines, or the modern 12 ... d4?

Answer:
I myself always played 12 ... Bd7, but that was just because 12 ... d4 had not become popular when I was active as a player. At the time of
writing, 12 ... d4 looks to be in pretty decent shape, and I think this would be my choice now.

164
Game 18
V.Atlas-S.Kindermann
Austrian League 2006

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Qg4 0-0

Question: This looks like a much safer way to play

than 7 ... Qc7. But isn’t Black ‘castling into it’?

Answer: Right on both counts! If Black can get away with it, then 7 ... 0-0 is really the move he would like to play. He does not sacrifice
any material, and he avoids having his kingside pawns wiped out. The problem is, he is castling into quite a strong attack.

Question: Is 7 ... Nf5 possible here?

Answer: It is another move Black would like to play, if he could, because it defends g7 without committing the king to the kingside.
However, the move has been out of favour for some 50 years now! After 8 Bd3, the threat is to take on f5 and g7, so the reply is forced: 8 ... h5
(the immediate 8 ... Qh4?? drops a piece after 9 Bxf5) 9 Qf4! cxd4 (9 ... Qh4 10 Ne2! Qxf4 11 Nxf4 Ne7 12 Be2! h4 13 Nh5 Kf8 14 Bg5 is
clearly better for White, M.Tal-T.Petrosian, USSR Championship, Moscow 1983, as here is 12 ... g6 13 dxc5!) 10 cxd4 Qh4 (this way, Black
avoids the dxc5 capture mentioned in the previous note, but on the other hand, White’s doubled pawns have been straightened out) 11 Qxh4!
(now this is strongest) 11 ... Nxh4 12 Bg5! Nf5 (12 ... Nxg2+ 13 Kf1) 13 Ne2 Nc6 14 c3 and Black never escaped from the dark-square chains,
D.Yanofsky-W.Uhlmann, Stockholm Interzonal 1962.

165
 This game was immensely influential, and the move 7 ... Nf5 has never really recovered.

Question: What about 7 ... g6, similar to the Portisch-Hook lines?

Answer: In this position, 7 ... g6? is just plain bad, I believe. As noted in the context of the Fischer-Hook line, when he plays ... g6 in such
positions, Black usually needs to leave his knight on g8 for some time, to cover the f6- and h6-squares. Combining ... g6 with ... Ne7, as here,
looks terrible, and indeed, there are no worthwhile practical examples, so bad does it look!
On the other hand, the alternative Fischer-Hook treatment, 7 ... Kf8, is also playable here, and had a brief
spell of popularity in the 1980s. However, I don’t think it is as good as in the Portisch-Hook Variation.

Question: Why?

Answer: Since here the black queen is not blockading the white queenside on a4, White can reply 8 a4! and bring his bishop to the strong
square a3. It is not the end of the world for Black, but I see no reason for him to play this way – if he wants to play a ... Kf8 variation, then the
Portisch-Hook system seems a much more logical form in which to do so.

166
 8 Bd3
This has emerged in recent years as definitely the most dangerous move, and the only one to pose Black
real problems. The older line was 8 Nf3 Nbc6 9 Bd3 f5 10 exf6 Rxf6 11 Bg5.

However, this lost ground rapidly in the 1980s, once it was realized that Black can reply 11 ... e5!

Question: Is that really playable? It looks as though it should lose to something!

Answer: That is why it took literally decades before black players realized it was feasible! But once one examines the tactics with an open
mind, it soon becomes apparent that Black is fine in all variations:
a) 12 Qh4 e4 is the first point. Black will win two pieces for the rook.

Question: But isn’t his king exposed?

Answer: Not seriously, because most of White’s developed minor pieces end up disappearing from the board! 13 Bxf6 gxf6 14 Qxf6 exd3
(14 ... Qf8 is also adequate, but the text is simpler) 15 cxd3 cxd4 16 Nxd4 Nxd4 17 Qxd4 Nc6 and Black is fine.
b) 12 Bxh7+ Kxh7 13 Qh5+ Kg8 14 Bxf6 gxf6 15 dxe5 Qf8! is no problem for Black.

167
 For example, 16 exf6 Qxf6 17 0-0 Bf5 and although Black’s king is bare of pawn cover, his pieces do an
excellent job of guarding His Majesty. White has a couple of extra pawns, but his queenside pawns are weak.
Chances are balanced.
c) 12 Qg3 looks strong, because 13 Bxf6 seems to be a powerful threat, but Black has the excellent (if
forced) reply 12 ... Rxf3! 13 gxf3 c4 14 Be2 Qa5 15 Bd2 Nf5 and Black is actually better.
8 ... f5
A major parting of the ways. As we will see, the absence of the knights from f3 and c6 favours White in
this line, by comparison with the variation just examined in the last note.

Question: Doesn’t 8 ... Nbc6 9 Nf3 f5 transpose to the last note?

Answer: Yes, but after 8 ... Nbc6, White has the much more dangerous try 9 Qh5!. Now the only move is 9 ... Ng6 (9 ... h6 was played quite
a few times at GM level, until White finally proved that it loses by force: 10 Bxh6! gxh6 11 Qxh6 Nf5 12 Bxf5 exf5 13 0-0-0 f4 14 Nh3 Ne7 15
Ng5 Bf5 16 g4! Be4 17 Rhe1 Qb6 18 e6 Bg6 19 Rd3 1-0 S.Maus-R.Hübner, Lugano 1989).

168
 Question: So what is going on here?

Answer: 9 ... Ng6 was popularised by the German Grandmaster Gerald Hertneck, and quickly became the main line. Black defends the mate
threat, without creating any structural weaknesses, and he always has the option of ... c4, driving the white bishop off the b1-h7 diagonal.

Question: It looks as though he should be fine.

Answer: The trouble is, after a huge amount of practical testing, White has perfected a plan, which offers him attacking chances, and the
analysis of which runs literally to move 35 and beyond. Theoreticians still argue about whether White is better or not, but it is clear that Black
is playing for a draw in this line, and from most players’ viewpoint, it is not terribly attractive to have to memorise 40-odd moves of theory, in
order to reach a possibly tenable position. The basic line runs as follows: 10 Nf3 Qc7 11 Be3! (the most subtle; White delays castling, since the
moment he castles short, Black closes the position with ... c5-c4: for instance, 11 0-0 c4 12 Bxg6 fxg6! and he has no real problems) 11 ... c4.

Question: Is this forced?

Answer: Not exactly, but Black is running out of waiting moves. 11 ... Nce7 was tried, but 12 h4 still keeps some advantage.
The main line goes on 12 Bxg6 fxg6.

Question: Why this recapture? What happened to capturing towards the centre?

Answer: In this line, Black almost always has to take with the f-pawn, because 12 ... hxg6? 13 Qh4 leaves him struggling with the threat of
Ng5.
After 13 Qg4 we reach the basic position of this variation. By leaving his king in the centre, White has
preserved the possibility of h4-h5, levering open the h-file. As I said above, the analysis of this position runs
extremely deep, and there is still no final verdict, but it is easy to see that Black’s position lacks dynamism
(he has no pawn breaks, now that he has had to take back with the f-pawn), and he lacks counterplay. It is just
a question of whether White can create serious attacking chances or whether Black will hold him off and
draw.
A recent example, featuring one of the staunchest defenders of the position on the black side, was the game
D.Svetushkin-A.Dgebuadze, Livigno 2011, which continued 13 ... Qf7 14 Ng5 Qe8 15 h4 h6 16 Nh3 Bd7 17
Qe2 b5 (this queenside pawn advance is the only chance of counterplay) 18 g4 Ne7 19 h5 gxh5 20 g5 hxg5
21 Nxg5 g6 22 Kd2 a5 23 Rag1

169
 23 ... Rf5?! 24 Nf3 Qf8?! 25 Nh4 Qg7 26 Bg5 Rxg5 27 Rxg5 and White eventually won.

Question: So are you saying the Hertneck System is bad?

Answer: No, but I just don’t think it very attractive in a practical sense. Black ends up defending a blocked position, in which White has
most of the winning chances that exist, and which has also been analysed almost to death. By all means play the line if you like it, but it is not
really my cup of tea.
We now return to 8 ... f5:
9 exf6 Rxf6 10 Bg5

The best move. 10 Qh5 allows Black to meet the mate threat with 10 ... h6! when he stands perfectly
satisfactorily.
10 ... Rf7
Now we see the difference from the knights not being on f3 and c6: 10 ... e5? is simply bad after 11 Qg3
when there is no exchange sacrifice on f3.
170
 11 Qh5!
Again the most precise. 11 Qh4 h6! is less of a problem for Black.
11 ... g6
Forced (or is it? – see below!), but severely weakening the dark squares.

Question: But why is this necessary? What is

wrong with 11 ... h6 as in previous variations?

Answer: The problem is tactical. White has 12 Bh7+! Kxh7 (the computer regards 12 ... Kf8 as perfectly playable, so maybe this line is
worth another look for Black) 13 Qxf7 hxg5 14 Nf3 Nbc6 15 Nxg5+ Kh6 16 Nxe6 Bxe6 17 Qxe6+ Ng6 which is regarded as good for White,
although admittedly, looking at it again, it also may not be so bad for Black. Perhaps 11 ... h6 is not as bad as has hitherto been assumed?
12 Qd1

This is more or less a tabiya for the line starting with 8 ... f5.

Question: And what is going on?

Answer: In many ways, it is a typical Winawer. White has the bishop pair-and Black’s dark squares are weak (especially so here, with the
move ... g6 played). On the other hand, White has the traditional weak queenside pawn structure. Most GMs would probably take the view that
White is somewhat better, but even if that is so, it is not clear that he has a huge amount. Winawer players should be reasonably happy with this
type of position.
12 ... Qa5
The alternative, which has been widely played here, is 12 ... Nbc6 13 Nf3 Qf8.

Question: That looks a slightly funny square for the queen.

Answer: It is rather unusual to see the queen on f8, certainly, but in this position, it makes some sense. Black covers the sensitive dark
squares around his king. He can follow up with ... h6, ... Qg7, and then ... Bd7 and ... Raf8, for example. One recent practical example, by a
regular practitioner of the line, went 14 0-0 c4 15 Be2 h6 16 Bc1 Kh7 17 a4 Bd7 18 Rb1 b6

171
 19 a5 bxa5 20 Ba3 Rd8 21 Re1 Bc8 22 Bc5 Qg7 23 Bf1 Nf5, with an unclear position in M.Zoldan-
M.Drasko, Cento 2011. Black went on to win, but he was far the stronger player. Overall, the ... Qf8 lines are
a respectable way to play, but White should be a little better, at least.
13 Bd2
13 Qd2 is also possible. Then Watson quotes a Kindermann recommendation 13 ... c4 14 Be2 Nf5,
intending ... Nd6. Once again, we have a typical Winawer struggle, where Black’s dark squares are horrible,
but he has the usual structural compensation.
13 ... Nbc6 14 Nf3

14 ... Qc7

Question: What is the point of this? It looks rather

odd to retreat to c7, rather than, say, a4.

172
 Answer: 14 ... Qa4 is a possible move and has also been played, but the text has been the most popular in recent years. Black plans to keep
his pawn on c5 and to advance ... e6-e5, not something he is usually able to do in the Winawer.

Question: Doesn’t that risk opening the position for White’s bishops?

Answer: Yes, it does, but on the other hand, Black frees his own pieces too, and also rids himself of the structural weakness on e6. In the
resulting positions, practice suggests that chances are fairly balanced.
14 ... c4 is once again a major alternative, with the standard blocked Winawer position.

Question: How do you assess the position?

Answer: White is probably a little better, but it is not a great deal, and in practice, most Winawer players should be happy enough with such
a familiar structure. Of course, if Black could play the move ... g6-g7, he would be in clover, but alas, the rules do not permit this!
15 dxc5

Question: Slightly surprising!

Answer: Well, we have seen similar ideas before in this book. White wins a pawn, at least temporarily, and hopes that, even if Black regains
it, the opening of the position will favour the white bishop-pair.
15 0-0 is the main alternative. Then after 15 ... e5 16 Nxe5 (16 Ng5 Rf8 17 c4 is a major alternative, when
17 ... exd4! 18 cxd5 Ne5 is unclear, but seemingly satisfactory for Black) 16 ... Nxe5 17 dxe5 Qxe5 18 c4 (18
Re1 Qf6 was equal in M.Bluvshtein-A.Barsov, Hastings 2004/05) 18 ... d4 19 f4 Qc7 20 Qf3 White was
maybe slightly better in D.Stellwagen-J.Timman, Malmo 2006, but Black held the draw without huge
problems, despite being pressed for quite a few moves.
15 ... e5 16 Be2?!
Rather passive. A more critical continuation is 16 Ng5 Rf8 17 c4. Now Watson quotes the game
Y.Pelletier-S.Kindermann, Biel 2003, which continued 17 ... e4 18 cxd5 (18 Be2 Qe5 is fine for Black) 18 ...
exd3 19 d6 Qd7 20 dxe7 Qxe7+ 21 Be3.

Here Kindermann has suggested 21 ... Rf5! 22 Nf3 Be6 23 0-0 Bd5 24 Nd2 Rh5 with an unclear position,
in which Black seems to have quite good chances.

Question: This is all pretty hairy, compared with the

usual blocked Winawer structures after ... c5-c4 by Black.

173
 Answer: That is true. The more fluid plan with ... e6-e5 in this line results in an altogether more dynamic position, but one in which Black
seems to be holding his own.
16 ... Bg4 17 0-0 Bxf3

Question: Why give up the bishop?

Answer: It is not forced, but Black prefers to eliminate the knight, which can otherwise cause trouble by jumping to g5 at some moment. In
the structure resulting from this exchange, Black feels his knights are the equal of White’s bishop-pair.
18 Bxf3 e4 19 Bg4 Nf5

Question: What is your assessment of this position?

White has an extra pawn and the two bishops, after all!

Answer: Yes, but I think chances are still about equal. The extra pawn is not very significant, in view of the trebled c-pawns, and Black’s
pieces are all active and well placed. It is not so easy to activate the white bishops, as the enemy knights and centre pawns control a lot of
squares. Black has to be careful, especially considering his somewhat draughty king, but with accurate play, I think he is fine.
20 Be3 Nxe3 21 fxe3 Rxf1+ 22 Qxf1 Qe7 23 Qf4 Ne5 24 Rf1 Re8
Black is comfortably holding here.
25 h4 h5 26 Bh3?!
This is the start of White’s downward spiral. He still harbours ambitions, and wishes to keep his bishop
active on the h3-c8 diagonal, but now the black knight gets the wonderful outpost c4, attacking a host of weak
white pawns. 26 Be2 looks better, dominating the knight.
26 ... Kg7 27 a4 Nc4 28 Rb1 Qxc5 29 Rxb7+ Re7 30 Rxe7+ Qxe7

174
 White still has an extra pawn, but it is not felt at all, and the black queen plus knight combination is known
to be effective in the endgame. The position should be a draw, but White soon goes wrong.
31 Kf2 a5 32 Qg5??
A dreadful misjudgement. White probably thought that, after the queen exchange, his bishop would be
better than the knight on an open board, but in actual fact, the bishop has no targets, whilst the white pawns
drop like ripe fruit. White should just mark time with 32 Ke2 and hope that his opponent cannot find a way to
strengthen his position.
32 ... Qxg5 33 hxg5 Nb6
Now the a4-pawn falls, and the passed black a-pawn will decide the game.
34 Be6 Nxa4 35 Bxd5 Nxc3 36 Bc4 a4
White is losing a piece.
37 Ke1 a3 38 Kd2 a2 39 Bxa2 Nxa2 40 c3
The knight is trapped, but while White is rounding it up, Black’s king sets up a winning pawn ending.
40 ... Kf7 41 Kc2 Ke6 42 Kb2 Kd5 43 Kxa2 Kc4 44 Kb2 Kd3

175
 45 c4
Desperation, but 45 Kb3 Kxe3 46 c4 Kd4 47 Kb4 e3 sees Black queen with check.
45 ... Kxc4 46 Kc2 h4
Zugzwang. Black takes the opposition.
47 Kd2 Kb3 0-1

Question: So what is the overall conclusion on the 7 ... 0-0 variation?

Answer:
I think Black is objectively okay in these lines. There is no doubt that 7 ... 0-0 is a much calmer and safer way to handle the black position,
although it lacks the fireworks of the 7 ... Qc7 lines. Much depends on the black player’s mood and preparation – although I played 7 ... Qc7
most of the time in my playing days, there were occasions when I just did not fancy the hair-raising struggles that ensue, and preferred to play a
quieter, less knife-edge position. On those occasions, I used 7 ... 0-0, which can be played with much less preparation (especially in the main
line with 8 ... f5) and is less stressful on Black’s nerves!

176
Game 19
V.Hort-T.Petrosian
European Team Championship, Kapfenberg 1970

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Qc7

We have already seen this variation in Game 14, but this is where we see the real point of the move order.
7 Qg4

7 ... f5
This is the point. Having the queen on c7 already allows Black to defend his g7-pawn by advancing the f-
pawn, giving him additional possibilities.

Question: Is 7 ... f6 possible instead?

Answer: This move did indeed become quite popular in the late 1960s, chiefly thanks to the impressive Black win in the game
A.Matulovic-R.Byrne, Sousse Interzonal 1967, which continued 8 Nf3 c4! 9 Be2 Nc6 10 0-0 Qf7! 11 Qh3 Nge7 12 a4 Bd7 13 Ba3 0-0-0 14 a5
h5, and Black stood clearly better, thanks to the poor position of the white queen, which will be a target for the advancing black kingside
pawns.
However, nowadays the line is regarded as rather dangerous for Black, thanks to some startling tactical
resources, beginning 8 Bb5+! Kf8.

