First

published in 2002 by Gloucester Publishers Limited, Northburgh House,

10 Northburgh Street, London EC1V 0AT

Copyright © 2002 Byron Jacobs

The right of Byron Jacobs 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.

Kindle ISBN: 978-1-78194-060-0

Ebook ISBN: 978-1-78194-061-7

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

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

Everyman Chess Series

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.
Contents
Bibliography
Introduction

1 e4 e6 2 d4 d5

3 Nc3 Bb4: Winawer Variation

1 The Winawer with Qg4
2 The Winawer: Positional Main Lines
3 The Winawer: Other Lines

3 Nc3 Nf6: Classical Variation

4 4 e5: Modern Variation
5 4 Bg5 dxe4: The Burn Variation
6 4 Bg5 Others: Classical and McCutcheon

3 Nd2: Tarrasch Variation

7 The Tarrasch with 3 ... Nf6
8 The Tarrasch with 3 ... c5
9 The Tarrasch: Others

Others
10 The Advance Variation
11 The Rubinstein Variation and Others

Index of Complete Games

Bibliography
Books
Encyclopaedia of Chess Openings volumes A-E (Informator 2001)
Nunn’s Chess Openings, John Nunn, John Emms, Graham Burgess and Joe Gallagher (Everyman Chess
1999)
The French Advance, Tony Kosten (Everyman Chess 1998)
The French Classical, Byron Jacobs (Everyman Chess 2001)
The French Tarrasch, John Emms (Batsford 1998)
The French Winawer, Neil McDonald (Everyman Chess 2000)

Periodicals and Databases

ChessBase (MegaBase 2002)
Informator 1-81
TWIC
Introduction
Over the past few years, the popularity of the French Defence has increased dramatically at all levels.
Hitherto, the most frequently seen defences against 1 e4 were 1 ... e5, leading to the complexities of such
openings as the Scotch Game, the King’s Gambit and, of course, the Ruy Lopez and 1 ... c5 which leads to
the vast theoretical edifice that is the Sicilian Defence. Both 1 ... e5 and 1 ... c5 have drawbacks for the
practical player, who likes to play ‘system-type’ openings. Playing 1 ... e5 can lead to a great variety of
positions and one must be prepared for all manner of outlandish gambit continuations. Playing the Sicilian
nearly always offers White the opportunity to turn the game into a wild fist fight from the earliest stages.
The French Defence is an attractive opening for players who want to start the game with a reasonably
solid structure but are also keen to maintain some dynamism in the position. Even if you don’t want to
play it, the French is an important opening to understand as positions featuring the typical central structure
with White having pawns on d4 and e5 versus black ones on e6 and d5 can arise from numerous other
opening lines. Witness the followng two games, one a Sicilian and one a Caro-Kann, where the
middlegame positions could easily have arisen from lines of the French.

Short-Miles
BBC TV, London 1980
Sicilian Defence

1 e4 c5 2 Nf3 d6 3 Bb5+ Bd7 4 Bxd7+ Nxd7 5 0-0 Ngf6 6 Qe2 e6 7 c3 Be7 8 d4 0-0 9 Rd1 Rc8 10 e5
Ne8 11 Nbd2 cxd4 12 cxd4 Qc7 13 Nf1 Qc4 14 Qxc4 Rxc4 15 Ne3 Rc7 16 Bd2 d5 17 Rac1 Nb8 18 h4
Rxc1 19 Rxc1 Nc6 20 h5 h6 21 Kf1 Nc7 22 Ke2 Rc8 23 Kd3 Kf8 (Diagram 1)

Diagram 1

24 Ng1
Short now begins a very impressive exploitation of his kingside space advantage. It is no coincidence
that Short has been a keen French player himself (as Black) since his earliest days and thus has a fine feel
for these positions.
24 ... Na8 25 Ne2 Nb6 26 f4 Ke8 27 g4 Kd7 28 Rg1 Na4 29 Nd1 Rh8 30 f5 Nb6 31 Nf4 Rf8 32 Ne3
Rc8 33 fxe6+ fxe6 34 Ng6 Ke8 35 Rf1 Nd7 36 Ng2 Bg5 37 N2f4 Bxf4 38 Rxf4 a5 39 Nh8 b6 40 Nf7
Ke7 41 g5 hxg5 42 Rf1 Nb4+ 43 Ke2 Rc2 44 Kd1 Rxb2 45 Bxg5+ Kf8 46 Nd8+ Kg8 47 Rf7 Nf8 48 h6
gxh6 49 Bxh6 Rg2 50 Rxf8+ Kh7 51 Bd2 Nd3 52 Nxe6 Nb2+ 53 Kc1 Nc4 54 Bc3 Rg1+ 55 Kc2 Rg2+
56 Kb3 b5 57 Nf4 Rf2 58 e6 b4 59 e7 Nd6 60 Nxd5 Re2 61 e8Q a4+ 62 Qxa4 Black resigns
White has the better structure

Nimzowitsch-Capablanca
New York 1927
Caro-Kann Defence

1 e4 c6 2 d4 d5 3 e5 Bf5 4 Bd3 Bxd3 5 Qxd3 e6 6 Nc3 Qb6 7 Nge2 c5 8 dxc5 Bxc5 9 0-0 Ne7 10 Na4 Qc6 11 Nxc5 Qxc5 12
Be3 Qc7 13 f4 Nf5 14 c3 Nc6 15 Rad1 g6 16 g4 Nxe3 17 Qxe3 (Diagram 2)

Diagram 2
Black has the better structure

White has misjudged the position. His aggressive posturing is inappropriate when so many minor
pieces have already been exchanged.
17 ... h5 18 g5 0-0 19 Nd4 Qb6 20 Rf2 Rfc8 21 a3 Rc7 22 Rd3 Na5 23 Re2 Re8 24 Kg2 Nc6 25
Red2 Rec8 26 Re2 Ne7 27 Red2 Rc4 28 Qh3 Kg7 29 Rf2 a5 30 Re2 Nf5 31 Nxf5+ gxf5 32 Qf3 Kg6
33 Red2 Re4 34 Rd4 Rc4 35 Qf2 Qb5 36 Kg3 Rcxd4 37 cxd4 Qc4 38 Kg2 b5 39 Kg1 b4 40 axb4
axb4 41 Kg2 Qc1 42 Kg3 Qh1 43 Rd3 Re1 44 Rf3 Rd1 45 b3 Rc1 46 Re3 Rf1 Black resigns
It will not be possible in this book to examine in depth every single nuance of the French. Instead I hope to give you a flavour of this
fascinating opening and help you to decide if this could be the opening for you. Alternatively, if you are a 1 e4 enthusiast this can help you to
choose a variation to meet the French

Byron Jacobs,
Brighton, November 2002
Chapter One
The Winawer with Qg4
Introduction
Welcome to the wonderful world of the Winawer. This variation is named after the Polish master Simon
Winawer who developed the ideas in the variation in the latter part of the 19th century. It is a dynamic
counterattacking system that was much favoured at the start of his career by the great Soviet World
Champion Mikhail Botvinnik. Rather than carefully building up his position Black immediately takes the
fight to his opponent by pinning the white knight. It can lead to a marvellous strategic and/or tactical
battle, which isn’t always appreciated by White players of a calm disposition. In fact it is the greatest
compliment to the Winawer that White often plays 3 Nd2 (the Tarrasch variation), blocking in his own
bishop on c1, because he is too afraid to play Nc3, precisely because of 3 ... Bb4.
White has three main philosophies against the Winawer:
1) Try for a big advantage with Qg4, when it might all go horribly wrong.
2) Settle for a small plus by leaving the queen at home and developing with moves like Nf3, as in
Chapter Two.
3) Attempt to bamboozle Black with one of the offbeat systems described in the third chapter.

Poisoned Pawn Variation

We’ll begin by looking at the bold and brash Qg4! main line, which is often referred to as the Poisoned Pawn Variation.
With 3 ... Bb4 Black breaks the rule that you should bring out your knights before your bishops. This
provoked Bobby Fischer into making the famous and rather dogmatic pronouncement that the ‘Winawer is
anti-positional and weakens the kingside’.
In order to punish Black the American would usually play his queen swiftly to g4 where it aims at g7.
This seems very logical, but ... just a minute ... isn’t White also breaking a basic positional tenet: don’t
bring your queen out too early?
In fact both players are avoiding the normal rules in order to achieve positional ends. It is this which
makes the variation so exciting and difficult to understand!
1 e4 e6 2 d4 d5 3 Nc3 Bb4
Besides the general idea of ... Bxc3+ inflicting doubled pawns on White there is also the threat of 4 ...
dxe4, winning a pawn.
4 e5
Already a major decision by White. He grabs a space advantage in the centre and deprives the black
knight of its best square on f6. On the other hand he is in some danger of overextending himself as his
centre is vulnerable to flanking blows.
Other moves are discussed in the third chapter.
4 ... c5!
This attack on d4 with ... c7-c5 is an essential strategical theme in all the main line French systems.
Black immediately begins the fight for control of the centre by undermining the base of White’s pawn
structure. Compared to the Tarrasch Variation the attack looks particularly favourable for Black here as
White cannot support his centre with c2-c3 as the knight is in the way.
5 a3!
A good moment to force the bishop to declare its intentions, to borrow a phrase from Nimzowitsch in
his book My System.
5 ... Bxc3+
The alternative 5 ... Ba5, declining the chance to capture, is discussed in the Positional Main Lines
section in Chapter Two.
6 bxc3
Now White has doubled c-pawns, but at least he has kept solid control of the d4-square.
6 ... Ne7
Black hurries to develop his kingside before he is hit by Qg4. Alternatives are the counterattacking 6 ...
Qa5 in Chapter Two and 6 ... Qc7 one move earlier than usual, which is discussed below.
7 Qg4 (Diagram 1)

Diagram 1
The critical poisoned pawn

More restrained alternatives such as 7 Nf3 are examined in the second chapter. Black now has three
methods of dealing with the threat to his kingside pawns:
1) Make a positional concession with 7 ... Kf8 (or maybe even 7 ... g6)
2) Take a deep breath and castle into it with 7 ... 0-0. White should now get a good initiative on the
kingside but Black retains his positional pluses.
3) Fight fire with fire ... jettison the kingside and go after White’s weakened queenside with ...
7 ... Qc7
The alternatives that hold onto the g7-pawn – 7 ... 0-0 (Games 4 and 5) and 7 ... Kf8 (Game 6) – are
discussed below.
8 Qxg7
The white queen decimates the black kingside. It is too late for normal development with 8 Nf3? as
after 8 ... cxd4 the threat of Qxc3+ forking the king and rook means that White has to play the ugly 9 Qxd4
when 9 ... Nbc6 is deeply unpleasant.
8 ... Rg8 9 Qxh7 cxd4 10 Ne2
The knight defends c3 against the fork and stops Black taking on e5 with check. On the other hand the
bishop on f1 is none too pleased at being shut in.
10 ... Nbc6 11 f4!
It is necessary to bolster e5 in this fashion as if 11 cxd4? Nxd4! 12 Nxd4 Qc3+ does not inspire
confidence.

TIP: The e5-pawn is nearly always a crucial central foothold for White in the French. It should
always be maintained unless there is a very good reason to give it up.

11 ... Bd7 12 Qd3 dxc3 (Diagram 2)

Diagram 2
A highly complex position

Now White has a variety of moves.

TIP: It is suggested that you play through games 1-3 quickly, without referring to the notes, in order
to get a feel for the dynamic features of this variation. Then return to the discussion of the
positional features here.

You will see that in Short-Kosten (Game 1) White took back on c3 immediately with 13 Qxc3. This move
was a favourite of former World Champion Anatoly Karpov and so deserves respect. Short wins in
convincing style, but you will see from the notes that Black had much better resources. In Game Two, 13
Be3 was well met with the sacrificial 13 ... d4, gaining time for an attack on the white king by attacking
the bishop. In Game Three, Svidler plays the safer 13 Rb1, but once again a bit of carelessness allows
Black to break in good style with ... d5-d4! and overwhelm his opponent.
So what conclusions can we draw from these games? The main one is that the position is full of
contradictions and paradoxes.
White destroys Black’s kingside, but ...
Black’s kingside is wrecked and White has a passed pawn on the h-file as well as the bishop pair. The
dark-squared bishop looks particularly menacing: Black will surely regret having given his away.
All this is undoubtedly true, but there is a further important point to remember: the point of a game of
chess is to mate the opponent’s king!
Where exactly is the white king going to find shelter? Surely not on the queenside. The centre is also
likely to become open. So the best bet is the kingside! Now we see that the white queen’s greed in
snaffling the pawns on g7 and h7 has been distinctly double edged – she has opened the g- and h-files for
the black rooks to attack her king’s future residence.
Black wants to open the position, but ... .
Here is another contradiction. Black wants to open the centre to get at the white king, or at least to
dismantle White’s pawn chain. But in opening the position he is breaking the rule:
If the opponent has the two bishops you should keep the position closed!
White’s dark-squared bishop is delighted when the position opens up, as it has no rival in the enemy
camp. So Black has to make sure when he blasts things open that the co-ordination of his pieces is so
superior to White’s that it outweighs factors such as the two bishops and the passed pawn on the h-file.
These contradictions explain why Black can have such mixed fortunes in this line. Thus Kosten suffers a complete disaster in Game 1
whereas Korchnoi has a convincing win over Spassky in Game 2. Here it should be mentioned that Black commits what many Winawer
players regard as a cardinal sin in Game 1: namely, he plays ... Rc8 rather than castling queenside. It is highly tempting to put immediate
pressure along the c-file, but Black is falling into the same strategical malaise as White: his rooks cannot be easily co-ordinated and his king
remains stuck in the centre. It is much better for Black to delay action on the c-file in favour of ... 0-0-0! when his rooks are joined and his
king is reasonably safe. The attacking plans in Games 2 and 3 hinge on the prospect of a central breakthrough with ... d5-d4! which wouldn’t
have been possible if the black king had been in the centre.

NOTE: Normal positional considerations do not always apply when positions become as wild as they do in the Winawer Poisoned
Pawn.

Black Holds on to the g7-pawn

Not everyone wants to see their kingside pawns massacred, even if it results in dangerous counterplay.
Black has tried three main methods to keep the white queen out of g7. Playing through the opening moves
again, after
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3
the first opportunity for Black not to play ball is 6 ... Qc7. The idea is that after 7 Qg4 Black can
respond with 7 ... f5! uncovering a defence of g7 by the queen along the second rank. This was played by
Nigel Short against Gary Kasparov in Game Seven. Short went the way of virtually all Kasparov’s
opponents as Black, but that had more to do with the workings of genius than any flaw in Black’s set up. In
fact, the World Number One wrote an analysis of the game in which he suggested that Black was at least
equal after the opening.

TIP: Just because Kasparov crushes a variation it doesn’t necessarily mean it is bad.

Another way to defend g7 is to play 6 ... Ne7 and after 7 Qg4 to give up castling rights with 7 ... Kf8 as in
Game Six. Black decides that as the centre is closed his king won’t have too much to fear on f8 – in fact
less to fear than if he castles with 7 ... 0-0, when White can launch a direct attack. There is a lot to be said
for this reasoning and when White overpresses Black scores a crushing victory in the illustrative game.
However, we should point out that there is one serious drawback: the rook on h8 becomes difficult to
activate.

NOTE: A king marooned in the centre is not only a target for the opponent, it also interferes with
the connection of the rooks and thus hinders development.

Why should connecting rooks be so important? Well, they are such commanding pieces that to leave one
out of your plans will surely mean failure. Just think how powerful rooks become when they are doubled
on the seventh rank. Irving Chernev once said that ‘when you double rooks you double their power’.
Looking back, you will see that this drawback to 7 ... Kf8 applies equally to the situation of the white
king in the Qxg7 Main Line. It isn’t only a target in itself – it also stops the white rooks from coordinating
and working together.
 Finally we come to 7 ... 0-0, which is the main way for Black to guard g7. At the same time it develops
the kingside and prepares to attack White’s centre with ... f7-f6 or, more typically, stabilise his position
with ... f7-f5. The downside is that White’s queen, two bishops and knight (after Nf3) are all poised to
launch an immediate and massive assault. Has Black castled into a fatal attack? Certainly he needs to be
extremely careful – look what happens to him in Game Four: just one inaccurate move and it all caves in.
In Game Five, Black quickly utilises the ... f7-f5 idea.
This puts White into a quandary, as if he moves his queen away then the bishop on d3 finds its diagonal
aiming at h7 blocked. So he usually plays 9 exf6 en passant, which maintains a promising attack but loses
the spearhead of his pawn structure. If he isn’t careful then Black might gain time with ... e6-e5
uncovering an attack on the queen. In the illustrative game, Timman made an inaccuracy which allowed
Anand to apply continuous pressure, but overall there is nothing wrong with the plan of castling kingside.
Warning: The Winawer Poisoned Pawn is highly theoretical. One correspondence player has dedicated
his life to unravelling the mysteries of this variation.

Illustrative Games

Game 1
Short-Kosten
Hastings 1988

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

Black develops rapidly sacrificing the kingside. The alternative main line is 7 ... 0-0 (Games 4 and 5),
provocatively castling into the attack. Black can also defend the pawn with 7 ... Kf8!? (Game 6). Nobody
wants to go 7 ... g6 – it’s very weakening of the dark squares.
8 Qxg7 Rg8 9 Qxh7 cxd4 10 Ne2 Nbc6 11 f4
On 11 cxd4 Nxd4! 12 Nxd4 Qc3+ wins.
11 ... Bd7 12 Qd3
Back to the thick of things.
12 ... dxc3 (Diagram 3)

Diagram 3
A critical Winawer position

Almost a starting position for this opening. Black has lost most of his king’s wing, but hopes to
generate enough play on the other side and in the centre to make up for it. Broadly speaking, unless Black
whips up counterplay swiftly, then the long term assets of White’s kingside pawns should decide. That is a
simplification, but it summarises the character of many 7 Qg4 Winawer middlegames.
13 Qxc3
White captures the pawn with the knight in Game 3, while 13 Be3 is seen in Game 2.
13 ... Rc8?!
Kosten deviates from his earlier game vs Balashov at Minsk 1986, which White won after 13 ... Nf5
14 Rb1 d4 15 Qd3 0-0-0 16 Rg1 f6 17 g4 Nh6? 18 exf6 Rxg4 19 Rxg4 Nxg4 20 f7 e5 21 Bg2 Be6 22 h3
Nf6 23 fxe5 Nxe5 24 Qxd4!. Black’s play was later improved by 17 ... Nh4! 18 exf6 e5 19 f7 Rxg4 20
Rxg4 Bxg4 21 Bh3 Qd7 22 Bxg4 Qxg4 23 Qg3 Qh5 24 Rb3 e4 25 Qg7 d3 with rough equality in Hellers-
Djurhuus, Gausdal 1992.
14 Rb1 Nf5 15 Bd2 a6 16 Rg1!
A strong move. After 16 g3 b5 17 Bh3 Qb6! 18 Qd3 Nce7 19 g4 Nh4 20 Rf1 Rc4! Black developed
nifty counterchances in Tal-Farago, Wijk aan Zee 1987, and 21 Rb4 Rh8! 22 Ng3 Nc6 23 Rb2 Qd4! 24
Qxd4 Nxd4 25 Kd1 Nhf3 26 Nh5 Nxd2 27 Kxd2 Ra4 28 Ra1 Ke7 29 c3 led to a draw.
16 ... b5
This just does not seem immediate enough.
17 g4!
 A simple and extremely powerful procedure. The pawns advance.
17 ... Nh4 18 Rg3 Qb6 19 Qd3 Rh8 20 Rh3
White rooks patrolling the third rank can be a feature of these middlegames.
20 ... Ne7 21 Nd4 Rc4 22 c3 Ra4
If Black can find no better ideas than this then he is already down the tubes. Short plays simple chess,
and just pushes the h-pawn.
23 Qg3 Nhg6 24 Rxh8+ Nxh8 25 Bd3 Nhg6 26 h4 Rxa3 27 h5 (Diagram 4)

Diagram 4
White’s kingside pawns march on

This rook’s pawn is a tad more significant than the one that Black just took. The black game is beyond
saving.
27 ... Nf8 28 h6 Neg6 29 f5 Ne7 30 Qh2 exf5 31 h7 Neg6 32 gxf5
Swamping him.
32 ... Nh8
The worst possible square.
33 Qh5 Qc7 34 Kf1 Rxc3 35 e6!
Accurately calculated. The counterplay that Black has come up with is quite insufficient.
35 ... Rxd3 36 exf7+ Kd8
Or 36 ... Nxf7 37 h8Q.
37 Qg5+ Kc8 38 Rc1 Rxd4 39 Rxc7+ Kxc7 40 Ba5+ Kb7 41 Qd8 Rf4+
Tony Kosten is notorious for disliking resigning, often carrying on token resistance in hopeless
positions. This policy once enabled him to save a game a whole queen down, when the Danish IM Iskov
accidentally stalemated him.
42 Kg1 Rg4+ 43 Kh2 1-0

Game 2
Spassky-Korchnoi
Candidates final, Belgrade 1977

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Qg4 cxd4 8 Qxg7 Rg8 9 Qxh7 Qc7 10 Ne2 Nbc6 11 f4 Bd7 12 Qd3
dxc3 13 Be3 d4!? (Diagram 5)
 Diagram 5
A sacrifice for the initiative

A big novelty at the time. Black activates his men at the cost of a pawn.
14 Bf2
He could just have taken it immediately.
14 ... 0-0-0 15 Nxd4 Nxd4 16 Qxd4 b6
Time out for this. He does not want to lose a7 and let the white queen and bishop in.
17 Bh4 Bb5!?
Discovering an attack on the white queen.
18 Qe4 Bxf1 19 Rxf1
19 Qa8+ Kd7 20 Rd1+ was a line leading to exchanges and after 20 ... Nd5 21 Rxd5+ exd5 22 Qxd5+
Kc8 23 Qa8+ Kd7 24 Bxd8? Rxd8 White stays a piece behind, e.g. 25 Qd5+ Ke8 and the checks run out.
He has instead to take the draw with 24 Qd5+ etc.
19 ... Rd5!
Spassky may have underestimated the effect of this blocking move.
20 Bxe7 Qxe7 21 Rf3 Kb8
A very important little move. The contrast in the safety of the two kings is what decides this game.
22 Kf1
On 22 Rxc3 Qh4+ is annoying.
22 ... Rd2 23 Rf2 Rgd8 24 Qf3 Rxf2+ 25 Kxf2 Rd2+ 26 Kg3 Qd8 (Diagram 6)

 Diagram 6
Black’s pieces take control

Korchnoi’s major pieces now start bobbing and weaving around Spassky’s king.
27 Qe4 Qg8+ 28 Kh3 Qh8+ 29 Kg3 Qg7+ 30 Kh3 Rd8
Threatening mate, and forcing White to loosen the pawn cover around his king.
31 g4 Rh8+ 32 Kg3 Qh6 33 Qg2 Qh4+ 34 Kf3 Rd8
The queen and the rook create penetration threats all over the place. Defence is very difficult, if not
already impossible.
35 Qg3 Qe7 36 g5 Rd2 37 Kg4 Qb7 38 Qxc3
Finally the pawn falls, but here it doesn’t matter.
38 ... Rg2+ 39 Kh3 Rf2 40 Kg4 Qe4 0-1
The white king’s defences collapse.

Game 3
Svidler-Ivanchuk
Linares 1999

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Qg4 Qc7 8 Qxg7 Rg8 9 Qxh7 cxd4 10
Ne2 Nbc6 11 f4 Bd7 12 Qd3 dxc3 13 Rb1
Another line.
13 ... 0-0-0 14 Nxc3 Na5!? 15 g3
An experiment which does not impress. A better try might be 15 Nb5!?, e.g. 15 ... Bxb5 16 Rxb5 Kb8.
Other moves are not very dangerous for Black, for instance 15 Rg1 Kb8 16 g4 Ng6 17 Nb5 Bxb5 18
Qxb5 Nc4 or 15 h3 Kb8 16 Ne2 Nec6 17 g4 Nc4.
15 ... Kb8 16 Ne2 Ba4! 17 c3?!
17 Nd4 was better.
17 ... Nf5!
17 ... Bb3 was also interesting.
18 Bh3
This allows a central breakthrough. Maybe it was better to play 18 Rb4 Bb3!? with an unclear
situation.
18 ... d4! (Diagram 7)

Diagram 7

19 Bd2
What else? On 19 Bxf5 dxc3 20 Nd4 (or 20 Qxc3 Rd1+ 21 Kf2 Rxh1 22 Qxc7+ Kxc7) 20 ... exf5
Black stands excellently, e.g. 21 Rb4 Bc6 22 0-0 Be4. On 19 Kf2 Bc6! 20 Rg1 Rh8! 21 Bxf5 Rxh2+ and
the roof caves in.
Maybe Svidler forgot that in case of 19 0-0 dxc3 20 Qxc3 Qxc3 21 Nxc3 he loses after the simple
combination 21 ... Nxg3! 22 Nxa4 (22 hxg3 Rxg3+ 23 Kh2 Rxc3 wins) 22 ... Nxf1+ 23 Kxf1 Rd1+ 24
Ke2 Rgg1 25 Rb5 (the only counter hope) 25 ... Rxc1 26 Rxa5 Rc2+ 27 Ke3 Rxh2 and the bishop has
nowhere to hide.
Taking on d4 leads to a problem with ... Bc2 at the end, winning the b1 rook.
The thematic central thrust

19 ... Nb3! 20 Bxf5 dxc3! 21 Qxc3 Nxd2 22 Qxc7+ Kxc7
Via a series of highly imaginative and active moves Ivanchuk has achieved a winning position, as soon
becomes clear.
23 Rc1+ Bc6 24 Bh3 Kb6
and White must lose the exchange, either to the bishop or a knight fork at f3.
25 Rxc6+
Choosing to lose it this way.
25 ... bxc6!
Ivanchuk makes a new asset out of it: a nice mobile c-pawn.
26 Kf2 c5 27 Bg2 c4 28 h4
The h-pawn is his last hope.
28 ... Nb3 29 h5 Rd2 30 Ke3 Ra2 31 h6 Rxa3 32 h7 Rh8 (Diagram 8)

 Diagram 8
Black has a winning endgame

The activity of the other black units means that despite the indignity of deploying his rook so passively
he can still win.
33 Nc3 Ka5! 34 Ne4 Nc5+ 35 Ke2
Unfortunately the king can’t move into the centre – 35 ... Nxe4 36 Bxe4 Kb4 with the threat of 37 ...
Rd8+
35 ... Nxe4?
An inaccuracy in a won position. 35 ... Ra2+! 36 Ke3 Rxg2 37 Nxc5 Rxg3+ 38 Kd4 Kb5 led to an easy
victory.
36 Bxe4 Rxg3 37 Rb1 Rh3 38 Kd2?
After 38 Rb7! Black must demonstrate good technique. After the text move the game is over.
38 ... Rd8+ 39 Kc2 Rd4 0-1
Such games indicate why Short and Adams acclaim Ivanchuk as a genius.

Game 4
Maus-Hübner
Lugano 1989

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Qg4 0-0 8 Bd3 Nbc6

Not 8 ... c4?! 9 Bh6! Ng6 10 Bxg6 fxg6 11 Be3 and White has a very strong attack with h2-h4-h5. The
proactive thrust 8 ... f5 is played in the next game.
9 Qh5 h6?
On 9 ... Ng6 10 Nf3 is hard to meet, e.g. 10 ... c4 11 Ng5 h6 12 Nxf7! and wins. But Black can play 10
... Qc7!? so that if 11 Ng5 h6 12 Nxf7 Qxf7 13 Qxg6 Qxg6 14 Bxg6 cxd4 regains the pawn.
10 Bxh6! (Diagram 9)
Diagram 9
A winning sacrifice

10 ... gxh6 11 Qxh6 Nf5
The only way to block mate. On 11 ... Ng6 something like 12 h4 Re8 13 h5 Nf8 14 Rh3 does him in.
12 Bxf5 exf5 13 0-0-0 f4 14 Nh3 Ne7
Of course on 14 ... Bxh3 comes 15 gxh3! and crush down the new g-line.
15 Ng5 Bf5 16 g4!
Bash! He carries on thumping.
16 ... Be4 17 Rhe1
Threatening to take the bishop and then mate on h7.
17 ... Qb6 18 e6 Bg6 19 Rd3 (Diagram 10) 1-0

Diagram 10
The black king will be mated

 Off to h3, and on 19 ... Bxd3 20 cxd3 it will be mate.
Simple chess from a mighty Maus.

Game 5
Anand-Timman
Dortmunder Schachtage 1999

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 Qg4 0-0 8 Bd3 f5!

Altogether a wiser choice.
9 exf6 Rxf6 10 Bg5
On 10 Qh5 both 10 ... g6 and 10 ... h6!? come into consideration.
10 ... Rf7 11 Qh5 g6
Regarded as best; 11 ... h6 creates more serious weaknesses.
12 Qd1
Having induced some damage in the black kingside, the queen returns home.
12 ... Nbc6 13 Nf3 c4
Black has also played the immediate 13 ... Qf8 here, or even 12 ... Qa5, e.g. 13 Bd2 (13 Qd2 c4!) 13 ...
Nbc6 14 Nf3 Qc7! 15 0-0 e5 with an active game.
14 Be2 Qf8 15 Qc1!?
An interesting novelty. Anand defers castling and plans to try to dominate the kingside and the dark
squares first. 15 0-0 h6 16 Bh4 Nf5 17 Bg3 g5 leaves Black comfortably placed.
15 ... Nf5 16 h4!? h6
On 16 ... Nd6 White plays 17 h5, causing problems around Black’s king.
17 Bf4 Bd7 18 Ne5 Nxe5 19 Bxe5 (Diagram 11)

Diagram 11
White controls the dark squares
A long term problem for Black is his inability to ‘shield’ his dark squares. In giving away the bishop at
move 5 he cedes domination of those squares to White. This is particularly chronic here when the black
pawns are fixed on light squares.
19 ... Ne7 20 Qe3 Nc6 21 Bg3 Re8 22 h5 g5 23 0-0
Late castling, but in blocked positions that is not always a problem. White has fixed the pawn structure,
and now aims to advance with f2-f4 at the right moment to open up the game and the black kingside for his
bishops and major pieces.
23 ... Ne7 24 Be5
And not the immediate 24 Be5 because of 24 ... Nf5.
24 ... Nc6 25 Bh2 Rg7
Now after 25 ... Ne7 26 g4 takes away the square f5.
26 Bg4 Qf6
Maybe Timman could have sacrificed a pawn with 26 ... e5!? 27 Bxd7 Rxd7 28 dxe5 (28 Bxe5 Qf5)
28 ... Qf5 29 Rae1 Re6 with compensation. I think that this would have been a better practical chance than
just sitting there and waiting for it.
27 Rae1 Qe7 28 a4! b6 29 Qh3 Nd8 30 Re2 Bxa4
When White plays the line of the Winawer with 7 a4 or 8 a4 then his a-pawn may actually be won by
force. Often it does not matter much as Black loses time and has to misplace pieces to get at it. Here’s
another instance where White just ploughs on regardless. Instead of saying ‘Black wins the a-pawn’, one
might say ‘Black acknowledges that action on the kingside is about to stuff him so helps himself to the a-
pawn anyway, arguing that he might as well be hung for a sheep as for a lamb.’
31 f4
After due preparation, here we go.
31 ... Bd7 32 f5 Rf7 33 Rfe1 Rf6 (Diagram 12)

Diagram 12
The battle for e6 rages

34 Be5 Rxf5
Ceding material, but on 34 ... Rf7 35 Bc7!? maintains the pressure.
35 Bxf5 exf5 36 Bc7
This is good enough, but 36 Bf6! Qxf6 37 Rxe8+ Bxe8 38 Rxe8+ Kf7 39 Re5 wins immediately.
36 ... Ne6
Forced. After 36 ... Qxe2 37 Rxe2 Rxe2 the knight at d8 hangs, as it would after 36 ... Qf8 37 Rxe8
Bxe8.
37 Qxf5 Qf7
 On 37 ... Nxc7 38 Rxe7 Bxf5 39 Rxe8+ Nxe8 40 Rxe8+ Kf7 41 Rh8 Kg7 42 Rd8 White wins the
ending.
38 Qxf7+ Kxf7 39 Bd6
Black’s material inferiority is slight but, notwithstanding the queen exchange, the activity of White’s
pieces is still great.
39 ... a5 40 Rf1+ Kg7 41 Re5 Nf4 42 Rxe8 Bxe8 43 Be5+!
Forcing the king to a more passive square.
43 ... Kg8 44 Bxf4 gxf4 45 Rxf4 Kg7 46 g4 a4 47 Kh2 1-0
The ending is lost, e.g. 47 ... a3 48 Rf1 Ba4 49 Ra1 Kf6 (49 ... Bxc2 50 Kg3) 50 Rxa3 Bxc2 51 Kg3
Kg5 52 Ra8.

Game 6
Hracek-M.Socko
MK Cafe Cup, Koszalin 1999

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Qg4 Kf8!? (Diagram 13)

Diagram 13
The king calmly defends g7

A risky, but interesting move that became very popular at the end of 80s. It is still not clear how White
can get a real advantage, and Black could be satisfied with results in practical games.
8 h4
The most direct and dangerous move. The alternative is 8 a4 which prevents Black’s following queen
manoeuvre.
8 ... Qa5
8 ... Qc7!? 9 Qd1 b6 10 h5 h6 11 Rh3 Ba6 12 Bxa6 Nxa6 13 Qd3 Nb8 14 a4 was Saulin-Mkrtchian,
Moscow 1996, where White went on to win.
9 Bd2 Qa4!
A most irritating move for White to cope with, as it pressures c2 and d4. Note also that the queen is not
unknown to go back to e8 from here, to sometimes assist in defence.
10 Rh3!?
10 h5 h6 11 Kd1 b6! 12 Nf3 Ba6 13 Bxa6 Nxa6 14 Qf4 Rc8 15 g4 Ke8 16 g5 cxd4 17 gxh6 gxh6 18
Nxd4 Nc5 19 Qf6 Rg8 led to a Black win in Drei-Lputian, Reggio Emilia 1998.
10 ... Nbc6
The problem with Black ever grabbing on c2, as we shall see in the game, is that White plays the
powerful developing move Bd3 with tempo.
11 h5 h6 12 Qf4 Bd7!?
White kept some initiative after 12 ... b6 13 Rf3 Nd8 14 dxc5 Qxf4 15 Rxf4 bxc5 16 Be3 c4 17 Rb1
Ndc6 18 Nf3 g5! 19 hxg6 Nxg6 20 Rg4 Kg7 21 Nh4!? in Atlas-Psakhis, Geneva 1992.
13 Rf3 Be8 14 Rb1
This was a new move. 14 g4 cxd4 15 Bd3 b5 16 cxd4 Nxd4 17 Rg3 a5!? 18 Ne2 Nxc2+ 19 Bxc2 Qxc2
20 Re3 Nc6 21 Qf3 Qc5 22 and now 22 Rc1! would have given White play. Instead, in Bobarykin-
Nikitin, Novosibirsk 1998, he chose 22 Nf4? and lost after 22 ... Nd4 23 Qd1 Nc2+.
14 ... b6 15 g4?!
Hracek believes in White’s attacking chances and declines to exchange queens. But this was a bit
optimistic. 15 dxc5 Qxf4 16 Bxf4 bxc5 left an unclear position.
15 ... Qxc2
Yes please. Black has three pieces near her king and reasons that she can withstand the onslaught.
16 Bd3 Qa4 17 g5
After 17 Bg6? Nxg6 18 hxg6 the black queen returns with 18 ... Qc2 winning.
17 ... c4 18 Be2 Qc2
Black could take a second pawn, and 18..Qxa3!? deserves serious attention, as did 18 ... Nf5!?.
19 Rc1 hxg5 20 Qxg5
20 Qxf7+ Bxf7 21 Rxc2 g4! 22 Rf4 Kg8 23 Bxg4 Nd8 24 h6 Bg6! leads to a clear advantage for Black.
20 ... Qh7
Dropping back to help the threatened sector.
21 Rh3 Ng8
On 21 ... Nf5 22 Bd1! reroutes effectively.
22 Nf3 (Diagram 14)

Diagram 14
White has compensation for the pawn

22 ... f6
Hereabouts Socko decides to relieve her cramp, but White’s pieces now gain activity.
23 exf6 gxf6 24 Qf4 Ke7 25 Qc7+ Kf8
On 25 ... Bd7 26 Nh4 is annoying.
26 Qf4 Ke7 27 Nh4
It looks like repetition with 27 Qc7+ was the best solution.
27 ... Kd7 28 Bg4 Nge7
She could have tried 28 ... Nh6!? when 29 Ng6? fails to 29 ... Bxg6 30 hxg6 Qxg6 31 Rxh6 Rxh6 32
Qxh6 Qxg4 etc.
29 Qxf6??
After 29 h6!? or 29 Ng2!? the result of the game was unclear, but Hracek makes a terrible blunder.
29 ... Qe4+ 30 Re3
What else? 30 Be2 Bxh5 31 Re3 (31 f3 Bxf3! 32 Qxf3 Rxh4) 31 ... Raf8!? (31 ... Qh1+ 32 Bf1 Bg4) 32
Qg5 Qh7! would lose as well.
30 ... Qxg4 31 f3
On 31 Qxh8 Bxh5! 32 Qxa8 Qg1 mate, may have been what White overlooked.
31 ... Qg8!? 32 Kf2 Rxh5 33 Qf4 Ng6 34 Nxg6 Bxg6 35 Rg1 Qf7 36 Ree1 Qxf4 37 Bxf4 Rg8 38 Rg4 Ne7 0-1

Game 7
Kasparov-Short
Novgorod 1997

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Qc7 7 Qg4 f5

Avoiding the mayhem of 7 ... Ne7 8 Qxg7 Rg8 9 Qxh7 cxd4 (Games 1-3), but weakening the kingside.
8 Qg3 cxd4 9 cxd4 Ne7 10 Bd2 0-0 11 Bd3 b6
As so often in the French Defence, Black hastens to exchange off the light-squared bishops when he
can.
12 Ne2 Ba6 13 Nf4 Qd7 14 h4 (Diagram 15)

Diagram 15
White begins a kingside attack

Staking out territory.
14 ... Bxd3 15 Qxd3 Nbc6 16 Rh3!?
No sense in castling. The rook pops out here.
16 ... Rac8 17 Rg3 Rf7 18 h5 Nd8 19 c3 Rf8 20 Kf1 Rc4 21 Kg1
Castled by hand.
21 ... Nf7 22 a4!?
A very interesting move. It is not out of the question that his bishop will yet pop up on the excellent a3-
f8 diagonal. Meantime, this pawn has a certain ‘bait’ factor.
22 ... Rfc8 23 Qb1 Nc6 24 Qd1 Ne7
 Nothing to do.
25 h6 g6 26 Qh5
The rook, the knight, the h-pawn and now the queen get in on the attacking act.
26 ... Rxa4 27 Rxa4 Qxa4 (Diagram 16)

Diagram 16
Can White break into Black’s fortress?
28 Nxe6
After 28 Qh4 Black defends by 28 ... Qd1+ 29 Kh2 g5! 30 Rxg5+ Nxg5 31 Qxg5+ Kf7 32 Nh5 (or 32
Be3 Qg4) 32 ... Rg8 33 Ng7 Rxg7 34 Qxg7+ Ke8 35 Bg5 Qh5+ with perpetual check.
28 ... Qc2?
The losing move. Black could have defended with 28 ... Rc6 29 Nf4 Kf8.
29 Qh4 f4 30 Bxf4 Nf5 31 Rxg6+! 1-0
The roof caves in, and yet another Kasparov attack crashes through against a world class grandmaster.
Chapter Two
The Winawer: Positional Main Lines
Introduction
In this section White avoids the crazy positions of Chapter One in favour of the relatively sedate lines that
begin after 7 a4 or 7 Nf3. Plays becomes much more strategic and long-term. Here we also consider 5 ...
Ba5 – a radical treatment of the position by Black who declines to give White doubled pawns.

Black Captures on c3
It is recommended that you play through games 8-12 in this section, without paying much attention to the
notes, to get a feel for the set up before reading the comments that follow. Then you should play through
the games again and look at the analysis in more detail. The games break down as follows:
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 In game 12 Black tries 6 ... Qa5!?. 7 Nf3 (Diagram 1) Fischer
preferred 7 a4 in Game 8.

Diagram 1
A positional approach

7 ... b6 Or 7 ... Qc7 8 h4, as in Game 9. 8 Bb5+ Bd7 9 Bd3 Ba4 This is seen in Games 10 and 11.
Games with 5 ... Ba5 are summarised below under the heading discussing this move.
The play in these variations revolves strongly around planning, rather than theoretical knowledge, so we will now consider various ideas for
both sides to get a feel for the positions that arise.

White’s Doubled c-pawns

It is another paradox in the Winawer that White’s central pawn complex can be weak and strong at the same time.
The pawn on c3 is part of a doubled pawn complex and cannot be supported by any other pawn; the
c4-square also cannot be defended by a pawn and so is an ideal outpost for a black piece.

NOTE: An outpost is a square in the other half of the board from which a piece cannot be evicted
by a pawn.

If the outpost square is in the centre then it can become of enormous strategic importance. Knights in
particular like centre square where they are safe from pawn attack, as they work best at short range and
cannot influence things like a bishop can from a distance. So, if you can bed a knight down on a central
outpost, it will generally be rather happy.
Note however that if Black were to put a knight on c4 it would unwittingly shield the c3-pawn from
frontal attack. So Black has to decide carefully whether it is advantageous to put the knight there
(normally the answer is yes).
The doubled pawns have also presented Black with an outpost square on a4, which can be used by the
bishop or queen to attack the vulnerable c2-pawn whilst simultaneously blocking the queenside.
Thus from a static point of view the c3-pawn is undoubtedly weak, but compare the overall central situation with that in the Classical after
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5. Already you will see that White is going to have to give up his pawn chain on d4: the capture
... c5xd4 cannot be stopped, and he will have to recapture Nxd4. In contrast, in the Winawer, White has the pawn on c3 ready to recapture on
d4 and keep a firm pawn centre.

White Plays a3-a4

Despite his contention in Chapter One that the Winawer is anti-positional, Fischer never had much luck with a direct attempt at refutation with
the brutal Qg4 treatment. By 1971, when he was at his peak, he had switched to more positional means. After 1 e4 e6 2 d4 d5 3 Nc3 Bb4 4
e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 he played 7 a4 against Larsen – see Game 8.
There are three ideas behind this move. The first, which is seen to perfection in this game, is to clear
a3 for the dark-squared bishop so that it can put intense pressure on the black centre.
Secondly, if Black castles queenside, the a-pawn might also become useful as a battering ram to
weaken the defences on the b-file.
And thirdly – and highly significantly – by playing a3-a4 White takes the a4-square away from Black’s light-squared bishop. Let’s see why
the bishop is so strong on a4.

Black Plays ... Ba4

Of course this move assumes that White has played 7 Nf3 or another positional approach that avoids a3-a4. On a4 the black bishop attacks the
c2-pawn and ties a white piece – perhaps the queen’s rook or more likely the queen – to its defence. Now it is highly ignominious for a rook or
queen to be tied down by a mere bishop. The plan of ... Ba4 to put pressure on c2 is seen in its most startling form in Game 10.

TIP: Good strategy depends as much on stopping your opponent playing good moves as playing good moves yourself!

In fact I recommend a close study of Game 10 as it shows a refreshing avoidance of ‘natural’ and planless moves. Have a close look at the
above diagram. Black has avoided the usual development of the queen with ... Qc7? in favour of the idea of ... h7-h6, ... Kd7, ... Qg8 and ...
Qh7 (see Diagram 2),
 Diagram 2

when the white queen adds to the attack on c2. Such a scheme exhibits the difference between moving
pieces around and really playing chess!
Avoiding stereotyped play

Open and Closed Positions

NOTE: Black could afford to play a move like ... Kd7 because the centre is blocked. In an open position, e.g. a main line
Sicilian, such a plan would almost always be suicidal.

In a closed position more thoughtful, strategic play is possible. Also noteworthy about Game 10 is that from the above diagram Black avoided
the normal and typical 15 ... Nbc6 in favour of 15 ... Na6! He has seen that this knight has a much more useful career on the kingside via e8
and f6 and it eventually arrives at g3! In some scenarios after ... Nc7 it would be good to play ... Nb5, when the knight attacks the c3-pawn
and ties down White even further. There was no need to hurry because White had no open lines with which to exploit this slow manoeuvre.
White is sometimes prepared to accept a disgusting pawn structure if it means that the position
becomes more open and he can exploit his dark-squared bishop or some other advantage in firepower.
You couldn’t imagine a more ugly move as far as the central pawn structure is concerned than d4xc5 by
White. But in Games 9 and 11 White plays it voluntarily. If you look at Game 9 (Short-Ivanchuk), you will
see that d4xc5 cleared the way for the imaginative manoeuvre Rh4 and later Rbb4, when the White rooks
became powerfully co-ordinated.
Remember what was said about doubling the rooks doubles their power in Chapter One? This is a very odd scenario but not even
Ivanchuk could survive the pressure! Note also that White has won the battle for the a4-square here: the black bishop has had to retreat to c6.

Black Plays ... b7-b6

For all the talk about the usefulness of a black bishop on a4, White would much rather tolerate the bishop there than have the light-squared
bishops exchanged. Thus if Black plays 7 ... b6 (Diagram 3) as in Games 10 and 11, with the positional threat of 8 ... Ba6:

Diagram 3
White wants to keep bishops on

White usually replies 8 Bb5+!. This check is highly disruptive as after 8 ... Bd7 (if 9 ... Nbc6 then the plan of ... Ba6 has been sidelined) 9
Bd3! White has succeeded in developing his bishop whilst ruling out ... Ba6. Nevertheless White hasn’t nullified all of Black’s weapons this
trick as, with the feint at playing ... Ba6, Black has gained the time to arrange ... Ba4.

Black puts the Queen on a4

It is curious how many of the ideas in this section involve the a4-square. In fact, one of the chief weaknesses of White’s doubled pawns is that
he can no longer play b2-b3 to cover the ‘soft underbelly’ of his position – the pawn on c2. In Game 12 Black puts his queen on a4 as soon as
possible. Although relatively little known, this method of play has a good track record for Black.

TIP: The a4-square is a useful one for Black to pressurise the white c2-pawn!

Black Plays 5 ... Ba5

This has been dubbed the ‘Winawer Declined’ as Black refuses the offer to capture on c3.
The key move after 1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Ba5 is 6 b4!
This temporarily breaks the pin on the knight and so wins time for aggressive play after 6 ... cxd4 with 7 Nb5 or 7 Qg4. It has been
known for a long time that if 6 ... cxb4?! 7 Nb5! is very strong – 7 ... bxa3+ 8 c3! and White is ready to launch a big attack involving Bxa3 and
Nd6+. In that case the knight would wreak a wonderful revenge on the bishop that dared to pin it!

The Aggressive 7 Qg4

Here there are features akin to the Poisoned Pawn Variation in Chapter One as White often goes hunting with Qg4 and Qxg7. He smashes up
the black kingside, but on the other hand his centre comes under intense pressure after the standard ... c7-c5 etc. Note that now the white
centre is no longer supported by that apparently feeble pawn on c3. Often Black ends up trading his dark-squared bishop for the knight on c3
(though not directly with Bxc3+).
It is worth comparing the position reached in Game 13 after 7 Qg4 Ne7 8 bxa5 dxc3 9 Qxg7 Rg8 10
Qxh7 Nbc6 11 Nf3 (Diagram 4),

Diagram 4
Another poisoned pawn variation!

with that in the Poisoned Pawn Variation.

White doesn’t have any serious problems down the c-file, which is blocked after ... d4xc3. Because of
this he hasn’t been obliged to play Ne2, so that the knight has gone to the more natural square f3, where it
defends e5 or can be used aggressively with Ng5. The drawback to this is that the e5-pawn isn’t as secure
as it would be after f2-f4. And in fact in Game 14 Anand does prefer 11 f4 to bolster e5 before
developing his knight to f3.
As the knight isn’t on e2, the bishop on f1 has freedom to be developed to d3 or possibly b5. On the other hand this too remains
problematical as it would leave the g2-pawn hanging to a potential ... Rxg2.

The Positional 7 Nb5

White decides he doesn’t want to get involved in the excesses of a queen foray on the kingside, at least not for the time being, and who can
blame him? Instead he takes the chance to free his knight while he has the chance. Black’s reply 7 ... Bc7 is practically forced as allowing a
knight check on d6 would be painful. It is ironic that these two pieces are still involved in their own personal war, though this time the knight is
in the ascendancy as it can decide whether they are exchanged by Nxc7+. In this variation White is more interested in establishing a positional
plus in the centre by getting firm control of d4 than in blowing Black’s brains out.
The theory in the 7 Qg4 variation runs 6 b4 cxd4 7 Qg4 Ne7 8 bxa5 In Game 15, White played 8 Nb5
dxc3 9 Qxg7 Rg8 10 Qxh7 Nbc6. Now 11 Nf3 is Game 13 and 11 f4 Game 14.
In the 7 Nb5 line, it goes 7 Nb5 Bc7 8 f4 when 8 ... a5 is Game 16 and 8 ... Bd7 Game 17.
These lines are much less critical than the Qxg7 line of Chapter One though, as will be seen in the
games, there are some sharp tactical variations involving pawn sacrifices. If you play this line as White,
unless Black chooses the double edged 5 ... Ba5, it is pretty much ‘theory proof’.

Illustrative Games

Game 8
Fischer-Larsen
Candidates match, Denver 1971

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

Preventing a black piece arriving on a4 (as occurred in Game 6) and preparing to develop the queen’s
bishop at a3. The less committal 7 Nf3 (Games 9-11) is more usual. Also, Black can insist on ... Qa4 by
playing straightaway 6 ... Qa5 as in Game 12.
7 ... Nbc6 8 Nf3 Bd7 9 Bd3 Qc7 10 0-0 c4 11 Be2 f6 12 Re1 Ng6 13 Ba3 (Diagram 5)

Diagram 5
White sacrifices the centre

This is a fairly typical theme. White gambits a pawn in the interests of opening the position, acquiring
the bishop pair and pinning down the black king – pretty good value for a measly pawn.
13 ... fxe5 14 dxe5 Ncxe5 15 Nxe5 Nxe5 16 Qd4 Ng6 17 Bh5 Kf7
A very brave decision from Larsen. 17 ... 0-0-0 was possible as 18 Qxa7 b6 gets the queens off and
keeps White’s advantage to a minimum, although also possible is 18 a5, keeping the initiative.
18 f4 Rhe8 19 f5 exf5 20 Qxd5+ Kf6
This unpalatable move is more or less forced as 20 ... Be6? allows 21 Rxe6! Rxe6 22 Qxf5+ Rf6 23
Qd5+ Re6 24 Rf1+ winning.
21 Bf3 Ne5 22 Qd4 Kg6 23 Rxe5 Qxe5 24 Qxd7 Rad8 25 Qxb7 (Diagram 6)

 Diagram 6
How should Black defend?

Black has achieved an impressive centralisation of his forces, but the most relevant feature of the
position, unsurprisingly, is his shaky king.
25 ... Qe3+?
Black opts for a forcing continuation but Fischer handles the tactics brilliantly and emerges in a
winning endgame. Black’s best chance was the cold-blooded 25 ... Qxc3, daring White to do his worst
after 26 Qc6+ Kg5. One possible line then is 27 h4+ Kxh4 28 Qc7 Qxa1+ 29 Bc1 Qxc1+ 30 Kh2 which
appears decisive for White but now Black has the amazing defence 30 ... Rd6!! which wins.
It is not surprising that Larsen shied away from such variations.
26 Kf1 Rd2 27 Qc6+ Re6 28 Bc5
The bishop rushes to the defence and now Black has nothing better than the game continuation.
28 ... Rf2+ 29 Kg1 Rxg2+ 30 Kxg2 Qd2+ 31 Kh1 Rxc6 32 Bxc6 Qxc3
Giving up the a-pawn is hopeless. Black had to try 32 ... a5.
33 Rg1+ Kf6 34 Bxa7 g5 35 Bb6 Qxc2 36 a5 Qb2 37 Bd8+ Ke6 38 a6 Qa3 39 Bb7 Qc5 40 Rb1 c3 41 Bb6 1-0

Game 9
Short-Ivanchuk
Horgen 1995

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 Nf3 Qc7 8 h4

8 a4 Nbc6 9 Bd3 Bd7 would transpose to the previous game, but Black more often prefers 8 ... b6
planning to exchange light-squared bishops by ... Ba6.
8 ... Bd7 9 h5 h6 10 Bd3 Ba4 11 dxc5
An anti-positional move by which White seriously compromises his pawn structure but hopes to
benefit from the active piece play that he generates with the opening of the position.
11 ... Nd7 12 Rh4
An unusual but powerful way to develop the rook. This piece will find great activity along the third
rank, especially from the g4-square.
12 ... Qa5 13 Be3 (Diagram 7)
 Diagram 7
Ignoring the attack on c3

13 ... Rc8 14 Rb1 Bc6 15 Qd2 Qxa3 16 Rg4 Kf8 17 Rbb4
The king’s rook has been having a successful career along the fourth rank and so his colleague decides
to join him. White’s structure is horrible but his pieces are tremendously active and Black is struggling to
coordinate his forces.
17 ... Nf5 18 Bxf5 exf5 19 Rgf4 Qa1+ 20 Qd1 Qxd1+
If 20 ... Qxc3+ then 21 Bd2 Qa3 22 Rxf5 and Rbf4 is coming.
21 Kxd1 Re8 22 Rxf5 Kg8 23 Rg4 (Diagram 8)

Diagram 8
Active rooks!

23 ... Re7
A curious variation here is 23 ... Nf8 24 Rf6! and suddenly Black cannot defend the h-pawn.
 24 Rg3 Kh7 25 Bd4 Rhe8 26 Rf4 Nf8 27 Rfg4 g5 28 hxg6+ Nxg6 29 Rh3 Bd7 30 Ng5+ Kg8 31 e6
White has been angling to force open the kingside for some time and finally achieves it with this pawn
advance.
31 ... Bxe6 32 Nxe6 Rxe6 33 Rxh6 Re1+ 34 Kd2 R8e2+ 35 Kd3 Kf8 36 Rh5 Re4 37 Rxe4 dxe4+
38 Kd2 Rb1 39 c6 bxc6 40 Bxa7 Ke7 41 Bd4 Rb5 42 Rxb5 cxb5 43 c4
Now with a bishop against a knight and an extra pawn, White wins easily.
43 ... bxc4 44 Kc3 Nf4 45 g3 Ne6 46 Kxc4 Kd6 47 Bf6 Kc6 48 g4 Kd6 49 c3 Kc6 50 Be5 Nc5 51 Kd4 Nd3 52 Bg3 Nc5 53
Bf4 1-0

Game 10
Kir.Georgiev-Yusupov
Las Palmas 1993

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 Ne7 5 a3 Bxc3+ 6 bxc3 c5 7 Nf3 b6 8 Bb5+ Bd7 9 Bd3 Ba4 10 h4
In the next game White tries 10 dxc5!? to lessen the following queenside clamp.
10 ... h6 11 h5 c4
The previous game demonstrated how active the white pieces can become when the game is opened up
and so here Black opts for a strategy of closing the position.
12 Be2 Kd7 (Diagram 9)

Diagram 9
Black has a queenside clamp

It looks peculiar to send the king wandering off in the middle of the board, but the blocked nature of the
position means that he is in no immediate danger.
13 Nh4 Qg8 14 Bg4 Qh7
This manoeuvre, relocating the queen on h7, is now well known in the Winawer but whoever thought of
it first must be a very imaginative player.
15 Ra2 Na6 16 Rh3 Nc7 17 Kf1 Raf8 18 Kg1 g5 19 Nf3 Nc6 20 Nh2 f5 21 exf6 Rxf6 22 Re3 Rhf8
23 f3
Until this point the game is balanced but here White makes a serious mistake in taking away this square
from his knight. Much better was 23 Re2, planning Nf3.
23 ... Qf7 24 Bd2 Ne8 25 Bh3 Qxh5 26 Ng4
Following his slip on move 23, White has now had to pay a price of one pawn to activate his knight.
 26 ... Rf5 27 Nf2 Qg6 28 Bxf5 exf5 (Diagram 10)

Diagram 10
The white rooks are impotent

It was perhaps not necessary for Black to give up the exchange here but the resulting position is very
instructive. It demonstrates how placing simple numerical values on pieces, without considering all the
features of the position, is unsatisfactory. White has a rook for a knight but the black knights are scarcely
less active than the white rooks in this blocked position.
29 Re2 Nf6 30 Qe1 Nh5 31 Ra1 Ng3 32 Re5
White’s rooks are so impotent that he is desperate to jettison one of them to free his position.
32 ... Qd6 33 Nh1 Nxh1 34 Kxh1 Nxe5 35 dxe5 Qe6 36 Be3 Qxe5 37 Qd2 Re8 38 Bf2 f4 39 Rd1 Kd6 40 Re1 Qf5 41 Rxe8
Bxe8 42 Bd4 Ba4 43 Qe1 Kd7 44 Qb1 h5 45 Qb4 Bc6 46 Qb1 g4 47 fxg4 hxg4 48 Qe1 Ba4 0-1

Game 11
Nunn-P.Nikolic
Belgrade 1991

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ 6 bxc3 Ne7 7 Nf3 b6 8 Bb5+ Bd7 9 Bd3 Ba4 10 dxc5
An antipositional move but White is prepared to wreck his pawn structure in the interests of active
piece play.
10 ... bxc5 11 0-0 c4 12 Be2 Ng6 13 Ng5 (Diagram 11)
 Diagram 11
Again White sacrifices his centre

13 ... Nxe5 14 f4 Nd3 15 Bxd3 cxd3 16 f5 e5 17 Qh5
White’s attacking scheme is rather crude, but Black’s lack of development makes it difficult for him to
organise a defence.
17 ... Qe7 18 Ne6
White is anxious to prevent Black from playing ... 0-0-0 but in fact he would do better to play 18 cxd3
as 18 ... Nc6 19 f6 gxf6 20 Nxh7 gives White a very good position.
18 ... Kd7 19 Bg5 f6 20 Be3 Na6 21 cxd3 g6 22 Qg4 Bc6 23 Nd4
A tremendous sacrifice from White that leaves Black with terrible long term problems on the dark
squares.
23 ... exd4 24 Bxd4 h5 25 Qh3 g5 26 Rfe1 Qf8 27 Re6 Rh6 28 Qe3 Re8 29 Re1
White’s pieces are so active that he is scarcely missing his knight.
29 ... Nc7 30 Bc5 Qf7 31 Be7 (Diagram 12)

 Diagram 12
A decisive block for a touchdown

Cleverly creating space on the queenside for the queen to infiltrate.
31 ... Rh7 32 Qc5 Bb7 33 Qxa7 Kc8 34 Rb1 Rxe7 35 Rxb7 1-0

Game 12
Nijboer-P.Nikolic
Dutch Championship, Rotterdam 1997

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

A rare move to play this so soon, but one gaining in popularity.
7 Bd2 Qa4 8 Qb1 c4 9 h4!?
White grabs some kingside space. 9 Nh3!? Is an interesting alternative, e.g. 9 ... Nc6 10 g3 Bd7 11 Nf4
0-0-0 12 Bh3 (12 h4!? is worth a try, too) 12 ... h5 13 Qd1 h4 14 g4 f6 15 0-0 Nge7 16 exf6 gxf6 17 g5
Rdg8 18 Nxe6 with unclear prospects in Krivoshey-Nikolaev 1995.
9 ... Nc6 10 h5 h6 11 Ne2 Nge7 12 Qc1 Bd7 13 g3?!
13 g4 was a more natural continuation.
13 ... 0-0-0 14 Bh3 Kb8 15 Be3 Ka8 (Diagram 13)

Diagram 13
Black’s king is very secure

Safely stowed. It’s difficult for anybody to achieve much in such blocked middlegames.
16 Qd2 Rc8 17 g4 Nb8 18 Ng3
18 0-0 Rc6 19 Rfb1 planning Rb4 may have been a preferable choice.
18 ... Rc6 19 0-0 Rb6 20 f4 g6 21 Kh2 a6 22 Rf2 Nbc6 23 f5?
A dubious decision. It is very difficult for White to improve his position without this thematic
advancement, but here it was much better to wait with active operations on the kingside. Now White gets
an unpleasant position.
23 ... exf5 24 gxf5 Bxf5 25 Bxf5 Nxf5 26 Nxf5 gxf5 27 Rxf5 Nd8
Planning to transfer the knight to the blockading square e6 and opening the way for the queen.
28 Bf2 Qd7 29 Qf4 Ne6 30 Qg4 Qa4 31 Qd1 Ng5! 32 Be3 Ne4 (Diagram 14)

 Diagram 14
The black knight stands proud

Now the consequences of White’s unfortunate operation on move 23 are quite evident: His forces are
totally uncoordinated, the black knight dominates over the white bishop, and White’s queenside pawns are
terribly weak.
33 Rxf7 Rb2 34 Rc1 Nxc3 35 Qf3 Ka7 36 e6 Re8 37 Rf8?
A mistake in a difficult situation. 37 e7 was better.
37 ... Rxe6 38 Qg3 Re8 39 Bxh6 Rxc2+ 40 Rxc2 Qxc2+ 0-1
Now 41 Rf2 Re2 is winning, as is 41 Kh3 Ne4.

Game 13
Shirov-Akopian
Merida 2000

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

Another move entirely, and it retains its adherents amongst strong GMs. This sharp line became very
popular in modern practice, mostly thanks to the efforts of Akopian and the other Armenian grandmasters,
Lputian and Vaganian.
6 b4!?
A gambit for activity. 6 Bd2 and 6 Qg4 are the main alternatives. At Hastings 1986-87, Chandler tried
6 dxc5 vs Lputian, but after 6 ... Bxc3+ 7 bxc3 the experiment brought him nothing, and he was lucky to
draw.
6 ... cxd4
On 6 ... cxb4 7 Nb5 White gets very good play for his pawn.
7 Qg4
The alternative main line, 7 Nb5, is seen in
Games 16 and 17.
7 ... Ne7 8 bxa5
White can play Nb5 here too – see Game 15.
8 ... dxc3 9 Qxg7 Rg8 10 Qxh7
There is much similarity with the lines from the normal 5 ... Bxc3+ Winawer here, but some nuances
too.
10 ... Nbc6 11 Nf3
In the next game White defends the pawn with 11 f4.
11 ... Qc7 12 Bf4 Bd7 13 Bg3
Shielding the g-pawn and so allowing the other bishop out to play.
13 ... 0-0-0 14 Bd3 Qxa5 15 0-0 Qc5
An attempt to improve on 15 ... Qa4 16 Rfe1 Ng6 17 Qh6 Rh8 18 Qe3 d4 19 Qe2 Nce7 20 Rab1 Bc6
21 Rb4 Qxa3 22 Rxd4 Bd5 23 Ng5 De Firmian-Lputian, Yerevan Olympiad 1996, when White had an
edge.
16 Qxf7
A not very successful try for the advantage in this double-edged line. White wins a pawn, but invites a
powerful exchange sacrifice down the new f-file. Alternatives are:
a) 16 Qh4 a6 17 Ng5 Be8 18 Qf4 Nd4 19 Rfe1 Nef5 was unclear in J.Polgar-Vaganian, Groningen
1993.
b)16 Qh4 Rg7 17 Rfb1 a6 18 h3 Rdg8 19 Kh2 d4 20 Be4 Nd5 21 Bxd5 Qxd5 22 Rd1 Nxe5 23 Nxe5
Qxe5 24 Rxd4 Rxg3 25 fxg3 Bc6 26 Rad1 f6 27 Rg4 Re8 28 Rc4 Qe2 29 Rxc6+ bxc6 30 Qd4 was a
White win in Lutz-Mueller, German Championship, Altenkirchen 1999.
c) 16 Rfe1 could be recommended; 16 ... Rh8 17 Qxf7 Rdf8 18 Qg7 Nf5 19 Bxf5 Rhg8 20 Qh6 Rxf5
21 Qe3 was Dolmatov-Lputian, Manila 1990. White won it.
16 ... Rdf8 17 Qh7 Rxf3! (Diagram 15)

Diagram 15
The French exchange sacrifice

18 gxf3 Nxe5 19 Be2
Black would be doing very well after 19 f4 Nxd3 20 cxd3 d4.
19 ... d4 20 Rfe1 Bc6
Akopian has great compensation for the exchange.
21 Kf1 Nxf3 22 Bxf3 Bxf3 23 Re5 Bd5 24 Kg1
Kd7 25 Rd1?
25 Qf7! was unclear.
25 ... Qxa3 26 Qh4 Qa4?
A terrible mistake, at the moment when Black could get a big advantage with the dynamic 26 ... d3!,
e.g.
a) 27 cxd3? c2 and the passed pawn is splendid;
b) 27 Rxd3? Qc1+ with excellent play;
c) 27 Qf4 dxc2 (or maybe 27 ... d2!?) 28 Rc1 Rg4!.
27 Rxe6! (Diagram 16)

Diagram 16
Black is punished for his mistake

The spectacular refutation of Black’s last move.
27 ... Rxg3+
On either 27 ... Bxe6 28 Rxd4+ or 27 ... Kxe6 28 Re1+ the new attack crunches through. So Black tries
this defensive method instead.
28 hxg3 Nf5 29 Qh7+ Kxe6 30 Qg6+ Kd7
30 ... Ke5 31 f4+ Ke4 32 Re1+ Kf3 33 Qxf5 wins.
31 Qxf5+ Kc6 32 Qf6+ Kc5 33 Qe7+ Kc4 34 Qe5
Kc5 35 f4!
Black’s queenside assets are frozen and White’s mobile passed pawns quickly decide the game.
35 ... b5 36 f5 Kc6 37 Qf6+
37 Qxd4 Qxd4+ 38 Rxd4 allows some counterplay after 38 ... a5.
37 ... Kc5 38 Qe5 Kc6 39 Rxd4! Qa1+ 40 Kf2 Qh1 41 Rxd5!
On 41 Rd3 b4! Is tricky.
41 ... Qxd5 42 Qxc3+ Kd6 43 Qd3 a5
He also loses after the exchange of queens, 43 ... Qxd3 44 cxd3 a5 45 Ke3 a4 46 Kd2 Ke5 47 g4 a3 48
Kc2 b4 49 Kb3 Kf6 50 d4 etc.
44 Ke3 a4 45 Qxd5+ Kxd5 46 Kd2 b4 47 Kc1 1-0
and White will win after 47 ... Ke5 48 g4 Kf6 49 Kb2 Ke5 50 c3.

Game 14
Anand-Khalifman
Linares 2000

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Ba5 6 b4 cxd4 7 Qg4 Ne7 8 bxa5 dxc3 9 Qxg7 Rg8 10 Qxh7
Nbc6 11 f4
The other way of hanging on to e5.
11 ... Qxa5
11 ... Bd7 is an option. 12 Rb1 Qc7 13 Ne2 0-0-0 14 Qd3 d4 15 Ng3 Ng6 16 Be2 Nh4 was the
continuation in Sulskis-Ivanov, Pardubice open 1996, where Black won.
12 Nf3 Bd7 13 Rb1
The critical move. 13 Ng5 invites the dangerous exchange sacrifice 13 ... Rxg5! 14 fxg5 0-0-0 15 Qxf7
Nf5 16 Bf4 Qa4 as in Martin Gonzalez-Comas Fabrego, Spanish Championship 1993, where Black won.
The white queen is a bit out of it and Black’s active pieces generate much play.
13 ... 0-0-0 14 Qd3 (Diagram 17)

Diagram 17
A similar position to Chapter One

Back home, after having inflicted her damage.
14 ... Nf5 15 Rg1
And certainly not 15 Rb5? Qa4 16 Qxc3 Rxg2! 17 Bxg2 Qxb5 with a big advantage for Black.
15 ... d4
Bidding for active play before White gets his act together.
16 g4!
A natural and strong move.
16 ... Nfe7 17 Rg3 Be8 18 h4 Nf5
 The bid for play with 18 ... f6 19 exf6 Bg6 fails to 20 fxe7! Bxd3 21 exd8Q+ Qxd8 22 Bxd3 and wins,
and 18 ... Nd5 19 Ng5 is good for White too.
19 Rg1 Ne3
Otherwise from where is the black counterplay to come? The white pawns are already steaming away
on the kingside.
20 Bxe3 dxe3 21 Qxe3 Ne7
Khalifman enjoys considerable activity, but it hardly looks like it is going to be quite sufficient for the
pawns, but 21 ... Qxa3 22 Rb3 Qa1+ 23 Kf2 isn’t good enough either.
22 Bc4 Bc6 23 Rb3 Kb8 24 Qxc3
Another one bites the dust. Note, again, the white major pieces all on the third rank.
24 ... Qc7
Nor does 24 ... Qc5 25 Be2 Qxc3+ 26 Rxc3 Nd5 27 Rc4 Ne3 28 Rc5 Nxg4 29 Ng5 promise Black too
many chances to survival.
25 Be2
On the immediate 25 h5 Ba4! causes annoyance.
25 ... Nd5 26 Qc4 Qa5+ 27 Kf2 Ba4
Maybe 27 ... Rc8!? was slightly stronger.
28 Rb2 Rc8 29 Qd4 Rc3
In the case of 29 ... Rxc2 30 Rxc2 Bxc2 31 h5 it is difficult to believe that it will be possible to stop
the strong white pawns on the kingside.
30 Bd3 Rgc8 31 Rc1 (Diagram 18)

Diagram 18
White consolidates his extra pawns

31 h5 Bxc2 32 Rxc2 Rxc2+ 33 Bxc2 Rxc2+ also looks very nice for White after 34 Kg3.
31 ... Rd8 32 Rcb1 Bc6 33 Rb3 Rxb3 34 Rxb3 Nf6 35 Qb4 Nxg4+?!
A tougher defence was 35 ... Qd5! when White counters best with 36 Be2! Nxg4+ 37 Ke1! Rh8, still
keeping a clear plus.
36 Kg3 Qd5 37 Bh7!
A nice move, quickly deciding the game in White’s favour.
 37 ... Qd1
37 ... Qd7 38 Rd3 Bd5 39 Kxg4 f5+ 40 Kh3 Qxh7 41 c4 Qh5 42 cxd5 Qg4+ 43 Kh2 Rg8 44 Qd2!
tidies up and wins.
38 Kxg4 Rd2 39 Kg5! Bxf3 40 Bd3 Rg2+ 41 Kf6
Safe in the hills.
41 ... a6 42 Rb1 1-0
And now 42 ... Qd2 43 Qxd2 Rxd2 44 Kxf7 wins easily.

Game 15
Yermolinsky-Vaganian
Tilburg 1993

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Ba5 6 b4 cxd4 7 Qg4 Ne7 8 Nb5

Another line, blending elements of 7 Nb5 and 7 Qg4.
8 ... Bc7 9 Qxg7 Rg8 10 Qxh7 a6
10 ... Bxe5 11 Nf3 is fine for White.
11 Nxc7+ Qxc7 12 Ne2 Bd7!?
12 ... Qxe5 is a major and obvious alternative. Then 13 Qd3 Nbc6 14 Bb2 Bd7 15 0-0-0 Qf6 16 Nxd4
Qxf2 17 Nf3 Nf5 was about equal in Chandler-Vaganian, Manila interzonal 1990.
After 13 Bb2 Qf6! leads to probably equal chances, as does 13 ... Qc7 14 f4 Nbc6 15 Qd3 Nf5 too.
13 Bb2
13 f4 fails to hold on to e5 because of 13 ... Ba4 14 Kd1 Rg6!, shutting the white queen out of the new
action zone at c2.
13 ... a5!? (Diagram 19)

Diagram 19
Black initiates counterplay

A clever move, bringing the other rook into it.
14 Nxd4
If 14 f4 axb4 15 axb4 Rxa1+ 16 Bxa1 Qb6!? and things will turn Black’s way after 17 c3 Qa7 18 Qb1
d3! 19 Nd4 Nbc6 20 g3 Nxd4 21 cxd4 Bb5.
14..Qxe5+ 15 Be2 Nbc6 16 Rd1
If 16 Qd3 Black may entertain the complexity of 16 ... axb4 (16 ... Rxg2 is also worth a go) 17 0-0
bxa3 18 Nb5 Qxb2 19 Nc7+ Kd8 20 Nxa8.
16 ... axb4 17 axb4 Ra2 18 Nxc6
Or 18 Bc3 Rxg2.
18 ... Qxb2 19 Nxe7 Rh8! 20 Qd3 Kxe7 (Diagram 20)

Diagram 20
The black pieces dominate

Taking stock, we see that White cannot hang on to his extra pawn whilst all of the black pieces are very
well placed.
21 h4 Rc8
He could also have gone in for 21 ... Qxb4+ 22 c3 Qc5 23 h5.
22 Rd2 Qxb4 23 0-0 Ra3!
Shifting the white pieces away from defence of c2.
24 Qd4 Qxd4 25 Rxd4 Rxc2 26 Bf3 b5 27 h5 Rc4 28 Rd2 f5 29 Rb2 Rac3 30 Be2 Rc2 31 Rfb1
Rxb2 32 Rxb2 Rc8 33 f4
On 33 Bxb5 the pin 33 ... Rb8 wins.
33 ... Rb8 34 Kf2 b4 35 Ke3 b3 36 g3 Rb4 37 Bd3 Bb5 38 Bb1 d4+ 39 Kd2 Kf6 40 Kc1 Rc4+ 0-1

Game 16
Almasi-Korchnoi
European Cup Final, Budapest 1996

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Ba5 6 b4 cxd4 7 Nb5

Theory also respects this line. The character of the middlegames which stem from it is altogether more
settled than those that arise after 7 Qg4.
7 ... Bc7 8 f4 a5
Chiselling. In the next game Black plays 8 ... Bd7.
9 Bd2 Nh6
A novelty. 9 ... Ne7 10 Nf3 Nf5 11 Bd3 0-0 12 Bxf5 exf5 13 Nfxd4 Nc6 14 Nxc6 bxc6 15 Nxc7 Qxc7
16 Be3 f6 was equal in Minasian-Lputian, World Open, Philadelphia 1994. Obviously there are
similarities with developing the knight at e7, especially if it ends up at f5 anyway.
10 Bd3 Bd7 11 Nxc7+ Qxc7 12 Nf3 Nf5 13 b5! (Diagram 21)

Diagram 21
The b5-pawn stifles Black

An interesting way of shutting up the queenside. The d7 bishop is now gasping. If White had set his
heart on it, he could have rounded up the d4-pawn. But instead he plays it like a gambit. The result is
spectacular.
13 ... Qc5 14 Qe2 0-0 15 a4 f6 16 0-0 Ne3 17 Rfe1
17 exf6!? deserved consideration, e.g. 17 ... Nxf1 18 fxg7 Rf7 19 Rxf1 Rxg7 20 Ng5 with an attack and
initiative.
17 ... Nc4 18 Bc1 b6 19 exf6 Rxf6 20 Ne5 Be8?
Definitely misguided. He ought to have taken on e5.
21 Ng4 Rf8 22 Qxe6+ Bf7 23 Nh6+! (Diagram 22)

 Diagram 22
An unexpected sacrifice

An inspired sacrifice, but 23 Qf5! was also very strong, e.g. 23 ... Bg6 24 Qg5 Ra7 25 f5 Bf7 26 Nf6+
Kh8 27 Nxh7! with a stomping attack.
23 ... gxh6 24 Qxh6 Bg6 25 Bxg6 Ra7
An admission that he is busted, but 25 ... hxg6 26 Qxg6+ Kh8 27 Qh6+ Kg8 28 Re6 finishes the job.
26 Bd3 Rg7 27 f5 Nd7
The knight’s first move of the game!!
28 Qe6+ Rgf7 29 Bh6 Rc8 30 f6 Nf8 31 Qg4+
Kh8 32 Re7 Rc7 33 Qg7+! 1-0
A nice finish to a beautiful game! Black resigned. The finale would be 33 Qg7+ Rxg7 34 fxg7+ Kg8 35
gxf8R mate.