177
 Question: That’s a shock!

Answer: For a long time, this move was regarded as a refutation of White’s last. The point is that White is now faced with three strong
threats: 9 ... cxd4, attacking c3, 9 ... Qa5, attacking both c3 and b5, and 9 ... c4, trapping the bishop on b5. It appears that White has no
satisfactory way to meet them all, but later, it became clear that he can simply develop with 9 Nf3! and then:
a) 9 ... cxd4? is met simply by 10 0-0 with a large development advantage. This position is extremely
dangerous for Black.
b) Likewise, the seemingly devastating 9 ... Qa5 is answered by the forced but very strong 10 Rb1!. For
example, 10 ... a6 11 Be2 (11 Bd3 c4 12 Be2 Qxc3+ is less clear) 11 ... Qxc3+ 12 Bd2 Qxc2 13 Rc1 Qe4 14
Qg3 and, once again, White has a huge development lead and a very strong attack.

The fact that Black has played ... f6, rather than closing the position with ... f5, makes the development
issue that much more serious for him, as the position can open up more after exf6. One recent all-GM
encounter saw Black massacred in typical style: 14 ... Nd7 15 exf6 gxf6 16 Qd6+ Ne7 (16 ... Kf7 is the only
178
hope, but White still has a massive attack after 17 0-0) 17 0-0! Rg8 (17 ... Qxe2 18 Rfe1 Qd3 19 Rxe6 is
decisive) 18 Rfe1 Qg6 19 Nh4 Qf7 20 Bh6+ Ke8 21 Bh5! Qxh5 22 Rxe6 Qf7 23 Nf5 Kd8 24 Rxe7 Qg6 25
Rxd7+ Bxd7 26 Qb6+ 1-0 N.Vitiugov-S.Dyachkov, Russian Team Championship 2008.
c) 9 ... c4 is the most critical, cutting off the bishop, but now there follows 10 a4! a6 11 Ba3+ Ne7.

Exercise: Find a strong move for White.

Answer: 12 Be8!!. It turns out that the bishop is immune, since after 12 ... Kxe8 13 Qxg7 Rf8 (13 ... Rg8 14 Qxh7 is no better) 14 exf6 Rf7
15 Qh8+ Rf8 16 Qxh7 Rxf6 17 Ne5 White has two pawns for the piece, and a clear advantage. Instead, 12 ... fxe5 13 Qh5 g6 14 Bxg6! is again
winning for White, which only leaves 12 ... Qd8. But then there follows 13 Bh5! Nbc6 (13 ... f5 14 Qh4 g6 still doesn’t trap the bishop, because
of 15 Bxg6! – the prelate certainly has a charmed life; 13 ... g6 14 Bxg6 hxg6 15 exf6 also wins for White) 14 exf6 gxf6 15 Ng5! (15 Ne5! is
also good) 15 ... Qa5 (or 15 ... fxg5 16 Qxg5 and Black is helpless against the check on f6) 16 0-0 f5 17 Qh4 1-0 E.Sutovsky-S.Dyachkov,
Moscow 2007. A devastating game, which ‘closes’ the variation with 7 ... f6 8 Bb5+ Kf8 pretty convincingly.
Returning to the safer 7 ... f5:
8 Qg3
This is one of two main lines. 8 Qh5+ is the other.

179
 Question: Can’t Black play 8 ... Qf7, avoiding the weakening of his kingside?

Answer: He can, but White is somewhat better after both 9 Qxf7+ Kxf7 10 c4!, and 9 Bb5+ Bd7 10 Qxf7+ Kxf7 11 Rb1.
Thus play tends to go 8 ... g6 9 Qd1.

Question: The dark squares on the black kingside look terrible!

Answer: Yes, this is another of these rather ugly variations, that one sees in the Winawer. However, as usual, the position is nothing like as
bad as it first looks. Although his dark squares are a bad accident, Black has his usual compensating factors: the weakened white queenside
pawns, in particular, which he can begin to pressure after 9 ... Bd7!.

Question: Why is this best?

Answer: It is a subtle idea, the point of which we will soon see. The bishop is aiming at the a4-square. Instead, 9 ... cxd4?! 10 cxd4 Qc3+ 11

180
Bd2 Qxd4 12 Nf3 Qe4+ 13 Be2 is a very risky pawn-grab. Black can probably survive with accurate play, but I cannot recommend this for
him.
Now White tends to go 10 Nf3.

Question: Isn’t 10 a4 a logical reply, to keep the bishop out?

Answer: It is, but now the pawn grab 10 ... cxd4 11 cxd4 Qc3+ 12 Bd2 Qxd4 13 Nf3 Qe4+ 14 Be2 Nc6 is more justifiable, since Black is a
move ahead in development, compared with the analogous line at move 9.
After 10 ... Ba4 we arrive at a key position.

White has several possibilities:

a) 11 Rb1 is regarded as the most testing of the various alternatives, although 11 dxc5 (see below) looks
more dangerous to me. After 11 ... Nd7 (11 ... cxd4 12 Rb4 is one point of White’s play) 12 Be2 (12 dxc5 is
less effective than a move earlier: 12 ... Nxc5 13 Be3 b6 14 Qd4 Ne7 15 Bd3 h6 16 0-0 g5 and Black was fine
in T.Wedberg-E.Berg, Malmo 2002) 12 ... h6 13 dxc5 (13 0-0 Ne7 is equal – Watson/Lahlum) 13 ... Nxc5 14
Rb4 Bd7 15 0-0 Ne4 Black had no particular problems in J.Timman-T.Shaked, Merrillville 1997.
b) 11 Bd3?! is natural, but amounts to a trap of sorts, since it allows 11 ... cxd4 12 cxd4 Qc3+ 13 Bd2
Qxd3! (one of the points of Black’s whole move order) 14 cxd3 Bxd1 15 Kxd1.

181
 Question: So what is so terrible about this for White?

Answer: It is not so terrible, but practice shows that Black has no problems and can even play for the advantage long-term. Play could
continue 15 ... Nc6 16 Rb1 (16 h4 h6 17 Ke2 Nge7 18 a4 Kd7 19 a5 Nd8 20 Rhb1 Nec6 21 Rb5 Rb8 22 Rg1 Nf7 23 g4 fxg4 24 Rxg4 Rhg8 25
Rb1 h5 26 Rg2 Rg7 27 Rbg1 Rbg8 28 Rb1 Kc7 with an unclear position in J.Egger-L.Johannessen, Istanbul Olympiad 2000, eventually won
by Black) 16 ... b6 17 Rc1 Kd7 18 h4 h6 19 Ke2 Nge7 20 Rc2 Rac8 21 Rhc1 Nd8 22 Rxc8 Nxc8 23 a4 Ne7 24 Be3 Nf7 25 a5 Nc6 26 axb6
axb6 and Black eventually won in J.Polgar-I.Morovic Fernandez, Buenos Aires 1992.
c) 11 c4 attempts to open the position, to favour the white bishops: 11 ... Ne7 12 cxd5 Nxd5 13 Bc4 was
E.Sutovsky-E.Berg, Internet blitz 2004, and now Fritz points out the cute trick 13 ... cxd4! 14 Bxd5 Qa5+ 15
Qd2 Qxd5 with equality.
d) 11 Ra2 deals with the tactical threats on the c-file, but the rook is ugly on a2. Black is fine after 11 ...
Nd7 12 h4 h6 13 Bf4?! (13 Be3 0-0-0 is equal) 13 ... Ngf6!, as the knight is coming to e4.

Black rapidly gained the upper hand and won in the following game: 14 Bxh6 Rxh6 15 exf6 Nxf6 16 Bd3

182
c4 17 Qc1 Qg7 18 Be2 Ne4 19 Qb2 g5 20 Qb4 b5 21 Ne5 a5 22 Qb2 gxh4 23 Bf3 Qg5 24 Kd1 Qd2# 0-1
K.Stokke-L.Johannessen, Bergen 2002.
It should be noted that this entire 6 ... Qc7 system is a major Scandinavian speciality. A really excellent
and thorough coverage of the whole line can be found in the 3rd edition of John Watson’s Play the French,
the chapter being guest-written by Norwegian player and organiser, Hans Olaf Lahlum.
e) 11 Bd2.

Question: What is the point of this?

Answer: It stops the check on c3, and thus rules out the Polgar-Morovic line. It also sets a nice trap: 11 ... Nc6 (11 ... cxd4? 12 cxd4 Qxc2??
13 Bb5+! is the trap; however, 11 ... Nd7 is also playable: for example, 12 Bd3 Ne7 13 0-0 h6 – a typical move in this variation, covering the
dark squares h6 and g5, and hoping to expand with ... g6-g5 – 14 h4 Kf7 15 Rb1 Rac8 16 Rb2 Nb6 with decent play for Black in F.Andersson-
A.Olssen, Hallstahammer, 2002, as quoted by Lahlum) 12 Bd3 h6 13 h4 Nge7 14 Qc1 0-0-0 15 Rh3 Rh7 16 Kf1 Rdh8 17 Kg1 c4 18 Be2 Nd8
19 g3 Nec6 20 Nh2 g5! with good play for Black, who went on to win in Z.Zhao-E.Berg, Athens 2001.
f) 11 dxc5.

Question: White wrecks his pawn structure again!

Answer: We have seen the idea quite a lot. White wants to open the position and expose the enemy’s weakened dark squares. He is
prepared to sacrifice his structure to do so. 11 ... Qxc5 (11 ... Nd7 12 Qd4 Nxc5 13 Bd3 h6 14 h4 b6 15 0-0 0-0-0 was fairly solid for Black in
J.Dworakowska-M.Socko, Ostrow 2002, but here 14 0-0 is possibly a sterner test: 14 ... Ne7 15 Be3 b6 16 Qh4 and White was better in
M.Chandler-A.Kinsman, British League 1998) 12 Bd2 (12 Bd3?! allows another Polgar-Morovic trick: 12 ... Qxc3+ 13 Bd2 Qxd3 14 cxd3
Bxd1 15 Rxd1 Nc6 and Black was fine in D.De Vreugt-B.Socko, Ohrid 2001) 12 ... Nc6 13 Bd3 0-0-0 14 0-0 h6 “looks solid” according to
Lahlum.

183
 However, 15 Rb1 Nge7 16 Qc1 g5 17 Nd4 appears rather dangerous to me.
Overall, 11 dxc5 looks like a critical line for this whole variation.
We now return to 8 Qg3:
8 ... cxd4
8 ... Ne7 has been under a cloud ever since the first game of the Tal-Botvinnik World Championship match
of 1960, which continued 9 Qxg7 Rg8 10 Qxh7 cxd4 11 Kd1, and White won in 32 moves. John Watson
attempted to revive the move, with some analysis in the second edition of his Play the French series, but did
not manage to convince himself of its viability. Readers armed with 21st century analysis engines may wish
to make another attempt.
9 cxd4 Ne7

Question: So what is happening here?

184
 Answer: By exchanging on d4 first, Black has prevented White from meeting this by taking on g7. He will follow up with ... 0-0, and then
... b6 and ... Ba6, exchanging off his bad bishop.

Question: That looks pretty solid. Isn’t Black doing well?

Answer: It is a solid line, but White’s chances should also not be underestimated. His doubled pawns have been dissolved, and he has hopes
of activating his dark-squared bishop. He still has serious weaknesses on the c-file, it is true, but these can be defended adequately, while he
builds up his attack on the other wing. Overall, chances are approximately equal, with a long positional battle in prospect.
10 Bd2
Covering c3 and so attacking g7.
10 ... 0-0 11 Bd3 b6 12 Ne2
This is the usual development plan. The knight heads for f4, from where it will attack the weakened e6-
pawn, and also have the possibility of jumping to h5.
12 Nf3?! seems less effective. One practical example saw White use the plan of taking with the pawn on
d3, which we will discuss later, but it brought him no great dividends here: 12 ... Ba6 13 0-0 Bxd3 14 cxd3
Nbc6 15 h4 Qd7 16 Qg5 h6 17 Qg3 Kh7 18 Qh2 Rf7 19 g3 Rc8 20 Kg2 Rff8 with equal chances in
M.Menacher-P.Haba, Leinfelden 2000. Black eventually won, but he outrated his opponent by some 200
points.
12 ... Ba6
Another Petrosian classic is worth a closer look, for the middlegame strategy: 12 ... Rf7 13 0-0 Ba6 14 h4
Bxd3 15 cxd3.

Again this recapture.

Question: What is the point?

Answer: White removes his backward pawn from the c-file and also covers the c4-square, which otherwise represents an outpost for the
black knight. However, the downside is that the d4-pawn becomes more vulnerable (it can no longer be defended by c2-c3),and Black has a
queenside pawn majority. 15 ... Nbc6 16 Be3 Qd7 17 Nf4 Ng6! 18 Nxg6 hxg6 19 Qf4 (19 Qxg6 is met by 19 ... f4 20 Bd2 Nxd4 with a good
game for Black) 19 ... Qe8 20 g3 Rc7 21 Kg2 Qf7 22 Rh1 and now Petrosian commenced a Long March with his king:

185
 22 ... Kf8! (evacuating the danger zone) 23 Qg5 Ke8 24 Rac1 Kd7 25 h5 gxh5 26 Rxh5 Rg8 27 Rh7 Kc8
28 Qh4 Qg6 29 Rh8 Rxh8 30 Qxh8+ Kb7, J.Diez del Corral-T.Petrosian, Palma de Mallorca 1969. The king
is extremely safe on the queenside, and such long strategical manoeuvres are typical of this variation.
Chances are still objectively equal, although Petrosian went on to win.
13 Nf4
After 13 Bxa6 Nxa6 we have another typical position.

Question: What is the assessment?

Answer: Chances are about equal. White can bring his knight to f4, tying Black down to the defence of e6, and then push his h-pawn, trying
to create kingside chances. Black, for his part, has play on the c-file: 14 Qd3?! Qc4! (Black is happy to exchange queens, because although he
plugs the c-file, he also gets the d5-square, and secures his king from attack) 15 Qxc4 dxc4 16 Bb4 Nxb4 17 axb4 Nd5 18 c3 a5 19 bxa5 Rxa5
20 Rxa5 bxa5 (now Black has a passed a-pawn, and he soon establishes a winning position) 21 Kd2 f4 22 f3 Rb8 23 Ra1 Rb2+ 24 Ke1 a4 25
Ra3 g5 26 g3 Rb1+ 27 Kf2 Rb2 28 Ke1 Ne3 29 Rxa4 Rb1+ 0-1 K.Goh-P.Haba, Internet (blitz) 2005.

186
 13 Bb4 is another option for White. A nice demonstration of the virtues of the black position was provided
by a well-known Winawer expert, in the following game: 13 ... Bxd3 14 cxd3 Rf7 (a typical prophylactic
idea; the rook unpins the knight on e7, defends g7 – anticipating Nf4-h5 – and can later swing across to c7, if
occasion permits, to double on the c-file) 15 0-0 Nbc6 16 Bd6 Qd7 17 a4

17 ... f4! (this tactical blow is a thematic idea in such positions) 18 Qg4 (18 Nxf4 Nxd4 is clearly better for
Black) 18 ... f3! 19 gxf3 Nf5 20 Ba3 h5! (winning the d4-pawn, after which White’s position soon collapses)
21 Qxh5 Nfxd4 22 Nxd4 Nxd4 and Black was winning in V.Vehi Bach-L.Psakhis, Manresa 1996.
Finally, 13 0-0 Bxd3 14 Qxd3 Rc8 15 Rfc1 Qc4 16 Qxc4 Rxc4 reaches a typical endgame for this
variation. Black has easy play on the queenside, although White’s position is hard to break down. Such
endings typically arise with the stronger player as Black, and he usually scores quite heavily. A characteristic
example was E.Gullaksen-R.Djurhuus, Norwegian Championship 1996: 17 a4 Nbc6 18 c3 Na5 19 Rcb1 Kf7
20 h4 Rg8 21 g3 Rc7 22 Kg2 Nc4 23 Ra2 Rb8 24 Kf3

24 ... a5 25 Bc1 b5 (Black’s last two moves are worth noting, as a typical way to create a passed a-pawn in
187
such positions) 26 Ba3?! b4! 27 cxb4 Nxa3 28 Rxa3 axb4 29 Raa1 Nc6 30 Ke3 Na5 31 Kd3 b3 32 Nc3 b2 33
Ra3 Nc4 34 Ra2 Rb3 35 Kc2 Na5 0-1.
13 ... Qd7 14 Bb4
14 h4 is another typical move. After 14 ... Bxd3 15 Qxd3 ...

... Black’s best is 15 ... Rc8!.

Question: Why does this deserve an exclamation mark?

Answer: This move of Kaidanov’s is generally regarded as the most accurate. Black’s idea is to play ... Rc4 and only then bring his knight
to c6. The older line was 15 ... Nbc6, which is probably less precise, though by no means bad. Two famous examples are worth seeing after 16
Rh3 Rac8 17 Rg3:
a) 17 ... Kh8 18 h5 Rf7 19 h6 g6 20 Rc1 Rff8 21 Ne2 Nb8 22 Kf1 Rc4 23 Kg1 Nbc6 (Black’s last three
moves explain the point of Kaidanov’s 15 ... Rc8; here, Black has lost a couple of tempi with his knight, in
order to reach the desired ... Rc4 and ... Nc6 set-up) 24 Bg5 Ng8 was S.Reshevsky-M.Botvinnik, The
Hague/Moscow 1948. White ought to be better here, but even though he seems to have achieved quite a bit of
progress on the kingside, he found it impossible to break through, and Black eventually won.
b) 17 ... Rf7 18 h5 Nd8 19 c3 was G.Kasparov-N.Short, Novgorod 1997.

188
 Now Kasparov gives 19 ... Qa4 as slightly better for Black. Even so, Short subsequently switched to 15 ...
Rc8, which suggests he thinks it better.
After the rook move and then 16 Rh3 Rc4 17 Rg3 Nbc6 18 c3 Rf8 (the other rook comes over to defend
g7) White has tried:

a) 19 Kf1 Rf7 20 Kg1 Na5 21 a4 Qc6 22 Ra3 Kf8 23 Qb1 Nb7 was agreed drawn in I.Teran Alvarez-
N.Short, Spanish Team Championship 2003.
b) So was 19 h5 Na5 20 h6 g6 21 Nh5 Kh8 22 Nf6 Qc8 23 Bg5 Rf7 24 Rf3 Nac6 25 Kf1 Nb8 in
R.Castellanos-L.Psakhis, Andorra 2003.
c) 19 Nh5 Ng6 20 Nxg7? f4! (once again, this typical tactical blow) 21 Rg5 Qxg7 and Black was winning
in M.Hoffmann-A.Rustemov, German League 2004.
Before we return to 14 Bb4, 14 Bxa6 Nxa6 15 h4 reaches another typical position: for example, 15 ... Rf7
16 h5 Rc8 17 c3 Nb8 18 h6 g6 19 0-0 Rc4 20 Nh5 Qc8 21 Nf6+ Kh8.

189
 Once again, White looks to have made great progress, but how does he break through? 22 Qh4 Ng8 23
Rfe1 Qf8 24 g4? (24 a4 is Fritz’s preference, but it is not clear what it is doing next after 24 ... Nc6) 24 ...
fxg4 25 Qxg4 Qc8 and Black was fine in I.Almasi-L.Psakhis, Vienna 1996. White’s game subsequently
collapsed with surprising speed: 26 Nxg8 Kxg8 27 Rac1 Nc6 28 Bg5 Rf5 29 Bf6 Qf8 30 Kg2 Qxh6 31 Rh1
Qf4 32 Qxf4 Rxf4 33 Rh6 Nxd4 34 Rch1 Rxc3 0-1.
14 ... Rf7 15 h4 Bxd3 16 Qxd3 Nbc6

Question: This is all starting to look very familiar!

Answer: Yes, the pattern of the game is much as usual. Black will bring his rook to c4 and pressurise the d4-pawn, whilst White attempts to
break through on the other wing. Chances are about equal. The rest of the game is worth seeing, though, as we get a classic Petrosian exchange
sacrifice.
17 Rh3 Rc8 18 Rg3 Nd8 19 h5 Rc4 20 h6
We have seen this scenario several times already. It appears that White is making decisive progress on the
190
kingside dark squares, but appearances are deceptive.
20 ... Nec6 21 Nh5 g6

22 Nf6+ Rxf6
Forced, of course, but all part of the plan. Black’s knights and central pawns will dominate the position,
while the white rooks prove ineffective, because of the lack of open files. We have seen several similar
examples in this book already, and it is a major positional theme of the Winawer (and indeed, the French
more generally).
23 exf6 Nf7 24 Qd2 Rxd4 25 Rd3 Rh4 26 Rh3 Rg4 27 Kf1 Nd6 28 Re1
28 f3 is possibly better, but White was presumably reluctant to cut his rook off on h3.
28 ... Kf7 29 Bc3 Ne4 30 Qd3 Nc5 31 Qd1 Rc4 32 Bb2 b5 33 Qe2 Qd6 34 Kg1 Ne4 35 Rd3 Qc5 36 Rc1
e5

The centralized black force is overwhelming and makes a striking visual impression.
37 Qe3 d4 38 Qe2 Nxf6 39 Rdd1 Nd5 40 Qd2 e4
191
 White is slowly pushed off the board.
41 Qg5 Nc7 42 Rd2 Ne6 43 Qh4 a5 44 Rcd1 Rxc2 45 Rxc2 Qxc2 46 Rc1 Qxb2 47 Rxc6 d3 48 Ra6
Qd4 0–1

Question: So do you recommend the 6 ... Qc7 move order?

Answer:
It depends on one’s taste. As the examples above show, it generally leads to a more strategical battle than the 6 ... Ne7 lines. Visually, the
positions arising after 7 Qg4 f5 8 Qh5+ g6 looks rather ugly, because of the weak dark squares, but once one shakes off the visual impression
and looks at the detail of the variations, one finds that Black can usually cover the weak squares with ... h6, and he has good play on the
queenside. I think 6 ... Qc7 is perfectly viable, as an alternative to the main lines with 6 ... Ne7 or the Portisch-Hook line with 6 ... Qa5.

192
Game 20
E.Najer-A.Korobov
Czech League 2011

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Ba5

A rather different interpretation of the Winawer than we have seen so far. Black retains his dark-squared
bishop and maintains the pin on the knight at c3. This move was invented by Botvinnik in the 1940s, and has
been especially associated in more recent years with the Armenian Grandmasters, Vaganian and Lputian.
Black does actually have one other viable move here, namely 5 ... cxd4. This was actually how he used to
play in the very early days of the Winawer, in the late 1920s, but the line quickly went out of fashion after the
discovery of 6 axb4 dxc3 7 Nf3! (7 bxc3 Qc7 is satisfactory for Black).

193
 Question: What is the point?

Answer: White’s idea is to sacrifice a pawn, to accelerate his development. This proved so dangerous that 5 ... cxd4 has been virtually
banished from master practice ever since, but in the 1990s, the Ukrainian GM, Vereslav Eingorn, revived it (Eingorn is a specialist in offbeat
French lines). The only thing most books quote is a game J.Nunn-V.Eingorn, Reykjavik 1990, where Black was whupped, but John Watson
gave the whole 5 ... cxd4 system a closer look in his Dangerous Weapons: The French volume, and suggests that improvements may be
possible. Interested readers should consult that volume, but briefly, the critical position arises after 7 ... Ne7 8 Bd3 Nd7 9 0-0 Nc6 10 Re1 Nxb4
11 bxc3 Nxd3 12 cxd3 0-0 13 Ra4 f6 14 Rg4 and now, instead of Eingorn’s 14 ... Nxe5!?, Black should investigate Watson’s suggested
improvement 14 ... Qc7!.
6 b4

Question: Crikey! What’s this?

Answer: This dynamic response was suggested by Alekhine, I believe, and is overwhelmingly the most popular line of play for White. With

194
the aid of a pawn sacrifice, he breaks the pin at once.
White does have a few alternatives:
a) 6 Qg4 mainly transposes to other lines: for example, 6 ... Ne7 7 Qxg7 (7 dxc5 Bxc3+ 8 bxc3 reaches
variation ‘c’) 7 ... Rg8 8 Qxh7 cxd4 9 b4 dxc3 10 bxa5 transposes to our main game.
b) 6 Bd2 is similar to 5 Bd2, but here, a simple equalizing line is 6 ... Nc6 7 Nb5 Nxd4! 8 Nxd4 (8 Bxa5
Qxa5+ 9 b4 Qb6 10 Nxd4 cxd4 11 Nf3 Ne7 12 Qxd4 Qxd4 13 Nxd4 Ng6 is equal) 8 ... cxd4 9 Bb5+ Bd7 10
Bxd7+ Kxd7! 11 Nf3 Bb6 with equality.

The king looks strange on d7, but White cannot really get at it.
c) 6 dxc5 looks like 5 dxc5, but here, the threat of 7 b4 limits Black’s options by forcing him to take on c3
at once. 6 ... Bxc3+ 7 bxc3 Ne7 (7 ... Qc7 8 Nf3 Nd7 9 Bd3 Nxc5 10 0-0 Bd7 was also satisfactory for Black
in M.Chandler-S.Lputian, Hastings 1986/87) 8 Qg4 and now Psakhis gave as simplest 8 ... Nd7 9 Nf3 (9
Qxg7 Rg8 10 Qxh7 Nxe5 was fine for Black in S.Reshevsky-M.Botvinnik, Radio match 1946) 9 ... Qc7 10
Qxg7 Rg8 11 Qxh7 Nxe5 12 Qh5 Nxf3+ 13 Qxf3 Bd7 14 Bf4 Qxc5 “with a promising position”.
6 ... cxd4

Question: But why not take the other way?

Answer: After 6 ... cxb4 7 Nb5 White has a very strong attack: for example, 7 ... Nc6 8 axb4 Bxb4+ 9 c3 Be7 10 Bd3 a6 11 Qg4 Kf8 12
Ba3 (Psakhis), and White has tremendous compensation for the pawn.

195
 7 Qg4
The parting of the ways. The text is the sharpest, and currently poses the 5 ... Ba5 variation its sternest
theoretical challenge.

Question: It all looks pretty odd – with an enemy piece

attacked and his own piece en prise, White ignores both!

Answer: It is a very sharp approach, but of course, we are already familiar with the general idea from other Winawer lines. Black must
always reckon with this queen raid, attacking his vulnerable spot on g7. The alternative is 7 Nb5, which leads to a rather less frantic game. Play
continues 7 ... Bc7 8 f4 and now the main line is 8 ... Bd7.

Question: That is a little surprising. What about “knights before bishops”?

Answer: Well, of course, the entire Winawer is a rebuke to that particular opening shibboleth! But the text is a very logical move in the

196
circumstances. Black wants to eliminate the annoying knight on b5, which hangs over his head in rather Damocletian fashion, and at the same
time, Black would also get rid of his traditionally bad French bishop.
White usually replies 9 Nf3.

Question: Doesn’t it make more sense to take the other bishop?

Answer: Prima facie, one would think so, yet in practice, 9 Nxc7+ is less popular here, and seems to pose Black fewer problems. After 9 ...
Qxc7 we reach an important position.

Question: What is going on?

Answer: We have another typical Winawer battle, albeit with some differences caused by the position of the white b-pawn. Black has lost
his dark-squared bishop, and has the usual issues on the dark squares, whilst his extra pawn is clearly only a temporary thing, as White will
soon recapture on d4. However, Black is ahead in development, and, most obviously, the advance of the white b-pawn has left him with
weaknesses down the c-file. Black’s remaining bishop is going to come to a4, attacking the backward c2-pawn, and generally exercising
unpleasant pressure on the white queenside. Then Black will develop his remaining pieces, castle short, and bring a rook to the c-file. Practice
suggests he has a decent game:
a) 10 Bb2 Ba4 11 Bd3 Nc6 12 Nf3 Qb6 13 Nxd4 Nge7 14 Qd2 a5! (an unpleasant blow to the white
queenside, which reveals another dark side to the advance b2-b4) 15 Bc3 with a further divide:

197
 a1) 15 ... 0-0 16 b5 Nb8 17 Ne2 Nf5

Question: What is the assessment?

Answer: Black is fine here. He needs to ensure his bishop on a4 does not get trapped or shut completely out of play, and he also need to
fight for the d4-square, to keep the white dark-squared bishop under control, but if he does these things, he should have no problems. The game
remained balanced after 18 h4 h5 19 Rh3 Nd7 20 Rb1 Qc7 21 Bxf5 exf5 22 Bd4 Rfc8 23 Rg3 Nf8 until White lost his head completely and lost
in just two more moves: 24 e6?! Nxe6 25 Be3? Qxc2 0-1, J.Friedel-V.Belov, Moscow 2008.
a2) 15 ... axb4 16 axb4 Kd7! was the highly creative course of the game M.Bluvshtein-S.Lputian, Mallorca
Olympiad 2004.

Question: What is Black playing at?

Answer: Lputian decides that his king is actually safer on d7 than it would be on the kingside. In French positions, with the centre pawns
locked together in the traditional fashion, this is frequently the case. After 17 Nxc6 Qxc6 18 0-0 Bb5 (once again this exchange of bishops) 19

198
Bd4 Nf5 20 Bc5 b6 21 Bxb5 Qxb5 22 c4 Qxc4 23 Bxb6 Rhc8 Black was fine, and eventually went on to win.
b) Another high-level encounter went 10 Nf3 Ba4 11 Bd3 Ne7 12 Bb2 Nbc6 13 Qe2 a6 14 0-0 Qb6 15 Qf2
Bb5 (this manoeuvre is worth noting; Black exchanges off White’s important light-squared bishop, after
which the black king will feel much more secure on the kingside) 16 Nxd4 Bxd3 17 cxd3 Rc8 18 Rfc1 0-0 19
Nxc6 was agreed drawn in P.Svidler-R.Vaganian, Moscow 2004.
Returning to 9 Nf3 and after 9 ... Bxb5 10 Bxb5+ Nc6 (10 ... Nd7 has also been played, aiming to keep the
c-file open for the black rooks; however, the text is more common, since Black can defend the d4-pawn and
make it more difficult for his opponent to recover that pawn) 11 Bb2 (11 0-0 Nge7 12 Bd3 Nf5 13 Bxf5 exf5
14 Bb2 Qd7 15 Kh1 0-0 16 Nxd4 Nxd4 17 Bxd4 Rfd8 was also perfectly satisfactory for Black in Ni Hua-
S.Lputian, Internet 2004, although he eventually lost a long game) 11 ... Nge7 12 Bd3 Bb6 13 0-0 a6 14 Kh1,
as in E.Schmittdiel-R.Vaganian, German League 2004, 14 ... Rc8 looks fine for Black.
All in all, it is clear that 7 Nb5 does not pose him any serious problems.
Now back to the position after 7 Qg4:

7 ... Ne7

Question: Is 7 ... Kf8 possible here?

Answer: It has not been played much at GM level, but both Botvinnik and Lputian have ventured it, if very occasionally (only once in each
case, as far as I can see). The move looks as playable as the similar move in other lines, such as the Portisch-Hook Variation. For example, 8
Nb5 (8 bxa5 dxc3 9 Nf3 Ne7 10 Bd3 Nd7!? 11 Qb4 was a bit better for White in W.Unzicker-M.Botvinnik, Amsterdam 1954, but here 10 ...
Nbc6 11 0-0 Qxa5 looks more natural) 8 ... Bb6 9 Nf3 Nc6 10 Bb2

199
 10 ... f6 (pugnacious, but simply 10 ... Nge7 11 Nbxd4 Nxd4 12 Nxd4 Qc7 looks reasonable for Black) 11
Nbxd4 Nxe5 12 Nxe5 fxe5 13 Nxe6+ Bxe6 14 Qxe6 Qf6 15 Qxf6+ Nxf6 was equal in N.Firman-S.Lputian,
Sochi 2005.
All things considered, 7 ... Kf8 looks as though it may be worth a closer look, especially if Black’s
theoretical problems in the main line persist.
8 bxa5 dxc3 9 Qxg7 Rg8 10 Qxh7

Question: So what is going on here? It looks as

though we are heading for another total mess!

Answer: That is absolutely right. We reach another position similar to the main line seen in Games 16 and 17. The main difference is that
Black’s pawn has already arrived on c3, somewhat earlier than in the other line. The effect of this is that Black is not threatening a queen check
on c3 in some lines, which allows White to develop his knight more actively on f3.
10 ... Nbc6 11 Nf3
200
 The alternative is 11 f4, which leads to positions quite similar to the main line Winawer. For example, 11
... Qxa5 12 Nf3. Here we see the difference referred to in the previous note. In the main line Winawer, the
threats of a check on c3 force this knight to develop to e2, where it is less active and does not threaten to raid
Black’s position with Ng5.

Question: Yes, but on the other hand, Black is usually a pawn

down in many of those normal Winawer lines, because his
c3-pawn drops off. Here, he has material equality.

Answer: Well spotted – glad to see you are paying attention!

Play may go 12 ... Bd7 13 Rb1.

Question: What about 13 Ng5 in this position? Isn’t that rather dangerous?

Answer: It is critical, but then Black has the interesting idea 13 ... 0-0-0!? 14 Nxf7 Nf5.

Question: What’s this? Isn’t he just losing the exchange?

Answer: He is sacrificing it, yes. After 15 Nxd8 Qxd8 he has a large development lead, in exchange for his material investment.
I once sat next to an extraordinary game between the English IM, Andrew Kinsman, who was an expert on
the 5 ... Ba5 Winawer, and Swedish GM and theoretician, Thomas ‘The Hitman’ Ernst. Both players rattled
out their moves to this point, and continued playing at express speed: 16 Qh3 Ncd4 (in poker terms, Black is
‘all in’) 17 Qxc3+ Kb8. Ernst now rattled out the move 18 a4 (18 g3 Qb6 is unclear), at which point Kinsman
thought for the first time in the game. He eventually played 18 ... Qh4+ and the game continued 19 g3 Rxg3
20 hxg3 Qxh1

201
 21 Kf2? (21 Ra3 is equal, according to Fritz) 21 ... Qh2+ 22 Bg2 Ne2! 23 Qf3? (by now even the relatively
best 23 Qa3 Nexg3 is much better for Black) 23 ... Nfd4 and Ernst resigned in T.Ernst-A.Kinsman, Gausdal
1995. The whole game lasted about 45 minutes!
Going back to the position after 15 ... Qd8, it is interesting to look at things with the computer. It starts off
by assessing the position as winning for White, but the longer you leave it, the lower its evaluation sinks.
White’s 16th and 17th moves are its first choice, yet once one reaches the position after 17 ... Kb8, the
assessment drops to zero! All in all, this is a fascinating variation, and one which is typical of many lines of
the 5 ... Ba5 Winawer. Black frequently invests an exchange for dynamic counterplay, relying on his active
minor pieces and the exposed white king.
Returning to 13 Rb1. Black goes 13 ... 0-0-0 14 Qd3 d4.