Game 17
Hector-Barsov
Vikings GM, York 1999

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Ba5!? 6 b4 cxd4 7 Nb5 Bc7 8 f4 Bd7!?

8 ... Nh6 9 Nf3 Bd7 10 Nxc7+ Qxc7 11 Bd3 a6 12 a4 Nf5 13 Qd2 Nc6 14 Bb2 0-0 15 Bxf5 exf5 16
Bxd4 Nxd4 17 Nxd4 led to a classic good knight against bad bishop scenario in Almasi-Lputian, Pula
1997, and White won.
9 Nxd4
The character of these middlegames is determined by which bishops take/are taken by which knights.
In Van den Doel-Lputian, Wijk aan Zee 1999, after 9 Nf3 Bxb5 10 Bxb5+ Nd7! 11 0-0 Ne7 12 g4 a6 13
Bd3 h5! 14 h3 hxg4 15 hxg4 Nb6 Black was very comfortable, and went on to win.
9 Nxc7+!? Qxc7 10 Nf3 would be another kettle of fish.
9 ... Nh6 10 Ngf3 Bb6
10 ... Nc6 11 Bb2 Nxd4 12 Bxd4 Nf5 13 Bf2 Bb6 14 Qd2 Bxf2+ 15 Qxf2 was better for White, with
her nicer bishop in Arakhamia-Shaw, Edinburgh 1999.
11 Bd3 Nc6 12 c3 Bxd4!?
An interesting novelty in a well-known position. Barsov prefers to keep a knight on the board. 12 ...
Nxd4 leads to a different balance of minor pieces, and after 13 cxd4 a6 14 a4 Nf5 15 Bb2 Ne3 16 Qe2
Nc4 17 0-0 Rc8 18 Bc1 Qe7 Black was doing fine in Minasian-Lputian, Lvov zonal 1990.
13 cxd4 Nf5 14 Bb2 Qb6 15 Be2
Black has nothing to worry about after 15 Bxf5 exf5.
15 ... Na5 16 Kf2?!
16 bxa5!? was worthy of attention as 16 ... Qxb2 17 Qc1 Qxc1+ 18 Rxc1 Ke7 19 Kd2 and White keeps
a small edge.
16 ... Nc4 17 Bc1 h5 (Diagram 23)

Diagram 23


18 g3 0-0-0 19 Rb1
Preparing to shift the f5 knight with 19 h3 would permit it to lodge elsewhere after 19 ... h4!? 20 g4
Ng3.
19 ... f6
It is clear now that it is only Black who can claim an advantage.
20 a4 Kb8 21 Re1 Be8! 22 Bd3 Bg6
A standard re-routing and activation of the bishop, which lands outside the constricting pawn chain.
The black knights are ideally placed

23 a5 Qc6 24 b5 Qd7 25 Bd2 Rc8 26 Bb4 Ka8
26 ... Qf7!? was an option.
27 Bc5 Qf7 28 Qa4?!
It was much better to play 28 Bxc4!? dxc4 29 Rc1 Ne7 30 Rxc4 Nd5. Black has great compensation
for the pawn, but White has something to suffer for at least.
28 ... h4!
Beginning of the crucial attack.
29 Bxc4
29 g4?! Nh6 leaves Black winning.
29 ... hxg3+ 30 hxg3 dxc4 31 Qxc4??
Final mistake. Good or bad White had to play 31 Rbc1! Rh3.
31 ... Nxg3! (Diagram 24)

Diagram 24
Removing the bodyguard

32 Kxg3
On 32 Rb3 Ne4+ 33 Kg1 fxe5 34 fxe5 Rh3! and Black wins.
32 ... Bxb1 33 exf6
33 Rxb1? Qg6+ wins the rook.
33 ... gxf6 34 Qxe6 Rhg8+ 35 Kf2 Qg6 36 a6
Too little too late.
36 ... Qg2+ 37 Ke3 Rge8 0-1
Chapter Three
The Winawer: Other Lines
This section contains an assortment of ideas for White and games with a quick ... b7-b6 by Black.

Black Plays an Early ... b7-b6

This fianchetto doesn’t come any earlier in the Winawer than after 1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 b6.
Black prepares the immediate development of his queen’s bishop. This may well be a sign that he intends
... Ba6 to exchange off light-squared bishops, though the former World Champion Tigran Petrosian, the
arch priest of manoeuvring in closed positions, was wont to play ... Qd7, ... Bb7, ... Nc6, etc. and then
castle queenside. Only later would he think about clearing the way for the c7-c5 advance.

TIP: The black plan with ... Qd7, ... Bb7, ... Nc6 etc, can sometimes be viewed as a mirror image of
the black strategy in the main lines of the King’s Indian Defence.

One question that now arises is how Black intends to answer 5 Qg4. In the style of Petrosian, Timman
comes up with the retrograde 5 ... Bf8 in his game with Shirov, which is Game 18. This looks faintly
ludicrous, but Black is arguing that the white queen is worse placed on g4 than d1 and so the time Black
spends undeveloping his kingside will be recouped later on when White is obliged to waste time re-
positioning his queen. Timman scores a great success against Shirov, but the jury is still out on this idea,
to say the least.

TIP: A loss of time is much less important in a closed position.

A refined version of Black’s idea is seen in Game 19. Black begins with 4 ... Qd7, so that he can answer
5 Qg4 with 5 ... f5! when his queen defends g7. You may recall this defensive motif from the game
Kasparov-Short given in Chapter One. As Black doesn’t need to play ... Bf8, it makes Qg4 less attractive
for White and he settles for the solid 5 Nf3, when 5 ... b6 was played. This looks okay for Black, but
Karpov, in a rare outing with the French, didn’t know what to do and quickly fell into a bind.

White Plays 4 e5 Ne7 5 Nf3

This was an old favourite of English GMs Hodgson and Gallagher. In fact Joe Gallagher recently
annotated one of his wins with it in New In Chess Magazine, so it is still garnering points him! Rather
than 5 a3, White makes a nonchalant developing move which, in fact, commits him to a pawn sacrifice if
he wants to keep the edge. Theory says it is okay for Black, but over-the-board it is offers White
attractive attacking options. It’s all there in Game 20.

White Avoids e4-e5

A considerable number of players avoid 4 e5, either because the rigid nature of the pawn structure doesn’t
appeal to them or because they are frightened off by the huge amount of theory.
White’s option range from extremely solid positional systems to violent attacking lines. None of these variations has ever achieved great
popularity, though you should certainly know something about them if you intend playing the Winawer as Black.
White Plays 4 Ne2
This is the most popular of the 4th move alternatives. White overprotects c3 so that if Bxc3+ he can recapture with the knight and keep his
queenside pawns intact. Hence in this variation queenside castling becomes an option for White – something very unlikely to be good after 5
e5. Because he can spirit his king away to the queenside, White can look for a direct attack on the kingside. This line will appeal to players
who like a ‘clean’ pawn structure and keep their pieces in a nice formation.
However, you cannot expect a big attack from the first moves if you are blocking in your king’s bishop with a knight on e2. There is also
the downside that the e4-pawn is being left undefended. After 4 ... dxe4 White will regain his pawn sooner or later, unless Black defends it
with the loosening ... f7-f5, but the fact remains that he can no longer cramp Black with e4-e5 as in the main lines. One very interesting idea
for Black is 4 ... Nc6!? when if 5 a3 Bxc3+ 6 Nxc3 (the opening loses all point if White recaptures 6 bxc3) 6 ... dxe4 the d4-pawn is attacked,
when White has to speculate with 7 d5!? There are three games with 4 Ne2: 4 ... dxe4 is Game 21, 4 ... b6 (highly unusual) Game 22, and 4 ...
Nc6 in Game 25.

White Plays 4 Bd2 or 4 Bd3

Similar comments to those on 4 e5 c5 5 Nf3 apply to 4 Bd2, which commits White to a pawn sacrifice. You should tread carefully as Black in
case you are meeting one of the specialists in this system! Check out Game 24.
In contrast 5 Bd3 is rather artificial looking and none too dangerous as long as Black is careful not to fall behind in development. This is the
subject of Game 25.

WARNING: These variations are tricky. If you want to play the Winawer as Black you must have a look at these white options.

Illustrative Games

Game 18
Shirov-Timman
Groningen 1996

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 b6
A more restrained system for Black, with the idea of ... Ba6 to exchange light-squared bishops. 4 ...
Qd7 (Game 19) leads to similar positions.
5 Qg4 Bf8
Looks daft when you first see it, but hanging on to the bishop has its points, just as does busting up the
white queenside with ... Bxc3. And 5 ... Bf8 is certainly not inferior to 5 ... g6 or 5 ... Kf8.
6 Bg5
As you might imagine, there are many different formations for White, and not much theoretical
consensus as to which is the best.
6 ... Qd7
6 ... f6 was certainly an option.
7 Bb5 Nc6 8 Nce2 a6 9 Ba4
There have been some games where the white bishop has been manoeuvred around to c2 in this
manner. This time it does not work out too well.
9 ... b5 10 Bb3 f6 11 exf6 Nxf6 12 Qh4 Ne7! (Diagram 1)
 Diagram 1
An imaginative pawn sacrifice

An imaginative and excellent continuation, involving the sacrifice of a pawn. 12 ... Be7 was more
staid.
13 Bxf6 Ng6 14 Qg5 gxf6 15 Qxf6 Bh6
Stopping castling and with the threat of 15 ... Rf8 trapping White’s queen.
16 Qf3
Only way out.
16 ... Qg7 17 Qg3
A better way of getting organised here might have been the line 17 c3!? Bb7 18 Bc2 0-0-0 19 Qg4 Qf6
20 Nf3 Rhg8 21 Qh5!, when White would avoid a lot of the embarrassing congestion that came his way.
17 ... Bb7 18 Nf3 0-0-0 19 0-0 Rhg8 20 c3 Qf6 21 Qh3 Bf8
Rerouting to d6. Keeping the bishop does not seem so silly now, does it?
22 Bc2 Bd6 23 Rae1?!
A mistaken plan. e6 is not vulnerable enough to warrant White attempting to hit it. 23 a4! was more to
the point.
23 ... Kb8
Black has tremendous play for his pawn, with Shirov in the unfamiliar situation of hardly having a
decent move.
24 Ng3?
A more serious inaccuracy. He had to play 24 Qh6!.
24 ... Nf4 25 Qh4 Qf8 (Diagram 2)

 Diagram 2
The white queen is in danger

Once again Timman hopes to trap Shirov’s queen. This time the threat is 26 ... Be7 27 Qxh7 Rh8, and
there is not much to be done about it.
26 Ng5
On 26 Ne5 h5 reinstates 27 ... Be7 as a threat.
26 ... Be7 27 Nxe6
This is insufficient, but taking three pawns for a piece was the best practical chance.
27 ... Bxh4 28 Nxf8 Rdxf8 29 Bxh7 Rg7 30 Bc2 c5 31 Re3
31 dxc5 d4 32 Be4 might have been a better try.
31 ... cxd4 32 cxd4 Bd8 33 Re5 Bc8 34 Rd1 Bg4 35 f3 Bd7 36 Kf2 Rh8 37 Bf5 Rxh2 38 Bxd7 Rxd7
39 Ke3 Nxg2+ 40 Kd3 0-1
A splendid performance from Jan Timman.

Game 19
Geller-Karpov
USSR Championship, Moscow 1976

1 e4 e6
A rare and quite unsuccessful outing with this defence for Karpov. He must have picked up the idea
from his own experience in making little headway against it in his 1974 match with Korchnoi, from which
he emerged as world champion.
2 d4 d5 3 Nc3 Bb4 4 e5 Qd7
A sideline with the same idea as in Game 7 – to answer Qg4 with ... f5 defending g7 along the rank.
5 Nf3 b6 6 Bd2
6 Bd3 was an alternative.
6 ... Ba6
The standard bishop swap.
7 Bxa6 Nxa6 8 0-0 Nb8
Now might have been a good time to take on c3.
 9 Ne2! Be7?!
Probably taking on d2 was better. Anatoly did not really know what he was doing.
10 Rc1! b5
Very weakening, but if c4 comes, White’s initiative would be very hard to meet.
11 Nf4 h5
Another ugly move. 11 ... Nh6 12 Nh5 Nf5 13 g4 is hard to meet, so he prepares the way into the game
for the knight. However, it is not to have much of a role in this game.
12 b3 Ba3
To discourage c4, but the bishop is offside here.
13 Rb1 a5? (Diagram 3)

Diagram 3
World champion at sea

Another inferior move. Better to have pulled the bishop back.
14 c4 c6 15 c5
Cutting off the bishop. Karpov hurries to rescue it, but it costs him a pawn. His game is already
critical.
15 ... Bb4 16 Bc1 a4
A sad necessity before White traps the bishop.
17 Nd3 Ba5 18 bxa4 bxa4 19 Qxa4 Qa7 20 Bg5 Bc7 21 Rxb8+!
Correctly preferring the initiative to just being a pawn up.
21 ... Qxb8 22 Qxc6+ Kf8 23 Nf4
For the exchange White has two pawns and a big attack. The vulnerability of e6 and g6 makes the black
position particularly bad.
23 ... Ra7 24 Nh4 Qe8 25 Qxe6!! (Diagram 4)

 Diagram 4
Spectacular and decisive

Wow!
25 ... fxe6
25 ... Qxe6 26 Nxe6+ fxe6 27 Ng6+ does not help either.
26 Nhg6+ Qxg6
After 26 ... Kf7 27 Nxh8+ Kf8 28 Nfg6+ Black will have to give up his queen anyway.
27 Nxg6+ Ke8 28 Nxh8 Ra4 29 Rd1 Ne7 30 Bxe7 Kxe7 31 Ng6+ Kf7 32 Nf4 Bxe5 33 dxe5 Rxf4
34 Rc1
They say that Geller thought for a long time before sacrificing his queen on e6. He was thinking through
the details of this rook ending.
35 ... Ke8 35 c6 Kd8 36 c7+ Kc8 37 g3 Ra4
What Geller most probably spent his time checking was the important line after 37 ... Rf5 38 f4 g5 39
a4! gxf4 40 a5! fxg3 41 a6! gxh2+ 42 Kh1 Rf2 43 Ra1 and White wins.
38 Rc6 Rxa2 39 Rxe6 g5 40 Rd6 Rd2 41 e6 Kxc7 42 e7 1-0
and as 42 ... Re2 43 Rxd5 Rxe7 44 Rxg5 leaves a trivially won ending, Karpov threw in the towel.
One of the great games in chess literature.

Game 20
Hodgson-Kupreichik
German Bundesliga 1998

1 d4 e6 2 e4 d5 3 Nc3 Bb4 4 e5 Ne7 5 Nf3

A deviation from the usual 5 a3 in the first two chapters.
5 ... c5 6 dxc5 Nbc6 7 Bd3 d4 8 a3 Ba5 9 b4 Nxb4 10 axb4 Bxb4 11 0-0 Bxc3 12 Rb1
A speculative gambit from Hodgson.
12 ... Qc7 (Diagram 5)
Diagram 5
What now for White?

What would you now play with White?
13 Ng5!?
Probably not that.
13 ... Qxe5 14 Qh5 g6 15 Qh6
The c3 bishop is not in the thick of things, you see? That’s one reason why Hodgson hoped he could
have fun over on the other side.
15 ... Ng8
Not what you want to do, but 15 ... Nf5 is met by 16 Bxf5 gxf5 17 Nxh7 or 16 ... exf5 17 Nxh7 and if
the c8 bishop moves Rxb7 will leave Black right in it.
16 Qh4 Kf8
So will castling by hand do the trick? Not really ...
17 Bf4 Qd5 18 Bd6+ Kg7 19 Nxf7!!
Splat!
19 ... Kxf7 20 Qd8!
Thud! The queen lands and threatens mate at f8.
20 ... e5
Only move.
21 Qf8+ Ke6 22 f4! (Diagram 6)

Diagram 6
The black king is doomed

Introducing the new attackers of f1 rook and the f-pawn itself.
22 ... e4 23 f5+ 1-0
A typical Hodgson game; wild, fantastical ... and probably not 100% correct.

Game 21
Kharlov-Beliavsky
Novosibirsk 1995

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 Ne2

This makes a change from 4 e5. Other ideas for White are 4 Bd2 (Game 24) and 4 Bd3 (Game 25).
4 ... dxe4
Black usually takes the pawn. More experimental options – 4 ... b6 and 4 ... Nc6 – are tried in the next
two games.
5 a3 Be7
Grabbing the pawn with 5 ... Bxc3+ 6 Nxc3 f5 has a very bad reputation, ever since Nimzowitsch lost
to Alekhine in 17 moves. But I would not be surprised if it turns out in reality to be one of Black’s better
lines. 5 ... Bxc3+ 6 Nxc3 Nc6 is well regarded as an equaliser.
6 Nxe4 b6?!
6 ... Nf6 was a simpler and superior move, similar to the Rubinstein variation (see Chapter 11) with
White’s king’s knight slightly misplaced on e2.
7 N2c3 Bb7 8 Qf3! (Diagram 7)

Diagram 7
The threat is Nf6+

8 ... Kf8
A very undesirable move to make. Forfeiting castling rights can hardly be right. He should have tried 8
... Nd7?! 9 Nd6+!? cxd6 10 Qxb7.
9 Qg3! Nc6
9 ... Qxd4 10 Qxc7! Leaves White better after either 10 ... Qd7 11 Qg3 or 10 ... Bxe4 11 Be3! Qd8 12
Qxd8+ Bxd8 13 Nxe4.
10 Be3 Bh4
10 ... Nf6 11 Nxf6 Bxf6 12 0-0-0 is pleasant for White and 10 ... Nxd4? 11 Bxd4 Qxd4 12 Rd1 wins
for him.
11 Qf4 g5
This way out is not adequate. White will have the bishop pair and the superior pawn structure. On 11
... f5 12 0-0-0! Leaves White nicely placed.
12 Nxg5 Bxg5 13 Qxg5 Qxg5 14 Bxg5 Nxd4 15 0-
0-0 c5 16 Nb5!
Offering the g2-pawn to activate his pieces. In particular White wants to get his rook to d7.
16 ... e5
White has massive compensation, in the form of initiative and development lead, after 16 ... Nxb5 17
Bxb5 Bxg2 18 Rhg1 Bf3 19 Rd7 a6 20 Bd3 Ke8 21 Rd6 b5 22 Bf4.
17 Nxd4 exd4
On 17 ... cxd4 White carries on opening up as quick as he can and stands clearly better after 18 f4! f6
(18 ... exf4 19 Bxf4 is nice) 19 Bh4 Kg7 20 fxe5 fxe5 21 Bc4.
18 c3! dxc3?!
Probably his last chance was 18 ... f6 and 19 ... Rd8. Now the white pieces enter.
19 bxc3 Kg7 20 Bc4 (Diagram 8)

Diagram 8
Black’s position is awful

20 ... b5
20 ... Bxg2 21 Rhg1 and 20 ... Nf6 21 Rd6 followed by 22 Rd7 both leave him in fearful trouble, but
losing the b-pawn like this, to deflect the white bishop away from control of f7, is not a lot better.
21 Bxb5 1-0
A sickened Beliavsky here threw in the towel.

Game 22
Zelcic-Zaja
Croatian Championship, Pula 1998

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 Ne2 b6

A rarity.
5 a3 Be7
5 ... Bxc3+!? 6 Nxc3 Ne7 7 Bd3 Ba6 8 Bxa6 Nxa6 9 Qe2 Nb8 10 Bg5 Nbc6 11 Bxe7 Nxe7 12 exd5
and here they agreed it drawn in J.Pedersen-J.Sorensen, Danish Championship 1996.
6 Nf4
6 e5?! led to no advantage for White after 6 ... Ba6 7 Be3 c5 8 Qd2 Nc6 9 0-0-0 Bxe2 10 Bxe2 c4 11
g4 Rb8 in Mach-Moravcik, Slovak Team Championship 1995.
6 ... dxe4 7 Nxe4 Bb7 8 Qf3! Bc6?!
8 ... Qxd4? 9 Nd6+ Bxd6 10 Qxb7 Bxf4 11 Qxa8 wins for White, but 8 ... c6!? was worth a go, with
interesting play after 9 Be3 Nf6 10 Bd3 Nbd7 11 Ng5!?.
9 Be3 Nf6 10 Nxf6+ Bxf6 11 Bb5! (Diagram 9)

Diagram 9
The bishop is pinned both ways
A very important move, giving White the initiative.
11 ... Qd7
On 11 ... Bxb5?! 12 Qxa8 Bc6 13 Qxa7 0-0 14 Rd1 Qd6 15 0-0 White escapes to safety with a big
material edge.
12 Bxc6 Qxc6 13 d5! Qb7
13 ... Qxc2 14 dxe6 exposes Black to a fierce initiative, and on 13 ... exd5?! 14 0-0-0 Qb5?! 15 Bd4!
Bxd4 16 Rhe1+! wins.
14 0-0-0 e5 15 Nh5 Nd7 16 Bh6!
Now his other bishop offers itself in surprising manner too.
16 ... Rg8 17 h4 0-0-0 18 Nxf6 Nxf6?!
Losing by force. A more persistent defence was 18 ... gxf6 19 h5 Kb8 20 g3.
19 Bxg7! Rxg7 20 Qxf6 Rxg2 21 Qxf7 Qa6 22
Qe6+ Kb7
Or 22 ... Kb8 23 d6 cxd6 24 Rxd6 Rc8 25 Qxe5 winning.
23 Qc6+! Kc8 24 d6 Qb7 25 d7+ Kb8 26 Qf6 Rgg8 27 Rhg1 1-0 (Diagram 10)

Diagram 10
A rook goes

Black loses a rook

Game 23
E.Berg-V.Potkin
World Under 18 Championship, Oropesa del Mar 1999

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 Ne2 Nc6!? 5 a3 Ba5!? 6 Be3 Nge7

Raymond Keene drew attention to such an idea in his biography of Nimzowitsch.
7 e5 Nf5 8 Ng3
8 b4 Bb6 9 g3 f6 would have been an interesting line.
8 ... Nxe3 9 fxe3 Qg5 10 Qf3 f6 11 h4 Qh6 12 exf6 gxf6 13 b4 Bb6 (Diagram 11)
Diagram 11
The bishop on b6 never moves again

14 Qh5+ Qg6
14 ... Qxh5 15 Nxh5 0-0 had to be okay.
15 Bd3 f5
The only consistent move, otherwise he has just lost a tempo on the line with capturing on h5 last
move.
16 Nce2
White will argue that the bishop at b6 is really a bit out of it in the given situation.
16 ... Ke7 17 Qf3 Rg8 18 Rh3 Bd7 19 Nf4 Qh6 20 c4!
Definitely the right plan. He opens up the game for his bishop and the black king is insecure.
20 ... dxc4 21 Bxc4 Rg4 22 Nge2 Rxh4 23 Rxh4 Qxh4+ 24 g3 Qg5 25 0-0-0 Re8 26 Rh1 h6?
Too ambitious. Returning it with 26 ... Kd8 was safer.
27 Rh5 Qg7 28 Bxe6! (Diagram 12)

Diagram 12
A demolition sacrifice

A spectacular refutation. The rook now constitutes an extra attacker, so White goes to work.
28 ... Bxe6 29 Nxe6 Kxe6 30 Qxf5+ Kd6 31 Qd5+ Ke7 32 Qe4+ Kd8
32 ... Kf7 loses to the same idea: 33 Rf5+.
33 Rd5+ 1-0
Black suffers massive material loss.
Despite the result in this game, the black opening experiment was not shown to be defective.

Game 24
Grabarczyk-B.Socko
Polish Championship, Plock 2000

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 Bd2

A quirky line, hardly ever seen at top levels.
4 ... dxe4 5 Qg4 Nf6
5 ... Qxd4 6 Nf3 h5!? is fun.
6 Qxg7 Rg8 7 Qh6 b6
7 ... Nc6!? 8 0-0-0 Bd7 9 f3 Nxd4 10 Nxe4 Rg6 11 Qe3 Bxd2+ 12 Rxd2 Nxe4 13 fxe4 e5 14 c3 Nc6
led to a swift draw in Yepez-Skalkotas, Nice Olympiad 1974.
8 0-0-0 Bb7 9 Nh3
A new idea, but it does not trouble Black much. 9 f3 Bxc3 10 Bxc3 Nbd7 11 Bb5 Qe7 12 d5!? 0-0-0!?
13 Bxd7+ Nxd7 14 fxe4 Rxg2 15 Qxh7 exd5 16 exd5 Qe3+ 17 Kb1 Bxd5! 18 Rxd5 Rxg1+ 19 Rd1 Rxh1
20 Rxh1 Qf3 21 Rc1 Nc5 22 Bb4 was equal in Leveille-Delisle, Quebec 1985.
9 ... Qe7 10 Nb5 Bxd2+ 11 Rxd2 Rg6 12 Qf4 Na6 13 Rg1
This move is an indication that White’s opening has not really worked.
13 ... Rg4 14 Qh6 Rg6 15 Qf4 Rg4 16 Qh6 Rd8!? (Diagram 13)

Diagram 13
Black declines the draw

Bravely passing over the draw by repetition.
17 Be2 Rg6 18 Qe3 Nb4 19 Nf4
Here if 19 Nxa7? Ra8 20 Qa3 Nbd5 21 Qxe7+ Kxe7 and g2 falls.
19 ... Rg7 20 a3 Nbd5 21 Nxd5 exd5 22 f3
22 Nxa7? c6 traps it.
22 ... a6 23 Nc3 Kf8?
Overplaying it. He should have preferred 23 ... exf3! holding equality in case of either 24 Qxf3 Ne4 25
Nxe4 dxe4 or 24 Qxe7+ Kxe7 25 Bxf3 Kd6 26 Re2 Rgg8.
24 fxe4 dxe4 25 Rf1 Rd6 26 g3 b5 27 Bd1 Rg6 28 Rdf2 Kg8 29 Rf5 Qd8 30 h4 Nd5 31 Nxd5 Bxd5
32 h5 Rge6 33 Qf4 (Diagram 14)

Diagram 14
The pressure mounts
33 ... Re8 34 h6 Kh8 35 Bh5 e3 36 Re5 Rg8
On 36 ... Bc4 37 Qxe3 Rxe5 38 dxe5 Rxh6 39 Qxh6 Bxf1 40 Bxf7 wins.
37 Qxe3 f6 38 Re7 Bc4 39 Rf4 f5? 40 Qe5+ 1-0

Game 25
Romanishin-Beliavsky
Belgrade 1993

1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 Bd3

Oleg Romanishin is always springing new opening ideas on his opponents. This one does not look very
bright at all.
4 ... dxe4 5 Bxe4 Nf6 6 Bg5 Nbd7 7 Nge2 h6 8 Bxf6 Nxf6 9 Bf3 0-0 10 0-0 c6
Perhaps White had hoped that his f3 bishop would be strong enough to give him good chances, but ...
c6 just puts it to sleep.
11 Ne4 Nxe4 12 Bxe4 Bd6
Romanishin’s opening experiment just handed Beliavsky the bishop pair and rather more than equality.
13 Qd3 f5
‘Threatening’ the h7-pawn, but Black reacts healthily by just advancing on the kingside.
14 Bf3 e5 (Diagram 15)

Diagram 15
Black frees his position

15 dxe5 Bxe5 16 Qc4+ Kh7 17 Rad1 Qf6 18 Nd4
On 18 c3 Be6 Black is doing great.
 18 ... Bd7 19 c3 Rae8 20 Qa4 a6 21 Nb3 Bc8 22 Qb4 Re7 23 Na5 g5!
Starting a strong and entirely justified attack.
24 g3 g4 25 Bg2 f4 26 Rfe1 c5! 27 Qc4
On 27 Qxc5 b6 forks two pieces.
27 ... Kg7
Black need only to organise his forces carefully. Beliavsky does the job with accuracy.
28 Nb3 b6 29 Kh1 h5 30 Nd2 f3 31 Bf1 h4 (Diagram 16)

Diagram 16
The infantry continue to advance

With his kingside overrun White has no hope of holding out for long.
32 Re4 hxg3 33 fxg3
If 33 hxg3 Rh8+ 34 Kg1 Qg6 with decisive attack.
33 ... Kh8 34 Bd3 f2! 35 Rxg4 Bxg4 36 Qxg4 Qh6 37 Nf1 Rg7 38 Qe4 Bxg3! 39 Nxg3 Rxg3 40
Qe5+ Rg7 41 Rf1 Qh4 42 Be2 Qh3 0-1
On 43 Qe4 Qxf1+! 44 Bxf1 Rg1 is mate, or 43 Qh5+ Qxh5 44 Bxh5 Rfg8 and termination at g1.
Chapter Four
3 Nc3 Nf6 4 e5:
The Modern Variation
Introduction
As its name suggests, this is one of the oldest variations in the French. It is a curious thing but the
Classical has always been less popular among amateur players than the Winawer, even though the
Classical is currently more popular than the Winawer at elite level!
Black attacks e4 with a sound developing move and goads the pawn forwards. Then the white centre
can be undermined with the traditional flanking blows ... c7-c5 and, not quite so often, ... f7-f6. The first
chapter in the Classical considers what happens if White advances straightaway with 4 e5; the two
chapters that follow consider the pin with 4 Bg5.
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4
After 4 e5 Nfd7 White normally bolsters the e5-pawn with 5 f4 in anticipation of Black’s attack upon its other supporting pillar with 5 ...
c5. In contrast to the Tarrasch Variation, after 5 ... c5 White can no longer maintain a pawn on d4: the knight on c3 gets in the way of c2-c3. It
is quite an achievement for Black to get rid of the d4-pawn as it frees the c5-square for his bishop and generally loosens White’s grip on the
centre. It also leaves White somewhat vulnerable along the a7-g1 diagonal.

Reasons to be Cheerful, Part One: Black

The d4-square is a magnet that draws pieces of both colour towards it. Black’s dark-squared bishop rejoices in having the c5-square where it
looks down White’s weakest diagonal. This weakness can become particularly acute if White has castled kingside, when a combination might
appear against the king sitting on g1.
The black queen belongs on b6, where she attacks d4 and – if for example White has moved his bishop away to e3 to defend a knight on d4 –
there might be an option of ... Qxb2. The best square for the queen’s knight is of course c6.

Reasons to be Cheerful, Part Two: White

So far it all sounds wonderful for Black: the queen, king’s bishop and queen’s knight all find excellent squares with no effort. But the question
arises as to what the black knight on d7 is meant to be doing and especially the perennially problematic ‘French’ bishop on c8. The knight has
the chance of exerting pressure on e5 or can have a square won for it with the sequence ... f7-f6, e5xf6, ... Nxf6. However, the bishop on c8
remains a problem for a long time, unless Black can free his game with ... e6-e5 (having first eliminated the e5-pawn with f7-f6) or – more
likely – it can be brought into the game with ... a7-a6, ... b7-b5, ... b5-b4, ... a6-a5 and then ... Ba6. A long-winded manoeuvre, but one
favoured by Gurevich amongst others.

TIP: The Classical Variation often develops as a battle between the minor pieces. Whoever can get their knights and bishops
working most effectively usually comes out on top.

The Greek Gift

It should also be remembered that even though Black has eliminated the d4-pawn, the pawn on e5 remains, which gives White an impressive
space advantage on the kingside and, most importantly, denies the black knight the f6-square. This means that Black is vulnerable to the Greek
Gift sacrifice (Diagram 1):
 Diagram 1
Beware of Greeks ...

Here Black appears to be doing well after 1 Nxc6 bxc6, but in fact White then has a decisive attack
with 2 Bxh7+! Kxh7 3 Ng5+ when 3 ... Kg8 4 Qh5 soon leads to mate e.g. 4 ... Rd8 5 Qxf7+ Kh8 6 Qh5+
Kg8 7 Qh7+ Kf8 8 Qh8+ Ke7 9 Qxg7+ Ke8 10 Qf7 mate, while 4 ... Kg6 also fails to survive after 5
Qd3+ f5 6 exf6+ and Black is soon mated.
The Greek Gift occurs in many different scenarios, though the basic mechanism – playing Bxh7+ then Ng5 followed by a queen entry,
usually on h5 – remains consistent. Sometimes it doesn’t work: for example if in the diagram position the black bishop had been on e7 rather
than c5 it would fail as after Ng5+ the bishop would take the knight. In fact the Greek Gift goes through every shade of value from decisive to
downright incorrect. If the mechanism for the combination exists, you have to check to see if it is good or bad before playing it or giving your
opponent the option of playing it.

The Modern Main Line

In some ways this variation is reminiscent of the Sicilian Defence:
1) White usually has a knight on d4 and more space but Black has a solid structure.
2) White usually castles queenside and Black usually goes the other way. A checkmating race often
ensues.
3) White begins a pawn attack on Black’s king whilst Black looks for counterplay along the c-file and
by utilising his queenside pawns.
In Game 26, Shirov won the battle of the wing attacks with Black against Polgar. The outcome was
similar in Game 27 between Topalov and Morozevich. In Game 28, Kasparov pulled one back against his
perennial ‘customer’ Shirov, though Black had a respectable opening. These games all began with the
aggressive set up (for both sides) 1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Be3 cxd4 8
Nxd4 Bc5 9 Qd2 0-0 (Diagram 2), with White castling queenside.
Diagram 2
The scene is set

In Game 29, Black played diverged from this with 7 ... Qb6 (rather than 7 ... cxd4) when after 8 Na4 Qa5+ 9 c3 cxd4 10 b4 he
was more or less compelled to sacrifice with 10 ... Nxb4 which doesn’t look very trustworthy.

The Tarrasch Treatment: 4 ... Nfd7 5 Nce2

White isn’t always happy at the idea of having his pawn forcibly removed from d4. Therefore, in recent years, he has started
playing 5 Nce2 so that he can answer 5 ... c5 with 6 c3 (Diagram 3)

Diagram 3
White holds the centre
when he keeps his pawn centre intact.

NOTE: This variation often transposes to a line of the Tarrasch played by Steinitz, which just goes to prove there us nothing
new under the sun!

The former FIDE World Champion Anand has used this technique as White to score some outstanding victories over players
such as Shirov and Morozevich. I guess this line would be an unpleasant surprise for any Classical player who didn’t also have
the Tarrasch in his repertoire. You can find examples of this line in Games 30 and 31.

Illustrative Games

Game 26
J.Polgar-Shirov
EuroTel Trophy, Prague 1999

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Be3 cxd4

7 ... Qb6 (Game 29) is the major alternative. The old main line with 7 ... a6 is not seen so often
nowadays.
8 Nxd4 Bc5 9 Qd2 0-0 10 0-0-0
In Game 28 White inserts 10 g3!? before
castling.
10 ... a6
It used to be thought that these middlegames were clearly in White’s favour, but after 7 ... Qb6 fell
under something of a cloud interest in the line with ... Bc5 revived and there are those who believe that
expansion on the queenside grants Black adequate chances.
11 h4 Nxd4 12 Bxd4 b5 (Diagram 4)

Diagram 4
Both sides begin their attacks

13 Rh3
White plays instead 13 h5 in the next game.
13 ... b4
A useful gain of tempo.
14 Ne2 a5 15 Qe3 Qc7 16 Bxc5 Nxc5 17 Nd4 a4
18 Kb1 a3!?
A fundamental decision as Shirov commits his pawns. Many would have waited until they had
developed further.
19 b3 Ba6 20 Bxa6 Rxa6
No bad bishop any more, but still an attacking build-up.
21 Qe1 Rb6
There is a large hole at e4.
22 c3 Qb7 23 Rc1 Ne4 (Diagram 5)

Diagram 5
Black’s knight is stronger

24 cxb4 Rxb4 25 Rd3 Rc4 26 Rxc4
On 26 Rdd1 Rfc8 he is under great pressure, but now Shirov gets him very quickly.
26 ... dxc4 27 Rd1 Nc5 28 Qc3 Qxg2 29 b4 Nd3 30 Qxa3 Ra8! 31 Rxd3 cxd3 32 Qxd3 Qxa2+ 33 Kc1 Qa1+ 34 Kd2 Rd8 35
Ke3 Qe1+ 0-1

Game 27
Topalov-Morozevich
Sarajevo 1999

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Be3 cxd4 8 Nxd4 Bc5 9 Qd2 0-0 10 0-0-0 a6 11 h4 Nxd4 12 Bxd4 b5 13
h5
This does not alter the character of the position much from the previous example.
13 ... b4 14 Ne2
Sometimes in this line they put the horse on a4, but there are obvious drawbacks to it living on the
edge, too.
14 ... a5 (Diagram 6)
Diagram 6
The pawns advance once more

15 Qe3 Qc7 16 Kb1 Ba6
The ‘bad’ bishop is once more.
17 Bxc5 Nxc5 18 Ng3 Rfc8 19 Rc1 a4
Here they come again.
20 Bxa6 Rxa6 21 Rhd1 a3!? 22 b3 Rc6
We have a situation very similar to the previous game The white knight now covers e4 but it is even
less well placed for defensive purposes on g3.
23 Qd4 Na6 24 Rd2 Qe7!
With two simple and powerful ideas; entry at h4 and ... Nc7-b5-c3.
25 Qd3 Qh4 26 Qf3 Rc3 27 Rd3 h6! (Diagram 7)

Diagram 7
White has no play

A typical ‘domination’ move. Apart from consolidating Black’s positional control of the game, it also
really rubs it in! Botvinnik said that he many times used the advice of Capablanca: ‘When you have the
advantage and your opponent has a passive piece position, do not hurry. With every move the likelihood
of an error from the defender increases.’
28 f5
Immediately the stymied Topalov weakens himself further.
28 ... Qg5 29 Qf1 Rxd3 30 Qxd3 Nc5 31 Qf3 Nd7 0-1
On 32 Re1 Qd2 and it drops apart.