Question: It is very similar to the lines seen in Game 17.

202
 Answer: Yes, and the general remarks made there apply here too. I think chances are balanced. For example, 15 g3 Kb8 (another Vaganian
game went 15 ... Qd5 16 Bg2 Qa2 17 Rb5 a6?! 18 Rc5 Kb8 19 0-0 Na7?! 20 Nxd4 and White was much better in A.Grischuk-R.Vaganian,
Izmir 2004, but 17 ... b6 looks a better choice) 16 Bg2 Bc8 (we have seen this manoeuvre in the main line Winawer too; Black plans ... b6 and
... Bb7) 17 Ng5?! (17 Rb5 Qa4 is unclear) 17 ... b6! 18 Bxc6 (a huge concession, but 18 Nxf7 Ba6 19 Qe4 Rd5 is tremendous for Black) 18 ...
Nxc6 19 Nxf7 Ba6! (once again, a black exchange sacrifice) 20 Nxd8 Rxd8 21 Qe4 d3! 22 Be3 (22 Qxc6 d2+ 23 Kf2 Bb7 wins) 22 ... dxc2 23
Rc1 Qb5 24 Rxc2 Bb7 25 Kf2 Nd4 0-1 K.Kulaots-R.Vaganian, Mallorca Olympiad 2004.
We now return to Najer’s 11 Nf3:

11 ... Qc7

Question: Why doesn’t Black take the a5-pawn?

Answer: The trouble with 11 ... Qxa5?! is that it expends a vital tempo, and allows White to develop his forces aggressively with 12 Bd3
Bd7 (12 ... Rxg2 is bad after 13 Ng5) 13 0-0 0-0-0 14 Bg5 (Psakhis) and Black’s game is difficult. We see the key problem here – if Black does
not pressurize the e5-pawn sufficiently, the enemy dark-squared bishop is free to develop great activity from g5.
12 Bf4
Here again, we see White exploiting the fact that he has been able to develop his knight to f3. Thanks to
this, he can defend the e5-pawn with pieces, avoiding the need to shut his important dark-squared bishop in
with f2-f4.
12 ... Bd7 13 a6!
This move had been known since the 1960s, but it was Volokitin’s revival of it that has placed the whole 5
... Ba5 Winawer under something of a theoretical cloud in recent years. Previous practice had gone 13 Bd3 0-
0-0 14 Bg3 Qxa5 (now Black is able to take on a5, even though he has lost a tempo in so doing) 15 0-0 Rh8
(this led to two outstanding victories for Khalifman, but Vaganian has also played 15 ... Nf5 which also
seems adequate: for example, 16 Bxf5 exf5 17 Qxf7 Rdf8 18 Qh7 f4 19 Bxf4 Bf5 20 Qh4 Rh8 21 Qg3 Rhg8
22 Qh4 Rh8 23 Qg3 Rhg8 24 Qh4 ½-½ A.Grischuk-R.Vaganian, Mallorca Olympiad 2004), with a divide:
a) 16 Qg7 d4 17 Qg4 Nf5 18 Rfb1 Nxg3 19 Qxg3 Qc7.

203
 White has achieved nothing at all, and is now outplayed in impressive style: 20 Kf1 Rhg8 21 Qf4 f5 22 h3
Kb8 23 Re1 Bc8 24 Ng5 Rd5 25 Nf3 Ne7 26 Rab1 Ng6 27 Qh2 Qh7 28 Rb4 Nh4 29 Nxd4? (29 Nxh4 Qxh4
was necessary, although here too, Black is for choice) 29 ... Nxg2 30 Reb1 Qd7 31 Ne2 Rxd3! (it just had to
happen! this exchange sacrifice is more immediately decisive) 32 cxd3 Qxd3 33 Rc1 c2 34 Rd4 Qf3 35 Nf4
b6! 36 Nxg2 Ba6+ 37 Kg1 Bb7 0-1 J.Polgar-A.Khalifman, Hoogeveen 2000.
b) In the same event, another Khalifman game went 16 Qxf7 Rdf8 17 Qg7 Rhg8 18 Qh6.

A.Galkin-A.Khalifman, Hoogeveen 2004, continued 18 ... Rxf3!.

Question: Another long-term exchange sacrifice?

Answer: Indeed. They are a feature of such positions. Play went on 19 gxf3 Nd4 20 Qf4 Nef5.

Question: What has Black got?

204
 Answer: Despite being an exchange and a pawn down, Black has decent practical chances, because his knights are just so much more active
than the enemy rooks. It is the same phenomenon we have seen many times in this book – the lack of open files reduces the effectiveness of the
rooks.
21 Kh1 Bb5 22 Rfd1 Bxd3 23 Rxd3 Qc5 24 Qc1 Kb8 25 Qd1 Rc8 (Black is extremely calm, not thrashing
about in the search for an immediate tactical solution, but just playing normal moves, on the basis that the
formal material count is irrelevant – his knights are at least the equivalent of the white rooks; in addition, the
white bishop on g3 is actually a very poor piece, having no mobility at all) 26 Bf4 Nc6 27 Qe1 Qc4 28 Bg5
Ncd4 29 Qd1 Qa4 30 Ra2 Qc4 31 Ra1 Qa4 32 Ra2 Nb5 (refusing the draw by repetition) 33 f4 Qe4+ 34 f3
Qa4 35 Qe1 d4 36 h4 Nc7.

Black just strengthens his position methodically, and now stands better. White now blundered with 37
Qe4?? and resigned without waiting for the reply, but he was already in trouble. It is again noteworthy how
the computer spends almost the whole game claiming a clear advantage for White, and approving of almost
all of his moves, yet at some point, it becomes clear that White has nothing.
These two great wins by Khalifman are well worth studying carefully, as they illustrate so many of the
themes of the 5 ... Ba5 Winawer.
Returning to the critical 13 a6:

205
 13 ... 0-0-0

Question: Why not just 13 ... b6 keeping the queenside closed?

Answer: The trouble with that is that, by comparison with the lines examined in the previous note, White has just secured his extra pawn,
and turned it into a thorn in Black’s side. Having the moves a6 and ... b6 included must be a significant benefit to White. Having said that,
though, even then the position is not so clear, especially in practice. The English IM, Thomas Rendle, for example, scored a notable win against
a GM opponent, after 14 Bg3 0-0-0 15 Bd3 Rh8 16 Qxf7 Rdf8 17 Qg7 Rhg8 18 Qh6 Rxf3 (all just as in the Khalifman textbook!) 19 gxf3 Nd4
20 Qf6 Nef5 21 Kf1 Qc5 and, once again, the silicon monster claims a white advantage, but in reality, it is all just totally unclear.

Black went on to win in S.Del Rio Angelis-T.Rendle, Gibraltar 2005.

14 Qd3
A strange choice. The logical follow-up is 14 axb7+ Kb8 (14 ... Qxb7?! leaves the black king exposed after
15 Qd3; Volokitin claims a decisive advantage, but even if that is a touch exaggerated, Black is not to be
envied: for example, 15 ... Qb2 16 Qd1!, as in D.Bryson-D.Wagener, Leon 2001) 15 Qd3.

206
 Now the key Volokitin game continued 15 ... Rg4 16 g3 Ng6 17 Qxc3 Nxf4 18 h3 Nxh3 19 Rxh3 Nxe5 20
Qxc7+ Kxc7 21 Nxe5 Re4+ 22 Kd2 Rxe5 23 Bd3 and White was just clearly better in A.Volokitin-S.Lputian,
Mallorca Olympiad 2004.
Volokitin himself suggested Black players should look into 15 ... d4, with the point that 16 Nxd4? Nxd4 17
Qxd4 Bb5 18 Qe4 Bxf1 19 Kxf1 Nf5 leaves Black with some initiative, although even here, three pawns is a
lot. However, practice has suggested Black still has problems: for instance, 16 Bg3 (Psakhis gives 16 Rb1
Rg4 17 Bg3 Nf5 18 Be2 as “also looks great for White”) 16 ... Nf5 17 Qc4 Nce7 18 Rb1 and White was
better in J.Moreno Carnero-A.Jerez Perez, Sanxenxo 2004.
All in all, Black is in some trouble in this entire variation. 13 ... b6 looks relatively best, but it is a
significant concession to leave the a-pawn alive. It is easy to see why the 5 ... Ba5 system is experiencing a
theoretical crisis at present.
14 ... Rg4 15 Bg3?!
White could still transpose into Volokitin-Lputian, by means of 15 axb7+ Kb8 16 g3.
15 ... Re4+ 16 Kd1 Na5 17 axb7+ Kb8
Now the position is much less clear. The poor position of the white king gives Black full compensation for
his material deficit.
18 Be2 Nf5 19 Re1 Qc5 20 Rb1 Ba4

207
 White is already defenceless against the various threats, notably 21 ... Nd4.

Question: It seems amazing that his position should collapse so quickly!

Answer: Maybe, but this is in the nature of the whole variation. White does not develop normally in this line, and with his king in the
centre, accidents can always happen. We saw the same thing repeatedly in the context of Games 16 and 17 – even when White is objectively
better in such positions, it only takes one error to bring the house crashing down around him.
21 Qa6 Qxa3 22 Bb5 Qa2 23 Rc1 Bxc2+! 24 Rxc2 Qb1+ 25 Rc1 c2+ 0-1

Question: So what is your verdict on the 5 ... Ba5 system?

Answer: It is very interesting. One merit is that positional lines, such as 7 Nb5, give White very little, so he has to enter the extremely sharp
main line with 7 Qg4, if he wants to play for an advantage. Such wild lines offer great practical chances for the black player, whatever their
objective theoretical problems, and many white players are not comfortable in such positions.

Question: But is it any good for Black?

Answer:
That is a harder question to answer. At the time of writing, Black is in some trouble in this main line with 13 a6, and he desperately needs an
improvement. As noted above, his best practical chance is probably 13 ... b6, and just play as in Del Rio Angelis-Rendle, but it is not pleasant
to have to allow White to keep his pawn on a6. If this does not hold water, maybe Black could look into the rare 7 ... Kf8, especially as a
surprise weapon.

208
Game 21
A.Karpov-J.Nogueiras
Rotterdam 1989

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 Bd2

Question: We have not seen this before! White

wants to prevent his pawns being doubled?

Answer: Yes. This is a move first popularized by Bogoljubow. It is not terribly ambitious, but is one of White’s most solid responses to the
Winawer.
White’s other major alternative here is 5 dxc5.

209
 This was brought to prominence after Reuben Fine used it to administer a severe hiding to Botvinnik at
AVRO 1938.

Question: What is the idea?

Answer: White abandons any attempt to establish a pawn centre and instead places his faith in rapid development. It is more dangerous than
it looks at first sight, and over the past 25 years, English GMs Joe Gallagher and Julian Hodgson have both used it with considerable success.
After 5 ... Nc6 (the main line, but among other options, Psakhis likes the solid set-up 5 ... Qc7 6 Nf3 Nd7 7 Bf4 Ne7 8 Bd3 Nxc5 9 0-0 Bxc3 10
bxc3 Nxd3 11 Qxd3 Bd7, which was an old game J.Gallagher-L.Hansen, London 1987; this looks like a good choice, especially if Black
wishes to keep more play in the position than in the “Knott endgame” below) 6 Nf3 Black has:
a) The main line is 6 ... Nge7 7 Bd3 d4 8 a3 Ba5 9 b4 Nxb4 10 axb4 Bxb4 11 0-0 Bxc3 12 Rb1.

Question: What is happening here? It looks very messy.

Answer: It is. Black has an extra pawn, but White’s attacking chances should not be underestimated. More to the point, theory runs very

210
deep here (Watson’s main line includes a suggested black improvement on move 19, for example), which may not be to every black player’s
taste, given that this is a sideline.
b) 6 ... d4 is a move barely covered in the books, but which has been played several times by English FM
and French aficionado, Simon Knott. It results in early simplification and an endgame, where Black seems to
be holding his own without undue problems: 7 a3 Ba5 8 b4 dxc3 (the point; White cannot avoid the exchange
of queens) 9 bxa5 Qxd1+ 10 Kxd1.

Question: What do you make of this position?

Answer: White is temporarily a pawn ahead, but a5 is dropping, of course. White has the two bishops and the usual potential pressure on
the dark squares, but his pawns are less secure than Black’s. All in all, a typical Winawer position. To my knowledge, Knott has reached this
position three times in rated games, and never lost, including one game against Joe Gallagher, the principal expert on the 5 dxc5 line: 10 ...
Nxa5 (10 ... Nge7 was twice chosen by Knott in later games, but looks less good after 11 Bb5) 11 Nd4 Bd7 12 Ne2 Rc8 13 Be3 Ne7 14 Nxc3
Nf5 15 Bd2? (15 Bf4 Rxc5 16 Ne4 is equal) 15 ... Rxc5 16 Ne4 Rxe5 17 Bd3 b6 18 g4 Bc6 19 f3 was J.Gallagher-S.Knott, London 1986, and
now 19 ... Nd4 is much better for Black.
5 Qg4 is also possible and was a favourite in his youth of English GM Mark Hebden.

211
 After 5 ... Ne7 White has two main options: 6 dxc5 (6 a3?! is well met by 6 ... Qa5!) and now Simon
Williams’ forcing suggestion 6 ... Bxc3+ (instead, 6 ... Nbc6 allows considerable complications after 7 Qxg7)
7 bxc3 Qa5 8 Bd2 Ng6 looks like a simple and economical way for Black to play; 6 Nf3 was Hebden’s
favourite, but then the simple sequence 6 ... cxd4 7 Nxd4 Ng6 8 Bd3 0-0 9 Bxg6 fxg6! once served me well
against the man himself.
Returning to Karpov’s 5 Bd2:
5 ... Ne7 6 Nb5
The alternative is 6 a3 Bxc3 7 Bxc3, insisting on the bishop-pair.

Question: This looks like a success for White. He has the two bishops, without
having incurred the usual mangled queenside pawns. Isn’t he better here?

Answer: No, not really. His centre cannot be maintained, and Black has easy development. The only problem for Black is that his best and

212
simplest path to equality is also rather drawish, so if he wants to play for a win, he must take slightly more risks. 7 ... Nbc6 8 Nf3 cxd4 9 Nxd4
(9 Bxd4 avoids the simplification, but offers nothing after, for example, 9 ... Nf5 10 c3 Bd7 11 Be2 Rc8 12 0-0 0-0, V.Savon-V.Hort, Skopje
1968) 9 ... Nxe5 10 Nxe6 Bxe6 11 Bxe5 0-0 12 Bd3 Nc6 13 Bg3 Qf6 has been played a great deal, and is really very equal.
7 ... b6 is the usual way to try to keep some play in the position. A recent example by a top player was 8
Bb5+ (the typical device to avoid a bishop exchange after ... Ba6) 8 ... Bd7 9 Bd3 Nbc6 10 Nf3 Ng6 11 0-0 0-
0 and Black had no problems in D.Mastrovasilis-A.Shirov, Porto Carras 2011. Instead, 8 b4 Qc7 9 Nf3 was
played in the famous game R.Nezhmetdinov-M.Tal, USSR Championship 1957, and now, instead of Tal’s 9
... Nd7, a clean solution is 9 ... cxb4 10 Bxb4 a5 (Uhlmann’s 10 ... Ba6 is also fine) 11 Bd2 0-0 12 Bd3 Ba6
with equality, as in J.Friedman-B.Gulko, Philadelphia 1993.

That game ended in short order, after White had a rush of blood: 13 Bxh7+?? Kxh7 14 Ng5+ Kg8 15 Qh5
Qxc2. Oops! White had obviously missed this. He resigned after 16 g4 Qd3.
6 ... Bxd2+ 7 Qxd2 0-0

8 dxc5

213
 This is a major crossroads. White has two main approaches – either exchange off the central pawns, as in
the game, relying on piece occupation of d4, or else aim to build a pawn centre, starting with 8 c3. The
intention is to follow up with f4.

Question: Is this good?

Answer: Practice suggests that neither plan poses Black any great problems. Building the pawn centre with c3 and f4 involves delaying
White’s development, and Black is able to attack the centre with all means at his disposal, such as ... f6, ... Nc6, ... Qb6, ... Nf5, etc. He should
always secure adequate counterplay.

Question: But what about the knight on b5? That is going

to entrench itself on d6, where it looks very strong.

Answer: Actually, it is no real threat. The knight can easily be challenged by ... Nc8 or ... Nf5, whereupon it will usually have little choice
but to exchange itself off.
Black has various sensible replies to 8 c3. Watson quotes an effective treatment from a recent game: 8 ...
Bd7!.

Question: Can’t White win a pawn with 9 dxc5

followed by a later b4, to defend the pawn?

Answer: He can, but after 9 ... Bxb5 10 Bxb5 Nd7 he must first surrender his bishop by 11 Bxd7 Qxd7 and then after 12 Nf3 Qc7 13 b4
Black has the effective (and thematic) pawn sacrifice 13 ... b6!. After 14 cxb6 axb6 he has excellent play on the queenside, against the
backward a- and c-pawns. Such pawn sacrifices often occur after dxc5 in this line, and practice shows that Black generally has good
compensation.
The alternative is 9 Nd6 cxd4 10 cxd4 Qb6 11 f4 Nbc6 12 Nf3 Nc8 (just as outlined above; the proud
knight on d6 is exchanged off almost immediately) 13 Nxc8 Raxc8.

214
 Question: How do you assess this position?

Answer: Black is just better. He is ahead in development and has serious counterplay against d4 and down the c-file. White’s pawn centre is
more of a liability than a strength. Play continued 14 Bd3 g6 (eliminating Bxh7+, and so threatening to take on d4) 15 Bb1 Rc7 16 0-0 Rfc8 17
Rd1 Nb4 with a comfortable advantage in B.Yildiz-N.Zhukova, Tbilisi 2011.
8 ... Nd7
8 ... Nbc6 is also possible, with the same pawn sacrifice idea of ... b6 that was seen in the last note.
However, the text seems simplest. Two pawns are attacked, so Black gets his pawn back at once, and the
Sicilian structure gives him typical queenside play, plus the possibility of breaking against the centre with f6.

9 f4?!
This is dubious, and a rather optimistic choice by Karpov, who was usually so careful about creating
weaknesses in his own position. Now e4 is weakened irretrievably, and Black’s knight will immediately take
aim at this juicy outpost.
215
 9 Nf3 is more cautious, but gives Black good play after 9 ... Nxc5 10 Bd3 Qb6 11 Rb1 Bd7 12 Nbd4
Nxd3+ 13 cxd3 Rfc8 14 0-0 Rc7 15 Rfc1 Rac8 16 Rxc7 Qxc7, with easy equality in H.Spangenberg-
O.Panno, Pena City 1996.
9 Qc3 is probably the best move.

Question: Doesn’t that make it hard for Black to regain his pawn?

Answer: Not really. He has easy equality after 9 ... a6 10 Nd6 Qc7 11 Nf3 (11 b4? a5 undermines the white queenside structure) 11 ... Qxc5
12 Qxc5 Nxc5, as in P.Garbett-I.Ibragimov, Turin 2006, the only drawback being that the position is rather dull. If Black is seeking a more
dynamic game, he can try, among others, 9 ... f6 10 exf6 Nxf6 which also offers comfortable equality, whilst retaining more play in the
position. For example, 11 Bd3 Nc6 12 f3 e5 13 0-0-0 Be6 14 Ne2 and now the thematic 14 ... b6 ensured him good play in C.Peptan-D.Poldauf,
Groningen 1995.
9 ... Nxc5 10 Nd4
10 Bd3 Nxd3+ 11 Qxd3 Ng6 12 g3 Qa5+ 13 c3 Bd7 14 Nd4 f6 15 Ngf3 fxe5 6 fxe5 b5 gave easy equality
in J.Meister-A.Yusupov, Osterburg 2006.
10 ... Qb6 11 0-0-0 Bd7 12 Ngf3 Rfc8

216
 Question: So what is the assessment here?

Answer: Black is very comfortable. He has a typical Sicilian pawn structure, with counterplay on the open c-file, whilst White has not
really got going on the other flank. In addition, White always has to worry about a possible enemy knight penetration on e4.
13 Qe3 Rc7 14 Kb1 Rac8 15 Rc1 a6 16 g4 Nc6 17 h4 Ne4 18 Rh2 Na5
Both sides are playing very logically, but it is clear that Black is dictating. White’s kingside pawn advance
is as much about clearing the second rank, so his rook can defend laterally, as it is about attacking the black
king.
19 Bd3 Nc5 20 c3 Na4 21 Ka1 Bb5 22 Bb1!?
22 h5 is more aggressive.
22 ... Nc4 23 Qe1

23 ... Ncxb2!

217
 Black exchanges two pieces for rook and pawn, but exposes the white king. Materially, the balance is
about even, but positionally, the exposed white king gives Black the advantage.
24 Rxb2 Nxb2 25 Kxb2 Be2+ 26 Ka1 Bxf3 27 Nxf3 Qa5 28 Nd4 Qa3 29 Nb3?
29 Ne2 was essential. Now Black establishes a winning advantage.
29 ... b5
The pin on the c-file means that the c3-pawn is doomed.
30 f5 b4 31 Rc2 Rxc3?!
31 ... bxc3 was more convincing.
32 Rb2 a5 33 Qd2 a4 34 Nd4 R3c4
34 ... Rc1 35 Rxb4 R1c4 was safer.
35 fxe6 fxe6?
And now 35 ... Qc3 was essential.
36 Nxe6

Suddenly, the tables have turned completely, and the black king comes under a decisive attack.
36 ... Qc3 37 Qxd5 Kh8 38 Ng5 Rf4 39 Ne4 Qc1 40 Rxb4
40 Rc2! would have ended the game at once, but Karpov’s move is still winning.
40 ... a3 41 Rb3 1-0

A narrow escape for Karpov and a tragedy for his opponent. However, it is clear that the opening was a success for Black, and he has nothing
really to fear from this 5 Bd2 system.

218
Game 22
A.Chistiakov-T.Petrosian
Moscow Championship 1956

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 b6

Question: I thought Black always played 4 ... c5 here?

Answer: This move initiates an altogether different interpretation of the Winawer. Rather than counterattack in the centre with ... c5, Black
adopts a much slower, more strategical plan – he intends to keep the position closed and develop within his own three back ranks. The move 4
... b6 has two ideas: either to exchange the bad bishop with ... Ba6, or else to fianchetto it with ... Bb7. Either way, Black usually holds back his
c-pawn for a considerable time.

Question: It looks rather passive.

Answer: There is indeed a danger that Black will find himself too cramped and passive in such lines. They are certainly not everyone’s cup
of tea. These lines had a vogue in the 1960s and early 1970s, with Petrosian being the past master. Bronstein and the young Ulf Andersson also
won many impressive games with such lines. This inspired others to take up the system, but in many cases, disappointment awaited these new
converts – they found that the resulting positions did not seem anything like as easy to play as Petrosian and others made them look! Lev
Psakhis, for example, is one of the world’s greatest French experts, but he comments, a propos the 4 ... b6 lines, in his book The Complete
French that “if this system were judged purely on the results of my games, it would have to be prohibited by law, so low is my score with it”.
4 ... Qd7 is another way to introduce the same sort of approach.

219
 Question: Wow! That looks a very strange move!

Answer: The idea is similar to the 6 ... Qc7 Winawer we have already seen. Black wants to be able to meet 5 Qg4 (other white moves may
well transpose to 4 ... b6, since the black queen usually goes to d7 in such lines anyway, so as to facilitate long castling; with the pawn still on
c7, the queen lacks another good square on the second rank: for example, 5 Bd2 b6, etc) with 5 ... f5 when his queen defends the g7-square.
Another classic Petrosian victory, over a very powerful opponent, is worth seeing, as it shows how
helpless even world-class players can sometimes look in these positions: 6 Qg3 b6 7 h4 Bb7!? (rather than
exchanging bishops, Black just fianchettoes and prepares long castling) 8 Bd3!? (8 Bd2 Nc6 9 Nf3 0-0-0 10
0-0-0 followed by Ne2-f4, was suggested by Petrosian biographer, P.H. Clarke) 8 ... Nc6 9 Nge2 0-0-0 10
Bd2 Nh6 11 a3?! (Petrosian clearly did not fear 11 Bxh6, but that is what Keres nonetheless recommended, to
be followed by 11 ... gxh6 12 Qe3, g3 and Nf4) 11 ... Be7 12 Bb5 (12 Qxg7? Ng4 leaves the white queen in
trouble) 12 ... Rdg8 13 Qd3 Nf7 14 0-0-0 Kb8 15 Nf4 Qc8!.

220
 Question: What is this all about?

Answer: It is typically deep Petrosian manoeuvring. He just unpins his knight on c6 and prevents Ba6, exchanging off the bishop that
defends his king.
Play continued 16 Nce2 Ncd8 17 Qb3?. A world-class player (double Candidate in 1959 and 1962!) is so
bamboozled by Petrosian’s in-depth manoeuvring that he just drops a pawn. Even so, after 17 c3 c5 18 Kb1
(Clarke), Black simply stands better on both flanks.

Question: So where did White go wrong?

Answer: It is hard to pinpoint a clear error. Rather, White has just made a series of superficially natural moves, ones which do not fit
together terribly well, whereas Black’s byzantine-looking moves turn out to harmonise perfectly. Such was often the experience of Petrosian’s
opponents in these positions! However, the thing to ask yourself, before you rush to take up these ... b6 systems, is “Can I play like Petrosian?”.
F.Olafsson-T.Petrosian, Bled 1961, concluded 17 ... c6! 18 Bd3 c5 19 dxc5 Bxc5 (two pawns are hanging,
and Petrosian goes on to win at a canter) 20 Nh3 Nxe5 21 Bf4 Ndf7 22 Bb5 Ka8 23 Nd4 Ng6

24 Qa4 Bxd4 25 Bd7 Qf8 26 Rxd4 e5! 27 Rb4 exf4 28 Rxb6 Nfe5 29 Rxb7 Kxb7 30 h5 Qd6 31 hxg6
Qxd7 32 Qxf4 Nxg6 0-1. An incredibly impressive advert for these ... b6 lines, but as the note to move 17
emphasises, it is a case of caveat emptor!
Returning to the immediate 4 ... b6:
5 Qg4
As usual, this is the most direct response, attacking g7, but here, it may not be best. Another standard anti-
Winawer option is 5 a3. The game G.Kasparov-V.Ivanchuk, Horgen 1995, a classic black victory which we
cannot ignore, then continued 5 ... Bf8.

221
 Question: Uh? Why not take on c3?

Answer: Black could take on c3, but the retreat to f8 is a typical idea in such lines. Black is playing very slowly and solidly. He keeps his
dark-squared bishop, defends g7, and prepares to develop the rest of his pieces with moves such as ... Bb7 (or ... Ba6), ... Nc6, ... Qd7, ... Ne7
and ... 0-0-0.
Kasparov played to seize kingside space, 6 Nf3 Ne7 7 h4, and now Black has:
a) 7 ... h5 fights for space, but concedes White an outpost on g5: 8 Bg5 Ba6 9 Bxa6 Nxa6 10 0-0 Qd7 11
Qd3 Nb8 12 Ne2.

Question: What is the idea?

Answer: White gets ready to advance c4. This is a typical plan in such positions. White has more space, but Black has got rid of his bad
bishop and has no real weaknesses. White needs to open some lines, in order to get at Black’s position.

222
 Following 12 ... Nbc6 13 c4 dxc4 14 Qxc4 Nd5 (the cost of White’s c4 advance has been to give Black an
outpost on d5 and a potential target on d4) 15 Rac1 Nce7 16 Bxe7 Bxe7 17 Nc3 0-0-0?! (17 ... c6 was better,
although White is still slightly better after 18 Nxd5 cxd5 19 Qc7) 18 Nxd5 exd5 19 Qa6+ Kb8 20 e6! Qc8 (20
... Qxe6? 21 Rxc7! Kxc7 22 Qxa7+ Kd6 23 Qxb6+ Kd7 24 Qb5+ Kd6 25 Ne5 wins, and if 20 ... fxe6 21 Ne5)
21 Qxc8+ Kxc8 22 exf7 and White was clearly better in A.Sokolov-L.Portisch, Moscow 1990.
b) Thus Ivanchuk preferred 7 ... h6 8 h5 a5.

Question: All these pawn moves! Why don’t the players develop their pieces?

Answer: This type of strategy is typical of the whole variation. Because the position is closed, rapid development is less of a priority than
normal, and more consideration can be given to longer-term strategical manoeuvring. The text is a typical move. Black intends ... Ba6,
exchanging bishops, and both sides are trying to out-tempo each other – Black wants to delay ... Ba6 until White has moved his king’s bishop,
whilst White is trying to avoid moving the bishop, hoping that any exchange of bishops will occur without him having wasted a tempo on a
move such as Bd3. Black’s 8 ... a5 is a waiting move, as part of this process, which also gains queenside space.
Kasparov continued 9 Bb5+.

Question: I thought you said White was trying not to move this bishop?

Answer: I did and he was, but he now recognizes that he lacks any other really useful waiting moves, so he switches his plan. He initiates a
manoeuvre, similar to what we saw in the lines with ... Qc7 and ... b6 – White gives a check on b5, hoping to upset the coordination of the
black pieces.
9 ... c6 10 Ba4

223
 This is the key. Now the exchange of bishops is avoided.

Question: Yes, but surely the white bishop is miserably passive where it stands now?

Answer: True, but White intends in the medium term to move his knight from c3, advance c2-c3 and then drop the bishop back to c2. That
way, his bishop finally reaches its desired diagonal, whilst avoiding the bishop exchange. The downside, however, is that the black bishop also
gets rather a nice diagonal, cutting through the white position.
10 ... Nd7 11 Ne2 b5 12 Bb3 c5
At last, Black makes the standard central pawn break. This is frequently delayed much longer even than
this.

Question: Why?

Answer: Well, as explained above, all the slow manoeuvring and delayed piece development in this line is often justified all the while the
position remains closed, but if it opens up, the lack of development can prove fatal. Black therefore usually has to be very careful not to allow
the game to open up too quickly, and ... c7-c5 tends to do that; hence, this central pawn break has to be delayed.

Question: So why does Ivanchuk play it here?

Answer: He judges that, in this position, his development lag is not so great, and so he can afford to commence pressure against the white
centre. It is the need for such careful and precise judgements that makes these slow ... b6/Qd7 systems so hard to handle as Black.
13 c3 Nc6 14 0-0 Qc7

224
 Question: How is this position?

Answer: I think White is objectively somewhat better. As usual in this line, he has more space, and I see nothing special on the black side to
compensate for this. However, in the further course of the game, Kasparov is outplayed completely, and loses one of his worst-ever white
games.
15 Re1
15 Nh2, intending f4-f5, was Tischbierek’s suggestion, but the following line shows that this plan is not so
easy to carry out: 15 ... Ba6 16 f4?! b4 (threatening 17 ... Bxe2, winning the d4-pawn) 17 Be3 bxc3 18 bxc3
Rb8 19 Bc2 Be7 20 f5 cxd4 21 cxd4 Bg5! and White has some problems. This is a typical example of how
difficult it can be to exploit White’s space advantage in these lines.
15 ... c4 16 Bc2 Nb6 17 Bf4 Be7 18 Bg3 Rb8

Question: This is all a bit obscure!

225
 Answer: Black is scratching around for something to do. Castling short is clearly out of the question (just Bb1 and Qc2 would then be very
strong), and it is just not clear what he should do. He therefore manoeuvres around, waiting for White to reveal his plans.
19 Nh2 Qd8 20 Ng4
20 Nf4 intending Qg4, looks better.
20 ... b4 21 axb4 axb4 22 cxb4?!
This looks strange. 22 f4 seems more logical.
22 ... Nxb4 23 Bb1 Bd7 24 b3?
Scratching a spot usually makes things worse. He should leave the queenside alone.
24 ... Ra8 25 Rxa8 Qxa8 26 bxc4 Nxc4

27 Nc1?
Completing a miserable collapse. 27 Bf4 is better, but by now Black has an obvious advantage anyway.
27 ... Ba4 28 Qe2 Qa7!
Amazingly, White simply cannot defend the d4-pawn.
29 Ne3 Qxd4 30 Nxc4 dxc4 31 Qf1 0-0 0-1
Kasparov had had enough. A shockingly bad game by the 13th World Champion, but one which points up
some of the characteristics of this line. Although objectively White reaches a middlegame where he must be
better (he just has more space for no compensation), it can be extremely hard to get to grips with the rather
slippery black position. Some white players seem to find such positions very hard to play.
Thus far, we have seen Black having all the fun, but lest you get the feeling that it is “Black to play and
win” in these lines, here is a more impressive treatment by White: 5 Bd2. This modest move is probably one
of White’s best choices.

226
 Question: It looks innocuous. What is the point?

Answer: The move does several things. Firstly, it prevents the pawns being doubled after a subsequent ... Bxc3+. Secondly, it sets up the
possibility of Nce2, offering the exchange of dark-squared bishops. In general, in this line, White’s queen’s bishop is his inferior bishop, since
the white centre pawns are fixed on dark squares, so an exchange of these bishops should suit him. Once Black has played ... Ne7, for example,
the reply Nce2 may well force the exchange. Finally, 5 Bd2 avoids a premature commitment of the white queen on the kingside, and waits for
Black to commit himself.
Play may continue: 5 ... b6 6 Bb5! (we saw this idea in Kasparov-Ivanchuk, White avoids the exchange of
light-squared bishops) 6 ... c6 7 Ba4 a5 8 a3 Bf8 9 Nce2 Ba6 10 c3 Ne7 11 Nf3 Nf5 12 0-0 Be7 13 Re1 Bxe2
(else the knight will come to f4, leaving the enemy queen’s bishop patrolling an empty diagonal) 14 Qxe2
Na6 15 Bc2 Nc7.

Thus far, we have been following the game J.Nunn-R.Hübner, Brussels 1986.

227
 Question: And what is the assessment? Black looks to be fine.

Answer: In fact, Nunn assesses the position as “extremely unpleasant for Black”. He may appear okay on the surface, but he has no real
counterplay, and his king has no safe place to hide. Black’s only active piece is the knight on f5, and this can always be driven away by g4.
Nunn quickly built up a winning advantage, although a blunder eventually cost him half a point.
Finally we return to Chistiakov-Petrosian, and 5 Qg4:
5 ... Bf8

Once again, this undeveloping move. As we have already said, Black takes the view that the bishop has
done its job on b4, by provoking the white e-pawn to advance to e5, and now he retains the bishop, covering
his dark squares. In this position, he also asks the white queen what she is now doing on g4 – as we will see,
the queen often proves a target here, for the onrushing black pawns.
6 Nf3 Qd7 7 a3 Nc6
Once again, there was nothing particularly wrong with 7 ... Ba6, to exchange off the bad bishop, but as in
the game against Olafsson, Petrosian prefers a more complicated approach, keeping pieces on. As we will
see, a glorious future awaits the supposedly ‘bad bishop’.
8 Be3 Bb7 9 Bb5?!