Game 28
Kasparov-Shirov
Astana 2001

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Be3 cxd4 8 Nxd4 Bc5 9 Qd2 0-0 10 g3!?
Qe7 11 0-0-0 Nb6 12 Nb3 Bxe3 13 Qxe3 Bd7 14 Kb1 Rfc8 15 g4!?
A curious shift of emphasis. Having preferred g3 to the more aggressive h4, Kasparov now goes on the
attack.
15 ... Nb4!? 16 Nd4
On 16 a3 there is the fascinating idea of 16 ... Nxc2!? 17 Kxc2 Ba4 18 Rd4 a5!? with a highly unclear
position.
16 ... Rc5 17 a3 Nc6?
He really had to go back to a6.
18 Ncb5?
Seeing ghosts. Why not just 18 Nb3 – ? Then 18..d4 is forced, but hardly fully adequate after 19 Nxd4
Nxd4 20 Rxd4 Rac8 when White may plug the black queen out with 21 Rd6! and the consequences of the
sacrifice on c3 are not sufficient for Black: 21 ... Rxc3 22 bxc3 Nd5 23 Qd4 Nxc3+ 24 Ka1. Garry let
himself be bluffed.
18 ... Nxd4 19 Nxd4 Rac8 20 Bd3 Nc4 21 Qh3 h6
He might also have played 21 ... g6!?, when the threat of the knight sacrificing itself at a3 is so
imminent that White has nothing better than to take it. Following 22 Bxc4 dxc4 23 Qc3 he has only a slight
edge.
22 g5 (Diagram 8)

Diagram 8
Coming, ready or not

22 ... Nxa3+
Of course. There is no going back, nor has there been for a long time. Opposite sides castling: attack
like crazy and may the best man win!
23 bxa3 Rc3 24 gxh6 g6
He cannot allow hxg7.
25 Qg2

A dual purpose move. He sets up a sac at g6 whilst bringing the queen back to have more bearing on
queenside events.
25 ... Rxa3 26 Nb3 Qb4 27 Bxg6! Rxb3+! 28 cxb3 Qxb3+ 29 Qb2 Qxb2+ 30 Kxb2 fxg6
The smoke clears and they find themselves in an unclear ending.
31 h4
Rooks need open lines. He sets about getting some more.
31 ... Kh7 32 h5 Kxh6 33 hxg6+ Kxg6 34 Rhg1+ Kf5??
After 34 ... Kf7 it was all far from clear.
35 Rd4 1-0 (Diagram 9)

Diagram 9
Whoops! Mate at g5

Shirov has never yet beaten Kasparov. The number of losses he has sustained is about 14. Hard to
believe when you compare ratings.

Game 29
Chandler-M.Gurevich
Leningrad 1987

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Be3 Qb6 8 Na4 Qa5+ 9 c3 cxd4 10 b4 Nxb4
11 cxb4 Bxb4+ 12 Bd2 Bxd2+ 13 Nxd2 (Diagram 10)
 Diagram 10


This dynamic variation, where Black gets three pawns for a piece, was very popular towards the end
of the 1980s. However, its popularity has waned since then, as other variations seem to give Black quite
decent counterchances without investing any material. The major problem with this line for Black is that
his king often struggles to find safety.
13 ... b6 14 Bd3 Ba6 15 Nb2
This knight is horribly placed on b2, contributing almost nothing to the white effort and sitting pretty as
a tactical weakness for Black to exploit. However, Black’s king is still a problem.
Three pawns for the piece

15 ... Nc5 16 Bxa6 Qxa6 17 Qe2 Qa3 18 Qb5+ Ke7 19 0-0 Qe3+ 20 Rf2 Rhc8 21 Rd1 g6 22 Nf1
Qa3 23 Rxd4 Qxa2 24 Ng3 Qb3 25 Qe2 Nd7 26 Rd1 Rc3 27 Nd3 Rac8 28 Rff1 a5
Superficially, the position seems promising for Black as he has very active major pieces and his
queenside pawns are set to roll. However, when White sets his attack in motion, the extra piece will
really make its presence felt.
29 f5 gxf5 30 Nf4 R8c5 31 Qh5 Nxe5 32 Qg5+ Kd7 33 Nxf5 Nc6
If 33 ... exf5 34 Qxf5+ Kd6 35 Qf6+.
34 Nxe6 (Diagram 11)

 Diagram 11
A crushing blow

White insists on giving up a knight. If now 34 ... fxe6 then 35 Qg7+ Kd8 36 Nd4 wins.
34 ... Kxe6 35 Rde1+ Kd7 36 Re7+ Kc8 37 Qg8+ 1-0

Game 30
Nijboer-M.Gurevich
Escaldes 1998

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 Nce2

In this variation White swivels the knight away to allow c3 to bolster his centre. We may get
transposition to lines beginning 3 Nd2 Nf6 4 e5, where White will also take three moves to get knights to
the squares e2 and f3, but different knights this time.
5 ... c5 6 c3 Nc6
In the next game Black assaults the centre with 6 ... cxd4 7 cxd4 f6.
7 f4 cxd4
Normally they do not reduce the central tension so soon, playing first 7 ... Qb6 8 Nf3 f6, but it makes
little difference.
8 cxd4 Qb6 9 Nf3
This position also arises via the Tarrasch variation – by which move order 9 ... Be7 10 g4!? f6 was
played in Game 55.
9 ... f6 10 a3
An intriguing little move, depriving Black of use of the b4-square. 10 exf6 Nxf6 11 Nc3 Bd6 12 Be2 0-
0 13 0-0 a6 14 Kh1 Bd7 15 a3 Rac8 was equal in Gostisa-Mohr, Slovak Championship 1992.
10 ... Be7
10 ... fxe5?! 11 fxe5 Be7 12 Nf4 and here Black conceived the ambitious plan of long castling in
Milicevic-Hadjiyiannis, Novi Sad Olympiad 1990: 12..Nf8!? 13 b4 Bd7 14 Be3 0-0-0 15 Nd3 and play
was unclear.
11 h4!?
A new idea. Usually 11 Nc3, which looks more reasonable, has been played. Morozevich-Moskalenko,
Moscow 1994 then continued 11 ... 0-0 12 Na4 Qd8 when the black queen has been shifted from her
active post.
 11 ... 0-0 12 Rh3!? fxe5?!
Preparing to blow up White’s bind by the sacrifice of a knight on e5, which is typical for this kind of
position. But in this instance the sacrifice is perhaps not fully justified. It was better to play 12 ... Na5!?
13 b4 Nc4 14 Nc3 a5 15 b5 fxe5 16 fxe5 Qc7 17 Qd3 with mutual chances.
13 fxe5 Rf5!
If you say ‘A’ you should also say ‘B’! But now, whatever White plays, Black is going to get two
central pawns for a piece. 13 ... Ndxe5? does not work: after 14 dxe5 Nxe5 15 Nxe5 Qf2+ 16 Kd2 White
wins.
14 Nf4 Ncxe5 (Diagram 12)

Diagram 12
Destroying the white centre
15 dxe5 Nxe5 16 Nd4?!
This move is the most active. But I would prefer the modest 16 Be2, e.g. 16 ... Bd7 17 Nxe5 Qg1+ 18
Kd2 Qd4+ 19 Nfd3 Rxe5 20 Qb3 with a very complicated position where White’s chances should be
preferred.
16 ... Rf7 17 Be3?! Qxb2 18 Be2 Nc6!
The material balance has been restored and now Black has a very strong initiative just for free.
19 Nf3
Also after 19 Nxc6 bxc6 20 Bd4 Qb8! 21 Nd3 Bxh4+! 22 Rxh4 Qg3+ 23 Bf2 Rxf2 24 Nxf2 Qxh4
White is in troubles.
19 ... Qc3+! 20 Bd2
20 Kf2? Bc5 21 Bxc5 Qxc5+ 22 Kg3 e5 wins for Black.
20 ... Qc5 21 Rg3
The best practical chance; ... e6-e5 is hard to prevent.
21 ... e5 (Diagram 13)

Diagram 13
Black’s pawns take over

White’s pieces are awkward and desperately lack coordination.
22 Rc1 Qd6 23 Rxc6!?
It’s hard to propose something else! If 23 Nxd5 then 23 ... e4 wins.
23 ... bxc6 24 Bb4 Qc7 25 Nh5 Bf5! 26 Ng5 Bxb4+ 27 axb4 Re7 28 b5 Bg6
Another possibility was 28 ... Rd8!? 29 bxc6 Qxc6 winning.
29 Rc3 Qb6?!
Black could already simplify the position by 29 ... Qa5 30 Qd2 h6! 31 Nh3 Bxh5 32 Bxh5 Qxb5
winning.
30 Ng3? Rd8?!
It was almost certainly mutual time trouble. Black had to continue 30 ... Rf8! The game degenerates
into a street fight.
31 bxc6 Qb4?!
31 ... Qa5 32 Qd2 h6 was more stubborn. 31 ... d4?! 32 Qb3+ Qxb3 33 Rxb3 d3 34 h5 dxe2 35 hxg6
hxg6 36 Kxe2 would have been equal.
32 Qd2 Qxh4 33 c7 Rxc7!? 34 Rxc7 Qxg3+ 35
Kf1 Rf8+?
Black had to repel the main threat of 36 Qxd5+! by 35 ... Qb3!, still with chances.
36 Bf3 e4??
Black missed a few possibilities to decide the game in his favour, but his last move is suicide! After
36 ... Kh8! 37 Qb4 Rg8 38 Nf7+ Bxf7 39 Rxf7 e4 40 Be2 Qe3 Black still kept the better chances.
37 Qxd5+ Kh8 38 Nf7+ Bxf7
On 38 ... Kg8 White has the neat win 39 Ne5+ Kh8 40 Qd6! Qf4 41 Nf7+! and the vulnerable back
rank decides after 41 ... Rxf7 42 Qxf4 Rxf4 43 Rc8+.
39 Qxf7! Rd8 40 Qd7 1-0
Presumably here Black lost on time. After 40 ... Rf8 he could still continue resistance.

Game 31
Bezgodov-Sakaev
Moscow 1999

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 Nce2 c5 6 c3 cxd4 7 cxd4 f6 8 Nf4

Both sides handle the opening in an unusual manner.
8 ... Bb4+ 9 Bd2 Qb6 10 Bxb4 Qxb4+ 11 Qd2 Qxd2+ 12 Kxd2 Ke7 (Diagram 14)

Diagram 14
A queenless middlegame

13 exf6+ gxf6 14 Re1 Nb6 15 Bd3 Nc6 16 Nf3 Kd6
Nicely placed here. No danger, and it can help the centre to roll.
17 Nh5 Rf8! 18 Bxh7 e5
The h-pawn was correctly offered as bait so that he can get the central pawns moving.
19 Ng3 e4 20 Nh4 Nxd4 21 f3 Rh8 22 Nhf5+ Bxf5
23 Nxf5+ Ke5! 24 Nxd4 Kxd4
The king is a strong piece, and here he is dominant.
25 Bf5 Nc4+ 26 Ke2 Rag8 27 Rd1+ Ke5 28 g4 Rh3! 29 fxe4 Re3+ 30 Kf2 dxe4 31 Rd7 Kf4
(Diagram 15)

Diagram 15
Black’s king is the master

32 Re1 Rh8 33 Bh7 Ne5! 34 Rxe3 Nxg4+ 35 Ke2 Nxe3 36 Rxb7 Nf5 37 Rxa7 Nd4+ 38 Kd1 Rd8
All of the black pieces now work in concert to support the advance of the two pawns. White is fast
swept away.
39 Rf7 f5 40 Kc1 e3 0-1
Chapter Five
3 Nc3 Nf6 4 Bg5 dxe4:
The Burn Variation
Introduction
Here a completely different pawn structure to that in the previous chapter is created: in fact it is more akin
to the Rubinstein Variation, to which it often transposes, than the typical Classical.
A player new to this variation may wonder why after 1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4
Be7 (Diagram 1)

Diagram 1
Capture with bishop or knight?

White takes on f6 with his bishop rather than the knight. Why does he grant Black the two bishops? The point is that the knight on e4 is in
fact the best minor piece on the board after 6 Bxf6 Bxf6. It cannot be dislodged by a pawn from its imposing centre square except by the
preparation of the terribly weakening move ... f7-f5. So White has come out best from the exchange. Nevertheless, the balance of power
between the knight and dark-squared bishop can suddenly alter. That is the inherent tension in the position: the knight can become enfeebled or
exchanged off, when Black’s ‘Dragon bishop’ has the last word. Mikhail Gurevich and Evgeny Bareev have won a huge number of games in
this line over the years as Black. For every game that Gurevich loses in this line he appears to win at least four or five – that is a great return
with Black at his level of play.
Black faces a choice on move six: 6 ... gxf6 or 6 ... Bxf6, which is best?

6 ... Bxf6 – The Solid Treatment

The safe recapture 6 ... Bxf6 keeps the black pawn structure intact but allows White easy piece play which can sometimes metamorphose into
a vicious attack. You can see Bareev score a crushing win with 6 ... Bxf6 against fatally slow play by White in Game 32, while in Game 33
Shirov demonstrates the potential long term power of the black bishop pair, if White’s initiative burns itself out.

6 ... gxf6 – The Dynamic Treatment

The young Russian Alexander Morozevich owes much of his fame these days to his skill with Black in the Classical French. He has done
much to show that some previously discarded variations are in fact fully playable.
A good example is the line
 1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Be7 6 Bxf6 gxf6 (Diagram 2)

Diagram 2
Dynamic play

In the early days of the Classical 6 ... gxf6 was frowned upon because of the damage to Black’s
kingside. However, Morozevich and others have shown that the doubled pawn complex of e6/f7/f6 is a
very tough nut to crack. Besides it control the important e5 square (thus the aggressive Ne5 is ruled out)
and can even expel the white knight from e4 by ... f6-f5.
If you have a solid centre you can quite reasonably castle kingside even with no g-pawn to protect your
king. Furthermore, from the diagram after White’s natural 7 Nf3 the Russian has championed 7 ... a6! The
intention is to play ... b7-b5, with influence over the c4-square, and then ... Bb7. In contrast, the older
move 7 ... b6?! is far less energetic and does nothing to stop White playing Bc4.

TIP: The shattered pawns on the kingside do not necessarily preclude Black from castling kingside
as the pawns and dark-squares bishop can provide good cover.

However, we should stress that there is nothing wrong with the alternative 6 ... Bxf6.
You can find examples of 7 ... a6 in Games 34-36. In Game 34 Gurevich gives the system a go instead of his usual 6 ... Bxf6. His
opponent, a certain G.Kasparov, comes armed with an improvement over an earlier game and manages to win after a very hard battle. In
Game 35 we see the earlier game referred to – a very smooth win as Black by Morozevich over Polgar. Finally there is Shirov-Topalov in
Game 36. This is a stupendous win for White though Radjabov has subsequently improved Black’s play with 14 ... Be6! getting the bishop out
before it is shut in by the passed pawn.

The Mighty Fallen

A line once regarded as highly reliable for Black is 1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Nbd7!? (rather than the immediate 5 ...
Be7) 6 Nf3. Now in Game 37 Black wins with 6 ... h6 after some inferior play by White, but the unbeatable move was thought to be 6 ... Be7.
If White follows the recipe recommended above with 7 Bxf6?! then 7 ... Nxf6 is already good for
Black – he ousts the knight from e4. Therefore a different approach is required. Topalov and others have
started playing 7 Nxf6+ Bxf6 8 h4! White keeps the tension on the kingside and prepares to castle
queenside with a sharp game. Kasparov himself takes up the baton in Game 38 and scored a complex but
direct attacking win over Shirov.
Illustrative Games

Game 32
Topalov-Bareev
Linares 1994

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Be7

6 Bxf6 Bxf6
The more ‘restrained’ recapture you might think. Wait and see. 6 ... gxf6 is the subject of Games 34-
36).
7 c3
Blunting the immediate effect of the black bishop. In the next game White plays the standard 7 Nf3.
7 ... Nd7 8 Qc2
8 f4!? Be7 9 Nf3 c5 and 8 Nf3 0-0 9 Qc2 e5 were alternative treatments, but Black equalises in either
case.
8 ... e5!
The right moment. On 8 ... 0-0 9 f4 might give White a slight edge.
9 dxe5
9 0-0-0 was more enterprising.
9 ... Nxe5 10 f4?
White is not developed enough for this sort of thing, but it calls for vigour and imagination to
demonstrate that. The quieter 10 Be2 was to be preferred. Note that in the line 10 Bb5+ c6 11 Rd1 Qe7
12 Nd6+ Kf8 13 Be2 Be6, although Black has had to give up the right to castle he is still doing very
nicely, thank you.
10 ... Ng6 11 g3 0-0 12 Bd3
On 12 Bg2 Re8 13 Ne2 Bf5 the pin is very nasty.
12 ... Qd5
White was threatening 0-0-0 with a slight edge.
13 a3?
One error too many, after which it all gets vicious. In retrospect, he ought to have jettisoned the a-pawn to get his king into safety with 13 0-0-
0 Qxa2 14 Nxf6+ gxf6 15 Ne2 or 15 Nf3, although White would be struggling to show full equality there.
13 Nxf6+? gxf6 14 0-0-0 Qxh1 15 Be4 gets crossed by the neat tactic 15 ... Bf5!.
13 ... Nxf4!! (Diagram 3)
 Diagram 3
White now wishes he’d castled

14 Nxf6+
Taking the horse does not help. On 14 gxf4 Bh4+ 15 Kf1 (15 Kd2 Rd8 wins, or 15 Ke2 Bg4+ 16 Nf3
f5 wins) 15 ... f5 is hard to cope with, e.g.16 Nf3 (16 Qe2 threatens to pin Black’s queen at c4, but by
sidestepping with the simple 16 ... Kh8! Black keeps a winning position. On 16 Nd2 Qxh1 17 Ndf3 Be7
and Black wins) 16 ... fxe4 17 Bxe4 Bh3+ 18 Kg1 Qc5+ 19 Nd4 Rxf4 and Black is killing him.
14 ... gxf6 15 Bxh7+ Kg7
Now 16 Be4 is met by 16 ... Re8, so Topalov has little option but to attempt to block with his queen.
16 Qe4
Now watch the fireworks!
16 ... Re8! 17 Qxe8 Bf5!!
Like the romantic games of the last century.
18 Qxa8
Two rooks won ... but a king is about to be lost. However, it’s too late to be back-pedalling as 18 Qe7
Nd3+ 19 Kf1 Bxh7 and 18 Qa4 Nd3+ 19 Kf1 Bxh7 leave Black easily winning.
18 ... Qe4+! (Diagram 4)

 Diagram 4
And mate in nine

Precise checking drives the white king into a mating net.
19 Kf2
On 19 Kd1 Qc2+ 20 Ke1 Nd3+ 21 Kf1 Qf2 mate.
19 ... Qg2+ 20 Ke3 Nd5+ 21 Kd4 Qd2+ 22 Kc5
Or 22 Kc4 22 ... Nb6+ 23 Kb4 Qxb2+ 24 Ka5 Nc4+ 25 Ka4 b5 mate.
22 ... Qe3+! 23 Kc4
23 Kxd5 Be6 is mate, and 23 Kb5 Qb6+ 24 Kc4 Ne3 is too.
23 ... Nb6+ 0-1
and it’s mate, e.g. 24 Kb5 c6+ 25 Kb4 Qe4+ 26 Kc5 Qe7+ 27 Kd4 c5, or 24 Kb3 Qe6+ 25 Kb4 Qc4+
26 Ka5 Qc5. This brutal game won the brilliancy prize.

Game 33
J.Polgar-Shirov
EuroTel Trophy, Prague 1999

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Be7 6 Bxf6 Bxf6 7 Nf3 0-0 8 Bc4 Nc6!?
A move so unusual that Judit is prompted to move her bishop again.
9 Bb5 Bd7
9 ... Nxd4? 10 Nxf6+ and 9 ... Bxd4 10 Bxc6 both cost a piece.
10 Qd2 Be7 11 0-0 Ne5!? 12 Nxe5 Bxb5 13 c4 Ba6!? (Diagram 5)

Diagram 5
Bishops versus knights

Far more ambitious than the retreat to e8.
14 Rac1 c5!?
To win, it may be necessary to take risks. Shirov stretches things to the limit in trying to make the most
of the pair of bishops.
15 dxc5 Qxd2 16 Nxd2 Rfd8
Taking back immediately permits the fork at d7, so he comes up with this scheme.
17 Nb3
If 17 Ne4 might have met with 17 ... Rd4!?. From b3 the knight is temporarily holding on to c5, but that
pawn cannot now get additional support, e.g. with b2-b4, and also the one behind it may become opened
up as another object of attack. (Perhaps this is how a mind wired like Shirov’s works!?)
17 ... Rac8 18 f4 b6!?
A move of great ambition. 18 ... f6 and 19 ... Bxc5 restored material equality without risk.
19 Rfd1
On 19 cxb6 axb6 ... f6 will cause the c4-pawn to fall.
19 ... Rxd1+ 20 Rxd1 Rc7!?
Keeping the rook out of the seventh rank.
21 Rd6!?
Judit tries to throw a spanner in the works, but Shirov refuses to be flustered and sticks to his plan of
regaining the pawn in a situation where the activity of his minor pieces will be superior to that of White’s.
21 ... Kf8! 22 cxb6 Bxd6 23 bxc7 Bxc7
White is still that pawn up, but the bishops are starting to weave their combined magical effect.
24 c5 g5! 25 Nc6 gxf4 26 a4 Bc4 27 Nd2 Bd5
Hitting its optimum square.
28 Nxa7 Ba5! 29 Nb1 Bb4 30 Nb5 Bxc5+ 31 Kf1 f3 32 gxf3 Bxf3 33 Nd2 Bd5 (Diagram 6)

Diagram 6
Victory to the clerics

Now the bishops totally outgun the knights.
34 Ke2 Ke7 35 Kd3 f5 36 Nc3 Bb7 37 Kc4 Be3 38 Kd3 Bg1 39 b4 Bxh2 40 b5 h5 41 a5 h4 42 Ke2
h3 43 a6 Bg2 44 Kf2 Bd6 45 Nf3 Bc5+ 46 Kg3 f4+ 47 Kg4 Bg1!
The long range bishops can cover the white pawns whilst simultaneously assisting with the advance of
their own passed pawns. Knights are not so nimble.
48 b6 Bxf3+ 49 Kxf3 h2 50 Kg2 f3+ 0-1
The finale might have been 51 Kh1 f2 52 a7 f1Q 53 a8Q Bc5+ 54 Kxh2 Bd6 mate.
Shirov’s imaginative efforts to extract the maximum potential from the position rewarded with the full
point.

Game 34
Kasparov-M.Gurevich
Bosna SuperGM, Sarajevo 2000

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Be7 6 Bxf6 gxf6

This capture leaves Black with weakened pawns, but he hopes for counter-chances.
7 Nf3 a6!?
Morozevichian stuff. It is more adventurous than the old 7 ... b6 main line. Black can also play 7 ...
f5!?.
8 g3
8 Qd2 f5 9 Nc3 b5 10 0-0-0 0-0 led to a double-edged middlegame in Watson-Anand, Palma de
Mallorca GMA 1989. White can prevent ... b7-b5 by playing 8 c4, for which see Game 36.
8 ... b5 9 Bg2 Bb7
Morozevich is another who has experimented with such formations.
10 Qe2 Nd7 11 0-0 0-0
An unusual, but by no means non-viable deployment by Gurevich.
12 Rfd1
This move is the novelty. Judit Polgar didn’t manage to find the correct plan and got nothing from the
opening after 12 Rad1?! Bd5 (see the next game).
12 ... Bd5 13 c3 f5 14 Ned2 c5 15 dxc5 Nxc5 16 Nf1
Heading for e3.
16 ... Qc7 17 Rxd5!? (Diagram 7)

Diagram 7
A positional exchange sacrifice

Garry sacrifices whenever he can! But here this may be the only way to fight for the advantage.
17 ... exd5 18 Ne3
White will get a pawn or two plus an active game for his invested exchange. But things are certainly
not 100% clear.
18 ... Bf6 19 Nd4 Bxd4 20 cxd4 Ne4 21 Nxd5 Qd6 22 Ne3 Qf6
And not 22 ... Qxd4 23 Nxf5 and the knight at e4 drops off.
23 Qh5 Rad8
23 ... Qxd4!? 24 Nxf5 Qxf2+ 25 Kh1 Qxb2 26 Rf1 was not clearly better for Black than the text.
24 Nxf5 Nd6 25 Ne3!? Qxd4 26 Rd1 Qg7
Black manages to hold the position after this, but 26 ... Qxb2 would have been my move. Grabbing a
pawn always makes sense if there is no clear refutation and I do not see one here.
27 Rd5 Kh8
Avoiding the threat of the pin.
28 Qd1 Nb7 29 b4!
Taking away important squares from the knight.
29 ... Rxd5!? 30 Qxd5 Nd8 31 Qd6 Ne6 32 Qxa6 Nd4 33 h4 f5 34 Nd5 Ne2+
34 ... Qe5!? deserved attention too.
35 Kf1 f4!
Taking advantage of the fact that white queen is on a6.
36 Kxe2 fxg3 37 Qd6!? (Diagram 8)

Diagram 8
Should Black take the draw?

37 ... Qb2+?
Black wants too much. There was a draw here with 37 ... Rxf2+ 38 Kd3 Rxg2 and White is forced to
deliver perpetual check by 39 Qd8+ Qg8 40 Qf6+ etc.
38 Kd3 Rxf2?
Better 38 ... Qb1+ 39 Kd4 Qb2+ 40 Kc5 Qxf2+ (if 40 ... Rc8+ 41 Kxb5 Qe2+ 42 Kb6 Qxf2+ 43 Kb7
Rg8 44 Nf4!) 41 Kxb5 Qe2+ 42 Kb6 Qf2+ 43 Qc5 Kg7! with chances still to save the game.
39 Qb8+ Kg7 40 Qxg3+ Kh8 41 Qb8+ Kg7 42 Qc7+ Kf8 43 Qe7+ Kg8 44 Qg5+ Kh8 45 Be4!
This move coordinates White’s game and so destroys Black’s hopes.
45 ... Qc2+ 46 Kd4 Qd2+ 47 Kc5 Qxg5 48 hxg5 Rxa2 49 Kxb5
The b-pawn will now be too strong for Black to cope with.
49 ... Re2 50 Nc3 Re3 51 Kc4 Rg3 52 b5 Rxg5 53 b6 1-0

Game 35
J.Polgar-Morozevich
Wijk aan Zee 2000

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Be7 6 Bxf6 gxf6 7 Nf3 a6 8 g3 b5 9 Bg2 Bb7 10 Qe2
Nd7 11 0-0 0-0 12 Rad1
This treatment does not prove so effective.
12 ... Bd5! 13 Rfe1 Kh8 14 Nfd2 c6 (Diagram 9)

Diagram 9


A remarkable plan. The cemented bishop at d5 becomes the lynch pin of the black position.
15 c4
On 15 b3 of course 15 ... b4.
15 ... bxc4 16 Nxc4 a5 17 Nc3 Re8 18 Ne3 f5! 19 Nc4
She is unsure what to do. I do not think I would have known either.
19 ... Qb8 20 Na4 Qb4 21 b3 Bf6 22 Qc2 Rg8
The first shade of attack.
23 Qc1 Rg4 24 Ne5 Nxe5 25 dxe5 Be7 26 Rd3 Rag8 27 Red1 f4 (Diagram 10)
An unusually ‘good’ bishop
Diagram 10
Storming the kingside fortress

Attack arriving.
28 Qc3 fxg3 29 hxg3 h5!
Not perturbed by the queen exchange Morozevich continues the deconstruction of White’s kingside.
30 Bxd5 cxd5 31 Rf3 Kg7!? 32 Qc7
Judit reckons that if there are any swindling chances then they are more likely to occur with queens on
the board, but Moro quickly finishes the job.
32 ... h4 33 Kg2 hxg3 34 fxg3 Qe4 35 Qc3 Qe2+ 36 Kh3 Rg5 0-1
Mate follows shortly.
A game of dizzy originality and an excellent win against a strong opponent.

Game 36
Shirov-Topalov
Sarajevo 2000

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Be7 6 Bxf6 gxf6 7 Nf3 a6 8 c4!? f5 9 Nc3 Bf6 10 Qd2
c5 11 d5 0-0 12 0-0-0 e5
Black has the bishops and a nice pawn centre. White has more space and the big d-pawn. Despite the
result in this game I am not convinced that Black’s chances were necessarily the worse out of the opening.
13 h4!?
Alexei, having castled long, seizes the first glimmer of attacking potential in the position.
13 ... b5 14 d6! (Diagram 11)

Diagram 11


Ignoring the gambit and pressing onwards.
14 ... Nc6 15 d7
A real pain, this.
15 ... Bb7 16 Qd6 e4
Advancing into enemy territory

16 ... Be7 shifts the queen off to another very nasty square, and after 17 Qh6 White already threatens an
instant win with 18 Ng5.
17 Nd5 Bg7 18 Ng5
For the fourth consecutive move a white unit lands in Black’s half.
18 ... Nd4
18 ... Be5? 19 Qh6 wins on the spot.
19 Ne7+ Kh8 20 Rh3 f4 21 Kb1 b4 22 Be2!! (Diagram 12)

Diagram 12
The white bishop is a decoy

If untaken then the bishop goes out to h5 and f7 falls. But the main significance of the move is to lure
the knight away from controlling f5 so that White’s queen can get there. Such is the cramp exerted by the
d7-pawn that the loss of a minor piece does not much matter.
22 ... f3
A desperate attempt to vacate f4 so that the knight has a way back, but it’s too late.
23 gxf3 Nxe2 24 Qxc5 Nf4 25 Qf5 Ng6 26 h5! Qxe7 27 hxg6 1-0
On 27 ... h6 28 gxf7 means that Black can only stop mate at h7 by surrendering his queen.
Superb play from Shirov. Some of the themes of the game, and especially the combination (which, in a
higher sense, one may say begins with the initial advance of the h-pawn) remind me of Kasparov’s win vs
Kramnik from Novgorod 1994.

Game 37
Lobzhanidze-Yagupov
St Petersburg 1997

1 e4 e6 2 d4 d5 3 Nd2 dxe4 4 Nxe4 Nd7 5 Nf3 Ngf6 6 Bg5

The game transposes here from a Rubinstein variation to the Burn line 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4
Nbd7 6 Nf3.
6 ... h6
In the next game Black breaks the pin with 6 ... Be7.
 7 Nxf6+ Nxf6 8 Bd2?!
8 Bh4 is more usual and more active. Jonathan Speelman then sprang 8 ... g6!? on Nigel Short in the
2nd game of their 1988 Candidates match. The idea apparently came to him in the back of a car whilst
waiting for tickets for an international soccer game in Oslo.
A decade later Speelman repeated the line against Plaskett, who went 9 Bc4, castled queenside, blew
his brains out with an attack, emerged two clean pawns up in a simple rook ending ... and drew it. The
failure to convert the advantage was described by IM Malcolm Pein as ‘almost unforgivable’. Fair
comment.
8 ... c5 9 Bd3 Qb6!
A rare chance for Black to initiate early complications in these quiet variations.
10 Qe2 Qxb2 (Diagram 13)

Diagram 13
Yum yum!

Black calls his bluff and takes the pawn.
11 0-0 Qb6 12 Rab1 Qc7 13 c4 Be7 14 Rfe1 cxd4 15 Nxd4 0-0 16 Nb5 Qc6 17 Rb3
This is about as far as White’s compensation gets ... and it’s not enough.
17 ... Rd8 18 Bf4 b6 19 Bb1 Bb7 20 Rg3 Ne4! (Diagram 14)

 Diagram 14
Black trades advantages

Excellent defence! He gives back the pawn to enter a queenless middlegame with the advantage of the
pair of bishops.
21 Bxe4 Qxe4 22 Qxe4 Bxe4 23 Bxh6 Bg6 24 Bf4 Rd7
Black has the better pawn structure too, and the White cause is critical.
25 Nc7 Rad8 26 h3 Rd1 27 Rxd1 Rxd1+ 28 Kh2 Rd3 29 Nb5 a6 30 Rxd3 Bxd3 31 Bd6 Kf8 32
Bxe7+ Kxe7
and the rest is simple technique.
33 Nc3 Bxc4 34 Kg3 b5 35 a3 Bd5 36 f4 Kd6 37 Kf2 Kc5 38 g4 Kc4 39 Nxd5 Kxd5 40 h4 Ke4 41 Kg3 a5 42 h5 b4 43 axb4
axb4 44 f5 exf5 45 g5 0-1

Game 38
Kasparov-Shirov
Fujitsu-Siemens Giants, Frankfurt 2000

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 dxe4 5 Nxe4 Nbd7 6 Nf3 Be7

A choice of opening which I find surprisingly passive for so enterprising a player as Shirov.
7 Nxf6+ Bxf6 8 h4!?
As ever, Garry looks for ways to sharpen the play.
8 ... 0-0 9 Bd3 c5 10 Qe2 cxd4 11 Qe4 (Diagram 15)
 Diagram 15
White intentions are unsubtle

Already a mate threat.
11 ... g6 12 0-0-0
When you play 8 h4, you probably go on to castle long. Note that it has been a few moves since Black
played ... cxd4 and White has still not recaptured. Such a lag is not uncommon in this line. But that pawn
will be nabbed.
12 ... Qa5 13 Bxf6 Nxf6 14 Qxd4 Nh5 15 a3
Kasparov permits himself one defensive move in the game.
15 ... Rd8 16 Qe3 Bd7 17 g4
In opposite sides castling middlegames it is usually the guy whose attack lands first who wins. Here
White is well ahead.
17 ... Nf6 18 Qf4!
A neat move to nudge away the knight and hence accelerate the attack.
18 ... Nd5
18 ... Kg7 19 h5 causes big problems.
19 Qh6
Already the attack is ferocious.
19 ... Nf6 20 Ng5 Bc6 21 Bxg6! (Diagram 16)

Diagram 16
Obliterating Black’s defences

Decisive.
21 ... hxg6
21 ... fxg6 22 Nxe6 Rxd1+ 23 Rxd1 Kf7 24 Ng5+ Kg8 25 h5 and the attack is still going strong, e.g. 25
... Be8 26 Ne6 Kf7 27 Qf8+ Kxe6 28 Rd6+ Ke5 29 Qxf6+ Ke4 30 Rd4 mate.
22 Nxe6! fxe6 23 Qxg6+ Kh8 24 Qxf6+ Kh7 25
Rhe1 Rxd1+ 26 Rxd1 Qc5 27 g5
Kasparov always emphasised that such a pawn is a legitimate piece and part of the attack.
27 ... Rf8 28 Qh6+ Kg8 29 Qxe6+
The fourth pawn for the invested piece.
29 ... Kg7 30 Qh6+ Kg8 31 Qg6+ Kh8 32 Qh6+ Kg8 33 Qe6+ Kg7 34 Rd6!
The introduction of this last attacking piece winds it all up.
34 ... Be8 35 Qe7+ 1-0
On 35 ... Rf7 36 Rg6+ wins the queen.
Chapter Six
3 Nc3 Nf6 4 Bg5 Others:
Classical and McCutcheon
The Classical
After 1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 Bxe7 Qxe7 7 f4 (Diagram 1)

Diagram 1
The bishops have gone

a position is reached that is similar to that in the 4 e5 main line, except that the dark-squared bishops
have been exchanged.

Who does the exchange of bishops favour? This is a tricky one to call.
Speaking in completely general terms the exchange of a piece benefits the player with less space, as he
then has one piece fewer to accommodate in the crowded confines of his pawn structure. This would
seem to give the nod to Black, but looking at things more concretely we see that Black has exchanged off
his ‘good’ bishop, the defender of his dark squares for White’s ‘bad’ bishop. So surely White has done
well to get rid of his bad bishop? However, who is to say that White’s bishop in the 4 e5 line was ‘bad’
when it sat on e3 supporting the knight on d4 and guarding the dark squares in general? It was surely
performing as good a function as Black’s bishop on c5 in these lines. A big empire requires a lot of men
to serve it, so perhaps the bishop exchange was to White’s disadvantage?
Bringing the actual position after 7 f4 into focus, you will see a couple of features that favour White. The black queen is on e7 rather than
d8, so Black has lost the useful developmental possibility ... Qb6 after ... c7-c5 which he had in the 4 e5 variation. Also, Black can’t play 7 ...
c5 here ‘in one go’ as then he is hit by 8 Nb5! with the double threat of 9 Nc7+ and 9 Nd6+. In the 4 e5 line this problem didn’t arise. So the
specific features tend to suggest that Black’s life is a little harder because of the bishop exchange. This opinion is supported by the fact that as
a rule elite players tend to answer 4 Bg5 with 4 ... dxe4, the Burn Variation, rather than go into this line with 4 ... Be7.
The Greek Gift
We have already discussed the famous bishop sacrifice in the previous chapter after 4 e5. It can occur in any Classical set up, or in similar
variations where White has played e4-e5 and driven the black knight from f6 so that it no longer defends h7 – for example the Advance
Variation or some lines in the 3 ... Nf6 Tarrasch. As the dark-squared bishops have been exchanged off, this line of the Classical is perhaps
peculiarly susceptible – there is no bishop on e7 to prevent the follow up Ng5+. Therefore the Greek Gift has been comprehensively examined
in the notes to game 39. As you will see, the sacrifice by no means always works, though White was making his task particularly hard by
choosing to play it against a great defender like Korchnoi!

A more Positional Approach by White

In game 39 White had a gung-ho attitude. In Game 40 he settles for a more positional approach. Having cleared the centre of his own pawns,
he seeks to exploit the dark squares on d4 and e5 as outposts for his pieces which can’t be controlled by black pawns (Diagram 2).

Diagram 2
Can White keep control?