Question: Why do you mark this as dubious?

I thought this was a standard idea in this line?

Answer: It is a standard idea if Black is trying to exchange bishops by ... Ba6, but here, Black has already committed himself against that
plan. As Clarke notes, White has no way of reinforcing the pin, so 9 Bd3 would be simpler.
9 ... 0-0-0 10 0-0 Nge7

228
 11 b4?!

Question: This looks sensible, trying to attack on

the queenside. What is wrong with it?

Answer: The problem is that, as Steinitz and Lasker pointed out, one should only attack when one has an advantage that justifies doing so.
Here, despite his extra space, White actually has to be careful, because his queen is misplaced and will soon come under attack from the black
forces on the kingside. Clarke recommends 11 Nd2 Nf5 12 Qe2.
11 ... f6!
Black starts his attack on the enemy centre. Now the knight on f3 is tied to defending e5, so White cannot
get his queen back from the kingside.
12 Rfe1
12 exf6 looks crackers, and I am not surprised that Chistiakov did not play it, but it may be the best try.
After 12 ... gxf6 13 Bd3 Fritz actually assesses the position as slightly better for White, but I do not believe
this. One thing readers should note about these ... b6 systems is that computer programs (much like many
human players of the white pieces!) tend to overrate the white position, because of the space advantage. In
this instance, the open g-file will surely come back to haunt White, in the medium term.
12 ... Nf5 13 Qh3 h5 14 g3
14 g4? fails to 14 ... hxg4 15 Qxh8 Bxb4 winning.

229
 14 ... a6 15 Ba4?
The bishop has no future on b3. Both 15 Bd3 Nxe3 (15 ... g5 is also good) 16 fxe3 (16 Rxe3 g5!) 16 ...
fxe5, and 15 Bxc6 Qxc6 16 Bd2, as given by Clarke, are somewhat better, but still pretty awful for White.
15 ... g5! 16 g4
Forced, else 16 ... g4 will win the d4-pawn and demolish the white position, but now the white monarch’s
residence suffers serious structural damage.
16 ... hxg4! 17 Qxh8 gxf3! 18 Qh5
18 ... Bxb4 is a big threat, of course, and 18 Bxc6 Qxc6 19 Qxf6 Be7 20 Qg6 and now the simple
prophylactic move 20 ... Kb8! (Clarke) ensures that e6 will not fall with check, and so 21 Nd1 Qxc2 is
curtains: 22 Qxe6 Qe4.
18 ... b5 19 Nxb5
Desperation, since the bishop is dead wood on b3.
19 ... axb5 20 Bxb5 fxe5 21 Kh1 Qg7

230
 22 Bxc6
The only way to hold his centre together, but now the ‘bad’ bishop will rule the light squares.
22 ... Bxc6 23 dxe5 Be7
23 ... d4 24 Bxg5 Be7 is also winning easily, but Petrosian does not even give White the g5-pawn!
24 b5 Bb7!
The b5-pawn, on the hand, he scorns – the queen’s bishop belongs on the long diagonal.
25 Qg4 Rh8
The rest is a horror show.
26 Qg1 d4 27 Bd2 g4 28 Bf4 g3

28 ... g3!
Forcing mate.
29 Bxg3 Nxg3+ 0-1
30 Qxg3 Qxg3 31 fxg3 f2+ is a finish to warm the heart of any French Defence player.

Question: That was wonderfully impressive!

These ... b6 systems look highly attractive.

Answer: Well, as I hope I have made clear, I would urge caution on the part of Black. When the systems go well, they look fantastically
impressive – one gets the idea that Black can just shut his eyes, play all these moves like ... b6, ... Qd7, ... Bb7/a6, ... Nc6, ... 0-0-0, etc, and
practically win by force. Needless to say, it is not so easy. At the start of this section, I quoted Lev Psakhis’ words about his own terrible results
in these lines. Another grandmaster who has written of a similar experience is Hans Ree. He tells of how he was blown away by games such as
the above, and took up the ... b6 systems himself as Black, yet found them incredibly hard to handle, and scored badly. There is a very fine line
between solidity and passivity, and it often takes a Petrosian to tell the difference. There really are not many Petrosians around, believe me!

Question: So are you saying “don’t touch these lines as Black?”

Answer: No, I am not saying that. If you like the positions, then give them a try, by all means. As we have seen, many white players also
find them difficult to play against, and you may achieve good results. If you play them and like them, that is great, but I would urge you not to
assume they are as good as Petrosian made them look. The Nunn-Hübner game, quoted in the notes above, is much more typical of the negative
side.

Question: Any final advice?

Answer:

231
Yes. If you do decide to investigate these lines further, then get hold of a copy of the book French Winawer: Modern and Auxiliary Lines, by
Moles and Wicker. It is a wonderful treatment of these systems, and the only decent one I know of. Although published in 1979, which makes
it positively prehistoric for an opening book, that matters little in this case, because the nature of these ... b6 systems means that concrete theory
plays a small role – it is ideas, rather than forcing lines, which dominate.

232
Game 23
R.Fischer-V.Kovacevic
Rovinj/Zagreb 1970

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 a3

Question: What is this? White is sacrificing a pawn?

Answer: Well, yes and no. In fact, as we will see, White can regain the pawn by force. What he is doing is trying to bring about a more
open game than usually occurs after 4 e5. He forces the exchange of bishop for knight, thereby gaining his traditional Winawer birthright of the
bishop-pair and unopposed dark-square bishop, but without closing the centre. He thereby hopes that he will be able to make more effective use
of the bishops.

Question: Does this get played much nowadays?

Answer: No. In recent years, it has lost popularity, but in the past, it was played regularly by Alekhine, Smyslov and Fischer, so it must be
respected. It is one of the sharpest replies to the Winawer, and the black player needs to be well armed against it.
4 ... Bxc3+ 5 bxc3

233
 5 ... dxe4

Question: Is this forced?

Answer: Not forced exactly, but no other move really makes much sense, in my opinion. 5 ... Ne7, for example, hopes to transpose into the
main line, if White plays a rapid e4-e5, but he has no reason to do so. After 6 Bd3 c5 7 exd5 exd5 8 dxc5, Psakhis claims that Black is not
really equalizing, and I agree. The two bishops are likely to be strong in such an open position.
6 Qg4
This is the point of White’s fourth move., With the aid of this typical anti-Winawer queen raid, White
regains his sacrificed pawn.
In recent years, especially at amateur level, the move 6 f3 has achieved a degree of popularity.

White makes it a real gambit, along the lines of the Blackmar-Diemer Gambit (1 d4 d5 2 e4 dxe4 3 Nc3
Nf6 4 f3).

234
 Question: What do you think of the line?

Answer: To be brutally honest, I think it is what is sometimes euphemistically called “a consignment of geriatric shoemakers”! Unlike the
Blackmar-Diemer, here White is not even threatening to capture on e4 (because of the reply ... Qh4+), so Black has a wide choice. 6 ... c5 looks
the most logical, but John Watson’s 6 ... Nd7 is also good. White has nothing for his pawn, in my opinion.
6 ... Nf6 7 Qxg7 Rg8 8 Qh6

This is the tabiya for the variation.

Question: So what is going on?

Answer: White has regained material equality, and has his two bishops, with the dark-squared bishop potentially being quite strong.
However, he has poor development, and it is also not entirely clear where his king will go. It is unlikely to be safe on the queenside, whilst
kingside castling takes a little time, in view of the need to keep the g2-pawn defended. Black has three main moves here, all of which are
equally respectable.
8 ... Nbd7
This is probably the most solid option. Black plans to develop with ... b6, ... Bb7, ... Qe7 and ... 0-0-0.
8 ... Rg6 immediately forces the white queen to determine her position, and prevents the pin on the knight
after Bg5. However, in view of the fact that the queen often finds herself somewhat stranded on h6, it is
arguably not so logical to force her back to a central post so early. 9 Qe3 (9 Qd2 is a major alternative: for
example, 9 ... b6 10 Bb2 Bb7 11 0-0-0 Qe7 12 c4 Nbd7 13 Ne2 0-0-0 J.Dukhin-M.Ustinov, Rybinsk 2001, as
quoted by Watson; Black is fully developed and stands well) 9 ... Nc6 10 Bb2 Qd6?! (10 ... b6) 11 f3 exf3 12
Nxf3 was another Fischer game, played just two rounds before the main game, in the same tournament. It is
worth seeing, as it shows how dangerous White’s dark-squared play can be in this line, if he gets going: 12 ...
Bd7 13 0-0-0 0-0-0 14 c4 Ng4?! 15 Qd2 f5?.

235
 This weakens the dark squares even further. R.Fischer-W.Uhlmann, Zagreb 1970, concluded: 16 d5! Nb8
17 h3 Nf6 18 Ne5 Ne4 19 Qd4 Rg3 20 Nf7 Qf4+ 21 Kb1 c5 22 Qe5 Qxe5 23 Bxe5 Rdg8 24 Bd3 (Fischer
begrudges giving his wonderful dark-squared bishop for the rook!) 24 ... Rxg2 25 Bxe4 fxe4 26 Nd6+ Kc7 27
Nxe4+ Kb6 28 Nf6 Ba4 29 Nxg8 Bxc2+ 30 Kc1 Nd7 31 Rdg1 1-0. A powerful display by Fischer, and an
example of what Black must be wary of in this variation.
Instead, 8 ... c5 9 Ne2 Bd7 (9 ... Nbd7 transposes to the note to Black’s 9th, below) is an interesting line,
recommended by John Watson in his latest French volume. Funnily enough, I saw it recommended by
Andrew Martin in a small pamphlet on the Winawer some 25 years ago (!), and once used it to good effect
myself: 10 Bg5 Rg6 11 Qh4 Bc6

12 Nf4? (naive, to say the least; critical is 12 d5! exd5 13 Nf4 Rxg5 – 13 ... Qa5? 14 Bxf6 defends c3,
which is the point of White’s 12th – 14 Qxg5 Nbd7 15 Be2 Qa5 when Black has a pawn for the exchange,
and a powerful pawn centre, with chances about equal) 12 ... Qa5! (refuting White’s last; the game continued
in amusing fashion) 13 0-0-0? Qxa3+ 14 Kb1 Nd5! (the move White had missed) 15 Nxd5 Bxd5 16 c4 e5!
236
and White did not last very long in J.Hickman-S.Giddins, Bradford 1988.
9 Ne2
Ljubojevic once surprised Korchnoi with 9 Nh3 and scored a striking victory.

Question: What is the point?

Answer: The move has a couple of points. Firstly, it defends g5, so the bishop will be defended additionally if it comes to that square.
Secondly, f2 is defended, which may be important if Black later plays ... Ng4 (a recurring theme in this line, as we will see). Korchnoi’s rather
bad defeat resulted in the move becoming quite popular for a while, but it seems that objectively, Black is fine:
a) Korchnoi chose 9 ... c5 10 Be2 Qa5 11 Bd2! Rxg2? (Ftacnik’s 11 ... cxd4 12 cxd4 Qf5 is probably still
fine for Black, who follows up with ... b6 and ... Bb7) but after 12 Ng5! (revealing another point of 9 Nh3 –
sometimes, the knight can jump to this square, rather than the bishop; the rook on g2 is now in trouble) 12 ...
cxd4 13 Qg7! Rxg5 14 Qxg5 Qxg5 15 Bxg5 he was just worse and went on to lose in L.Ljubojevic-
V.Korchnoi, Tilburg 1986.
b) 9 ... b6 (sticking to the standard plan) 10 Be2 (10 Ng5 is more active: for example, 10 ... Rg6 11 Qh4
Bb7 12 Nxh7 Nxh7 13 Qxh7 Qf6 with an unclear position; White has won a pawn, but still lags in
development, and will have to lose further time with his queen) 10 ... Bb7 11 0-0 Qe7 led to an impressive
black win in G.Lane-J.Tisdall, Gausdal 1987: 12 a4 0-0-0 13 Ba3 c5 14 Qh4 Qd6 15 Rfd1 e3 16 f3

237
 16 ... Rxg2+! 17 Kxg2 Rg8+ 18 Ng5 (18 Kh1 Ng4 wins) 18 ... h6 19 Rg1? (19 dxc5 was essential, when
19 ... hxg5 20 cxd6 gxh4+ 21 Kf1 Ne5 is unclear, according to the computer) 19 ... hxg5 20 Qg3 Qd5 21 Kf1
Nh5 22 Qg2 Nf4 23 Qg3 Nxe2 24 Kxe2 Qc4+ 0-1.
9 ... b6
9 ... c5 is also worth considering, and has been the choice of several very strong GMs in recent years.

Now:
a) 10 a4 Qc7 (10 ... b6 11 a5 Bb7 12 dxc5 bxc5 13 Ng3 Ng4 14 Qf4 Qf6 with equality, was an old game
M.Basman-R.Vaganian, Hastings 1974/75, but it seems a little odd to give the white a-pawn a target like this)
11 dxc5!? Qxc5 12 Qd2 Nb6 13 a5?! (13 Qd4 Qxd4 14 cxd4 Bd7 is equal) 13 ... Nbd5 with an unclear
position, where Black’s active pieces and superior pawn structure at least compensate for his dark-square
weaknesses, D.Andreikin-N.Vitiugov, Saratov 2011.
b) 10 h3 is an important option.

238
 Question: What is the point?

Answer: White stops the knight raid ... Ng4, and also prepares an extended fianchetto of his king’s bishop with g4 and Bg2. 10 ... Qa5 11
Bd2 b6 (11 ... Qa4 was a later Ponomariov choice: 12 Ng3 b6 13 dxc5?! bxc5 – 13 ... Rg6 14 Qh4 Nxc5 looks strong – 14 Be2 Ba6 15 0-0 Rg6
16 Qh4 Bxe2 17 Nxe2 Qxc2 and Black was doing very well in I.Nepomniachtchi-R.Ponomariov, Moscow (blitz) 2010) 12 c4 Qa4 13 Nc3
Qxc2 14 Nb5 cxd4 15 Bb4 was B.Jobava-R.Ponomariov, Warsaw (blitz) 2010.

Now the computer’s 15 ... a5! looks better for Black, the point being 16 Nd6+? Ke7! 17 Nxc8+ Kd8 when
Black has a winning advantage.
c) Psakhis quotes one of his own games, K.Georgiev-L.Psakhis, Sarajevo 1986, which continued 10 g3 b6
11 Bg2 Bb7 12 0-0 Qe7 13 a4 Ng4! 14 Qf4 f5, and concludes that “the opening has ended quite pleasantly for
Black”.
10 Bg5 Qe7
This move conceals several points. As well as preparing ... Bb7 and ... 0-0-0, it also prevents White from
castling long, and sets a tactical trap.
11 Qh4
11 Ng3? would walk into the trap: 11 ... Ng4! 12 Bxe7 Nxh6 and the bishop on e7 is trapped: 13 Bh4 (or
13 Bb4 a5) 13 ... Rg4.
11 ... Bb7
Black develops according to the now familiar plan. He intends long castling.
12 Ng3

239
 12 ... h6!
White was threatening 13 Nh5.
13 Bd2

Question: Can’t White take the pawn?

Answer: Taking with the queen is impossible, because of 13 Qxh6?? Ng4! (this tactic is a recurring theme in this line) 14 Bxe7 Nxh6 and
the white bishop on e7 is again trapped: 15 Bh4 (if 15 Bb4 a5) 15 ... Rg4. He can take the other way, but after 13 Bxh6 Rg4 14 Qh3 0-0-0, his
queen is rather awkwardly placed. Fischer tries to avoid this, but his move also fails to pose Black any serious problems.
13 ... 0-0-0

Question: How can we assess this position?

Answer: Black has a fine game. He has completed his development, his pawn on e4 is securely defended, and the white queen is somewhat
vulnerable on h4. Objectively, the game is no more than equal, as White has certain trumps too (the dark-squared bishop, most notably), but in
practice, White’s position is probably the more difficult to play.
14 Be2 Nf8!
Underlining the slight vulnerability of the white queen.

240
 15 0-0!?
15 Nh5 was suggested by Suetin, and is probably better, although not with Suetin’s suggested follow-up:
15 ... N8h7?! (15 ... N8d7 16 Ng3 is probably best play) 16 Bf4 (White should capture on h6) 16 ... Rxg2 17
Be5 Nd5 (17 ... Rg5! 18 Nxf6 Nxf6 19 Qxh6 Rg6 20 Qh4 Rdg8 also gives Black a decisive advantage) 18
Qxe7 Nxe7 19 Bg3, which he described as “enlivening White’s pieces”. However, 19 ... e3 is just winning for
Black: for example, 20 Rf1 Ng5 21 fxe3 Nf5 22 Bf4 Ne4.
15 ... Ng6 16 Qxh6 Rh8 17 Qg5
17 Qe3 looks more sensible. After 17 ... Nd5 18 Qg5 Nxc3 19 Bh5 Qxg5 20 Bxg5 Rdg8 the position is
unclear/equal. Of course, one must bear in mind that, with the white pieces against a much weaker player,
‘equals’ was not what Fischer was looking for. It is also alleged that the Soviet players in the tournament
(Smyslov, Korchnoi, and Petrosian) had prepared Kovacevic before the game, and even went so far as to feed
him moves during it, so Fischer’s defeat is perhaps not as shocking as at first it might seem.
17 ... Rdg8 18 f3!
Setting a nice trap.

241
 18 ... e3!!
A brilliant riposte. 18 ... Nh4 looks strong, but then there follows 19 fxe4!! Rxg5 20 Bxg5 and White is at
least equal.
19 Bxe3
19 Qxe3 is no better after 19 ... Nd5 20 Qf2 Qh4.
19 ... Nf8 20 Qb5 Nd5 21 Kf2
21 Bd2 a6! 22 Qd3 Qh4 wins.
21 ... a6 22 Qd3 Rxh2
The attack is overwhelming.
23 Rh1 Qh4 24 Rxh2 Qxh2 25 Nf1 Rxg2+ 26 Ke1 Qh4+ 27 Kd2 Ng6!

The Yugoslav master continues to conduct the attack with great verve.
28 Re1 Ngf4 29 Bxf4 Nxf4 30 Qe3 Rf2 0-1
Precise to the very end. There is no defence to the threat of 31 ... Nxe2.

242
Game 24
A.Stripunsky-Y.Shulman
US Championship 2010

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 Nge2

Another of the more popular fourth move alternatives.

Question: It looks a bit clumsy! What is the idea?

Answer: White simply wishes to prevent his pawns being doubled. Given the chance, he will follow up with 5 a3, and force the enemy
bishop either to retreat, or to take on c3, in which latter case, White will recapture with the knight. However, it is true that the move is rather
clumsy, and clogs up White’s development. For that reason, although it has always enjoyed a degree of popularity, especially at amateur level,
it does not pose Black any serious problems.
4 ... dxe4
This is the most natural move, of course, but it could be argued that, to some extent, it plays into White’s
hands. An alternative approach is to argue that the knight is poorly placed on e2, and that Black should
therefore just carry on developing, rather than assisting White in unravelling. After 4 ... Nf6 (even 4 ... Nc6
has been tried, with a similar philosophy in mind, but this seems a little extravagant) White has:
a) 5 Bg5 transposes into a line of the McCutcheon, which is considered harmless, after 5 ... dxe4 6 a3 Be7
7 Bxf6 Bxf6 (7 ... gxf6 is also possible) 8 Nxe4 0-0, etc.

243
 Question: This looks like a Burn French!

Answer: Indeed, it is quite similar to the line 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Be7 6 Bxf6 Bxf6. In fact, White has gained a tempo over the
Burn, since he has played both Ne2 and a3, whilst Black has only played ... 0-0.

Question: Does that mean White has an improved version?

Answer: Not really, because both of White’s moves are sub-optimal. His king’s knight would much rather be on f3, and the pawn being on
a3 rather than a2 is of no real value. Indeed, if White were to castle long, as he usually does against the Burn, it could even be argued that a3 is
a weakness, which may make it easier for Black to open lines on the queenside.
b) 5 e5 Nfd7 has not been played that much over the years, but in recent times, it has been used by GM
French experts such as Moskalenko and Rustemov. For example, 6 Qd3 (6 a3 Be7 7 f4 0-0 8 g3 c5 9 Be3 Nc6
10 Bg2 b6 11 0-0 Ba6 was comfortable for Black in H.Asis-V.Moskalenko, Montcada 2006) 6 ... 0-0 7 a3
Be7 8 Qg3 Kh8 9 Be3 c5 10 0-0-0 Nc6 11 f4

244
 11 ... b5! 12 Nxb5 (12 dxc5 b4 13 axb4 Nxb4 also gave Black a nice initiative in J.De la Riva Aguado-
V.Moskalenko, Sitges 2006) 12 ... Rb8 13 Nec3 c4 gave Black very dangerous play in R.Perez-A.Rustemov,
Spain 2002.
All in all, 4 ... Nf6 looks like a perfectly respectable way for Black to play, and also has the merit of
avoiding the quite heavily-analysed main lines. Given that 4 Ne2 is often played by weaker players, this may
be an attraction for the higher-rated black player, looking to get his opponent ‘out of book’.
5 a3 Bxc3+
A major crossroads. The text is the critical move, aiming to hang on to the gambit pawn.
5 ... Be7 is a safer way to play. Black returns the pawn, but retains the bishop-pair, and again reaches a
type of Burn French structure, where the white knight on e2 is less than optimally placed. After 6 Nxe4 Nf6
the main line is 7 Qd3.

Question: That looks a little strange. What is the point?

245
 Answer: White brings his queen to an aggressive position and prepares long castling. Even so, Black is rock solid, and has no great
problems: for example, 7 ... Nbd7 (7 ... 0-0 8 Bf4 Nd5 9 Bg3 Nc6 10 c4 f5 11 Nd2 Nf6 12 0-0-0 e5 13 d5 f4 14 Bh4 Ng4 was unclear and
satisfactory for Black in D.Campora-B.Gulko, Oropesa del Mar 1996) 8 Bf4 0-0 9 0-0-0 b6 10 N2c3 Bb7

11 Nxf6+ Nxf6 12 Be2 Nd5 13 Nxd5 Bxd5 14 Bf3 Bg5 15 Qe3 Bxf4 16 Qxf4 Bxf3 ½-½ E.Ghaem
Maghami-V.Korchnoi, Port Erin 2004, was another solid treatment by a super-GM. The only issue for Black
in this whole 5 ... Be7 line is generating winning chances if White does not do anything crazy.
6 Nxc3 Nc6
Hanging on to the pawn with the greedy 6 ... f5? has been discredited ever since the famous miniature
A.Alekhine-A.Nimzowitsch, Bled 1931, which saw Black splattered after 7 f3! exf3 8 Qxf3 Qxd4? (this is
larceny on a grand scale; the computer may claim that Black is objectively okay after 8 ... Nc6 but it is still
hard to recommend this plan to Black in good conscience) 9 Qg3 Nf6 10 Qxg7 Qe5+ 11 Be2 Rg8 12 Qh6
Rg6 13 Qh4 Bd7 14 Bg5 Bc6

246
 15 0-0-0 Bxg2 16 Rhe1 Be4 17 Bh5 Nxh5 18 Rd8+ Kf7 19 Qxh5 1-0.
7 Bb5
7 Be3 is an interesting gambit that was played a number of times by the late Dutch IM, Johan van Mil.

Question: What is the idea?

Answer: White simply develops with Qd2 and 0-0-0, and keeps his bishop-pair. Johan won some nice games with it, although objectively, 7
... Nf6 8 Qd2 Ne7! 9 Bg5 Ned5 is fine for Black.
7 d5 is another option. White trades central pawns and regains his sacrificed pawn, with a symmetrical
structure, in which he still has the bishop-pair: 7 ... exd5 8 Qxd5 Be6 9 Qxe4 (9 Qxd8+ Rxd8 is fine for
Black, because ... Nd4 will be annoying) 9 ... Nf6 10 Qh4. Even so, he seems to have no objective advantage
after 10 ... Bf5!
7 ... Nge7

247
 8 Be3
Although in the present game, it makes no difference, White usually flicks in 8 Bg5 here, and only after 8
... f6 does he play 9 Be3.

Question: Why is that?

Answer: His argument is that Black’s position is slightly weakened by the inclusion of the move ... f7-f6. If Black follows up with ... f6-f5,
it transposes anyway and makes no difference, but if Black prefers the plan with ... e6-e5 (see the next note), then White hopes the presence of
the pawn on f6 will harm Black slightly.

Question: Why doesn’t White simply recapture his pawn on e4?

Answer: There is nothing wrong with the move, but it offers no advantage: 8 Nxe4 0-0 9 c3 e5 and Black is comfortable. The text is more
ambitious.
8 ... 0-0 9 Qd2 f5
9 ... e5 is a perfectly good, if less ambitious, alternative.

White has:
a) 10 d5 Nd4! 11 Bc4 (11 Bxd4? exd4 12 Qxd4 c6 13 Bc4 Nf5 is better for Black and 14 Qd2? e3! 15 fxe3
Qh4+ even wins) 11 ... Nef5 12 0-0-0?! (12 Nxe4 b5 13 Ba2 Bb7 is comfortable for Black) 12 ... Nd6 13 Ba2
N4f5 14 h3 Nxe3 15 Qxe3 f5 and Black was simply much better in S.Garcia Martinez-W.Uhlmann, Leipzig
1983.
b) 10 dxe5 Qxd2+ 11 Bxd2 Nxe5 12 Nxe4 Bf5 is just equal, despite the white bishop-pair, as he cannot
avoid a weakness on e4: 13 f3 (not 13 Ng3? Bxc2 14 Rc1 Bd3) 13 ... Bxe4 14 fxe4 a6 15 Be2 N7c6 with
equality, K.Mokry-R.Knaak, Bratislava 1983.

248
 Question: But why doesn’t the bishop pair give White the better ending?

Answer: Not only is the e4-pawn weak, but, more importantly still, the black knights have such a great, secure central outpost on e5.
Usually, when the bishop-pair outplays knights in the endgame, it is because the knights lack a secure central outpost. When they have one, as
here, the knight is not inferior to the bishop.
Even with the move ... f7-f6 included (8 Bg5 f6 9 Be3 0-0 10 Qd2), 10 ... e5 is still perfectly good here,
and, indeed, it is Watson’s first choice for Black in his latest French volume.

After 11 Bc4+ (other moves do not really try to exploit the move ... f7-f6 – compare the note to 9 ... f5 in
the main game) 11 ... Kh8 12 d5 Nd4 13 0-0-0 Bg4! 14 Rde1 c5 15 dxc6 Nexc6 16 h3 (16 Nxe4 Rc8 17 Kb1
Na5 is equal, as given by Watson) 16 ... Bf5 17 g4 Bg6 18 Rd1 Rc8 White has nothing for the pawn,
A.Tukhaev-M.Rodshtein, Herceg Novi 2005.
Going back to the position after 9 ... f5:

249
 Black decides to keep the gambit pawn, which is the most ambitious continuation.

Question: So isn’t Black just a pawn up here?

Answer: White is not entirely without compensation. He has the bishop-pair, with the dark-squared bishop being especially valuable here,
given the way the black central pawns are all on light squares. Of course, as Winawer players, we should be used to seeing this as Black, but
here, the position is more open than usual (there is no fixed e5-d4-c3 vs. e6-d5 pawn chain). White will play f3, trying to open the g-file. With
accurate play, Black is fine, but he needs to be careful – one inaccuracy and things can go very wrong for him, as I found to my cost a couple of
times when playing the black side.
10 0-0-0
The immediate 10 f3 is also possible. Then 10 ... exf3 (10 ... f4 11 Bxf4 Qxd4 12 fxe4 e5 13 Be3 Qxd2+
14 Bxd2 Nd4 is a comfortable equalizer, Y.Solodovnichenko-M.Feygin, German League 2003) 11 0-0-0?! is
a speculative gambit, which Black can accept or decline: 11 ... e5 (Fritz is having none of it, and assesses 11
... fxg2 12 Qxg2 f4 as “winning for Black” – a little optimistic, perhaps, but he should be much better, even
so) 12 Bc4+ Kh8 13 d5 Nd4 is equal, Y.Solodovnichenko-C.Marzolo, Illkirch-Graffenstaden 2004.
10 ... a6
Widely regarded as best, although 10 ... Nd5 is also possible, and conceals a very nasty trap.

250
 11 Bg5?? e3 12 Bxe3 f4 13 Nxd5 Qxd5 and Black wins a piece, M.Pavlov-W.Uhlmann, Halle 1981.

Question: That looks great! So why is 10 ... Nd5 not the main line?

Answer: The trouble is that, after 11 Nxd5! exd5 12 Bxc6 bxc6 13 Bf4 we reach a position, about which Psakhis writes: “although Black
may not have much chance of losing, he has absolutely no chance of winning; the difference in strength of the bishops is too great”.
11 Bxc6 Nxc6

12 f3
12 Bg5 is the alternative: 12 ... Qd7 13 d5 Ne7 (13 ... exd5 14 Nxd5 Qf7 15 Bf4 Rd8 16 Bxc7 Rd7 was
equal in J.Hector-C.Matamoros Franco, Copenhagen 2002; 13 ... Ne5 also looks perfectly okay for Black) 14
dxe6 Qxe6 15 f3 h6 16 Bxe7 Qxe7 17 fxe4 fxe4 with equality, H.Westerinen-I.Thompson, Gausdal 2006.
12 ... e5
12 ... exf3 13 gxf3 e5 is the alternative equalizer: e.g., 14 d5 Na5 15 Qe2 b5 16 f4 exf4 17 Bxf4 Bd7 18

251
Rhg1 Rf7 with equality in N.Neelakantan-P.Harikrishna, Kolkata 2000.

However, it is worth running this on a few moves, to show some of the dangers that can befall Black in
these lines, if he is guilty of even a moment’s inattention: 19 Rg3 Qf6 20 Rdg1 Nc4 (20 ... Re8!) 21 Qh5
Ne5?? 22 Ne4! (suddenly Black is lost) 22 ... Nd3+ 23 cxd3 fxe4 24 Be5 Qf5 25 Qxf7+ Kxf7 26 Rxg7+ Ke8
27 Rg8+ Qf8 28 dxe4 and White won.
13 d5
13 dxe5 Qxd2+ 14 Bxd2 (14 Rxd2? f4 15 Bc5 e3 is much better for Black) 14 ... exf3 15 gxf3 Be6 is a
well-known equality, seen many times in practice.

13 ... Na5
Another possibility is 13 ... Ne7 14 Bg5 h6 15 Bxe7 Qxe7 16 fxe4 fxe4 17 Nxe4 Bf5, again with equality.

Question: An awful lot of these variations end in equality!

252
 Answer: That is true. This is one drawback of the whole 4 ... dxe4 and 5 ... Bxc3+ variation. It is deeply analysed, and many variations
fizzle out to a fairly dead equality.
14 Qe2
14 Bc5 Nc4 15 Qe2 Nd6 16 fxe4 Qg5+ 17 Kb1 f4 also gave Black no problems in G.Hartmann-
G.Hertneck, German League 1989.
14 ... b6 15 fxe4 f4

Question: So how should we assess this position?

Answer: It is just balanced. White has regained his pawn, and we have opposite-coloured bishops. However, the fact that so many other
pieces are on the board, and the kings are castled on opposite sides, means that, although equal, the position is far from drawn.
16 Bd2 Nb7
The knight heads to the natural blockading square d6.
17 Rdf1?!
The start of a plan of breaking up the black kingside pawn chain with g3, but White had obviously missed
his opponent’s surprise 19th move.
17 ... Nd6 18 g3 Bh3 19 Rf2?
Objectively, the exchange sacrifice 19 gxf4 was better, but White perhaps can be forgiven for missing the
next move.

253
 19 ... Bg2!
Very imaginative! Now Black secures his pawn on f3 and has a clear advantage.
20 Re1 f3 21 Qe3 h6 22 Ne2 Qe8 23 Ng1 Qh5 24 h3 Bxh3?!
A strange decision, which squanders most of Black’s superiority. He would still have a clear plus after 24
... Nc4 25 Qc3 Nxd2 26 Qxd2 Rf7.
25 Rxf3?!
25 Nxh3 Qxh3 26 Rxf3 is better.
25 ... Bg2 26 Rxf8+ Rxf8 27 Bb4 Qg4 28 Bxd6 cxd6
Following the mutual lapses at moves 24-25, Black retains a nice advantage. White’s pieces are passive
and the e4-pawn weak.
29 Kb1
This loses a pawn, but 29 Qxb6 Qxg3 is even worse.
29 ... Qxe4! 30 Qxe4 Bxe4 31 Rxe4 Rf1+ 32 Ka2 Rxg1 33 Rb4
Passive play is no better.
33 ... b5 34 a4 Rxg3 35 axb5 axb5 36 Rxb5 h5

254
 Now Black is winning. His kingside pawns are too many and too fast.
37 Rb8+ Kh7 38 b4 h4 39 Rf8 h3 40 Rf1 e4 41 b5 Rf3 42 Rg1 e3 43 b6 e2 44 Re1 h2 45 Rh1 Kg6 0-1

255
Game 25
A.Bezgodov-A.Kornev
Russian Championship, Krasnoyarsk 2003

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 Bd2

Question: Doesn’t this lose a pawn?

Answer: Well, once again, it is a sharp attempt by White to obtain a more open position than would arise after 4 e5. The move involves a
pawn sacrifice that is at least temporary, and may become permanent.

Question: Is the move any good?

Answer: Objectively, probably not. It has never achieved the level of popularity amongst GMs that 4 a3 has enjoyed, for example. The main
problem here is that White ends up losing his d-pawn.
This is probably a convenient place to round up White’s other fourth moves, that we have not already
covered:
a) 4 Qg4 is the most direct way to implement the traditional Qg4 idea. However, without any preliminary
supporting of the white centre, it comes up short after 4 ... Nf6 5 Qxg7 Rg8 6 Qh6 Rg6 (Watson suggests that
the immediate 6 ... c5! is even better: for example, 7 e5 cxd4 8 a3 Bf8 9 Qxf6 Qxf6 10 exf6 dxc3 11 bxc3
Nd7 and Black has the superior structure, with no attendant compensation for White) 7 Qe3 c5!.