Unfortunately for White he eventually lost control of the position and was overwhelmed by the inherent dynamism in an unopposed pawn
centre. It is a thin line between scoring a smooth strategical victory by blockading the opponent’s centre and being destroyed when it manages
to break free.

NOTE: In this variation White often has to relinquish the pawn on e5 and the struggle then revolves around control of the
central dark squares.

Game 41 is more Morozevich magic when Black ends up castling queenside!

The Alekhine-Chatard Attack

This is a tricky gambit line for which Black must be prepared. After 1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 h4, he can accept
the pawn with 6 ... Bxg5 7 hxg5 Qxg5. However, after 8 Nh3 chasing the black queen back to e7 followed by a build up such as Nf4 and Qg4
White has a very useful initiative on the kingside as compensation. Note that if Black tries ... g7-g6, intending h7-h5 to gain space, White will
clamp down on the h7-pawn with Rh6! In Game 42 we see the World Number One in a rather experimental mood – I guess the fact that it
was a rapidplay game made him avoid the beaten track against the veteran Korchnoi.
Rather than accept the offer, Black has a promising alternative in 6 ... c5! as in Game 43. After 7 Bxe7 he is compelled to play 7 ... Kxe7,
but the fact that White feels obliged to play his knight to an awful square on a4 to hold his centre together more than offsets the slight
discomfort of the black king. In contrast, the attempt by Black to play it ultra solid in Game 44 rebounds.

The McCutcheon Variation

This variation begins 1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4. Note that Black is unafraid of the pin on f6 as after 5 e5 h6 6 Bh4 g5 he
avoids losing a piece. The McCutcheon is named after a US Amateur who played it against the reigning World Champion Steinitz in a
simultaneous display more than a hundred years ago. Another World Champion, Capablanca, regarded it as the most dynamic line for Black in
the French, though this was in the 1920s before Botvinnik and others had refined the Winawer Poisoned Pawn.
The McCutcheon itself has many features of the Winawer: notably the willingness of Black to give up
his important dark-squared bishop to inflict doubled pawns on White’s queenside. Furthermore the
Winawer move Qg4 putting pressure on the g7-pawn plays an important part in White’s strategy as we see
in the usual sequence 5 e5 h6 6 Bd2 Bxc3 7 bxc3 Ne4 8 Qg4.
Looking over these moves, you will see that White rejected the chance to keep his pawns intact with 7
Bxc3. This is because after the reply 7 ... Ne4 White is threatened with doubled pawns anyway and if he
wastes time moving his bishop again then Black will get quick counterplay with ... c7-c5 etc. Note that in
contrast to the Classical main line, in which his king’s knight ends up on a dead square after 4 e5 Nfd7,
Black can be delighted here that his knight has got to an excellent central square and can be exchanged off
when necessary for the bishop on d2.
In the diagram position Black has to make an important choice as 8 ... 0-0?? loses to 9 Bxh6. Should he
defend g7 with 8 ... Kf8, when the rook on h8 stays shut in, or accept a slight kingside weakness with 8 ...
g6, when he can castle queenside later? The latter plan is the more popular these days.
In Game 45 Bobby Fischer takes apart Black’s position after 8 ... g6, though it is by no means as bad for him as the great World Champion
makes it appear!

The Modern Handling of the McCutcheon

In Game 46, Black comes up with a refined piece deployment. Because he has a dark square hole on f6, he plays his bishop to c6 and knight to
d7 to cover it – a role reversal for the minor pieces compared to the standard development of ... Nc6 and ... Bd7. The slight drawback is that
Black is exerting less pressure on d4 than after Nc6, but overall this method looks the most promising.
Games 47 and 48 feature 8 ... Kf8. It is noteworthy to see how Korchnoi tries to solve the problem of
the misplaced king. In Game 49 White tries out 7 Bxc3 but it doesn’t give him any particular advantage.
Finally in Game 50 we look at what happens if White offers the c3-pawn as a gambit. The inherent
solidity of Black’s position triumphs over his slight vulnerability on the dark squares.

Illustrative Games

Game 39
Van der Wiel-Korchnoi
Euwe Memorial, Amsterdam 1991

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 Bxe7 Qxe7 7 f4

A clamp on the dark squares.
7 ... 0-0
Not yet 7 ... c5? when 8 Nb5! causes problems. If Black does not want to commit the king so soon, then
7 ... a6 is sometimes preferred (see Game 41).
8 Qd2 c5 9 Nf3 Nc6 10 0-0-0
Opposite sides castling. Expect attacks. A different idea – 10 dxc5 f6 11 exf6 Qxf6 – is seen in the next
game.
10 ... Nb6
10 ... a6 11 dxc5 Qxc5 12 Bd3 b5? (no time for this) 13 Bxh7+! (get used to this idea) 13 ... Kxh7 14
Ng5+ Kg8 15 Qd3 Re8 16 Qh7+ Kf8 17 Qh5! Nd8 18 Nh7+ Kg8 (his problem is that 18 ... Ke7 runs into
19 Qg5+ f6 20 Qxg7+ with destruction) 19 Rd3! (the arrival of this major piece ensures the success of the
attack) 19 ... Qe7 20 Rh3 f6 21 Nxf6+! Nxf6 22 exf6 and because his queen cannot abandon protection of
the rook at e8, Black resigned in Chandler-Agnost, London 1989.
11 dxc5 Qxc5 12 Bd3 Bd7 13 Bxh7+!? (Diagram 3)
 Diagram 3
The classic bishop sacrifice

This time the consequences are not so clear, even though, logically, one would presume that
substituting ... Nb6 for ... a6 could hardly help Black’s defensive chances as he has moved a defender
away from his king. But it turns out that he has also cleared a path.
Gyula Sax played the more restrained 13 Kb1 a couple of times, e.g. 13 ... Rfc8 14 Nb5! and White
was better in Sax-Korchnoi, Rotterdam 1988, and 13 ... Rac8 14 Bxh7+!? Kxh7 15 Ng5+ Kg8 16 Qd3
Rfe8 17 Qh7+ Kf8 18 Qh5 Ke7 19 Nxf7 Na5! 20 Nd6 and a draw eventually resulted in Sax-Timman,
Rotterdam 1989.
The same bishop sac occurred in Sieiro Gonzales-Paneque, Cuban Championship 1989: 13 ... a6 14
Bxh7+!? Kxh7 15 Ng5+ Kg8 16 Qd3 Rfe8 17 Qh7+ Kf8 18 Qh4!? Rec8 19 Qh8+?! Ke7 20 Qxg7 Kd8 21
Qxf7 Qe7 with obscurity.
13 ... Kxh7 14 Ng5+ Kg8 15 Qd3 Rfe8
15 ... Rfc8!? is the alternative.
16 Qh7+ Kf8 17 Rhe1
As in the previously cited games with the Bxh7+ sacrifice, the way forward hereabouts is hardly clear.
Consider 17 Qh4!? Rec8 18 Rd3 for a speculative continuation, or 17 Qh5 Ke7 (17 ... Qe3+ 18 Kb1 Ke7
19 Qxf7+ Kd8 20 Rhe1) 18 Nxf7 Na5 is unclear too, but here not 18 ... Qe3+ 19 Kb1 Qxf4? 20 Rhf1 and
White is winning.
17 Qh8+ is not so hot as after 17 ... Ke7 18 Qxg7 Qe3+ 19 Kb1 Qxf4 20 Rhf1 Rg8! Black wins. Such a
line illustrates why prefacing the fireworks with Kb1, as Sax preferred, has its points.
17 ... Qb4!
Setting up the strong threat of ... Nc4.
18 Qh8+
On 18 Nce4 dxe4 19 Qh8+ Ke7 20 Qxg7 Kd8 and the king scurries away. 18 Qh4!? may be worth a
look.
18 ... Ke7 19 Qh4 Kd8!? (Diagram 4)

 Diagram 4
The black king runs for cover

Victor simply legs it! 19 ... f6? was not viable because of 20 a3! Qa5 21 exf6+ gxf6 22 Nxe6 Bxe6 23
f5 and the attack steams ahead. 19 ... Rh8 20 Nh7+ f6 21 exf6+ Kf7 22 fxg7 Rxh7 23 Qxh7 Rg8 24 Qh5+
Kxg7 is stronger. Because of the two threats of Nc4 or Qf4+ White would have to be satisfied with a
draw by perpetual check.
Korchnoi, the great counter-attacker, wants to win this game. I would not be surprised if he fully
appreciated that he was pushing things a bit far when passing over the objectively superior 19 ... Rh8.
20 Nxe6+ Kc8 21 a3
On 21 Nxg7? Nc4! Is too strong, e.g. 22 Na4 Qxa4 23 Nxe8 Bxe8 and Black wins.
21 ... Qe7 22 Ng5
Or 22 Qxe7 Nxe7 23 Nxg7 Rg8 24 Nh5 Rxg2.
22 ... f6! 23 e6?!
He ought to have played 23 exf6 Qxf6 24 Rxe8+ Bxe8 25 Nxd5 Nxd5 26 Rxd5 Bd7 27 Qh8+ Kc7 28
Qxa8 Qxf4+ 29 Rd2 Qxg5 30 Qf8 and White has a slight edge.
23 ... fxg5 24 exd7+ Qxd7 25 Qxg5
25 Rxe8+ was more exact as after 25 ... Qxe8 26 Qxg5 White would win the d5-pawn with unclear,
perhaps equal, play.
25 ... Rxe1 26 Rxe1 a6 27 f5
Better 27 Rd1 and White still has about equal chances. The balance of the game now starts to shift in
Korchnoi’s favour.
27 ... Kb8 28 Re6 Ka7 29 Rg6 Re8! 30 Rxg7 Re1+ 31 Nd1
31 Kd2 Qe8 32 b3 d4 wins.
31 ... Qe8 32 f6 32 Rg8 Qe4! 0-1
32 ... Qe4 33 Qd2 Nc4 34 Qf2+ Ne3! and White resigned.
One of many splendid fights between these two Grandmasters.

Game 40
Mrdja-Gleizerov
Turin 2000
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 Bxe7 Qxe7 7 f4 0-0 8 Nf3 c5 9 Qd2 Nc6 10 dxc5 f6
11 exf6 Qxf6 12 g3 Nxc5 13 0-0-0 Rd8 (Diagram 5)

Diagram 5
A typical Classical position

14 Nd4 Bd7 15 Qe3 Rac8 16 Be2 Be8 17 Bg4 Bf7 18 Rhe1 Nxd4 19 Qxd4 Rc6 20 Kb1
Obviously, White does not want to capture on f6 as this will only strengthen the black centre.
20 ... a6 21 a3 h6 22 Re5 Rdd6
This is an interesting position. Superficially, White would appear to be well on top as he has a
stranglehold on the key central dark squares. However, the black position is extremely solid and, although
the white position looks pretty, it is not easy to see where it can go. Meanwhile Black will try to free his
pieces and has a good long term target in the form of the white king. The c2-pawn in particular is a
sensitive spot in the white camp.
23 h4 Nd7 24 Re3 Bg6 25 Bh3 Qf7 26 Bf1 Nc5 27 Ree1 Nb3 28 Qf2 Na5 29 Bd3
Although the white bishop is, in terms of the pawn structure, superior to Black’s, it is the black one
which is creating more trouble and White hastens to exchange it.
29 ... Nc4 30 Qd4 Bh5 31 Ne2 Qc7 32 Ka2 Rd8 33 Rc1 e5 (Diagram 6)

 Diagram 6
Black achieves the ... e7-e5 break

Black has manoeuvred skilfully and the time is now ripe for this central break.
34 fxe5 Nxe5 35 Rf1 Nxd3 36 Qxd3 d4 37 Nf4 Bf7+ 38 Kb1 Bc4 39 Qf3 Bxf1 40 Qxf1 d3 41 Nxd3 Qxg3 42 h5 Rf6 43 Qh1
Qf3 44 Qh4 Rd5 45 Qc4 Kh7 46 Rg1 Rxh5 47 Qc7 Rf7 0-1

TIP: A position may sometimes appear attractive but if it is not going anywhere, its attractions may be short lived.

Game 41
Shirov-Morozevich
Fujitsu Siemens Giants, Frankfurt 2000

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 Bxe7 Qxe7 7 f4 a6!? 8 Nf3 c5 9 Qd2 Nc6 10 0-0-0 c4!? (Diagram 7)

Diagram 7
Flank attacks are pending

Short observed that Morozevich has a unique style. Here he clamps up the queenside.
 11 f5 Nb6
Of course 11 ... exf5? 12 Nxd5 would be daft.
12 fxe6 fxe6 13 h4 Bd7 14 h5 0-0-0 15 h6 gxh6 16 Rxh6
Shirov is one of the most dynamic of Grandmasters and so he sets about seeking the most active course
available. But in this game White’s advances lead nowhere whilst Black delivers a deadly assault from
long range.
16 ... Rdg8 17 Qf4 Be8 18 Qf6 Bg6
A known rerouting. The bishop reaches its best square.
19 g4
In retrospect he probably wishes that he had exchanged the queens.
19 ... Qe8 20 Bg2 Rf8 21 Qh4 Nb4 22 Rd2 Na4!
The fourth black piece to join in the attack. Already the white game is critical as if the knight at c3
moves a2 falls.
23 a3 Nxc3 24 bxc3
On 24 axb4 Qa4! is devastating.
24 ... Qa4! (Diagram 8)

Diagram 8
Black’s onslaught reaches critical mass

25 cxb4
If 25 axb4 Qa1 mate.
25 ... Qxa3+ 26 Kd1 Rxf3
Another attacker drops in. The game is over.
27 Qe7
27 Bxf3 Qxf3+ and the attack rages on, with Black suffering no material deficit either.
27 ... Re3 28 Qxe6+ Kb8 29 Qd6+ Ka8 30 Re2 Qa1+ 31 Kd2 Qc3+ 0-1
An astonishingly smooth despatch of one of the world’s best.

Game 42
Kasparov-Korchnoi
Zürich rapid 2001

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 h4

The Alekhine-Chatard Attack, which often leads to very sharp play. Taking the pawn grants White
tangible compensation and initiative.
6 ... Bxg5
Black accepts the offering. Of the numerous ways to decline the pawn, 6 ... c5 and 6 ... a6 are seen in
Games 43 and 44.
7 hxg5 Qxg5 (Diagram 9)

Diagram 9

8 Qd3
8 Nh3 Qe7 9 Nf4 is possible.
8 ... Nc6 9 Nf3 Qg6 10 Qxg6 fxg6 11 Nb5 Ke7
If instead 11 ... Kd8 then White hurls the other knight in with 12 Ng5.
12 Nxc7 Rb8 13 Nb5 Nb6 14 c3 Bd7 15 Bd3 Na5 16 b3 Bxb5 17 Bxb5 h6
Black accepts the gambit pawn

Black naturally wants to gain some space on the kingside with this move and ... g6-g5. However, it
falls prey to a typically brilliant Kasparov combination.
18 Nh4 Rhc8 19 Rh3 g5 20 Ng6+ Kf7 (Diagram 10)

 Diagram 10
Is the white knight trapped?

The knight looks stuck – has White overplayed his hand?
21 Rf3+!!
No. The black king is now lured to destruction.
21 ... Kxg6 22 Bd3+ Kh5 23 Rh3+ Kg4 24 f3+ Kf4 25 Kf2 g4 26 g3+ 1-0

Game 43
Nataf-Ulibin
Stockholm 1999

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 h4 c5 7 Bxe7 Kxe7 (Diagram 11)

Diagram 11
Why capture with the king?

If 7 ... Qxe7 8 Nb5 is very strong. However, the discomfiture of the black king is not a major factor in
the position.
8 f4 Qb6 9 Na4 Qa5+ 10 c3 b6
It is a close call as to which is the worst-placed piece – the black king on e7 or the white knight on a4.
Curiously, White now feels obliged to wander around with his own monarch.
11 Kf2 Ba6 12 Nf3 Bxf1 13 Rxf1 Nc6 14 Kg1 g6 15 b3 Rac8 16 a3 cxd4 17 cxd4 b5 18 Nc5 Nxc5
19 b4 Qa4 20 bxc5 Qxd1 21 Rfxd1 Rb8 22 Rdb1 Rb7 23 Kf2 Rhb8 (Diagram 12)

Diagram 12
Preparing a queenside breakthrough

A typically promising French endgame for Black. The white initiative has burnt out and almost all his
pawns are vulnerable.
24 Ra2 a5 25 Ke3 b4 26 Rab2 a4 27 axb4 a3 28 Rb3 Rxb4 29 Rxb4 Rxb4 30 Ra1 Rb3+ 31 Kd2
Nb4 32 Ne1 a2 33 Nc2 Rb1 34 Kc3 h5 35 g3 Kd8
Black can inevitably pick up a piece with ... Nxc2 but he first waltzes across with his king to prevent
subsequent white counterplay with Rxa2 and Ra7, hitting the f7-pawn.
36 Kd2 Kc7 37 Kc3 Kb8 38 Kd2 Nxc2 39 Rxa2 Nxd4 0-1

Game 44
Degraeve-M.Gurevich
Belfort 1997

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Be7 5 e5 Nfd7 6 h4 (Diagram 13)

 Diagram 13
A romantic gambit

6 ... a6
Taking time out to prevent the white knight sally to b5, but I think this popular move is already a
mistake, as it’s now hard to see from where the black counterplay will be coming.
7 Qg4
So often the most active post for the queen in the French.
7 ... Bxg5 8 hxg5 c5 9 Nf3
9 g6 f5! 10 Qf4 h6 stops the fun and games, but 9 dxc5 Nxe5 10 Qg3 is a path forward.
9 ... Nc6
Another line was 9 ... cxd4!? 10 Qxd4 Nc6 11 Qf4 Qb6! (and not 11 ... Qc7?! 12 0-0-0 Ndxe5 13 Nxe5
Qxe5 14 Qxe5 Nxe5 15 Re1 and White will win the d5-pawn with a good game, nor 11 ... d4?! 12 Nxd4
Nxd4 13 Qxd4 Qxg5 14 f4 Qg3+ 15 Kd2!? f6 16 Re1 fxe5 17 fxe5 0-0 18 Bd3 Qf2+ 19 Qxf2 Rxf2+ 20
Ke3 and White went on to win in Hector-Upmark, Stockholm 1994) 12 0-0-0 (12 Rb1?! d4 13 Ne4
Ndxe5! 14 Nd6+ Ke7 15 Nxe5 Qa5+ 16 Kd1 Qxe5) 12 ... Qxf2 13 Rd2 Qc5 14 g6!? fxg6 15 Bd3 Rf8
(this hits the white queen but weakens h7. Instead, 15 ... Nf8!? 16 Ng5 Qe7 17 Rf2 h6 18 Nf7 g5 19 Nd6+
leaves White with lots of compensation and continuing initiative) 16 Qg3 and White had interesting
compensation in Mirumian-Supatashvili, Ankara 1995.
10 dxc5
10 g6!? f5 11 Qg3 h6 12 Ng5! Nxd4 13 Nf7 Qa5 14 0-0-0! led to fantastical complications in the 1993
correspondence game Passos-Dutra Neto, viz. 14 ... Rg8 15 Kb1 b5 16 Qh4 b4 17 Rxd4!? cxd4 18
Nxd5!!.
10 ... Ndxe5 11 Nxe5 Nxe5 12 Qg3 Ng6
After 12 ... Nd7!? 13 Qe3 simply hanging on to the pawn is best.
13 0-0-0
White stands nicely out of the opening.
13 ... Bd7?!
13 ... Qe7!? deserved attention, e.g. 14 Bd3!? Qxc5 15 Bxg6 fxg6 16 Qe5 0-0 17 Rd2! with Black
retaining a dangerous, but defensible position. Might as well have a pawn for your troubles ...
14 Bd3 Qb8
14 ... Rc8?! 15 Bxg6 fxg6 16 Nxd5! exd5 17 Rhe1+ Kf7 18 Qf4+ Kg8 19 Rxd5 and the white attack
smashes through. Black will not be able to survive it.
15 Qe3 Ne7
15 ... Qe5? 16 Nxd5! exd5 (16 ... Qxd5 17 Bxg6 Qxa2 18 Bxf7+! Kxf7 19 Rxd7+ Ke8 20 Rxb7 wins)
17 Qxe5+ Nxe5 18 Rde1 f6 19 gxf6 gxf6 20 f4 and the black structure is so mauled that following the
regaining of his piece White was winning in Hector-Hansen, Gausdal 1987.
16 Bxh7! g6 17 Bxg6!
White gets for the piece three pawns and crushing attack against the uncastled black king.
17 ... Rxh1 18 Rxh1 Nxg6 19 Nxd5 Bc6 20 Nb6 Bxg2
Maybe 20 ... Ra7 was more persistent, but it’s very difficult to voluntarily put a rook on a7 where it
simply can’t move!
21 Rh6! Ke7
21 ... Ra7? now fails to 22 Rxg6! fxg6 23 Qxe6+ Kf8 24 Nd7+.
22 Qd4! (Diagram 14)

 Diagram 14
Black is busted

22 ... Qd8
It was still possible to put up some resistance after 22 ... e5 23 Qd6+ Qxd6 24 cxd6+ Kxd6 25 Nxa8
Kc6, but 23 Qg4! Bc6 24 Qf5 is decisive.
23 Qf6+ Ke8 24 Nxa8 Qa5 25 Qc3
The game is over: White has the exchange and two pawns and Black has no counterplay.
25 ... Qxa2 26 b3 Be4 27 f3 Bxc2 28 Qxc2 Qa1+ 29 Kd2 Qd4+ 30 Qd3 Qf2+ 31 Kd1 Qg1+ 32 Kc2 Qxc5+ 33 Qc3 Qxg5 34
Rh2 Kf8 35 Nb6 Qb5 36 Nc4 Ke7 37 Kb2 Nf4 38 Rd2 Nd5 39 Qe5 Qc5 40 Qd4 Qc7 41 Ne3 1-0

Game 45
Fischer-Rossolimo
US Championship 1965

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4

The McCutcheon Variation.
5 e5 h6 6 Bd2 Bxc3 7 bxc3
7 Bxc3 looks more natural but it fails to gain an advantage (see Game 49).
7 ... Ne4 8 Qg4 g6
The main line of the McCutcheon. If Black prefers not to weaken the dark squares there is also 8 ... Kf8
(Games 47 and 48).
9 Bd3 Nxd2 10 Kxd2 c5 (Diagram 15)
 Diagram 15
‘Both sides stand worse’

11 Nf3 Nc6
The classical French development. Nowadays Black often prefers a different scheme: ... Bd7-c6, ...
Nd7, ... Qe7 and ... 0-0-0, as in Game 46.
12 Qf4 Qc7 13 h4 f5
Black exploits the position of the white queen to block the kingside.
14 g4 cxd4 15 cxd4 Ne7
Black would do better to concentrate on development with 15 ... Bd7
16 gxf5 exf5 17 Bb5+ Kf8 18 Bd3 Be6
White’s problem is that his knight is not contributing to the game, so ...
19 Ng1
... angling for f4.
19 ... Kf7 20 Nh3 Rac8 21 Rhg1 b6 22 h5 (Diagram 16)

 Diagram 16
White begins the assault

22 ... Qc3+ 23 Ke2 Nc6

Black’s problem is that 23 ... g5 is demolished by 24 Nxg5+ hxg5 25 Qxg5.
24 hxg6+ Kg7 25 Rad1 Nxd4+ 26 Kf1 Rhe8 27 Rg3 Nc6 28 Qh4 Nxe5 29 Nf4 Ng4 30 Nxe6+ Rxe6 31 Bxf5 Qc4+ 32 Kg1 1-
0

Game 46
Sutovsky-Zifroni
Israel 2000

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4 5 e5 h6 6 Bd2 Bxc3 7 bxc3 Ne4 8 Qg4 g6 9 Bd3 Nxd2 10 Kxd2 c5 11 Nf3 Qe7 12 h4 Bd7
13 Qf4 Bc6 14 Nh2 Nd7 (Diagram 17)

Diagram 17
The modern black deployment

15 Ng4
 White has formed a plan, as so often occurs in the French, based on gaining control of the dark squares.
Black counters with an imaginative pawn sacrifice.
15 ... 0-0-0 16 Nxh6 f5 17 exf6 Nxf6 18 Ng4 Ne4+ 19 Bxe4 dxe4 (Diagram 18)

Diagram 18
Trouble brews on the d-file

Black’s clever play has shown up the seamy side of the white king position on d2.
20 Ne5 cxd4 21 Nxc6 bxc6 22 Rab1 dxc3+ 23 Ke1 Qd6
With both kings exposed, the player now have to bale out to a draw.
24 Qxd6 Rxd6 25 Rh3 e3 26 Rxe3 Rxh4 27 Rb3 Rh1+ ½-½

Game 47
J.Polgar-Korchnoi
Wijk aan Zee 2000

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4 5 e5 h6 6 Bd2 Bxc3 7 bxc3 Ne4 8 Qg4 Kf8 (Diagram 19)
Diagram 19
Compare this with Game 6

A common displacement in this line. Neither king castles.
9 Bd3 Nxd2 10 Kxd2 c5 11 h4 Nc6 12 Nf3 c4 13
Be2 b5 14 a3 Bd7 15 Qf4 Ke7
Korchnoi, one of the greatest French Defence players, elects to have his king live in the middle.
16 h5 Be8 17 Nh4 a5 18 Qg3 Rg8 19 Qe3 Bd7
In anticipation of the coming f5 he places this piece to control that square.
20 f4 b4 21 Rhb1
I think that I would have taken twice on b4.
21 ... bxc3+ 22 Qxc3 Qc7 23 g4 Rgb8 24 f5 Qa7
The critical moment. The only defence of d4 is 25 Nf3 (25 Ke3? Nxe5), when I judge White to be
slightly better. Instead Judit, as is her wont, sacrifices for attack.
25 f6+!? gxf6 26 exf6+ Kxf6 27 Rf1+ Ke7
The d4-pawn still hangs.
28 Rf4 a4!
A key defensive motif appears: the queen exchange.
29 Bxc4!? dxc4 30 d5 (Diagram 20)

Diagram 20
Can the black king survive?

30 ... Qa5!
In the nick of time. Definitely not 30 ... exd5? as 31 Qf6+ Ke8 32 Re1+ Be6 33 Rxe6+ fxe6 34 Qf8+
Kd7 35 Rf7+ wins, and 30 ... Nd8 meets with 31 Qf6+ Kd6 32 Nf5+ and the attack should triumph. Viktor
bales out.
31 dxc6 Bxc6 32 Qxa5 Rxa5 33 Re1
White’s activity and superior co-ordination mean that it will be hard for Black to make anything of the
extra pawn here.
33 ... Bd5 34 Kc3 Kf8 35 Ref1 Rb7 36 Ng6+ Kg8 37 Ne5 Bg2 38 Re1 Rc7 39 Kb4 Ra8 40 Rxc4 Rb8+ 41 Kc3 Rxc4+ 42
Nxc4 Kg7 43 Ne3 Bc6 44 Rf1 Rb5 45 Rf4 Re5 46 Kd3 Ra5 47 Kd2 Rc5 48 Kd3 ½-½

Game 48
De Firmian-Kaidanov
US Championship 1999

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4 5 e5 h6 6 Bd2 Bxc3 7 bxc3 Ne4 8 Qg4 Kf8 9 Bd3 Nxd2 10 Kxd2 c5 11 Nf3 Nc6 12 Qf4 c4
Again Black chooses to block it all up.
13 Be2 b5 14 a3 Rb8
This looks inferior to 14 ... Bd7, which would transpose to the previous game after 15 Qf4.
15 h4 a5 16 h5 b4 (Diagram 21)

Diagram 21
An over-hasty pawn break

17 axb4 axb4 18 cxb4 Rxb4
Black just does not look equal here to me. His bishop is still bad and his dark squares weak.
19 Qe3 Qb6 20 Ra8 Qb7 21 Rha1 Kg8 22 Bd1
With the intent of c3 and Bc2.
22 ... Rb2
Holding that up, but only temporarily.
23 Qa3 Kh7 24 Kc1 Rb6 25 c3 Re8 26 Ba4
Another plan. He wants to show up the guy at c8 as an embarrassment.
26 ... Kg8 27 Bxc6 Qxc6 28 Qc5 Bd7
On 28 ... Qxc5 29 dxc5 Rc6 30 R1a5 and Nd4 will follow whereupon the knight is fabulously placed
to help the new c-pawn’s advance.
29 Qxc6 Bxc6 30 Rxe8+ Bxe8 31 Ra8 Kf8
Black is not able to get his bishop outside the pawn chain, so this will be another good knight versus
bad bishop scenario. Nick de Firmian sets about expanding.
32 Nh4 Rb3 33 Kc2 Rb7 34 f4 Ke7 35 f5 (Diagram 22)

Diagram 22
A kingside attack in the endgame

35 ... Bc6 36 fxe6 Kxe6
Forced. 36 ... fxe6 37 Ng6+ Kf7 38 Rf8 mate, or 37 ... Kd7 38 Rg8 and the pawns go.
37 g4 Rc7 38 Rd8 Bd7 39 Nf5
Black is now demonstrably indefensible.
39 ... f6 40 Nxg7+ Ke7 41 Rh8 fxe5 42 dxe5 Bxg4 43 Rxh6 Kf7 44 Ne6! Re7
As 44 ... Bxe6 45 Rh7+ wins the rook.
45 Nd8+ Kg8 46 Rg6+ Rg7 47 Kd2 Rxg6
47 ... Bxh5 48 Rxg7+ Kxg7 49 Ke3 will lead to the capture of the remaining black pawns.
48 hxg6
White will be two pawns up in the minor piece ending and that’s a simple win.
48 ... Kg7 49 Ke3 Kxg6 50 Kd4 Kf5 51 Kxd5 Bf3+ 52 Kd4 Bh5 53 Nb7 1-0

Game 49
Sutovsky-Psakhis
Tel Aviv 1999

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4 5 e5 h6 6 Bd2 Bxc3 7 Bxc3 Ne4 8 Ne2

Fischer once stuck this bishop out at a5 against Petrosian. He never did so again. Note the trick 8
Bb4?! c5 9 dxc5? Nxf2! 10 Kxf2 Qh4+ and 11 ... Qxb4.
8 ... Nc6
Igor Glek’s recommendation. 8 ... c5 9 dxc5 Nxc3 10 Nxc3 Qa5 11 Qd2 Qxc5 12 f4 0-0 13 0-0-0 Nc6
14 h4 a6 15 Kb1 b5 also worked out okay for Black in Berzinsh-Achoundov, World Under 18
Championship 1992.
9 Nf4!?
An experiment.
9 Ng3 Nxc3 10 bxc3 g6 11 h4 Qe7
11 ... Ne7 12 Be2 c5 13 0-0 c4 14 h5 Nf5 15 Bg4 Qg5 was satisfactory for Black in Krakops-Glek,
European Team Championship 1997.
9 ... Qe7 10 Bd3 Nxc3 11 bxc3 Bd7 12 Nh5 Rg8 (Diagram 23)

Diagram 23
A tacit draw offer

On 12 ... 0-0-0 13 Nxg7 Rdg8 14 Nh5 Rxg2 15 Nf6 the occupation of the hole at f6 is very
disagreeable for Black. Alternatively 13 ... Qg5 14 Nh5 Qxg2 15 Ng3 Qh3 (or 15 ... Rdg8 16 Bf1) 16
Qh5 Qxh5 17 Nxh5 leads to similar problems. But 12 ... Qg5?! was not out of the question.
13 Bh7
Critical complications follow 13 Qg4 0-0-0 14 Bh7 Rh8 15 Qxg7 when Black mounts an original
counter-initiative by 15 ... Qa3! 16 Kd2 (or 16 0-0 Qxc3 17 Nf6 Nxd4 and Black is okay) 16 ... Qa5! 17
Nf6 Nxd4 18 Nxd7 Kxd7.
13 ... Rh8 14 Bd3
14 Nxg7+? Kf8 and White loses one of his minor pieces.
14 ... Rg8 15 0-0!?
Taking the draw may have been objectively better, as the long-castled black king proves hard to get at.
A knight on h5 is not effective in an attack on a king at c8.
15 ... 0-0-0 16 a4 Na5 17 f4 c5 18 Bb5 Kb8 19 Bxd7 Qxd7 20 Qd3 Qc7 21 Rab1 Rc8 (Diagram 24)

Diagram 24
Piling down the c-file

It is now difficult to organise defence of both pawns on the vulnerable c-file.
22 f5 cxd4 23 cxd4 Qxc2 24 Qb5 Qc3
Psakhis accurately organises his game and stays on top of the situation.
25 Kh1 Rc6 26 h3 a6 27 Qe2 Qxd4 28 Rf4 Qc3 29
fxe6 fxe6 30 Rf7
This counter-demonstration comes too late.
30..d4 31 Rd7 Rgc8 32 Qg4 d3 33 Qxg7 R6c7 34 Rxc7 Rxc7 35 Qg8+ Ka7 36 Qxe6
Material equality, but the d-pawn is a monster. Note how the h5 knight is still out of it.
36 ... Nc4 37 Nf4 d2 38 Rg1 Qd4 0-1
On 39 Rd1 Ne3 decides.
A game very much in the counter-attacking vein of Lev Psakhis at his best.

Game 50
Lanka-Morozevich
Kishinev 1998

1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 Bg5 Bb4 5 e5 h6 6 Be3!?

A sideline and a speciality of the Russian trainer Dvoretsky.
6 ... Ne4 7 Qg4 Kf8
On 7 ... Nxc3? 8 Qxg7 there is no satisfactory way out for Black, as 8 ... N (any check) is met by 9 c3,
and after he saves the rook White takes his bishop. Alternatively 8 ... Rf8 9 a3 Ba5 10 Bd2 means that
White gets his piece back and terrible damage has been inflicted on the black kingside.
8 a3
Continuing vigorously. 8 Ne2 is more sedate.
8 ... Bxc3+ 9 bxc3 c5!
Taking the pawn leaves White with loads of play after 10 Nf3.
10 Bd3 Nxc3
But now White has the irritant ... c4, shifting his bishop, to contend with.
11 dxc5 Nc6 12 Nf3 f5!? 13 exf6 Qxf6
Threatening 14 ... e5 and 15 ... e4, forking.
14 Qh5 e5 15 Bg6 (Diagram 25)

Diagram 25


15 ... Be6 16 0-0 Bf7 17 Nh4 Re8 18 Rae1 Kg8 19
f3 Re6
Forcing the exchange of bishops.
20 Bxf7+ Qxf7 21 Qg4 Rf6 22 Bd2 Nb5 23 f4 e4
24 c4
Frenetically trying to create an initiative before Black consolidates.
24 ... Nxa3!?
Pawn grabbing has cost some games through the opponent developing an initiative, but it has also won
many. Play through some of Short’s games to see what I mean.
A critical position in the 6 Be3 line

25 cxd5 Qxd5 26 Re3 h5!? 27 Qe2 Nc4 28 Rd1 Nd4 29 Qf1 Kh7
Connecting his rooks at move twenty-nine.
30 Rc3 Nxd2 31 Rxd2 Rd8 32 f5 (Diagram 26)

Diagram 26
How can Black best consolidate?

Moro has achieved a winning position and there must be several ways of steering the ship in safely to
harbour. But he now, for reasons that escape me, went into a storm. Maybe he had visions of the knight
dropping into g6, setting up swindles!?
32 ... g5?!
Certainly not necessary.
33 fxg6+ Kg7 34 Qe1 Rdf8 35 Rc1 Kg8 36 g7 R8f7 37 h3 Qe5 38 Qe3 Nc6 39 Re2 Kxg7 40 Qxe4
Qg5
Completing his unusual consolidation process.
41 Qc4 Qg3
The h4 knight is revealed as stranded.
42 Qa4 Rf4
Winning it.
43 Qa1+ R7f6 44 Rd1 Qxh4 45 Rd7+ Kh6 46
Kh2
Resigning was an option.
46 ... Rf2 47 Qc1+ Qf4+ 48 Qxf4+ R6xf4 49 Re8 Rf7 50 Re6+ Kg7 0-1
Chapter Seven
The Tarrasch with 3 ... Nf6
Introduction
After 1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 White is already facing a critical choice. Should he seize a
lot of space in the centre with 5 f4 or settle for the more modest 5 Bd3, making rapid development his
priority? Note that if he wishes White can defer the choice a move with 5 c3, but then it is necessary to
make up his mind. Let’s look at both moves in turn.

5 Bd3 – White’s Usual Move

In most cases White plays the direct developing move 5 Bd3, when play normally goes 5 ... c5 6 c3 Nc6 7
Ne2 cxd4 8 cxd4 (Diagram 1).

Diagram 1
The battle for the centre begins

All these moves are self-explanatory: Black attacks d4 and White defends it. Now, however, because
White has avoided 5 f4, Black can effectively undermine the e5 point with 8 ... f6! when if 9 f4?! (this is
almost never played) 9 ... fxe5 10 fxe5 Nxd4! (a little tactical trick) 11 Nxd4 Qh4+ wins back the knight a
pawn up. So unless he decides to play the very sharp and unclear 9 Nf4!? White has to acquiesce in the
line 9 exf6 Nxf6 (interesting is 9 ... Qxf6) 10 0-0 Bd6 11 Nf3 (Diagram 2).

 Diagram 2
White tries to control e5

White hopes to get the advantage by exploiting the hole on e5 and the backward e6-pawn, which makes
the bishop on c8 a feeble piece. Black meanwhile dreams of striking a tactical blow along the half open f-
file after ... 0-0 or against the h2-square after ... Qc7 etc. Alternatively he may try to achieve the advance
... e6-e5 under favourable circumstances and so activate his queen’s bishop. Taking play a little further, if
11 ... 0-0, White often takes the chance to neutralise any potential threat to h2 and generally strengthen his
grip on the dark squares with 12 Bf4. This is a clear plan, but it leads to highly murky consequences! In
order to keep his dark squared bishop Black can defer castling for a move in favour of 11 ... Qc7!?.
Besides these plans, Black can add to the pressure on d4 by playing ... Qb6 at some point. This idea is examined in games Game 51 and
52. Meanwhile an idea for White is to avoid playing Ne2 in favour of Ngf3!?. This proves highly effective in Kasparov’s hands in Game 53,
though giving up a pawn for long term compensation won’t be to everyone’s taste. Also noteworthy is the direct 4 e5 Nfd7 5 Ngf3 c5 6 c4!?
of Game 56.