256
 This vigorous attack on the centre is best, and highlights the fact that White’s house is built on sand, rather
than rock. 8 Bd2 Ng4 9 Qd3 Nc6 with enormous pressure against the white centre.
b) Even 4 Qd3 has been played here!

Question: That looks pretty beginner-like!

Answer: It is a rather odd move, to say the least. The main idea is that 4 ... dxe4 5 Qxe4 Nf6 6 Qh4 may see the white queen in an
aggressive position. Even here, Black should have no real problems, but it seems illogical to play into White’s hands like this. Instead, 4 ... Ne7
is a simple developing move, which asks White how he intends to develop his pieces. Typical play then is 5 Bd2 b6, when a top-level
encounter V.Anand-N.Short, Wijk aan Zee 1990 saw Black with no problems after 6 Be2 (6 Qg3!? is proposed by Psakhis) 6 ... 0-0 7 a3 Bxc3
8 Bxc3 a5 9 Nf3 Nd7 10 exd5 exd5.
c) 4 exd5 exd5 is a sub-line of the Exchange Variation.

257
 Question: Isn’t this just very drawish?

Answer: That is its reputation, but it is not entirely deserved. Plenty of practice has shown that both sides can make something of the
position, if they are determined to do so. Indeed, Magnus Carlsen recently used the line successfully as White, and commented afterwards that
he believes it should be possible to fight for a small advantage, even here.
White has:
c1) Larsen’s old move 5 Qf3 has long been known to be well met by 5 ... Qe7+!.

Question: Why is this good? It looks a bit odd.

Answer: The problem for White is that he has no good way to parry the check. Whichever minor piece he puts on e2 will block in the other
minor piece, whilst 6 Be3 also has its problems, as we will see: 6 Nge2 (6 Be3?! leads to problems after 6 ... Nf6 7 Bd3 c5!) 6 ... Nc6 7 Be3
Nf6 and White’s pieces are awkwardly placed and vulnerable to moves such as ... Bg4, and ... Ng4.
c2) 5 Bd3 with a further divide:
258
 c21) 5 ... Nc6 (Black of course has a very wide choice of how to develop here, but the text is probably the
main line) 6 a3 is the nearest thing to a critical line in this variation. White keeps open a path for his queen to
emerge on the kingside. It should not be underestimated by the black player, as White can often create a
dangerous initiative, in the face of inaccurate black play. By contrast, the old 6 Nge2 offers less than nothing:
6 ... Nge7 7 0-0 Bg4 (it is useful to provoke the weakening f3, before exchanging bishops) 8 f3 Bf5 with
equality.
After 6 ... Bxc3+ (6 ... Be7 is a solid treatment, but the text is more normal) 7 bxc3 ...

... now 7 ... Nge7 is the most common, but Black has several alternatives:
c211) 7 ... Qf6 anticipates the white queen’s emergence, as well as preparing ... Bf5, to exchange light-
squared bishops: 8 Ne2 (8 Rb1 Nge7 9 Qf3 Qxf3 10 Nxf3 f6 11 0-0 b6 12 Re1 Kd8 13 Re2 Bf5 was equal in
A.Domont-M.Ulibin, Biel 2007) 8 ... Bg4 9 h3 Bh5 10 0-0 Nge7 11 Qd2 Bg6 12 Nf4 Bxd3 13 Nxd3 0-0-0 14
a4 Na5 and Black again had no real problems, eventually winning a long game, A.Vuilleumier-M.Apicella,
Lausanne 2006.
c212) 7 ... Nce7 is a favourite of the veteran Hungarian GM and French expert, Ivan Farago.

259
 Question: What is the idea?

Answer: Black admits that his queen’s knight is not ideal on c6, and re-routes it to the kingside. He intends ... Nf6 and ... Bf5. Another point
is that, by keeping the king’s knight ready to jump to f6, he deprives White of the move Qh5. Play may go 8 Qf3 (8 Nf3 Bg4 9 Rb1 b6 10 0-0
Nf6 11 Re1 0-0 12 Bg5 Qd6 with a small edge for White was C.Varga-I.Farago, Zalakaros 2008) 8 ... Nf6 9 Bg5 Ne4.

Question: Eh? Doesn’t this lose a pawn?

Answer: It sacrifices one, yes. In compensation, Black exchanges off White’s good bishop and reaches a position with opposite-coloured
bishops, and the better pawn structure. Farago has played this several times, so he clearly believes in it: 10 Bxe4 dxe4 11 Qxe4 f6 12 Bf4 (12
Be3 0-0 13 Ne2 c6 14 Qd3 Bf5 15 Qd2 Be6 16 0-0 Bc4 offered similar compensation in I.Jakic-I.Farago, Bibinje 2006, where Black eventually
won) 12 ... 0-0 13 Ne2 Re8 14 Be3 Nd5 15 Qd3 Qe7 16 0-0 Nxe3 17 fxe3 Qxe3+ 18 Qxe3 Rxe3 with equality, but again, Black went on to win
against a lower-rated opponent in M.Pacher-I.Farago, Austrian League 2008.
c213) 7 ... Nge7 8 Qh5!? reveals the main point of White’s play here. He takes advantage of the fact that

260
Black has committed his knight to e7, to occupy a strong square with his queen, and prevent the exchanging
manoeuvre ... Bf5. Black has to be a little careful here, although he can equalize if he is: 8 ... Be6 9 Nf3 Qd7
10 Ng5 0-0-0 11 Nxe6 (11 Nxf7?? Qe8) 11 ... Qxe6+ 12 Be3 g6 13 Qf3 Nf5 is equal.
c22) There is nothing at all wrong with 5 ... Nf6: for example, 6 Nge2 Bg4 7 0-0 0-0 8 f3 Bh5 9 Nf4 Bg6
10 Nxg6 hxg6 11 Ne2 Re8 12 Bg5 Be7 13 Ng3 Nbd7 14 f4 Nh7 15 Qf3 c6 with equality, M.Carlsen-
F.Vallejo Pons, Bilbao 2012.

Now Magnus got a bit ambitious with 16 h4?! and would have been struggling after 16 ... Qb6!.
c23) 5 ... c6 is another solid option: for instance, 6 Qf3 and now 6 ... Qf6 (6 ... Nf6 keeps more tension) is a
rock-solid approach, which been played twice by Ivanchuk: 7 Bf4 (7 Qxf6 Nxf6 8 Nge2 Nbd7 9 a3 Be7 10 f3
h5 11 h4 Nf8 12 Nf4 Bd7 13 Kf2 0-0-0 led to a quick draw in E.Alekseev-V.Ivanchuk, Biel 2009) 7 ... Ne7 8
Qg3 Nd7 9 Nge2 0-0 10 0-0-0 Ng6 11 Bg5 Qd6 12 Qxd6 Bxd6 with dead equality in J.Moreno Carnero-
V.Ivanchuk, Calvia Olympiad 2004.

261
 Question: But how does Black win such a position?

Answer: By playing on! Ivanchuk went on to win this game, for example, despite playing an opponent rated over 2500. Even such
symmetrical positions can be won, if the player is sufficiently determined and does not just give his opponent a draw.
Returning to 4 Bd2:
4 ... dxe4
This is probably the best move, but it is quite heavily analysed, and players seeking fresher pastures might
like to look at 4 ... Nc6.

Question: What is the point?

Answer: Black continues developing, and also attacks the white d-pawn, which is quite hard to defend, with the bishop stuck on d2:
a) My only experience of this line was a highly enjoyable one, albeit against some rather feeble white play:
5 Bb5?! Nge7 6 e5 Bd7 7 Qg4?! Nf5 8 Bxc6 Bxc6 9 Nf3 Be7! 10 0-0? h5 11 Qf4 Rg8 12 h3 g5 13 Qh2 g4 14
hxg4 hxg4 15 Qh7

and now 15 ... Kd7! and White did not last much longer.
b) 5 Qg4 has also been tried: 5 ... Nf6 6 Qxg7 Rg8 7 Qh6 Rg6 8 Qe3 and now 8 ... Ng4! (Watson) looks
fine (at least) for Black.
c) 5 a3 Bxc3 6 Bxc3 Nf6 7 e5 Ne4 8 Bd3 Nxc3 9 bxc3 Qe7 and Black went queenside, with a perfectly
good position in L.Ljubojevic-U.Andersson, Palma de Mallorca 1971.
5 Qg4
This standard queen raid on the black kingside is the point of White’s play. 5 Nxe4? basically just loses a
pawn after 5 ... Qxd4. Admittedly, Alekhine did go on to beat no less a player than Salo Flohr from this
position, after 6 Bd3 Bxd2+ 7 Qxd2 Qd8 (A.Alekhine-S.Flohr, Nottingham 1936), but in his tournament
book, he admitted that his fifth move had been a “lapsus manus” and that he had intended 5 Qg4.
5 ... Nf6 6 Qxg7 Rg8 7 Qh6

262
 7 ... Qxd4
Grabbing the second pawn is the most logical, and probably objectively best.

Question: Are there any alternatives?

Answer: Yes, 7 ... Rg6 being the principal one. John Watson analyses this in some detail in Play the French 3, and it is a valid way to play,
but taking on d4 strikes me as a simpler and stronger solution.
8 0-0-0

Question: Black seems to be a pawn up!

Answer: Neither your eyes nor your arithmetic can be faulted!

Question: So what has White got in compensation?

Answer: Just as with the 4 a3 lines, he has broken up the black kingside by taking on g7, and he also has hopes that the black queen will
prove exposed on d4. However, there are two crucial differences from the 4 a3 line. Firstly, White does not have any central pawn structure on
c3 and d4. Secondly, and even more importantly, he has not actually forced the exchange of Black’s Winawer bishop, and so his usual hopes of
developing serious dark-square compensation are much less realistic here. Furthermore, Black now has a highly unpleasant retort:
8 ... Bf8!

263
 Question: Undeveloping the bishop? Why is this so good?

Answer: “Reculer pour mieux sauter”, as they say down my local. The bishop retreats from an exposed position (avoiding ideas such as
Nb5, for example), and, most importantly, drives the white queen into a rather poor position. With the black queen on d4, her white opposite
number is unable to retreat to the centre on e3, and so has to retreat down the h-file, where she is much less well placed. I believe this move
was first played by the great David Bronstein, back in 1950.
9 Qh3?!
This is actually not the best move. The critical line is 9 Qh4 when play continues 9 ... Rg4 10 Qh3.

Question: Why is this better than the game?

Answer: White has lured the black rook to g4, where he hopes it will prove exposed. Now play usually continues 10 ... Qxf2, nicking
another pawn, and also threatening to embarrass the white queen.

11 Be2 (in the stem game, White tried 11 Nb5?!, but after 11 ... Na6 12 Kb1 Bd7 13 Be3 Qf5 14 Nd4 Qg6
264
15 Nb3 Nb4 Black had two extra pawns and the initiative, I.Boleslasky-D.Bronstein, 14th matchgame,
Moscow 1950) 11 ... Rh4 (11 ... Rg6 12 g4 Qc5 is a suggestion of Psakhis, which has only been tried in one
or two amateur games, as far as I can see, whilst Watson analyses 11 ... Qxg2 12 Bxg4 Qxg4 13 Qxg4 Nxg4
14 Nxe4 as equal).

Question: Er ... it looks as though White has blundered his queen away?

Answer: In this case, appearances are deceptive! White has the clever trick 12 Qxh4! Qxh4 13 g3 and, lo and behold, the black queen is
trapped!

Question: Very neat! So is White better?

Answer: Alas, no. His ingenuity only brings him an ending which is at best unclear, and probably a bit worse for him after 13 ... Qh6! 14
Bxh6 Bxh6+ 15 Kb1.

265
 Question: So what is going on?

Answer: White has an extra exchange, but Black has two extra pawns and the bishop-pair. His e-pawn(s) are passed, although the e4-pawn
may be a tad vulnerable. There have been a lot of practical tests of this position, which suggest that Black probably has the edge, although an
assessment of ‘unclear’ is perhaps the fairest. A recent game, quoted by Watson, continued 15 ... Nc6 (15 ... Bd7 is equal) 16 Nh3 (16 Nb5 Ke7
17 Nxc7 Rb8 is presumably the idea, when Fritz prefers Black after 18 Nb5 e5) 16 ... e5 17 Nf2 Nd4 and Black was just better in Z.Ivekovic-
S.Martinovic, Zagreb 2011.
9 ... Qxf2
Now Black is simply better, because without the rook on g4, White has far fewer tactical tricks, and cannot
develop his kingside rationally.
10 Nge2 Bd7

11 g4?!

Question: Wow! That looks a bit crazy!

Answer: It is a sign of desperation. Normal moves are not much better: for example, 11 Ng3? e3 (11 ... e5 looks good, but is less convincing
after 12 Ncxe4 Nxe4 13 Nxe4 Bxh3 14 Nxf2 Be6 although even here, Black is just a pawn up) 12 Be1 Qf4 13 Bd3 (13 Kb1 e5 traps the queen)
13 ... Qh6 and after the queens come off, White is two pawns down.
11 Be3 is probably objectively best, but White’s position is cheerless after 11 ... Qf5.
11 ... Rxg4 12 Ng3 e3 13 Be1 Qf4 14 Bg2
14 Be2 Rh4 15 Qf1 Bc6 is just hopeless, so White tries to fish in troubled tactical waters.
14 ... Qh6
Fritz’s 14 ... e2+ 15 Rd2 Bc6 16 Bxc6+ Nxc6 17 Ngxe2 Qh6 looks even more convincing.
15 Qxh6 Bxh6 16 Bxb7 Bc6 17 Bxa8 Bxa8

266
 Question: What is happening here?

Answer: The point of Black’s play is that he will regain the exchange, thanks to the dual threats of ... Bxh1 and ... e2+. At worst, he will
emerge with an extra pawn and what should be a technically winning ending.
18 Rg1 e2+ 19 Rd2 Be3 20 Ncxe2 Nbd7 21 Rf1 Bg2 22 Rf2 Bd5
Rather an amusing picture! I suspect Black was enjoying himself at this point!
23 Nc3 c6 24 Rfe2 Bxd2+ 25 Kxd2 h5
After having his bit of fun, Black has finally taken back the exchange, and is a pawn up. The remainder is
a technical task, which Black carries out without undue difficulty, and which is not of great interest to our
theme, so we will pass through it fairly briefly.
26 b3 Kf8 27 Nf1 Rg1 28 Ne3 Bf3 29 Rf2 Ne5 30 Rf1 Rxf1 31 Nxf1 Bg2 32 Ne3 Ne4+
It seems rather odd to head for opposite-coloured bishops. 32 ... Nf3+ 33 Ke2 Nxe1 34 Kxe1 Bh3 looks
simpler.
33 Nxe4 Bxe4 34 Bg3 Nf3+ 35 Ke2
35 Kc3 offers better defensive chances.
35 ... e5 36 Nc4 f6 37 Ke3 Bd5 38 Nd6 Nd4 39 c4 Be6

267
 40 Bf2 Nf5+ 41 Nxf5 Bxf5 42 b4 a6 43 a4 Bd7
Black’s connected passed pawns in the centre should decide things, providing he is careful not to get them
blockaded on the wrong colour squares.
44 Ke4 Kf7 45 b5 cxb5 46 cxb5 axb5 47 axb5 Ke6
47 ... Bxb5 was good enough, but Black prefers to avoid letting the white king in after 48 Kd5.
48 b6 Bc6+ 49 Ke3 f5 50 Kd3 Kd5 51 Bh4 f4 52 Bd8 e4+ 53 Kc3 e3 54 Kd3 Bb5+ 55 Kc3 Ba6 0-1

268
Table of Contents
Title 2
Copyright 3
About the Author 4
Contents 5
Series Foreword 6
Bibliography 7
Introduction 8
Game 1 Smyslov-Botvinnik (USSR Championship 1944) 11
Game 2 Tolush-Botvinnik (USSR Championship 1945) 20
Game 3 Timman-Korchnoi (Leeuwarden match 1976) 27
Game 4 Van Seters-Korchnoi (Skopje Olympiad 1972) 36
Game 5 Suetin-Uhlmann (Berlin 1967) 41
Game 6 Byrne-Korchnoi (London 1979) 46
Game 7 Byrne-Vaganian (Moscow 1975) 54
Game 8 Ragozin-Botvinnik (Training match 1944) 60
Game 9 Hartston-Uhlmann (Hastings 1972) 71
Game 10 Bogdanovich-Uhlmann (Sarajevo 1965) 77
Game 11 Fischer-Uhlmann (Buenos Aires 1960) 81
Game 12 Felgaer-Korchnoi (Bled Olympiad 2002) 90
Game 13 Vidarsson-Ward (Reykjavik 1998) 98
Game 14 Kanafsck-Atalik (Mar del Plata 2003) 110
Game 15 Inarkiev-Vitiugov (Russian Championship 2008) 121
Game 16 Cheparinov-Grischuk (Baku Grand Prix 2008) 131
Game 17 Kuipers-Stellwagen (Dutch League 2011) 147
Game 18 Atlas-Kindermann (Austrian League 2006) 165
Game 19 Hort-Petrosian (European Team Ch’ship, Kapfenberg 1970) 177
Game 20 Najer-Korobov (Czech League 2011) 193
Game 21 Karpov-Nogueiras (Rotterdam 1989) 209
Game 22 Chistiakov-Petrosian (Moscow Championship 1956) 219
Game 23 Fischer-Kovacevic (Rovinj Zagreb 1970) 233
Game 24 Stripunsky-Shulman (US Championship 2010) 243
Game 25 Bezgodov-Kornev (Russian Championship 2003) 256

269