5 f4 – White’s most Enterprising Move

With 5 f4 White strengthens his grip on the e5-square. The move f2-f4 is also positionally desirable as preparation for the breakthrough f4-f5.
If you look at games in the Winawer and Advance Variations you will see that White often makes a big effort to get in f4-f5, but to do so he
first of all has to move the knight from f3. Besides wasting time this can be problematical if Black has already begun putting pressure on the
d4-square. So, strategically speaking, it saves White a lot of hassle if he can play f2-f4 before Nf3. So why doesn’t everyone automatically
play 5 f4 here? The reason is that White is neglecting his development for a move and also loosening his pawn structure. This last statement
may sound strange as we have just said that the move strengthens e5. However, there is now a potential weakness down the a7-g1 diagonal
and the d4-pawn is in the firing line.
Black has two distinct responses to 5 f4.

Black Plays Solidly

First of all, Black can arrange to play ... f7-f5 himself (Diagram 3).
 Diagram 3
A typical ... f7-f5 position

This is aimed at preventing the f4-f5 breakthrough discussed above. It also blocks a white bishop from attacking h7 after Bd3, so makes
Black’s king safer after he castles kingside. This is an essentially defensive approach, but having secured the kingside Black can look for
activity on the other wing by utilising his queenside pawns. It is reasonable to assume that White will play c2-c3, so Black will have a target to
undermine with the eventual pawn thrust ... b5-b4. Meanwhile White won’t be idle. He should try to break down Black’s defensive bastion in
the kingside with a well prepared g2-g4 advance.

Black Counterattacks
Instead Black could try to exploit the move 5 f4 in a more direct way by putting immediate pressure on d4. Note that besides obvious attacking
moves such as ... c7-c5, Nc6 and Qb6, this can involve undermining the defenders of the pawn (see Diagram 4).

Diagram 4
White has a big centre

In a sense this is achieved by Qb6 as the bishop on c1 can’t go to e3 without dropping the b2-pawn.
However, in Game 54, Vaganian – who played the French like a true maverick in his early days –
preferred 7 ... Qa5!? which causes discomfort to d4-pawn by pinning its defender on c3. This provoked 8
Kf2, when he later exploited the displaced white king to win in glorious style.
Other equally subtle methods are also possible. For example, from the diagram play could go 7 ... Qb6 8 g3 cxd4 9 cxd4 Bb4+!
This is already highly disruptive as blocking with 10 Bd2 loses the d4-pawn, so White must move his king. 10 Kf2 g5! Now the
threat is 11 ... g4 chasing away the knight and winning the d4-pawn. If 11 Nxg5 Qxd4+ is a good swap for Black, so White might
play 11 fxg5 when 11 ... Ndxe5 disposes of his proud e5-pawn with unclear play. You can find analysis of these lines in Game 55.

NOTE: In a number of these blocked French positions, the kings, especially the white one, wander around in the centre of the
board.

You will see from the diagrams that White regards the move Ndf3 as the best way to consolidate his centre rather than the
natural Ngf3. This is because the king’s knight already has a decent square ready for it on e2 where it bolsters d4 – so it would
be silly to take away the only good square available to the other knight by putting it on f3. Also it becomes apparent that White
has to accept that his king might end up remaining in the centre, perhaps on f1 or f2. I imagine this is what puts off a lot of
players off this line as White – at least if you play the alternative 5 Bd3 you get to castle!

Illustrative Games

Game 51
Korneev-Matamoros Franco
Elgoibar 1997

1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 Bd3 c5 6 c3 Nc6 7 Ne2

The usual development – f3 is left free for the queen’s knight. 7 Ngf3 is seen in Game 53.
7 ... cxd4 8 cxd4 f6 9 exf6
The reserved line. There are wild complications after 9 Nf4 Nxd4 10 Qh5+ Ke7 11 Ng6+ hxg6 12
exf6+ Nxf6 13 Qxh8.
9 ... Nxf6 10 0-0 Bd6 11 Nf3 Qb6 12 b3 0-0 13 Bf4 (Diagram 5)

Diagram 5
Exchanging Black’s good bishop

The exchange of the dark-squared bishops, as we have already seen in several games, fits in well with
White’s strategy. Very often, in this line too, White effects it by the route g5-h4-g3.
13 ... Bxf4 14 Nxf4 Bd7
14 ... Ne4 is seen in the next game.
15 Re1 Rae8 16 Qd2 Kh8
Black is failing to play vigorously enough.
17 Bc2 Qb4 18 Rad1
White has no objection to exchanges. His bishop is the better and that will tell in an ending.
18 ... Qxd2 19 Rxd2 Ne4 20 Bxe4 Rxf4 21 Bc2 Re7 22 Re3 Bc8 23 g3 Rf6 24 Ne5 Nxe5 25 Rxe5
(Diagram 6)

Diagram 6
White blockades the backward e-pawn

Almost a classic position for this variation. White prosecutes his advantage systematically.
25 ... Rc7 26 Kg2 Kg8 27 h4 Bd7 28 h5 Rf8 29 Bd1 Rfc8 30 Bg4 Kf7 31 Ree2 Rc6 32 Re3 R8c7 33
Rf3+ Kg8 34 Rf4 b5 35 Rd3 Rc8 36 h6!
Forcing further concessions.
36 ... gxh6 37 Rf6 Kg7 38 Rdf3 Be8 39 Rxe6 Rxe6 40 Bxe6
Doubling and isolating the black h-pawns in return for winning the much healthier e6-pawn constitutes
real progress for White.
40 ... Rc2 41 a3 Rd2 42 Rf4 Ra2 43 Bxd5 Bg6 44 b4 Rxa3 45 Bc6 a6 46 d5 Rd3 47 Kf1 Bh5 48 g4
Bg6 49 Ke2 Kg8 50 Bd7!
Switching to a better square with the aid of the tactic 50 ... Rxd5? 51 Be6+ winning.
50 ... Kh8 51 Be6 Kg7 52 f3 Kh8 53 Rf8+ Kg7 54 Rg8+ 1-0
and 55 Rxg6+ will win the rook at d3, so Black resigned.

Game 52
S.Hansen-Brynell
Gothenburg 1998

1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 Bd3 c5 6 c3 Nc6 7 Ne2 Qb6 8 Nf3 cxd4 9 cxd4 f6 10 exf6
Nxf6 11 0-0 Bd6 12 b3 0-0 13 Bf4 Bxf4 14 Nxf4 Ne4
A bid for activity, but it does not change things much.
15 Ne2 Nd6 16 Qd2 Bd7 17 Rad1 Nf5 18 Bb1 a5 19 Nc3 Nce7 20 Ne5
Such a nice outpost.
20 ... Be8 21 Rfe1 (Diagram 7)

Diagram 7
White takes a grip on e5

21 ... Rc8 22 Ne2 Nc6 23 Nxc6 Rxc6 24 Nf4 Bd7 25 g3
Solidifying, and placing the pawns on the right coloured squares for the bishops.
25 ... a4 26 bxa4 Rc4 27 Bxf5!
Good knight versus bad bishop again!
27 ... Rxf5 28 a5 Qc6 29 Rc1 Rf8 30 Nd3 Qd6 31 Nc5
Another good outpost.
31 ... Bc8 32 Qb2 h6 33 Re5 Rxc1+ 34 Qxc1 (Diagram 8)

Diagram 8
Black is without counterplay
34 ... b6 35 axb6 Qxb6 36 a4
 A clear pawn ahead with the much superior minor piece. Black could resign here.
36 ... Qb4 37 Qe3 Rf7 38 h4 Bd7 39 Nxd7 Rxd7 40 Rxe6 Qxa4 41 Re8+ Kh7 42 Qd3+ g6 43 Re6
Winning the g6-pawn.
43 ... Rg7 44 h5 1-0

Game 53
Kasparov-Bareev
World Cup of Rapid Chess, Group A, Cannes 2001

1 e4 e6 2 d4 d5 3 Nd2 c5 4 Ngf3 Nf6 5 e5 Nfd7

5 ... Ne4!? is more fun.
6 c3
Now the game transposes to the 3 ... Nf6 4 e5 Nfd7 5 Bd3 variation with 7 Ngf3. The experimental 6
c4!? is tried in Game 56.
6 ... Nc6 7 Bd3 Qb6 8 0-0
First chance, Garry gambits. Nunn popularised this gambit in the late 1970s.
8 ... g6
Bareev had a thing about fianchettoing the king bishop in the French. Here taking the pawn with 8 ...
cxd4 9 cxd4 Nxd4 10 Nxd4 Qxd4 11 Nf3 Qb6 might have been better, or something more exotic like 8 ...
Be7 9 Re1 g5!?.
9 dxc5 Nxc5
9 ... Qc7, winning the e5-pawn, looks better.
10 Nb3 Nxd3 11 Qxd3 Bg7 12 Bf4 0-0 13 Qd2 Bd7 14 Rfe1 a5 15 Bh6 (Diagram 9)

Diagram 9
Exchanging dark-squared bishops again

Just like Shirov in the earlier game, here too Black finds that the exchange of the dark-squared bishops
weakens his kingside and leaves him almost strategically lost. He has no counterplay and the greatest
attacker ever is about to go to work on his king’s position.
15 ... a4 16 Bxg7 Kxg7 17 Nbd4 Na5 18 Rab1
Not the only way of continuing the build up, but Garry takes things calmly.
18 ... Nc4 19 Qf4 Qd8 20 h4
A dismantler en route.
 20 ... h6 21 Qg3 Qe7 22 Ne2
Off to help in the assault.
22 ... Kh7 23 Nf4 Rg8 24 Re2 Raf8 25 Rbe1
Is this an example of Nimzovitschian ‘overprotection’ of the strongpoint at e5 from Kasparov?
25 ... Rc8 26 Nh2
Eyeing up the horrible hole at f6.
26 ... g5 27 Nh5 gxh4 28 Qh3 Rg5 29 Nf6+ Kg7 30 f4
Hereabouts Bareev could have resigned with a clear conscience.
30 ... Rg6 31 Nhg4 Rh8 32 Nh5+ Kf8 33 Ngf6
The knights swarm in.
33 ... Bc6 34 Qxh4 a3
The only active demonstration that Black has managed in the game, but it is quite ineffectual.
35 b3 Nb2 36 Kh2 Qc5 37 Re3 d4 38 Rg3 dxc3? (Diagram 10)

Diagram 10
White wins by force

39 Rxg6 fxg6 40 Nd7+! Bxd7 41 Qf6+ 1-0
Mate next move.

Game 54
Reshevsky-Vaganian
Skopje 1976

1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 f4 c5 6 c3 Nc6 7 Ndf3 Qa5

The main line – 7 ... cxd4 8 cxd4 Qb6 – is seen in the next game.
8 Kf2
It’s a closed sort of opening, so moving the king early like this is something you can get away with. But
not always. 8 dxc5, 8 Be3, 8 Bd3 and 8 a3 are all viable alternatives.
8 ... Be7 9 Bd3 Qb6 10 Ne2 f6 11 exf6
On 11 Kg3 it’s a safe bet that 11 ... g5!? would have followed, as indeed it did in a game between
Adorjan and Vaganian a few years before. Black won it.
11 ... Bxf6 12 Kg3
Looks a bit odd to you? Does to me too.
12 ... cxd4 13 cxd4 0-0 14 Re1?
This is the real error. 14 h4! would stop all the brouhaha to which we are now about to be treated and
might yet leave White with real hopes of an advantage, as he may yet drop the king back to h2. But it is
now out in the open, and Vaganian, like a Brazilian eyeing an errant Seaman, spots that this grants him a
marvellous chance.
14 ... e5!! (Diagram 11)

Diagram 11
The first of multiple sacrifices

For the rest of the game Reshevsky gets blasted. Of course Black could not have taken on d4 because
of Bxh7+ at the end winning his queen.
15 fxe5 Ndxe5!
Sac number one.
16 dxe5 Bh4+!
Number two.
17 Kxh4
17 Nxh4 Qf2 mate.
17 ... Rxf3!
Number three.
18 Rf1
18 gxf3 Qf2+ 19 Kg5 h6+ 20 Kf4 g5 is mate, so Reshevsky scrambles out another defender, but it is
already too late. The white king is indefensible out in the fields.
18 ... Qb4+! 19 Bf4 Qe7+ 20 Bg5 Qe6! (Diagram 12)

 Diagram 12
The white king is indefensible

Threatening mates at h3 and g4. After the shackles were thrown off them at move fourteen the black
pieces have become rampant.
21 Bf5
Tantamount to an acceptance of defeat, but 21 h3 fails to 21 ... Qxh3+!! 22 gxh3 Rxh3 mate.
21 ... Rxf5 22 Nf4 Qxe5 23 Qg4 Rf7 24 Qh5 Ne7 25 g4 Ng6+ 26 Kg3 Bd7 27 Rae1 Qd6 28 Bh6
A pointless move. Vaganian just ignores it.
28 ... Raf8 0-1
A magnificently conducted attack, suggestive of the great talent of Rafael Vaganian. With more
discipline he would have become a serious contender for the world championship. But in late 1986 when
Karpov was asked if Vaganian would be helping him in his next match, he replied: ‘He can’t even help
himself.’

Game 55
Kalichkin-Gavrjushin
Russian Cup, Perm 1997

1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 c3 c5 6 f4 Nc6 7 Ndf3 cxd4 8 cxd4 Qb6 9 Ne2

Reaching a position we have seen before in Chapter Four (Game 30). Instead, 9 g3 is the true Tarrasch
continuation, when play can continue 9 ... Be7 (9 ... Bb4+10 Kf2 g5 is another way to play) 10 Bh3 0-0 11
Ne2 f6 12 Rf1 (not 12 0-0? fxe5 13 fxe5 Ndxe5! wins a pawn, while after 12 Bxe6+ Kh8 13 Bxd5? fxe5
14 fxe5 Ncxe5 15 dxe5 Nxe5 Black had a big attack in Savon-Gleizerov, Cheliabinsk 1991) 12 ... Ndb8
13 Rf2 Bd7 14 Kf1 a5 15 Kg2 Na6 16 Nc3 Nc7 17 Be3 Qa7 18 Na4 fxe5 19 fxe5 Be8 20 Rc1 Bg6 with a
complicated game in McShane-Kindermann, Lippstadt 1998.
9 ... Be7 10 g4 f6 11 Bh3 0-0 12 Rf1 fxe5 13 fxe5 Bb4+ 14 Bd2 Rxf3! (Diagram 13)
 Diagram 13
The familiar exchange sacrifice

Yet another example of the ‘French Exchange sacrifice’ ( ... Rxf3): in this case Black’s goal is not to
weaken White’s pawn formation on the kingside, but to destroy White’s pawn centre. In this particular
line of the French Defence this sacrifice is a fairly common theme.
15 Rxf3 Nxd4 16 Nxd4 Qxd4
For the sacrificed exchange Black has oodles of play.
17 Bxb4 Qxb4+ 18 Qd2 Qxd2+ 19 Kxd2 Nxe5 20
Rf2?!
It was probably better to have gone in for 20 Rc3 Nc4+ 21 Kc2.
20 ... g5!
Black blocks up the h3-bishop and establishes an outpost on f4; from that square the knight will support
a further advance of the central pawns.
21 Re1 Ng6 22 Kc3 Bd7 23 Bf1 (Diagram 14)

Diagram 14
The central pawns give good play

23 ... Kg7
He could also have set the pawns going now with 23 ... e5!?.
24 Kd4 Rc8 25 Re3
He might well have considered the line 25 Bd3 e5+ 26 Rxe5 Nxe5 27 Kxe5 Bxg4 28 Rg2 h5 29 h3
Bxh3 30 Rxg5+.
25 ... Rc1 26 Rc3 Rd1+ 27 Rd3 Rc1 28 Ra3?!
Wiser to have repeated with 28 Rc3.
28 ... a6 29 Rc3 Rd1+ 30 Kc5
30 Rd3 e5+ 31 Kxd5 Nf4+ 32 Rxf4 Rxd3+ 33 Kxe5! Rd1 34 Rd4 Rxd4 35 Kxd4 Bxg4 36 Bg2 b6 37
Kd5! would have left White at least equal in the bishop ending due to his active king.
30 ... e5 31 Kb6 Bc6
31 ... Bxg4!?.
32 Bxa6 d4 33 Rxc6?
A big error. 33 Rcc2 d3 34 Rcd2 Rxd2 35 Rxd2 e4 36 Bxd3 exd3 37 Rxd3 would have been not so
clear, and also 33 Rc5 was an option.
33 ... bxc6 34 Kxc6 e4
Now the black passed pawns are unstoppable.
35 b4 d3 36 Rb2 e3 37 a4 d2 0-1

Game 56
Yurtaev-Driamin
Russian Cup, Tomsk 1998

1 e4 e6 2 d4 d5 3 Nd2 Nf6 4 e5 Nfd7 5 Ngf3 c5 6 c4!?

Something of an experimental approach.
6..cxd4
6 ... Nc6!? was intriguing.
7 cxd5 exd5 8 Bb5 a6
I might have tried 8 ... Nc6!?.
9 Bxd7+ Bxd7 10 Nxd4 Nc6 11 N2f3 Be7 12 0-0 0-0 13 h3! (Diagram 15)

Diagram 15


Removing the one active option for Black’s queen’s bishop. A comfortable middlegame has resulted
for White, with Black lacking the activity for which one hopes when there is an isolated queen’s pawn.
The isolated pawn is blockaded

13 ... Rc8 14 Be3 Nxd4 15 Qxd4 Bb5 16 Rfc1 Qd7 17 Bg5!
Ridding Black of his bishop pair and amplifying the white advantage.
17 ... Bxg5 18 Nxg5 h6 19 Nf3 Rfd8 20 b3
Taking the c4-square away from him.
20 ... Be2 21 Nd2 Qf5 22 Re1 Qd3 23 Nf3! a5 24 Qxd3 Bxd3 25 Nd4 (Diagram 16)

 Diagram 16
The IQP is blockaded again

The horse arrives at the blockading square. Now it’s good knight versus bad bishop time.
25 ... Re8 26 Rad1 Ba6 27 Re3 g6 28 f4 b6 29 f5
Effective expansion.
29 ... Rc5 30 fxg6 fxg6 31 e6
With these pieces on the board, the new passed pawn is a very frightening weapon.
31 ... Re7 32 Rde1 Kg7 33 Rg3 Kf6 34 Kf2 Rc8 35 h4 g5 36 h5
A massive hole at f5 has opened up.
36 ... g4 37 Rxg4 Rg7 38 Rf4+ Kg5 39 Rf5+ Kg4 40 Rxd5 Bb7
Nothing to be done, so he ends it this way.
41 Re4 mate
Chapter Eight
The Tarrasch with 3 ... c5
Introduction
1 e4 e6 2 d4 d5 3 Nd2 c5
This is a radically different way to handle the Tarrasch. White isn’t allowed the time to play e4-e5 (as he
is after 3 ... Nf6) and consolidate his space advantage in a closed centre. Instead Black attacks d4
forthwith and cannot be prevented from dissolving White’s pawn centre. Therefore there is no danger of
his pieces suffocating behind a fixed pawn chain in the centre; his minor pieces cannot be denied strong
centre squares like f6 and c5, as is the case in most other French systems.
This all sounds wonderful for Black, but in chess, as in life, everything comes at a price. Black frees
himself from the tyranny of White’s pawn chain, but to do so he must either accept an isolated pawn or be
prepared to fall behind in development and face an attack by White’s pieces.
Black has to make his choice at move four after 4 exd5
Black plays 4 ... Qxd5
After 1 e4 e6 2 d4 d5 3 Nd2 c5 4 exd5 Qxd5 (Diagram 1)

Diagram 1
Black avoids an IQP

Black avoids an IQP (Isolated Queen’s Pawn) position so if nothing dramatic happens he will
complete his development with equal chances. He hopes that the fact that White has blocked in his
queen’s bishop with 3 Nd2 will prevent him taking advantage of Black’s early queen excursion.
This line is highly theoretical, as White can’t just sit back and trust in his centre pawns to give him a
nice edge – he doesn’t have any! In fact, White’s only chance for a long term, ‘quiet’ advantage would be
if he could somehow stop Black developing his queen’s bishop. This however isn’t very promising.
So White has to look for an attack on the kingside which has led to analysis that goes continually deeper. If White can’t find a decisive
blow then he will have to accept that Black has fully equalised. Thus in Game 60 Adams scores a tremendous win as White thanks to a
novelty he had prepared in a sharp line, whereas Kasparov tried to play it positionally (for a change!) against Anand in Game 61, and not only
did he fail to gain any advantage he even lost – an astonishingly rare thing to happen to the World Number One as White. In Game 62 Black
takes too many liberties and allows Adams to wrap up another win in this system by castling queenside and then going all out against the black
king. Still 1/3 isn’t bad for Black against Kasparov and Adams (twice)!

TIP: After 4 ... Qxd5, White must try to develop some sort of initiative as Black holds the long term trumps.

Black plays 4 ... exd5

1 e4 e6 2 d4 d5 3 Nd2 c5 4 exd5 exd5 (Diagram 2)

Diagram 2
Black takes on the IQP

is an entirely different way to play it. Black doesn’t fall behind in development, frees his light-squared
bishop and adopts a set up almost free from sharp tactical variations. The drawback is the blemish on his
pawn structure – the isolated pawn on an open file. If you want a text book lesson in how to play against
the IQP, have a look at Karpov wins in Games 57 and 58. Features that are good for White include getting
a firm grip on the d4-square and stationing a knight there. Knights love central squares where they can’t
be ousted by pawns. Another useful idea for White is to exchange off dark-squared bishops to consolidate
his hold on the d4-square. As a rule every exchange in an IQP position favours the attacker, as there will
be fewer pieces to inject dynamism into the situation and so distract him from the task of exploiting the
weak pawn.
Despite the static weakness of the IQP, it can still be proud to be centre pawn. Indeed, it controls the
c4- and e4-squares and can support a knight, bishop or even rook on one of these squares. At the same
time, Black’s pieces have a lot of freedom of action, and White has to watch out for combinations against
his f2-square. This is the weakest point in his camp and has become a potential target due to the
disappearance of his centre pawns. If you doubt Black’s counter attacking chances in the IQP, have a look
at Game 59.
You will see that Bareev adopted a different scheme to Karpov’s opponents. Namely after 3 Nd2 c5 4
exd5 exd5 5 Ngf3 rather than 5 ... Nc6 when the knight is pinned by 6 Bb5 he played 5 ... Nf6! whereupon
6 Bb5+ Bd7 7 Bxd7+ Nbxd7 avoided any pin and seemed quite promising for Black.

Illustrative Games

Game 57
Karpov-Uhlmann
Madrid 1973

1 e4 e6
Wolfgang Uhlmann always plays the French Defence.
2 d4 d5 3 Nd2 c5 4 exd5 exd5 5 Ngf3 Nc6 6 Bb5 Bd6 7 dxc5 Bxc5 8 0-0 Nge7
Black lands himself with an isolated queen pawn, but he hopes that the activity of his pieces will
compensate for that. When Black does not want an IQP then 4 ... Qxd5 is preferred instead (see Games 60
and 61).
9 Nb3 Bd6 10 Bg5 0-0 11 Bh4
Setting up a later exchange of the bishops at g3. With every exchange the possibilities for active play
for Black may well reduce and so the d5-pawn may become more of a problem.
11 ... Bg4
11 ... Qc7 is the next game.
12 Be2 Bh5 13 Re1 Qb6 14 Nfd4 (Diagram 3)

Diagram 3
Exchanges favour White

14 ... Bg6 15 c3 Rfe8 16 Bf1 Be4 17 Bg3 Bxg3 18 hxg3 a5
This looks like a sensible bid for activity, but it meets with an excellent and accurate response after
which Uhlmann probably regretted the weakening of b5.
19 a4! Nxd4 20 Nxd4 Nc6
On 20 ... Qxb2 21 Nb5 threatens a fork at c7 and also trapping the queen with 22 Re2.
21 Bb5 Red8 22 g4!
Already thinking about drowning out the black bishop.
22 ... Nxd4 23 Qxd4 Qxd4 24 cxd4
The position simplifies, but the black game is critical. White’s rooks are coming in.
24 ... Rac8
24 ... Kf8 looks better.
25 f3 Bg6 26 Re7 b6 27 Rae1 h6 28 Rb7 Rd6 29 Ree7 (Diagram 4)

 Diagram 4
Rooks on the seventh

Two in. Pure Karpovian domination. Doubled rooks on the seventh rank and the better bishop too.
Contrast his methods with those of Kasparov.
29 ... h5 30 gxh5 Bxh5 31 g4 Bg6 32 f4 Rc1+ 33 Kf2 Rc2+ 34 Ke3
and Black is indefensible.
34 ... Be4 35 Rxf7 Rg6 36 g5 Kh7 37 Rfe7 Rxb2 38 Be8 Rb3+ 39 Ke2 Rb2+ 40 Ke1
Dodging the checks.
40 ... Rd6 41 Rxg7+ Kh8 42 Rge7 1-0
Machine-like demolition.

Game 58
Karpov-Kuzmin
Leningrad Interzonal 1973

1 e4 e6 2 d4 d5 3 Nd2 c5 4 exd5 exd5 5 Ngf3 Nc6 6 Bb5 Bd6 7 dxc5 Bxc5

Karpov likes playing against this isolated queen’s pawn. Players like Korchnoi like to try to use it to
gain active middlegame chances.
8 0-0 Nge7 9 Nb3 Bd6 10 Bg5 0-0 11 Bh4 Qc7 12 Bg3 (Diagram 5)
 Diagram 5
Again White plays for exchanges
The familiar trade route.
12 ... Bxg3 13 hxg3 Bg4 14 Re1 Rad8 15 c3 Qb6
Often an active square for the queen in this line.
16 Bd3 Ng6 17 Qc2 Bxf3
This does not really work out in his favour. They advise never to give bishop for knight without good
reason.
18 gxf3 Rd6 19 f4 Rfd8 20 a3! h5?!
This escapade rebounds on him. His kingside is weakened.
21 Kg2 h4 22 Re2 Nf8 23 Nd2 Rh6 24 Nf3 hxg3 25 fxg3 Nd7 26 Rae1 Kf8 27 g4!
An unexpected and powerful advance.
27 ... Qc7 28 g5 Rh8 29 Kg3! (Diagram 6)

Diagram 6
The white king steps up

The king is a strong piece.
29 ... Nc5 30 Bf5! g6
It looks like White must start going backwards, but ...
31 b4! Ne4+
On 31 ... Nd7 White comes crashing in with 32 Bxg6! fxg6 33 Qxg6 with big pressure, but giving up a
pawn like this is not good enough either. 31 ... gxf5 32 bxc5 leaves f5 hard to defend.
32 Bxe4 dxe4 33 Qxe4 Kg7 34 b5 Na5 35 Qe7 Qxe7
On 35 ... Qxc3 36 Re3! wins.
36 Rxe7 Rd3 37 Rc7 Nb3 38 Kg4 Rf8 39 Ree7 1-0

Game 59
I.Gurevich-Bareev
Biel 1993

1 e4 e6 2 d4 d5 3 Nd2 c5 4 exd5 exd5 5 Ngf3 Nf6 6 Bb5+ Bd7 7 Bxd7+

7 Qe2+ Be7 8 dxc5 is another important line.
8 ... Nbxd7 8 0-0 Be7 9 dxc5 Nxc5 10 Nb3
10 Nd4!? Qd7 11 N2f3 0-0 12 Bf4 Rfe8 13 Re1 Bf8! was equal in Akopian-Bareev, Debrecen 1992.
10 ... Nce4 11 Bf4
11 Nbd4 0-0 12 Nf5 Bc5 13 Be3 Rc8! Rozentalis-Dokhoian, Klaipeda 1988.
11 ... 0-0 12 Qd3 Re8 (Diagram 7)

Diagram 7


In the isolated queen’s pawn middlegame Black often looks to make use of outposts for his pieces at e4
or c4. White often seeks simply to squash the opponent’s activity and get into a better ending.
13 Be5!? Qd7 14 a4
He might just have put the bishop on the blockading square: 14 Bd4!?.
14 ... a6 15 Rfd1 Rac8 16 c3 Bd8!
Preparing to re-post at a more active square, such as c7 or b6.
A typical IQP position

17 h3 h6 18 a5
Again, 18 Bd4 may have been better. The a5-pawn can become a liability now.
18 ... Rc6 19 Bxf6
On 19 Nbd4 Rc5 20 Nb3 Black might entertain avoiding the repetition with 20 ... Rb5?! 21 Bxf6 Bxf6
22 c4 Rb4.
19 ... Nxf6 20 Nfd4 Rc8
A phase of pure piece play now passes.
21 Nc2 Bc7 22 Ne3 Rcd8 23 Nd4
White pirouettes his knights in search of effective play against the isolani, but he does not quite get it.
23 ... Re4! 24 Nf3 Qc6 25 b4 Rde8!
25 ... Bf4 was an active alternative.
26 Kf1
On 26 Rac1 Bf4, and 26 Nxd5? loses to 26 ... Nxd5 27 Qxd5 Re1+! 28 Nxe1 Rxe1+ 29 Rxe1 Qxd5.
26 ... Rd8 27 Rab1 Bf4 28 Nf5
28 b5 axb5 29 Qxb5 might have been a better try for equality.
28 ... Qc7 29 N5d4 Rc8 30 Rb3
30 g3 allows the very dangerous incursion of 30 ... Bxg3! 31 fxg3 Qxg3.
30 ... Ree8!
Maybe the knight would like the e4 outpost!?
31 Ne2 Be5 32 Rc1 Ne4!
Seizing a tactical chance.
33 Qxd5?
Taking the bait, but he should have preferred the line 33 Nxe5 Qxe5 (aiming at h2) 34 Kg1 Rc6! 35
Qd4 (on 35 Nd4 Nxf2! 36 Nxc6 Nxd3 37 Nxe5 Nxc1 wins, or 36 Kxf2
Qf4+) 35 ... Qg5.
33 ... Bg3!! (Diagram 8)

Diagram 8
An unexpected incursion

Bang! Some loose pieces show up in the white camp and thus combinational possibilities exist.
34 c4
A sad move from Ilya, who by now must have been regretting taking the d-pawn, but on 34 Nxg3 Rcd8!
forces the queen away and 35 Qf5 Qc4+ then picks up the b3 rook. Or if 34 fxg3 Rcd8 35 Qf5 Qc4!
threatens the rook and a knight fork at g3.
34 ... Rcd8 35 Qf5
Or 35 Qh5 Qxc4 36 Re3 Bxf2 wins.
35 ... Qxc4!
35 ... g6 36 Qg4 h5 also did the trick.
36 Re3
36 Rxc4 Rd1+ 37 Ne1 Nd2+ 38 Kg1 Rxe1 is checkmate.
36 ... Bb8 37 g3
Taking the queen still leads to mate.
37 ... Qxb4
First fruits.
38 Kg2 Nd6 39 Qd3 Rxe3 40 Qxe3 Re8
Last move of the time control, but 40 ... Qxa5! just picked up a second pawn.
41 Qd3
41 Qc3 was a better try, but 41..Qxc3 42 Nxc3 Rc8 still leaves Black comfortably winning.
41 ... Qxa5
Winning.
42 Nf4 Qf5 43 Qb3 Nb5 44 Nd5 h5 45 h4 Bd6 46
Ng5 g6 47 Qc4 Kg7 48 f4 Kh6
Bareev made rather a meal of the technical process. I am sure that there were simpler ways than this of
winning the game.
49 Nc3 Rc8 50 Qb3
On 50 Qxf7 Qxf7 51 Nxf7+ Kg7 52 Nxb5 (or 52 Nxd6 Nxd6 53 Rd1 Rxc3 54 Rxd6 Rb3 and the b- and
a-pawns win trivially in the rook ending.) 52 ... Rxc1 53 Nbxd6 Rd1 54 Kf3 b5 and the knights are so tied
up looking after each other that Black can just drift the two passed pawns down towards victory.
Alternatively 50 Nxf7+ Kg7 51 Nxd6 Rxc4 52 Nxf5+ gxf5 was
winning.
50 ... Kg7 51 Nxb5 Rxc1 52 Nd4
52 Nxd6 Qc2+ 53 Qxc2 Rxc2+ 54 Kf3 a5 is an easy win, as is 52 Qb2+ f6.
52 ... Qf6 53 Qxb7 Be7 54 Qd5 Rc8 55 f5 Rd8 56 Nde6+ fxe6 57 Nxe6+ Kh6 58 Nxd8 Bxd8 59
fxg6 Kxg6 60 Qd3+ Kg7 61 Qd1 Qc6+ 62 Kh2 Bc7 63 Qd2
On 63 Qxh5 Qg6 wins.
63 ... Qc5 64 Kh3 Qe5 65 Qd7+ Kf6 0-1

Game 60
Adams-Dreev
Wijk aan Zee 1996

1 e4 e6 2 d4 d5 3 Nd2 c5 4 exd5 Qxd5 5 Ngf3

Adams enters the 4 ... Qxd5 main line since he has an improvement prepared. The inferior 5 dxc5 is
Game 61.
5 ... cxd4 6 Bc4 Qd6 7 0-0 Nf6 8 Nb3 Nc6 9 Nbxd4 Nxd4 10 Nxd4 a6 11 Re1 Qc7 12 Bb3 Bd6 13
Nf5!
Entering a maelstrom of complications in one of the sharpest and most theoretical lines in the French
Defence.
13 ... Bxh2+ 14 Kh1 0-0
14 ... Kf8 15 g3 exf5 16 Bf4 Qc6+ 17 Kxh2 leaves White well placed.
15 Nxg7! (Diagram 9)

Diagram 9
A critical sacrificial variation

The difficulties Black will experience in extricating the h2 bishop validate this sacrifice.
15 ... Rd8
On 15 ... Kxg7 16 Qd4! is very hard to meet.
16 Qf3
Adams had already beaten the same opponent in a game from 1992 with 16 Qe2! Kxg7 17 g3 Bxg3 18
fxg3 b5 19 Bf4 Bb7+ 20 Kh2 Qc5 21 Be5!.
16 ... Kxg7 17 Bh6+!
Some games saw the trapping of the bishop with 17 g3, but after 17 ... b5 Black developed counterplay
quickly enough, e.g. 18 Bf4?! Bb7! 19 Bxc7 Bxf3+ 20 Kxh2 Ng4+ 21 Kg1 Rd2 22 Rf1 Rc8 0-1 Ye
Rongguang-Dolmatov, Moscow 1992, or 18 Kxh2 Bb7 19 Qf4 Qc6 20 Rg1 Rd1! 21 Be3 Rxa1 22 Qh6+
Kg8 23 Qg5+ Kf8 24 Qc5+ Ke8 25 Qxc6+ Bxc6 26 Rxa1 Ng4+ 27 Kh3 Nxe3 28 fxe3 Rd8 Wolff-Gulko,
USA Championship 1992. Note in these games how it became apparent that the white king is also not
100% secure.
17 ... Kg6
Unsurprisingly after 17 ... Kxh6? 18 Qxf6+ Kh5 19 g4+ Kxg4 20 Rg1+ Bxg1 21 Rxg1+ it will be mate.
18 c3
Vacating c2 for the bishop.
18 ... Nd5
18 ... Rd5 19 Bc2+ Rf5 20 g4 is also going to end in White’s favour, but 18 ... Nh5 may be an
improvement.
19 Rad1!
Adams brings out an extremely important novelty (discovered by Speelman) that changed the
evaluation of this terribly complex line. Khalifman-Gulko, Lucerne 1993, saw White bring up a rook to
help with the attack, but this proved not quite enough to finish the job: 19 Re4 Kxh6 20 Rg4! (threatening
21 Qh3 mate) 20 ... Nf4! 21 g3 (On 21 Kxh2 Black tidies up with Ng6, whereupon he is winning) 21 ...
Rd3! 22 Qxf4+ Qxf4 23 gxf4! (On 23 Rxf4 Bxg3 24 fxg3 f5 etc.) 23 ... Rh3! 24 Kg2 Rh5 25 Rh1 Bd7?! 26
Rxh2 Bc6+ 27 Kg1 Rxh2 28 Kxh2 Rd8 29 f5! and a draw resulted. Alternatively 19 g3!? Kxh6 20 Bxd5
Kg7 21 Be4 Bxg3 is not clearly in White’s favour.
19 ... f5
On 19 ... Kxh6 20 Rxd5! grants White access to the crucial checking square at f6, so 20 ... Rxd5 21
Qf6+ Kh5 22 Re4 Bg3 23 fxg3 Qxg3 24 Bxd5 exd5 25 Re5+ wins for him. If Black tries to scurry back to
shelter with 20 ... Kg7 21 Qg4+ Kf8 22 Rxd8+ Qxd8 23 Kxh2 grants White a huge edge. Or 19 ... Bf4? 20
Bxf4 Nxf4 21 Qg4+ and 22 Qh4+ will pick up the rook at d8.
20 Bc1
Mickey decides that Black now just might be able to get away with pocketing his nice bishop, so he
pulls it back.
20 ... Bd6?!
In effect the losing move. Black had only one chance to put up resistance with 20 ... Be5!?, when it is
not out of the question that he could survive, e.g.
a) 21 c4 Nb4 22 Rxd8 Qxd8 23 Rxe5 Nd3 24 Qg3+ Kf7 25 Bg5 Nxf2+ 26 Kg1 Qd4.
b) 21 Bxd5 Rxd5 22 Rxd5 exd5 23 Qe3 Kf7 (his only hope is to give back the piece, 23 ... Bf6 failing
to 24 Qh6+ Kf7 and 25 Qh5+! does the king in) 24 Qxe5 Qxe5 25 Rxe5 Be6 26 Be3 when White’s
superior pawn structure gives him the edge, but it’s not much.
21 Bxd5 exd5 22 Rxd5 (Diagram 10)

Diagram 10
The black king is wide open

Still only the one pawn for the bishop, but a big development lead and the black king open to the four
white attackers.
22 ... Bd7?!
22 ... Bf8 offered more hope.
23 Qh3!
Nasty.
23 ... Bf8 24 Re3
Another big gun wheels up.
24 ... Kg7
If 24 ... f4 25 Qh5+ Kg7 26 Rg5+ (or 26 Red3) 26 ... Kf6 (26 ... Kh8 27 Qf7 mates) 27 Re4 Be8 28
Qh4 wins, or 24 ... h6 25 Rg3+ Kh7 26 Qh5.
25 Rg3+ Kh8
If 25 ... Kf7 26 Qxh7+ Ke6 27 Qxf5+ Ke7 28 Bg5+ Ke8 29 Qg6 mate, or 25 ... Kf6 26 Bg5+ Kf7 27
Qxh7+ Bg7 28 Bxd8.
26 Qh4
Black struggles to avoid the mate.
26 ... Be6
On 26 ... Bg7 27 Bh6! Rg8 28 Bxg7+ Rxg7 29 Rxg7 Kxg7 30 Qe7+ finishes him off.
27 Bf4 Be7 28 Bxc7 1-0
Since 28 ... Bxh4 29 Be5+ is curtains.

Game 61
Kasparov-Anand
Reggio Emilia 1992

1 e4 e6 2 d4 d5 3 Nd2 c5 4 exd5 Qxd5 5 dxc5

An experiment which he never repeated.
5 ... Bxc5 6 Ngf3 Nf6 7 Bd3 0-0 8 Qe2 Nbd7 9 Ne4 b6!
Perhaps Kasparov had not anticipated this. Anand gives up the pair of bishops, but achieves a swift
development.
10 Nxc5 Qxc5! 11 Be3 Qc7 12 Bd4 Bb7 13 0-0-0
White has the bishop pair, but Black is active and sees a way to create ingenious counterplay.
13 ... Nc5! (Diagram 11)

Diagram 11
Eliminating the bishop pair

14 Be5
On 14 Bxf6 Nxd3+ 15 Rxd3 Qf4+ 16 Kb1 Qxf6 Black is fine.
14 ... Nxd3+ 15 Rxd3?!
The other recapture, 15 Qxd3, was better when the situation is left unclear after 15 ... Qc6 16 Bxf6
gxf6 17 Nd4.
15 ... Qc4!
Beginning a remarkable adventure.
16 Nd4?!
On 16 Kb1?! Be4 17 Re3 Qxe2 18 Rxe2 Bxf3 19 gxf3 Nd5 leads to a very cosy ending for Black.
Once again on 16 Bxf6 Qf4+ avoids any dents in his pawns as the queen will take on f6.
16 ... Be4!
 16 ... Qxa2!? 17 Bxf6 gxf6 (17 ... Qa1+? 18 Kd2 Qxh1 19 Rg3 g6 20 Qh5)
17 Re3 Qxa2!
17 ... Qxe2?! 18 Rxe2 Bxg2? would lose to the pin on the g-file after 19 Rg1.
18 Bxf6
Black can survive the white initiative after 18 Rxe4 Nxe4 19 Qxe4 Qa1+ 20 Kd2 Qxh1 21 Qg4 f6 22
Nxe6 Rf7 23 Bxf6 Re8!, when the exposed white king is about to get hit, e.g. 24 Bxg7 Rxf2+ 25 Kd3
Qf1+ and wins. So Kasparov plays what at first seems like a strong alternative.
18 ... Bg6!
Cool as you like, Anand drops the bishop back and stays, temporarily, a piece down against the great
Kasparov. Instead, grabbing out here with 18..Qa1+?! 19 Kd2 Qxh1 would leave Black under real
pressure after 20 Rxe4 gxf6 21 Qg4+ Kh8 22 Qh4.
19 Ra3 Qd5 20 h4?!
There is no saving the bishop. On 20 Bh4 Qxd4, or 20 Be5 f6 and the knight or bishop must go. But that
last line might have been better than Kasparov’s actual choice.
20 ... gxf6! 21 h5 Qxd4 22 hxg6 hxg6 23 Rah3 f5
Only move, and an elegant way of covering the mate threat at h8.
24 Rh4 f4!
Once again the most accurate. Defending by evacuating the king was not so good as White can switch
the attack to the other side, viz. 24 ... Qf6 25 Qe3 Rfd8 26 Qh3 Kf8 27 Rh8+ Ke7 28 Qa3+ Kd7 29 Rd1+
Kc6 30 Qa4+ Kc7 (30 ... Kb7? 31 Rd7+ Rxd7 32 Qxd7+ Ka6 33 Rxa8 wins) 31 Rdxd8! Rxd8 32 Qxa7+
Kc6 33 Rh3 and the exposed king continues to be battered. I suspect it was such a line Kasparov had in
mind when advancing his h-pawn, but the Indian holds his nerve and shows great, and necessary cheek,
against the world champion.
25 Qf3 Rac8
The white king is also a legitimate target.
26 Rxf4
On 26 Qh3 the policy of audacity would still be correct, e.g. 26 ... Qxf2! 27 Rh8+ Kg7 28 Rh7+ (or 28
Qh6+) Kf6 and Black is winning.
26 ... Qc5 27 c3 Kg7 (Diagram 12)

Diagram 12
 Black consolidates his extra pawn

Tidying up. Anand has the extra pawn.
28 Rhh4 Qe5 29 g3 Qe1+ 30 Kc2 Rcd8 31 Rd4 Qe5 32 Rhf4 Qc7 33 Qe3 e5! 34 Rxd8 Rxd8 35
Re4 Rd5
Kasparov’s compensation is clearly insufficient. He starts to fish.
36 g4 b5 37 g5 Qd6 38 f3 a5
Anand aims to peel away the remaining defences of the white king.
39 Qe2 Qe6 40 Qh2 Qf5 41 Qg3 Qd7 42 Qe1 b4! 43 cxb4
On 43 Rxe5 Qa4+ 44 Kc1 bxc3! 45 bxc3 (or 45 Rxd5? Qa1+ and wins the queen) 45..Qf4+! 46 Re3
Rd3 and wins the rook.
43 ... Qa4+ 44 b3
There is no hope. If 44 Kc1 axb4 45 Rxe5 Rd6 wins, or 44 Kc3 Qc6+ 45 Rc4 axb4+ 46 Kxb4
Rb5+ 47 Kc3 Qxf3+ 48 Kc2 Qb3 wins.
44 ... Qa2+ 45 Kc3
Or 45 Kc1 axb4 46 Qxb4 Qa1+ 47 Kc2 Qd1+ wins.
45 ... a4!
Exposing the king to the full attack.
46 bxa4 Qa3+ 47 Kc2 Qxa4+ 48 Kc3 Qa3+ 49 Kc2 Rd3 0-1
A famous win for Anand en route to winning a big tournament. This game, in conjunction with his
overall performance there, lifted him into the front ranks of the world’s GMs.

Game 62
Adams-Levitt
Dublin zonal 1993

1 e4 e6 2 d4 d5 3 Nd2 c5 4 Ngf3 cxd4 5 Nxd4 Nf6 6 exd5 Qxd5

An unusual line.
7 Nb5!? Qd8 8 Nc4! (Diagram 13)

Diagram 13
An early cavalry charge

8 ... Nd5
On 8 ... Qxd1+ 9 Kxd1 Nd5 10 Ne3 White is a shade better, with his queenside pawn majority.
9 Ne3 Nc7 10 Bd2 Nxb5 11 Bxb5+ Bd7 12 Bxd7+
Qxd7 13 Qg4!?
Holding up his kingside development.
13 ... Nc6 14 0-0-0
Threatening a discovery, like 15 Bb4 and win of the g7-pawn.
14 ... Qd4 15 Qe2 Qa4 16 Kb1 Be7 17 Bc3 0-0 18 Rd7 Nb4
An escapade which rebounds on him, but his game was already uncomfortable.
19 Qg4! g6 20 Qd4! Qxa2+ 21 Kc1 f6
Quite forced.
22 Bxb4 Bxb4 (Diagram 14)

Diagram 14
White to play and win
23 c3!
Not 23 Qxb4? Rad8! 24 Rxd8 Qa1+ 25 Kd2 Rxd8+ 26 Ke2 Qxh1 27 Qe7 Rf8 28 Ng4 (as in the game)
when Black can defend by 28 ... Qc1.
23 ... Qa1+ 24 Kc2 Qxh1 25 Qxb4 Rad8 26 Rxd8 Rxd8 27 Qe7 Rf8
27 ... Rb8 28 Ng4 is the same.
28 Ng4 1-0
and there is no defence to 29 Nh6+ winning the rook and/or mating. Well calculated by Adams. ‘Very smooth’ – A.J. Miles.
Chapter Nine
The Tarrasch: Others
1 e4 e6 2 d4 d5 3 Nd2
Here we consider some alternatives for Black to the usual 3 ... Nf6 or 3 ... c5.

3 ... Be7 – Romanishin’s Move

The talking point in this chapter is 3 ... Be7!?. In the 1970’s or so it was used by Oleg Romanishin, but
mainly as a hobby or eccentricity. It has taken the efforts of the creative Russian GM Alexander
Morozevich to show that it is a complete and fully viable system. Nowadays it is accepted as a valuable
weapon for Black, right up there with 3 ... Nf6.

The Idea Behind 3 ... Be7

There is a very subtle point to the move. After the alternative 3 ... Nf6 White can establish a strong centre
with 4 e5 Nfd7 5 Bd3 or 5 f4 etc. In contrast after 3 ... Be7 4 e5 the pawn on e5 isn’t attacking a knight, so
Black can get on with attacking d4 straightaway with 4 ... c5 5 c3 Nc6. Here you will see that White
doesn’t have time to complete his development with 5 Bd3 intending 6 Ne2 as the d4-pawn is already
hanging! Nor does he have time for 5 f4 and 6 Ndf3. So White is a tempo short of building the centre that
proves so potent against 3 ... Nf6.

White Tries to Prove that 3 ... Be7 is a Mistake

It could be argued that Black has lost a tempo if White plays d4xc5 at a judicious moment and Black
recaptures ... Bxc5. Of course the manoeuvre ... Bf8-e7xc5 isn’t pretty, but then the move d4xc5 by White
isn’t the most critical idea in the Tarrasch. Winning a tempo makes it more respectable, but giving up the
centre in this way doesn’t make a very thematic impression.
Another way for White to try to exploit 3 ... Be7 is by lunging at g7 with 4 e5 c5 5 Qg4. This is Adams’
choice in Game 63 and is hugely successful when Short allows the arteries of his position to harden with
5 ... g6 (the heroic 5 ... g5!? has done well here for Morozevich who used it once to beat Adams himself)
6 dxc5 f5? when his dark squares were left terribly weak.
Morozevich and Adams have a ‘feud’ in the 3 ... Be7 system and one of Michael Adams’ best results was in Game 64. This was a
startlingly quick demolition of one of the best players in the world. In contrast, Game 65 is a lesson in how not to take on 3 ...
Be7. With the insipid 6 exd5 White lost control of the centre and suffered a terrible defeat.

3 ... a6 – Eingorn’s Move

This move is a favourite of the Ukrainian grandmaster Vereslav Eingorn. With 3 ... a6 Black prepares ... c7-c5 as in the 3 ... c5
Tarrasch, but first of all prevents the reply Bb5+. Whether he should spend a move to achieve this is a moot point. Have a look
at Game 66.

3 ... Nc6 – The Guimard Variation

The Guimard variation, 3 ... Nc6, also has its adherents even though Black is breaking one of the fundamental rules in the
French that you shouldn’t obstruct the c-pawn from applying pressure on the opponent’s centre (if you compare this to 1 d4 d5,
the move 2 Nc3 has few fans as it blocks the c2-c4 advance). Having played ... Nc6, it is imperative that Black find counterplay
somehow and this can be achieved with ... f7-f6. Still, attacking a pawn chain from the front isn’t considered as strong as hitting
its support, as with ... c7-c5, so generally Black’s set up gets the thumbs down. Of course this doesn’t stop it being a good
surprise weapon. You can check out Game 67.

Illustrative Games

Game 63
Adams-Short
Sarajevo 1999

1 e4 e6 2 d4 d5 3 Nd2 Be7

A curio, used by Romanishin amongst others. IM Joksic observed ‘After this move White is in
zugzwang!’ Not quite.
4 e5
A little unusual. The next year Plaskett beat Short with 4 Ngf3 Nf6 5 Bd3!? c5 6 c3 Nc6 7 0-0 dxe4 8
Nxe4, although this had little to do with the position which arose from the opening. The most testing line
is 4 Bd3 c5 5 dxc5, as played in Games 64 and 65.
4 ... c5 5 Qg4!? g6 6 dxc5!?
An interesting treatment by Adams, who straightaway tries to make something of the slightly weakened
dark squares in the black camp.
6 ... f5
I do not like this move, and am confident that later Short disliked it too. It further aggravates the dark
square weaknesses.
7 Qg3 Nd7 8 Bb5!? (Diagram 1)
 Diagram 1


All out to weaken Black’s dark squares.
8 ... Qa5 9 Bxd7+ Bxd7 10 c3 Qxc5 11 Ne2 Bd8 12 Nb3 Qe7 13 h4!
To weaken his pawns now.
13 ... Bc7 14 h5 gxh5 15 Bg5
The key to the weakening process: the exchange of the dark-squared bishops. The black game is
already critical.
15 ... Qg7 16 f4 h6 17 Bh4 Rh7 18 Nc5 b6 19 Nxd7 Qxg3+ 20 Nxg3 Kxd7 21 Nxh5 b5 22 Rh3
Setting off towards the black pawns.
Black is weak on the dark squares

Diagram 2
Black is horribly passive

22 ... Rf7 23 Rg3 Ne7 24 Bxe7 Rxe7 25 Rg6 Rh8 (Diagram 2)
 26 Rd1 Bb6 27 Rd3 a5 28 Nf6+ Kc7 29 Rxd5! Be3 30 Rxb5 Bxf4 31 Rxa5 Kb7 32 Rb5+ Ka7 33
Nh5 Bg5 34 Rg7 Rhe8 35 Rxe7+ Rxe7 36 Rb4
And the rest is just tidying up.
36 ... Rc7 37 Nf4 Rc5 38 Nd3 Rd5 39 Ke2 1-0

Game 64
Adams-Morozevich
Sarajevo 2000

1 e4 e6 2 d4 d5 3 Nd2 Be7 4 Bd3 c5 5 dxc5 Nf6 6 Qe2

The inferior 6 exd5 is seen in the next game.
6 ... Nc6 7 Ngf3 Nb4?!
I do not think this quite works. Better just to have taken the pawn back.
8 Nb3 Nxd3+ 9 cxd3 a5 10 Bg5
Simple moves, very much in the style of Adams, secure White a big edge.
10 ... a4 11 Nbd2 h6 12 Bxf6 Bxf6 13 e5 Be7 14 Rc1 (Diagram 3)

Diagram 3


14 ... Ra5 15 Qe3 0-0 16 0-0 Bd7
Black has no compensation at all for his lost pawn.
17 Rc2 Qa8 18 Rfc1 Rc8 19 Nf1 Qa7 20 d4 Qa6
21 Ng3 b6 22 c6 Be8 23 Nh5
Adams switches his attentions to Moro’s king. With Black still tied up dealing with the monster at c6, it
does not last long.
23 ... Qb5 24 g4!
Crude, but very effective.
The bishops have no scope

24 ... Ra7 25 g5 hxg5 26 Nxg5 Rac7 27 Kh1 (Diagram 4)

Diagram 4
Black cannot defend the kingside

Clearing the way for the rooks to the new g-file.
27 ... Rxc6 28 Rxc6 Rxc6 29 Rg1 Qxb2 30 Nxg7!
Crash!
30 ... Qc3 31 Nxe8 1-0

Game 65
Kudrin-Short
Monarch Assurance, Port Erin 1999

1 e4 e6 2 d4 d5 3 Nd2 Be7 4 Bd3 c5 5 dxc5 Nf6 6 exd5

Probably not as good as 6 Qe2.
6 ... Qxd5 7 Ngf3 Nbd7 8 Nb3 a5!
Better than just taking the c-pawn. He aims to boot the knight from defence of c5, or provoke White
into weakening himself in dealing with it.
9 c4 Qh5 10 Be2 e5! 11 0-0 a4 12 Nbd2
12 Nfd4 Qg6 13 f4!? might have been tried.
12 ... Nxc5 13 Re1 0-0 (Diagram 5)
Diagram 5


14 Nf1 Rd8 15 Qc2 e4
The opening has worked out marvellously for Short.
16 Ng3 Qg4 17 Ng5 Qh4
Now just threatening 18 ... h6, stranding the knight. Kudrin is probably already lost.
18 Be3
To meet 18 ... h6 with 19 Bxc5 and e4 falls.
18 ... Nd3
A splendid outpost for the horse.
A fine position for Black

19 Red1 Ng4 20 Bxg4 Bxg4 21 N5xe4
Giving up the exchange for a pawn to reduce the pressure, but it does not help much.
21 ... Bxd1 22 Rxd1 Nb4 23 Rxd8+ Rxd8 24 Qxa4 f5 25 Bc5?
A final blunder. 25 Nc3 was necessary.
25 ... Nd3 (Diagram 6) 0-1

Diagram 6
White loses material

White must lose material, since if 26 Bxe7 Qxe7 27 Nc3 Qe1+ 28 Nf1 Qxf2+ and mates.

Game 66
Dvoirys-Dolmatov
USSR Championship, Odessa 1989

1 e4 e6 2 d4 d5 3 Nd2 a6
One idea of this move is to play ... c5 and respond to dxc5 with ... Bxc5 and then slot the bishop back
to a7. It also prevents the Bb5 ideas in the 3 ... c5 4 exd5 exd5 lines (Games 57-59). 3 ... a6 still retains a
reasonable reputation.
4 e5 (Diagram 7)
 Diagram 7


It makes sense to take play into something akin to the Advance Variation, as ... a6 is not that functional.
Other treatments include leaving the pawn at e4, e.g. 4 Ngf3 c5 5 c3!?.
4 ... c5
Black might consider 4 ... Bd7!? followed by ..Bb5 to exchange light-squared bishops.
5 c3 Nc6 6 Ndf3 Nge7 7 Bd3 cxd4 8 cxd4 Nf5 9 Ne2 Be7 10 0-0 Qb6 11 a3 Bd7
A sort of Advance variation

Now that the check on b5 is no longer there the d4-pawn really is en prise.
12 Bc2 Rc8
12 ... a5!? was worth a look.
13 b4 a5
He could have castled. Black’s king is still in the middle. Ready for an attack? OK: here we go!
14 bxa5 Nxa5 15 Bxf5!? exf5 16 Nf4 Qc6 17 e6!? (Diagram 8)

 Diagram 8
A sacrifice to open lines

Opening him up for just the one pawn.
17 ... Bxe6 18 Ne5 Qa6 19 Nxe6 Qxe6
He would rather take back with the pawn, but that fails to 20 Qh5+.
20 Bd2 Qa6 21 Bb4! Nc6 22 Nxc6 Rxc6 23 Re1 Re6
A last desperate attempt to block the e-file, but his lag in development is too great.
24 Rc1 Kd7?
On 24 ... 0-0 25 Bxe7 wins. Black had to try 24 ... Rxe1+ 25 Qxe1 Qe6 though White is clearly better
in the endgame.
25 Rxe6 fxe6 26 Bxe7 Kxe7
The exchanges leave the king in the town square in his nightshirt.
27 Rc7+ Kd8 28 Qc1 1-0
He could not hope to have lasted out for long. Rare to see Dolmatov crushed like that. The moral of the
story is: Castle Next Time!

Game 67
Khalifman-Timman
Reykjavik 1991

1 e4 e6 2 d4 d5 3 Nd2 Nc6 4 Ngf3 Nf6 5 e5 Nd7 6 Be2 f6 (Diagram 9)

 Diagram 9
Black targets e5

7 exf6 Nxf6 8 0-0 Bd6 9 c4
This is typical play in the Guimard Variation. It is White who tries to break up the black centre with
c2-c4 rather than Black trying to do it to White with ... c7-c5.
9 ... b6
White would consider c4-c5 if Black did not prevent it. However, this weakening of the light squares
now allows White to mount an early initiative.
10 cxd5 exd5 11 Bb5 Bd7 12 Re1+ Ne7 13 Bxd7+ Qxd7 14 Ne5 Qf5 15 Qa4+ Kf8
This king move now makes it very difficult for Black to coordinate his position. In particular his king’s
rook remains a spectator to the end of the game.
16 Nf1 Bxe5
This is really asking for it. He had to try and play around the knight with a plan such as 16 ... Ng4 17 f4
h5.
17 dxe5 Ng4 18 Bf4 c5 19 Ng3 Qe6 20 Bg5 Nh6 21 f4 (Diagram 10)

 Diagram 10
The black rooks are spectators

Black has a temporary hold over f5 but this will soon be broken.
21 ... Nef5 22 Bxh6 gxh6 23 Nxf5 Qxf5 24 Qc6
Material is level but the draughty position of the black king renders his position hopeless.
24 ... Kg7 25 Qxd5 Qxf4 26 Re4 Qf5 27 Qb7+ Kg6 28 Rf1 1-0
Black never managed to move either rook.

NOTE: The problem with having to make an enforced early king move is often not the immediate danger to the king, but the
consequent inability to coordinate the rooks.
Chapter Ten
The Advance Variation
Introduction
In the Tarrasch and Classical White plays 4 e5, but it is with gain of time by attacking a black knight on
f6. Meanwhile in the Winawer he plays 4 e5 as he hopes to prove the black bishop is misplaced on b4
and, besides, the pawn is threatened with capture by ... dxe4. In contrast, White plays 3 e5 voluntarily in
the Advance and without gain of time. In some sense White isn’t asking enough of the position: Black
immediately knows how the pawn centre will look and can deploy his pieces accordingly.
For this reason the Advance, although a fully respectable opening line, has never enjoyed the
popularity of the other variations above. As usual in the mainline French, the d4-square is one of the focal
points of the strategic battle. A typical main line runs as follows:
1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Bd7 6 Be2 Nge7 7 Na3 cxd4 8 cxd4 Nf5 9 Nc2 Qb6 10 0-0
(Diagram 1)

Diagram 1
The d4-pawn is the focus

You will notice that White played the rather tortuous manoeuvre with his queen’s knight from a3 to c2 to defend the d4-pawn. And this
was just as well, as Black has been able to attack it with the queen and two knights. Because the centre is blocked, Black has been in no hurry
to castle – indeed his bishop is still sitting on f8. Wisely he preferred to use the time to bring his knight to the active square on f5.

Plans and Themes in the Advance

The Advance Variation is another line where the key to the play is very much based on an understanding of the plans for both sides. Here we
will discuss typical themes that arise from the diagram as these are characteristic of play in this line.
With his pieces tied down to the defence of d4 you might think White has a poor position. In reality in
the centre it is a ‘draw’ though the status quo favours White as it locks his space advantage in place. On
the queenside Black has more chances of doing some attacking. For example he could play ... Rc8 seizing
the open file (though there isn’t much to do there at the moment) or he could play ... Na5 with the idea of
... Bb5 to exchange off his ‘bad’ bishop for White’s good bishop as well as possibly playing ... Nc4 in
some scenarios. Another plan is ... Nb4 to exchange off knights to weaken White’s defences on the
queenside, as well as preparing ... Bb5 as above. Finally, he could consider a pawn assault with 10 ... a5.
Of course, he might combine various elements from these ideas.
Meanwhile White has to decide what he is going to do. He could try Bd3 and Bxf5, but despite
doubling Black’s pawns this seems ineffective; Black plays ... Be6 at some point with a very solid centre.
It is worth mentioning that Bd3 does at least set a good trap. Say in the diagram position play went 10 ...
Rc8 11 Bd3. Now it seems like Black can win a pawn with 11 ... Nfxd4? but this fails to 12 Nfxd4! (the
correct way to capture, so that Black can’t save himself next move by taking on f3 with check) 12 ... Nxd4
13 Be3 Bc5 (Diagram 2)

Diagram 2
White has a nasty trick

14 b4! and Black ends up in trouble however he plays, for example 14 ... Nxc2 15 Bxc5 Qc7. Now, not
16 Bxc2 b6 when Black survives, but instead 16 Bd6! Qc3 17 Qg4! which attacks the weakness on g7 and
causes huge problems for Black. In fact Black always has to think twice before capturing the d4-pawn as
traps like this abound in the Advance.
The classic trap in the French is 1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Qb6 6 Bd3 cxd4 7 cxd4
Nxd4?? 8 Nxd4 Qxd4?? 9 Bb5+ and Black loses the queen.
That was a pleasant interlude for White, but returning to the diagram we still need to find a good plan for him. Rather remarkably, the best
line involves g2-g4! Now when we first learn to play chess experience soon teaches us that advancing pawns in front of your king in this
fashion is bound to lead to disaster. And this is perfectly true for open positions. But in semi-closed positions – for example in certain lines of
the main line King’s Indian or the Ruy Lopez – different rules apply. With g2-g4 White gains space on the kingside and ejects the black knight
from f5, which takes some of the pressure off the d4-pawn.

The Milner-Barry Gambit

After 1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Bd7 White’s most natural development for his bishop is 6 Bd3 which leads to a pawn
sacrifice after 6 ... Qb6 7 0-0 cxd4 8 cxd4 Nxd4 9 Nxd4 Qxd4 10 Nc3 (Diagram 3).
 Diagram 3
A dangerous gambit

White has such a substantial lead in development that Black can be thankful that the opening is the
French – he wouldn’t last another ten moves after such extravagant play in the Open Sicilian! Now 10 ...
Qxe5 grabs a second pawn, but gives White a useful initiative after 11 Re1 Qb8 12 Nxd5. Instead the
preventive 10 ... a6! to rule out Nb5 looks good for Black. For example 11 Qe2 Ne7! 12 Kh1 (preparing
f2-f4 to defend e5 before Nc6) 12 ... Nc6 13 f4 Bc5! and Black has a solid position and the upper hand.
After the moves 1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Bd7 6 Be2 Nge7 7 Na3 cxd4 8 cxd4 Nf5 9
Nc2 Black played the immediate 9 ... Nb4!? in Game 68. This should have been enough to equalise but he
made an instructive error in the early middlegame. Games 69 and 70 feature 9 ... Qb6 when, according to
how aggressive he feels, White can opt for 10 0-0 or 10 h4!? In Game 71 the slower approach 3 e5 c5 4
c3 Nc6 5 Nf3 Bd7 6 a3 is made to look much worse than it is after the active 6 ... f6. In fact the idea of
preparing the b2-b4 advance makes more sense when Black has played 5 ... Qb6 rather than 5 ... Bd7, as
Black can’t so easily switch play to a kingside attack. You can find examples of 5 ... Qb6 6 a3 in Games
72 and 73. Finally in Game 74. Black foregoes ... Bd7 in favour of a quick ... Nge7. The point is to get the
queen to b6 and both knights attacking d4 a tempo sooner than usual before White has had time to organise
his defence of the pawn. It is interesting to see how Timman meets this threat.

Illustrative Games

Game 68
Anand-M.Gurevich
Manila Interzonal 1990

1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Bd7

This has been the most popular move recently. 5 ... Qb6 (Games 72 and 73) is the traditional main line,
and 5 ... Nge7 (Game 74) is also played occasionally.
6 Be2
The alternative 6 a3 is seen in Game 71.
6 ... Nge7
 6 ... f6 is another important line.
7 Na3
Going to c2 from where it supports d4.
7 ... cxd4 8 cxd4 Nf5 9 Nc2 Nb4
Major alternatives are 9 ... Qb6 (as in the next two games) or 9 ... Be7.
10 Nxb4 Bxb4+ 11 Bd2 Qa5
A better way of handling it here might have been 11 ... Qb6! when after 12 a3 Bxd2+ 13 Qxd2 Black
gets the chance to trade off, not the queens, but his ‘bad’ bishop with 13 ... Bb5 with equality.
12 a3! Bxd2+ 13 Qxd2 Qxd2+ 14 Kxd2 f6 15 Rac1 (Diagram 4)

Diagram 4
White has an endgame initiative

Vishy Anand has the open c-file, a bit more space and the slightly superior bishop. But still, it is not all
that much.
15 ... Ne7 16 b4 Kd8
On 16 ... Nc6 17 b5!? Na5 18 Kc3 a6 19 a4 leaves White slightly the more comfortable.
17 Bd3 Rc8 18 Rxc8+ Nxc8
18 ... Kxc8 was viable too.
19 g4 h6?!
19 ... Nb6! may have been better, when after 20 exf6 gxf6 21 g5 Nc4+ 22 Bxc4 dxc4 23 gxf6 Bc6
Black has active play, and 20 b5 Ke7 21 Rc1!? is not clear.
20 Nh4 Ne7 21 f4
White’s space advantage on the kingside increases.
21 ... a6
On 21 ... fxe5 White would have ‘broken the rules’ with the recapture 22 dxe5!, thereby granting
himself a nice pawn roller.
22 Rf1 Bb5 23 f5?
Anand proves a little too precipitate in making his advances. 23 Re1, just maintaining the superiority in
space, was superior.
23 ... h5?!
It might have been better to seek immediate active counterplay with 23 ... Bxd3! 24 Kxd3 h5 25 Ng6
Rh6 26 Nf4 hxg4 (or even 26 ... exf5!?) 27 Nxe6+ Ke8 28 Nxg7+ Kf7.
24 Ng6?! Nxg6?
And here too there was the opportunity to go into the line given in the previous note with 24 ... Bxd3!.
25 exf6!
A neat zwischenzug.
25 ... gxf6 26 fxg6 Ke7?!
26 ... Bxd3 27 Kxd3 Ke7 (27 ... hxg4 28 Rxf6 Ke7 29 g7 Rg8 30 Rg6 is winning for White as the pawn
ending following 30 ... Kf7 31 Rxg4 Rxg7 32 Rxg7+ Kxg7 33 Ke3 is lost because of the outside passed h-
pawn.) 28 g5! fxg5 29 Rf7 and the rook’s arrival at the seventh rank is terminal, or 28 ... f5 29 Ke3 like
the actual game.
27 g5! f5
27 ... Bxd3 28 gxf6+.
28 Bxb5 axb5 29 Rc1! (Diagram 5)

Diagram 5
Now the endgame is winning

29 ... Kd6 30 Ke3 Rg8 31 Kf4 b6 32 Rc3 Rxg6 33 Rh3
Now the g- and h-pawns are just too strong.
33 ... Rg8 34 Rxh5 Rc8 35 g6 Rc4 36 Rg5! Rxd4+ 37 Ke3
Not 37 Kf3? Rg4! and after 38 Rxg4 fxg4+ 39 Kxg4 Ke7 the black king gets back in time.
37 ... Re4+ 38 Kf2 1-0
The g-pawn is going home.

Game 69
Grischuk-Gulko
Esbjerg 2000

1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Bd7 6 Be2 Nge7 7 Na3 cxd4 8 cxd4 Nf5 9 Nc2 Qb6 10 0-0
The ambitious 10 h4 is seen in the next game.
10 ... Rc8 (Diagram 6)
Diagram 6


10 ... a5!? is a move here too.
11 Kh1
This was a new move. 11 a3, 11 h4 h5 12 Bg5 and even 11 g4!? were all known, e.g. 11 g4 Nfe7 12
Nfe1 h5 13 gxh5 Nf5 14 Be3 Na5 15 b3 Bb5 and chances were equal in Grosar-Poldauf, Austria 2000.
A standard Advance position

11 ... Na5
11 ... h5!? would prevent White’s next.
12 g4
Typical in this line.
12 ... Ne7 13 Nfe1!? Bb5 14 Nd3!
Protecting the so important light-squared bishop from exchange.
14 ... h5!? 15 gxh5 Nf5 16 Be3 g6?!
It looks that the simple 16 ... Nc4 was stronger than this optimistic move. Black’s anticipated play
down the h-file proves insufficient.
17 hxg6 fxg6 18 Rg1 Qc7 19 Nde1
Now switching the knight back for defence. After the natural 19 Rc1?! Qh7 20 Rg2 Black could make a
draw immediately with 20 ... Nh4 21 Rg3 (not 21 Rg4? Nf3!) 21 ... Nf5 and they repeat.
19 ... Bxe2 20 Qxe2 Qh7 21 Nf3 g5
With the sneaky intent of 21 ... Rxc2! 22 Qxc2 Ng3+ winning the queen.
22 Rg2! (Diagram 7)

Diagram 7
Black’s counterplay is insufficient

22 ... Nh4
Now on 22 ... Rxc2? 23 Qxc2 Ng3+ 24 fxg3! protects the queen with the rook.
23 Nxh4?!
After the simple and strong 23 Qb5+ Nc6 24 Nxh4 (or even 24 Nxg5!?) 24 ... gxh4 25 Ne1 White
could keep a solid advantage. The text gives Black a glimmer of hope in the unnecessary complications.
23 ... Rxc2! 24 Qb5+ Nc6 25 Nf3 Bh6! 26 Bxg5 a6?!
The final mistake in time trouble, but even after the best move, 26 ... Rg8, White’s advantage is
obvious, e.g. 27 Rag1 Qe4 28 Bxh6 Rxg2 (28 ... Qxf3 29 Qb3 Qe4 30 Qxc2! Qxc2 31 Rxg8+ Kd7 32
R1g7+ Ne7 33 Rf7! will win) 29 Kxg2 Qg4+ 30 Kf1 Qh3+ (on 30 ... Qxf3 31 Rg8+ Kf7 32 Qxb7+ wins)
31 Rg2 Qxh6 32 Ne1, etc.
27 Qb3 Bxg5 28 Nxg5 Qf5
If 28 ... Nxd4 29 Qa4+ wins.
29 Nxe6! Ke7 30 Ng7
30 Qxd5! did the trick too.
30 ... Qe4 31 Qxb7+ Kf8 32 Ne6+ Ke8 33 Qc8+ Ke7 34 Qxh8 Rxf2 35 Qg7+ 1-0
On 35 ... Kxe6 36 Qg6+ wins.

Game 70
Movsesian-M.Gurevich
Sarajevo 2000

1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Bd7 6 Be2 Nge7 7 Na3 cxd4 8 cxd4 Nf5 9 Nc2 Qb6 10 h4
Ambitious.
10 ... f6!?
An interesting novelty in a well-known position. 10 ... Bb4+ 11 Kf1 h5 12 a3 Be7 13 b4 Rc8 14 g3 f6
15 Bd3 fxe5 16 Bxf5 exf5 17 dxe5 led to a White win in Topalov-Nikolic, Monte Carlo 1997.
I am not sure that after 10 ... Nb4 11 Nxb4 Qxb4+ 12 Kf1 Bb5 13 g4 Bxe2+ 14 Kxe2 Qc4+ Black has
too many problems.
11 g4 Nfxd4!? (Diagram 8)

Diagram 8
A typical sacrifice for Black

A typical and strong knight sacrifice. Black gets a few pawns and the initiative for his piece.
12 Ncxd4
On 12 Nfxd4 Black gets his play in the event of 12 ... Nxd4 13 Qxd4 Bc5 14 Qf4 fxe5 15 Qxe5 Bxf2+
16 Kd1 0-0, or 12 ... Nxe5!?, or 12 ... fxe5!?, or 12 ... Bb4+!? 13 Kf1 Nxe5. Mix it up the way you like it!
13 g5
If 13 Nxe5 fxe5 14 Nb3 Black has loads of play after 14 ... Bd6.
13 ... Bc5! 14 0-0
On 14 gxf6 gxf6 15 Nxe5 fxe5 16 Bh5+ Ke7 17 Bg5+ Kd6 the black king is snug and the compensation
patent.
14 ... Nxf3+
14 ... 0-0 is not so accurate as after 15 gxf6 gxf6 16 Bh6 Rf7 17 Nxe5 fxe5 18 Bh5 White develops
excellent activity.
15 Nxf3 Qb4!?
Eyeing g4.
16 Ne1!?
Movsesian aims to shield his king with a knight on g2, and so offers a third pawn for the piece, but 16
Kh1 Qg4 (or even 16 ... Bc6!?) 17 Rg1 Qh3+ 18 Nh2 Qxh4, or 16 Be3 Qg4+ 17 Kh1 Bd6!? look too
dangerous for White.
16 ... Qxh4 17 Ng2 Qb4 18 Bd3 0-0-0 19 a3 Qb6 20 b4
Movsesian tries to organise a counter-attack against the black monarch.
20 ... Bd4 21 Be3 Kb8 22 Qf3
On 22 Qg4 Bxe3 23 fxe3 e5 24 Qh4 Rdf8 keeps the kettle boiling.
22 ... Bxe3 23 fxe3 Rhf8
23 ... Rdf8!? planning 24 ... h6, was also to be considered.
24 Qg3+?!
If 24 Bxh7 Bb5!? 25 Rf2 d4 with a steaming initiative.
24 ... e5 25 Bxh7 25 ... d4!
Gurevich plays the whole game with great energy.
26 Be4
On 26 exd4 Rh8!? 27 g6 (or 27 Bf5? Bxf5 28 Rxf5 Qxd4+ wins) 27 ... Bb5 28 Rf2 Rxd4 with 29 ...
Rhd8 to follow is splendid.
26 ... Rh8 27 gxf6
Opening all lines, but White is in a pickle. 27 exd4 Rh3 28 Qf2 Rdh8 29 Ne1 Qe6 doesn’t look too
good either.
27 ... gxf6 28 Qf3
On 28 exd4 Rh3 29 Qf2 Rdh8, or 28 Qg7 dxe3 29 Rae1 (29 Qxf6 e2+ 30 Rf2 Qd4 wins) 29 ... e2+ 30
Rf2 Rh1+! 31 Kxh1 Qxf2 and Black breaks in.
28 ... Rdf8 29 exd4 Rh3 30 Qd1 f5 31 Qd2 Rfh8 (Diagram 9)

Diagram 9
The white king is too exposed

The finish of the game is very close.
32 Nf4!?
Desperation. None of 32 Bc2 Bc6 33 Bxf5 Rh2, or 32 Bxf5 Bxf5 33 Rxf5 Rh1+, or 32 Bd5 Qd6 33
Bc4! (33 dxe5 Qb6+ or 33 Nf4! Rg3+! 34 Bg2 Rg4) 33 ... Bc6 34 d5 Bb5 was any better.
32 ... fxe4! 33 Nxh3 Bxh3 34 Qh2
Or 34 Rf2 e3 35 Qxe3 Qg6+ wins.
34 ... Qxd4+ 35 Kh1 e3 0-1
A splendid victory from Mikhail Gurevich.

Game 71
Adams-Epishin
Ter Apel 1992

1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Bd7 6 a3 f6

Epishin seeks immediate action.
7 Bd3
7 b4 fxe5 8 bxc5 e4 9 Ng5 Nf6 is a weirder line.
7 ... Qc7 8 0-0 0-0-0
8 ... fxe5 9 Nxe5 Nxe5 meets with 10 Bf4 Bd6 11 dxe5 and Black cannot capture again on e5 because
of the check at h5.
9 Bf4?! c4!
And not 9 ... g5? 10 Bg3 g4 because of 11 exf6. Instead Epishin shuts things down on one wing and
goes rapidly to work on the other.
10 Bc2 h6! (Diagram 10)

Diagram 10
Black prepares kingside action

11 h4
Very weakening, but 11 Bg3 f5 12 h4 Be8! 13 Nbd2 Bh5 14 Bf4 Be7 15 Qb1 Rf8 leaves him facing a
quick ... g5 anyway, possibly first prepared by ... Qd8. One is left with the impression that Mickey did not
really know what he was doing in this variation.
11 ... Be8!
On 11 ... f5 Adams may have had the (necessary) cheek to play 12 h5. But now the bishop nips outside
the pawn chain to join the attack.
12 b3 cxb3 13 Bxb3 Bh5 14 Nbd2 fxe5?!
Simply 14 ... g5! and then 15 ... f5 was tremendously strong. At the position’s crossroads, Epishin
elects to attack with the pieces and not with the pawns, but you ought always to provide yourself with a
good reason should you not choose the obvious move in any position.
15 dxe5 Bc5 16 Qb1
16 Qe1 Nge7 17 Be3 looks stronger.
16 ... Nge7 17 c4 Rhf8 18 Bh2 Bxf3 19 Nxf3 Rxf3! (Diagram 11)

Diagram 11
The thematic sacrifice again

An obvious but very strong sacrifice.
20 gxf3 Nd4 21 Kg2
On 21 Bd1 dxc4 the black initiative is too powerful to be withstood.
21 ... Nef5 22 Bg3 Qf7! 23 f4
Neither 23 Bd1 Qg6 nor 23 cxd5 Nxh4+ 24 Bxh4 Qxf3+ 25 Kh2 Qh5 26 Qe4 Nxb3 is adequate either.
23 ... g5 24 cxd5 gxf4 25 dxe6
25 Bxf4 Nxh4+ wins.
25 ... Qh5 26 e7 Nxh4+ 0-1
Now 27 Bxh4 Qg4+ 28 Bg3 f3+ 29 Kh2 Qh5+ 30 Kg1 Ne2 is mate.
Adams annihilated.
Game 72
L.Hansen-Antonsen
Farum 1993

1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Qb6 6 a3

Not even interested in the Milner-Barry Gambit: 6 Bd3 cxd4 7 cxd4 Bd7 8 0-0!?.
6 ... Nh6
If Black wants to prevent b2-b4 then 6 ... c4 should be played – see Game 73.
7 b4 cxd4 8 cxd4 Nf5 9 Be3
9 Bb2 is possible.
9 ... f6 (Diagram 12)

Diagram 12
Black attacks the centre

10 Bd3 Nxe3 11 fxe3
This exchange is occasionally seen in the French. It looks as if it ought to be good for Black, who has
gained the bishop pair and doubled the white pawns. However, the f-file is now dangerously open and the
pawn on e3 serves the useful function of bringing additional defence to the white e4-pawn. Who is most
favoured by this transaction often revolves around whether Black can continue to undermine the position
by getting rid of the white pawn at e5. This is the theme of the subsequent play here.

11 ... fxe5 12 b5
White is fighting for e5.
12 ... e4 13 bxc6 Bd6 14 0-0 0-0 15 Ng5 h6 16 Rxf8+ Bxf8 17 Nxe4 dxe4 18 Bxe4 bxc6 19 Nd2
Ba6 20 Qg4 (Diagram 13)

 Diagram 13
Too many black weaknesses

The complications are at an end and White has emerged on top thanks to the numerous weaknesses in
the black position.
20 ... Qb2 21 Rd1 Qxa3 22 Qxe6+ Kh8 23 Qg6 Kg8 24 Nf3 Bd6 25 Qe6+ Kh8 26 Nh4 Qxe3+ 27
Kh1 Be2 28 Ng6+ Kh7 29 Qf5 Kg8 30 Qe6+ Kh7 31 Ne7+ Kh8 32 Qg6 1-0
After 32 ... Qxe4 33 Qxe4 Bxd1 Black can wriggle but White eventually wins even more material, e.g.
34 Ng6+ Kh7 35 Ne5+ Kg8 36 Qxc6 Rd8 37 Qc4+ Kh738 Qd3+ Kg8 39 Qxd1 Bxe5 40 Qb3+ and wins.

Game 73
Grischuk-Korchnoi
Biel 2001

1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Qb6 5 Nf3 Nc6 6 a3 c4 7 Nbd2 Na5 (Diagram 14)

 Diagram 14


8 g3
Another scheme of development is 8 Be2 Bd7 9 0-0.
8 ... Be7 9 h4 Bd7 10 Bh3 f5 11 exf6 gxf6 12 0-0 h5 13 Re1 Nh6 14 Nh2 0-0-0 15 Qxh5 Rdg8 16
Qe2 f5 (Diagram 15)

Diagram 15


Black has a clamp on the white queenside and useful open files on the opposite wing. All in all, good
compensation for a pawn.
Fighting for the b3-square
Black has full compensation

17 Ndf3 Nb3 18 Bxh6 Rxh6
18 ... Nxa1 would be a bad mistake after 19 Bf4 with Qe5 coming.
19 Rad1 Bxh4 20 Nxh4 Rxh4
White has returned the pawn in the hope of making something out of his dark square control but he
never really gets anywhere.
21 Bg2 Rh7 22 Nf3 Na5 23 Qd2 Qd6 24 Ne5 Ba4 25 Rb1 Nc6 26 Qf4 Kc7 27 Re2 Nxe5 28 Qxe5 Qxe5 29 Rxe5 Kd6 30
Rbe1 Rh6 31 R5e2 f4 ½-½

Game 74
Timman-Andersson
Sigeman & Co, Malmoe 2000

1 e4 e6 2 d4 d5 3 e5 c5 4 c3 Nc6 5 Nf3 Nge7 6 Bd3 cxd4 7 cxd4 Nf5 8 Bc2

Alternatives here are:
a) 8 Bxf5 exf5 9 Nc3 Be6 10 Ne2 h6 doesn’t promise an advantage to White – 11 h4 g6 12 Bd2 Qb6
13 Rb1 a5 14 Kf1 Be7 15 Kg1 Kd7 16 a3 Qa6 17 Be3 a4 was equal in Adams-Vaganian, Lucerne 1997.
b) 8 Be3 Be7 9 Nc3 Nxe3 10 fxe3 0-0 11 0-0 Bd7 12 a3 f5 13 exf6 Bxf6 14 Qc2 h6 15 Rad1 Rc8 and
Black was very comfy in Jonkman-Glek, Vlissingen 1998.
 8 ... Qb6 9 Bxf5 exf5 10 Nc3 Be6 11 0-0
After 11 Ne2 h6 12 h4 0-0-0 13 Kf1 Kb8 14 Kg1 Be7 15 Rb1 Black sacrificed a pawn to activate his
bishop and gain interesting counterplay with 15 ... f4! 16 Nxf4 Bf5 17 Ra1 Be4 in Benjamin-Shaked, US
Open 1998.
11 ... h6 (Diagram 16)

Diagram 16
Black is solid in the centre

12 Na4
A new move, exploiting Black’s "extra" ... Qb6. Benjamin-Edelman, New York 1992, saw instead 12
h4 Be7 13 Ne2 0-0-0 14 Rb1 and Black did not get enough when he gambitted 14 ... g5 15 hxg5 hxg5 16
Bxg5 Bxg5 17 Nxg5 Rh5 18 f4 Rdh8 19 Nf3 Kb8 20 Kf2 and White consolidated his advantage.
12 ... Qb5 13 Ne1 Be7 14 Be3 Rc8 15 Nd3 b6 16 Nc3 Qc4
Looks a little bit risky but after 16 ... Qa6 17 Nf4 the pawn on d5 is weak.
17 Rc1 g5!? 18 f4 g4
Black loses a piece after 18 ... Nxd4? 19 b3.
19 a4 Nb4?!
But now I do not see how White can get an advantage after 19 ... Nxd4!?, for example 20 Nb5 Nxb5 21
Rxc4 dxc4 22 axb5 Rd8 23 Qa4 Rxd3 24 Kf2 Rd7 with reasonable compensation for the queen.
20 Nf2 Qa6?!
Andersson might have tried 20 ... Na2!? 21 Nxa2 Qxa2 22 Rxc8+ Bxc8.
21 Nb5 Rxc1 22 Bxc1 Kd7? (Diagram 17)

Diagram 17
Now it’s not as solid as it seems

The last mistake. 22 ... Qb7! was much stronger.
23 Nxg4!
The black queen is far away from the kingside, so it’s logical for White to attack there.
23 ... fxg4 24 f5
With a winning initiative.
24 ... Rg8
Or 24 ... h5 25 fxe6+ fxe6 26 Rf7 wins.
25 fxe6+ fxe6 26 Bd2!
26 Bxh6!? was also fun and worth a punt.
26 ... Nc6
On 26 ... h5 27 Bxb4 Bxb4 28 Qc2 is winning.
27 Bxh6 Qc8 28 Rf7 Rh8 29 Nd6 Qg8 30 Bg7 1-0
Now 30 ... Rh7 31 Qxg4 Nxd4 loses to 32 Rxe7+! Kxe7 33 Bf6+ Kf8 34 Qxd4.
Chapter Eleven
The Rubinstein Variation and Others
The Rubinstein Variation
The Rubinstein Variation opens 1 e4 e6 2 d4 d5 3 Nc3 (or 3 Nd2) 3 ... dxe4 4 Nxe4 Nd7.
This variation has much in common with the Burn variation of the Classical which runs 3 Nc3 Nf6 4 Bg5
dxe4 5 Nxe4. However, the Rubinstein move order is more flexible for White as he hasn’t committed his
bishop to the g5-square. Therefore he has alternative schemes, including a quick g2-g3 to put pressure on
the b7-square. On the other hand, by adopting the Rubinstein move order Black has avoided the Classical
4 e5 which, judging from recent games, is something to be pleased about!
In fact, one of the best things about the Rubinstein is its universality – you can play it against both 3 Nd2 and 3 Nc3. Also there is relatively
little theory to learn.

TIP: The Rubinstein is a good choice if you are in a solid mood and are facing a highly aggressive opponent.

Plans
With 4 ... Nd7 Black prepares to challenge White’s excellently placed knight on e4 with 5 ... Ngf6. Once the knight has been ousted or
exchanged, Black will be ready to attack the white centre with ... c7-c5. If he succeeds in dissolving the d4-pawn without incurring any
weakness elsewhere or falling a long way behind in development, Black will have fully equalised. White for his part will try to prove that rather
than freeing his game with ... c7-c5 Black is clearing the centre to his detriment as his opponent’s pieces are better able to utilise the resulting
open lines. White will try to saddle Black will a positional weakness somewhere, for example a doubled pawn on f6 or difficulties in developing
his queen’s bishop.
In an epic battle in Game 75, Kasparov tries a different method: he maintains a pawn on d4 with c2-c3 and looks for an immediate attack.
This proved good enough to bamboozle the surprised Ponomariov, but you can bet that theory soon began to draw its fangs. Instead of 8 ...
Nd7?! the solid 8 ... a6! was good enough. Game 76 is worth looking at if only for the marvellous trap that White sets at move eight.

The Fort Knox

Chess doesn’t come much more solid than this! Without any more ado Black plays 1 e4 e6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Bd7 and then 5 ...
Bc6. The reasoning behind this is hard to argue with: Black often struggles for the whole game to find a decent diagonal for his queen’s bishop,
so why not put it on c6 at once? In combination with a knight on f6 the bishop either drives away or eliminates White’s proud knight on e4. In
such a quiet position Black can afford a slow but strategically well motivated plan. The drawback to Black’s bishop manoeuvre is that he is
more or less compelled to give up the bishop for a knight and so grant White the bishop pair. However, Black remains very solid in his Caro-
Kann like carapace of pawns after the bishop exchange and then ... c7-c6, etc. Note that as White has the two bishops Black shouldn’t be so
keen for the centre break with ... c7-c5 as he is in the Rubinstein – such a move becomes extremely double edged and has to be carefully
considered.
In both Rubinstein games given here Black becomes too complacent inside his fortress with the result that he is both ground down (Game
77) and blown away (Game 78). The moral is clearly that no position, no matter how safe from attack by the opponent, can survive our own
carelessness!

TIP: The Fort Knox is a good choice if you are in a very solid mood and are facing an extremely aggressive opponent.

The Exchange Variation

1 e4 e6 2 d4 d5 3 exd5 exd5
When Mikhail Gurevich, the great French Defence expert, played a match against a computer some years ago the computer was instructed by
its programmer to play in as ‘Un-French’ a manner as possible against 1 ... e6. The reasoning was that Gurevich was more likely to go wrong
in unfamiliar territory. The Computer duly came up with the French Exchange. Indeed the position after 3 ... exd5 is more akin to that in a
Queen’s Pawn Opening. Black no longer has to worry about how he is going to develop his queen’s bishop, and after 4 Nf3 he can play 4 ...
Bg4 without even having to accept an isolated pawn in return for the privilege, as occurs in the Tarrasch 3 Nd2 c5 4 exd5 exd5 variation. Nor
does the black king’s knight find itself deprived of the f6-square by a pawn on e5. White’s sole advantage is the extra tempo born of moving
first. Yet for all this many French players don’t like playing against the Exchange. I guess it is partly the ‘Computer’ factor above of being in
unusual surroundings, but mainly it is because the positions reached are somewhat sterile. Indeed, Black often answers 4 Bd3 with 4 ... c5
willingly accepting an IQP to make it a more dynamic game. Note however that this is in response to the inferior position of the bishop at d3:
much less appealing is 4 Nf3 c5?! when the bishop can be developed aggressively without loss of time with 5 Bb5+.
The computer’s choice against Gurevich was also a shrewd psychological choice (not a characteristic normally associated with computer
play), as the Russian GM had lost a heart breaking game as White against Nigel Short in the Exchange – have a look at Game 79. You will see
that the black bishop really enjoyed its ‘Un-French’ freedom on g4!

The King’s Indian Attack

In this line White avoids the main lines after 2 d4 with either 2 d3 or 2 Qe2. A structure is often reached similar to that in the Closed Sicilian.
Besides adding a lot of ideas to the French as Black, Alexander Morozevich has almost single-
handedly transformed the King’s Indian Attack into a dangerous attacking weapon. The Russian GM
Kiriakov is one of the most difficult players to beat but he is wiped out in Game 80 by the supposedly
innocuous 2 Qe2. Morozevich has shown that it pays for White to delay the development of his queen’s
knight rather than play the routine 3 Nd2. This worth of this idea is seen if Black adopts the following
popular set up:
1 e4 e6 2 d3 d5 3 Qe2 Nf6 4 Nf3 c5 5 g3 Nc6 6 Bg2 Be7 7 0-0 b6 8 c3 Qc7 (Diagram 1)

Diagram 1
A typical King’s Indian Attack

and here White has the option of 9 Na3!, planning either to embarrass the black queen with Nb5 or retreat back to c2 and
support a b2-b4 advance to attack on the queenside. The knight would be far less effective on d2.

A Solid System Versus the King’s Indian Attack

However, Black has a solid system against which Morozevich comes a cropper in Game 81. The pattern of black pieces is well
worth remembering. On the queenside he plays ... c7-c5 and ... Nc6 and combines this with ... g7-g6 and a kingside fianchetto.
And to avoid having his knight attacked by e4-e5 and shutting in the bishop Black develops his king’s knight on e7. The upshot
of all this is that Black exerts enormous pressure on the d4-square. This square is important because if White ever plans to gain
space in the centre he needs to play d3-d4, when the pawn comes into the killing zone of the black pieces. The pressure along
the a1-h8 diagonal also makes a White pawn advance on the queenside problematical, and encourages Black to launch his own
pawn storm against the target on c3.

Illustrative Games

Game 75
Kasparov-Ponomariov
Linares 2002

1 e4 e6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Nd7 5 Nf3 Ngf6 6 Nxf6+

6 Bd3 would transpose to the next game after 6 ... Be7 7 Qe2, though Black might prefer 6 ... Nxe4 7
Bxe4 Nf6 8 Bg5 Qd6!? 9 Bxf6 gxf6 10 0-0 f5 as in Anand-Speelman, Linares 1991.
White can also play 6 Bg5 entering the Burn variation with 5 ... Nbd7 (Games 37 and 38).
6 ... Nxf6 7 c3
Certainly less popular than 7 Bd3.
7 ... c5 8 Ne5
Unusual. Garry hits upon a plan involving a sharp gambit.
8 ... Nd7 9 Bb5 Bd6 10 Qg4!
Very disruptive.
10 ... Kf8
Not what you want to do, but 10 ... 0-0 loses the exchange to 11 Bxd7 Bxd7 12 Bh6.
11 0-0!?
Yet another Kasparov sacrifice. This gambit yields compensation in the form of Black’s uncastled king
and development lag. But it is still speculative.
11 ... Nxe5 12 dxe5 Bxe5 13 Bg5 Bf6
13 ... f6 14 Rad1 Qb6 15 Be3 yields real compensation too.
14 Rad1 Qc7 15 Qh4 Bxg5 16 Qxg5
Threatening mate.
16 ... f6 17 Qh5 g6 18 Qh6+ Kf7
Not 18 ... Qg7 because of 19 Rd8+ Kf7 20 Be8+! Rxe8 21 Qxg7+ Kxg7 22 Rxe8 winning. How to
continue? Has White run out of steam?
19 Rd3!? a6 20 Rh3!?
No he has not. 20 ... axb5 is impossible because of 21 Qxh7+ Rxh7 22 Rxh7+ and 23 Rxc7.
20 ... Qe7 21 Bd3
Now attacking g6.
21 ... f5 22 g4 (Diagram 2)
 Diagram 2
Hack attack!

He keeps hacking away.
22 ... Qf6 23 Rd1 b5 24 Be2
Preventing Black from developing his bishop at b7 because of 25 Rd7+ winning it.
24 ... e5 25 Rhd3 Ra7 26 Rd6 Qg7 27 Qe3 Rc7 28 a4
Prising his way into Ponomariov’s defences.
28 ... e4
28 ... bxa4 29 Bc4+ is dreadfully strong.
29 axb5 axb5 30 Bxb5 Qe5 31 Qg5
There are just too may white men swarming all over the black camp. Something will surely give.
31 ... Qe7 32 Qh6 Be6 33 Qf4 Bc8
White threatened 34 Rxe6! Kxe6 35 Bc4+ Kf6 36 Rd6+ Kg7 37 Re6 and 38 Qe5+ follows
with a decisive penetration.
34 Qh6 Be6 35 gxf5 gxf5 36 Be2 (Diagram 3)

 Diagram 3
Black’s position finally crumbles

The final switch of direction.
36 ... Qf6 37 Bh5+ Ke7 38 Rxe6+! 1-0
Black resigned, as 38 ... Kxe6 loses to 39 Rd6+ and 38 ... Qxe6 to 39 Qg7+. An important win for
Kasparov, teaching the teenage new World champion who’s boss.

Game 76
Gross-Blatny
Czech Championship 1998

1 e4 e6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Be7 5 Nf3 Nf6 6 Bd3 Nbd7 7 Qe2 Nxe4 8 Bxe4 (Diagram 4)

Diagram 4
Careful. There may be a trick!

8 ... a6!?
There are tactics even in quiet positions. Many players have fallen for the trick 8 ... Nf6? 9 Bxb7! and
White wins a pawn after 9 ... Bxb7 10 Qb5+ and 11 Qxb7. Black’s move 8 ... a6 prevents this.
8 ... c5!? is more common when 9 Be3 0-0 10 0-0-0 Nf6 11 Bd3 Qc7 12 dxc5 Bxc5 13 Bg5 Be7 14
Kb1 left White more active in Keres-Pavey, USSR-USA 1954.
9 0-0 Nf6 10 Bd3
Of course 10 Bxb7 doesn’t work now.
10 ... b6 11 Rd1
It’s always a question where to put the rooks and in this case it seemed more suitable to place this rook
on e1 and the second one on d1.
11 ... Bb7 12 c4 c5 13 dxc5 Bxc5 14 Bg5 Qe7
14 ... Qc7!? was another development.
15 Nd2! Rd8 16 Be4 h6?!
It was probably more accurate here to have played 16 ... 0-0 17 Bxf6 gxf6 18 Bxb7 Qxb7 19 Ne4 Be7
(19 ... Bd4? 20 Rxd4! Rxd4 21 Nxf6+ Kg7 22 Qe5) 20 Qf3 Kh8 when after 21 b4 White has at best only a
slight pull.
17 Bxf6 gxf6 18 Qf3?!
A serious inaccuracy letting Black equalise. After 18 Bxb7! Qxb7 19 Nf3 Be7 20 Qe3 Kf8 21 Rxd8+
Bxd8 22 Rd1 Be7 23 Qf4 White’s advantage is quite significant.
18 ... Bxe4 19 Nxe4 Bd4
It’s now more or less equal.
20 Rd2 f5 21 Ng3 0-0 (Diagram 5)

Diagram 5
Black has equalised

22 Rad1 Bc5 23 Qf4 Rxd2 24 Rxd2 Kh7 25 h3 Rd8 26 Nf1 a5 27 b3 Rxd2 28 Qxd2 Qf6 29 a4?!
Unnecessarily creating a weakness for himself at b3.
29 ... Qe5 30 g3 Qe4 31 Kh2?
Simply blundering a pawn! After the correct 31 Ne3 Kg6 32 Qe2 e5 33 g4! White still has good
drawing chances.
31 ... Qb1
Winning.
32 Qe2 Qxb3 33 g4 Qxa4
33 ... fxg4!? 34 hxg4 Qxa4 35 Qe4+ Kg8 would also have done the trick.
34 gxf5 Qd7 35 fxe6 Qxe6
The rest is easy: White simply can’t stop the a-pawn.
36 Qd3+ Qg6 37 Qf3 a4 38 Ng3 a3 39 Ne4 Bd4!?
39 ... a2 40 Nf6+ Kg7 41 Ne8+ Kf8 42 Qa8 a1Q was another way to finish him off.
40 h4 a2 0-1

Game 77
Adams-Rozentalis
Belgrade 1999

1 d4 e6 2 e4 d5 3 Nd2 dxe4 4 Nxe4 Bd7

The Fort Knox variation, so called as Black plans to erect an unambitious but super strong structure.
5 Nf3 Bc6 6 Bd3
The adventurous 6 Neg5!? is seen in the next
game.
6 ... Nd7 7 c3 Ngf6 8 Nxf6+ Nxf6
8 ... gxf6!? 9 Bf4 Bd6 10 Bxd6 cxd6 11 Qe2 Qa5 12 0-0 Qh5 13 Be4 Bxe4 14 Qxe4 d5 led to equality
in Schaefer-Dobosz, Cuxhaven 1993.
9 0-0
9 Qe2 Bxf3 – the swap of the Fort Knox variation – 10 Qxf3 Qd5 11 Qxd5 Nxd5 12 0-0 0-0-0 13 Rd1
Be7 14 Bd2 was then agreed drawn in Ivanov-McDonald, Maidstone 1994.
9 Ne5!? was well worth a go.
9 ... Bxf3
9 ... Bd6!? 10 Qe2 Bxf3 11 Qxf3 c6 12 Bg5 Be7 13 Rfe1 Nd5 14 Bxe7 Nxe7 was equal in Van de
Velden-Beals, Soest 1996.
10 Qxf3 Qd5 11 Qe2 Be7
11 ... Qh5!? practically forces White to exchange the queens, and on 12 Qxh5 (after 12 Qc2?! Bd6 13
h3 0-0-0 Black is the one who fights for the opening advantage) 12 ... Nxh5 13 g3 Nf6 14 a4 Bd6 15 a5
any white advantage is symbolic.
12 Re1 0-0 13 Bf4 c6 14 a4 (Diagram 6)

Diagram 6
Black is solid but passive

This type of position is quite well-known in modern chess: it arises from different variations of the
Caro-Kann defence and out of some variations of the French. The plans of the both sides are clear: White,
who has some advantage in space and the pair of bishops, will try to attack the black king, while Black,
who has a rock solid but passive position, will try to find his chances in a counter-attack or with some
break in the centre.
14 ... Rfd8 15 Bc7!? Re8
15 ... Rd7?! is met by 16 Be5 and the black queen is in danger.
16 Be5 a5!? 17 Qc2 g6 18 Rad1 Qd8 19 Qb3 Qd7
20 h4!? Ng4?!
This logical move proves to be a serious inaccuracy: the knight plays a very important part in the
defence of the kingside and without it White’s attack pretty soon becomes decisive. It was better to defend
with 20 ... h5 21 g3 Ng4 22 Bf4 Bf8 23 Be4 Bg7 24 Bf3 Nf6 with a solid position.
21 h5 Nxe5 22 dxe5
Opposite coloured bishops reduce winning chances in endings, but often enhance them in attacks.
22 ... Qc8
Sad necessity. On 22 ... Qc7? Adams could break in with 23 hxg6 hxg6 24 Bxg6! fxg6 25 Qxe6+ Kh8
26 Rd7 Qc8 27 Red1.
23 Re3 Bg5 24 Rh3 Qc7 25 Re1 Rad8
25 ... Bf4 26 Qc2 Bxe5 does not work because of 27 hxg6 hxg6 28 Bxg6! fxg6 29 Qxg6+ Qg7 30 Qh5
and the attack is too strong, e.g. 30 ... Bc7 (30 ... Bf4 31 Rh4 wins) 31 Rh4 (31 Re4 Rad8!) 31 ... Kf8 32
Rg4 Qf6 33 Rg6 and White is winning.
26 Qc2 (Diagram 7)

Diagram 7
White intends a later Bxg6

Threatening to take on g6.
26 ... Re7 27 Re4!
Bringing the last piece to the attack.
27 ... Red7 28 Rg3 Bh6 29 hxg6 hxg6 30 Rh4! Bg7 31 f4!
White’s play is very simple, logical and at the same time lethal! Now the situation in the centre is
stable and he may prepare for the decisive blow.
31 ... Kf8 32 Kh2! Rd5 33 Rh7 R8d7 34 Bxg6!
All the white pieces take the ideal positions and it’s time to decide the game with this sacrifice.
34 ... fxg6 35 Qxg6 Rf7 36 Qh5! Bxe5
36 ... Rxf4 37 Rgxg7 Qxg7 38 Rxg7 Kxg7 39 Qg5+ or 36 ... Rd2 37 Rhxg7! Rxg7 38 Qh8+ both win.
37 Qh6+ Rg7
37 ... Ke7 changes nothing: 38 Rxf7+ Kxf7 39 Qh7+ Kf6 40 fxe5+ Qxe5 41 Qg7+ Kf5 42 Qg6+ Kf4 43
Qg4 mate.
38 fxe5 Qxe5 39 Qxg7+ 1-0
Typically accurate play from Adams.

Game 78
Kotronias-Anastasian
Yerevan 2000

1 e4 e6 2 d4 d5 3 Nd2 dxe4 4 Nxe4 Bd7 5 Nf3 Bc6 6 Neg5!? Be7

6 ... Nd7 7 Nxf7!? Kxf7 8 Ng5+ Ke8 9 Bc4 Bxg2 10 Rg1 Bd5 11 Nxe6 gave a tremendous attack for
the piece in Hector-Lyrberg, Geneva 1995.
7 Bd3 Nd7
Others:
a) 7 ... h6? was really asking for it and in Hübner-Komenda, Hessenliga 1994, White won after 8 Nxf7!
Kxf7 9 Ne5+ Kf8 10 Ng6+ Ke8 11 Nxh8.
b) White also had a strong attack with another kind of sacrifice after 7 ... Nf6 8 Qe2 h6 9 Nxe6 Bxf3 10
Nxg7+ Kf8 11 gxf3 Kxg7 12 Bd2 Nc6 13 c3 Qd7, Kartmann-Czech, Bruchkoebel 1993. The g-file is
draughty.
8 Qe2 Ngf6?! (Diagram 8)

Diagram 8


It’s amazing that so strong a player as Anastasian didn’t know that this variation leads to an almost lost
position right after the opening. 8 ... Bxf3 was necessary.
9 Nxf7! Bb4+?!
If 9 ... Kxf7 10 Ng5+ Ke8 11 Nxe6 Qc8 12 Nxg7+ Kf8 13 Ne6+ Black is getting hammered. Or if 9 ...
Bxf3 10 Qxe6 Bxg2 11 Rg1 Bd5 12 Nxd8 Kxd8 13 Qh3 Re8 14 Be3 g6 15 0-0-0 and White won in
Namyslo-Rabl, Wuerttemberg 1997.
What is the knight doing on g5?

10 c3 Bxf3 11 Qxe6+! Qe7 12 Qxe7+ Bxe7
12 ... Kxe7 13 Nxh8 Bxg2 14 Rg1 and wins.
13 Nxh8 (Diagram 9)

Diagram 9
White is already winning

White already has a rook and two pawns for the pair of minor pieces and if he wins one more he will
have the material advantage. In the game Black experiences enormous problems in rounding up the ‘dead’
knight at h8.
13 ... Bd5 14 0-0 Bg8
On 14 ... 0-0-0 15 Re1 Bf8 16 c4 Bg8 17 Bg5 g6 18 Bxg6 hxg6 19 Nxg6 White is winning.
15 Re1 Kf8 16 h3 Bd6 17 Bg5 Nb6
On 17 ... c5 18 dxc5 Nxc5 19 Bc2 Rd8 20 Rad1 Bxa2 21 Bxh7! wins.
18 b3 Nbd5 19 c4 Nb4 20 Bf5 Rd8 21 d5 a5 22 Re3!
Preparing the attack against black king. White’s pawns restrict all the black pieces and in a couple of
moves he creates strong threats on the kingside.
22 ... b6 23 g4 Na6 24 Rf3! Nc5 25 Bd2! 1-0

Game 79
M.Gurevich-Short
Manila Interzonal 1990

1 d4 e6 2 e4 d5 3 exd5
Mikhail Gurevich had led the Interzonal all the way and now needed only to draw this game to qualify
for the Candidates matches. Short needed a win.
3 ... exd5 4 Nf3 Bg4
An active continuation. More usually Black is happy to equalise with 4 ... Bd6 or 4 ... Nf6.
5 h3 Bh5 6 Be2
6 Qe2+ Qe7 7 Be3 is more testing, planning Nc3 and 0-0-0.
6 ... Bd6
6 ... Nc6 7 0-0 Bd6 8 Nc3 Nge7 9 Be3 f6 10 Qd2 Bf7 11 Rae1 was Marshall-Capablanca, St.
Petersburg 1914.
7 Ne5 (Diagram 10)

Diagram 10
White plays for a draw

He exchanges as many pieces as possible.
7 ... Bxe2 8 Qxe2 Ne7 9 0-0
9 Qb5+?! does not win a pawn after 9 ... Nbc6 as d4 is attacked.
9 ... 0-0 10 Bf4 Re8 11 Qg4 Bxe5 12 Bxe5 Ng6
With simple moves Short has brought about full equality. But he had to win ...
13 Bg3 Nd7 14 Nd2 Nf6 15 Qf3 c6 16 Qb3 Qb6!?
17 Qxb6
Queens off. He must have been over the moon. But Short keeps trundling.
17 ... axb6 18 a3?!
18 a4! would have stopped any of the subsequent queenside shenanigans.
18 ... Ne4
18 ... Re2 19 Rfd1 Ne4 20 Nxe4 dxe4 21 Rac1 led to nothing. 18 ... Nh5 19 Bc7 Nhf4 20 Bxb6 Ra6 21
Bc5 b6 22 Bd6 Ne2+ 23 Kh1 Nxd4 24 c4!? with balanced chances was proposed as an odd sideline.
19 Nxe4 Rxe4 20 Rfd1 b5 21 Kf1 f6 22 f3 Re6 23
Re1 Kf7 24 Rxe6
Another exchange, but Short has not run out of steam yet.
24 ... Kxe6 25 Re1+
25 Be1!?, off to b4, was another set up.
25 ... Kd7 26 Ke2
26 Bh2!, meeting 26 ... h5 with 27 g4, might have led to a more desirable arrangement of the kingside
pawns.
26 ... h5
Probing.
27 Kd3 h4 28 Bh2 Ne7 29 Bf4 Nf5 30 Bd2 b6 31
Re2
If 31 Bb4 Rc8 (intending ... c5) 32 Re2 c5 33 Bc3 Ra8.
31 ... c5 32 Be3 b4! (Diagram 11)

Diagram 11
Black breaks in on the queenside

33 axb4 c4+ 34 Kc3 Nd6 35 Re1 Ra4 36 Kd2 Rxb4 37 Ra1 Rxb2 38 Ra7+ Ke6 39 Rxg7 b5 40 Bf2
b4 41 Kc1 c3 42 Bxh4 Nf5 0-1
After 43 Rh7 Ne3 is terminal.
‘Short played this right down the line!’ – M.Chandler. Nigel regards this as his most memorable game.

Game 80
Morozevich-Kiriakov
St Petersburg 1997

1 e4 e6 2 Qe2 (Diagram 12)

 Diagram 12
An unusual line for White

Chigorin’s move, but not too critical an idea. If you have read somewhere about not developing the
queen too early and not putting it in front of your king’s bishop, then we may have read the same books.
2 ... d5
The Sicilianesque 2 ... c5 is the next game.
3 d3 dxe4!?
3 ... Ne7!?.
4 dxe4 b6 5 Nf3 Bc5
5 ... Ba6!? looks more logical, for example 6 c4 c5 7 Nc3 Ne7 8 Bf4 Ng6 9 Rd1 Qc8 10 Bc1 Be7 11
h4 Nc6 with equality in Bakhtadze-Rabinovich, Tallinn 1997.
6 c3 a5
Kiriakov is determined to gain a tempo by developing his bishop on the white queen from a6. He also
secures the position of his c5 bishop, but it does not look right there to me, and he soon shifts it.
7 Qc2 Nf6 8 Bg5 h6 9 Bh4 Be7 10 Na3!?
A cute move, but the normal 10 Nbd2 was good too. Moro lures the black king’s bishop off to smash up
the queenside pawns. But that costs an important potential defender.
10 ... Nfd7 11 Bg3 Bxa3
On 11 ... 0-0 12 Nb5 Na6 13 0-0-0 White has a big advantage in space.
12 bxa3 Ba6 13 c4 Bb7 14 Rd1 Qc8 15 Be2 0-0 16 0-0 Nc6
Black’s game looks comfy enough.
17 e5 Ne7 18 Nd4 Nc5 19 f3 Nf5 20 Bf2 Nxd4?!
It was stronger to play 20 ... Rd8! and Black has enough compensation for a pawn after 21 Nxf5 exf5
22 Rxd8+ Qxd8 23 Bxc5 bxc5 24 Qxf5 Qd2.
21 Rxd4!
A new attacker makes his appearance.
21 ... Rd8 22 Rg4 (Diagram 13)

Diagram 13
White goes for the king

22 ... Rd7
It would have been wiser to have supplied himself with another defender by 22 ... Nd7! when he is
holding out after, e.g. 23 f4 Nf8 24 Bh4 Rd7 25 f5 exf5 26 Qxf5 Qe8.
23 Bh4 Kh8 24 Qc1 Qf8
On 24 ... Kh7 White may immediately do the business with 25 Bf6! gxf6 26 exf6 Qf8 27 Qf4 Rad8 28
Rg7+ Qxg7 29 fxg7 Kxg7 30 h4, etc.
25 Qe3 Na4
Hoping to trade queens, but now the knight’s even further away from the scene of the action.
26 Kh1 Qc5 27 Qf4 Rg8 28 Rd1 Nc3
What else?
29 Rxd7
Four attackers now.
29 ... Nxe2 30 Qxh6+! 1-0
30 ... gxh6 31 Bf6+ Kh7 32 Rxf7+ Rg7 33 Rfxg7+ Kh8 34 Rg8+ mates.

Game 81
Morozevich-Dolmatov
Russian Cup 1997

1 e4 e6 2 Qe2 c5 3 g3 Nc6 4 c3 g6 5 Bg2 Bg7 6 f4 Nge7 7 Nf3 d6 8 0-0 0-0 9 d3 (Diagram 14)

Diagram 14
From the French to the Sicilian

Play has veered back into something like a Closed Sicilian.
9 Na3 Rb8 10 Kh1 d5 11 e5 Qa5 12 Nc2 b5 13 a3 d4!? 14 cxd4 cxd4 15 Nfxd4 Nxd4 16 Nxd4 Bb7 17
b4 Bxg2+ 18 Qxg2 Qb6 led to chances for both sides in Fedorov-Zakharevich, Minsk 1997.
9 ... b5 10 Be3
On 10 a3 Black would keep on trundling with 10 ... a5.
10 ... b4
Dolmatov has already solved all openings problems.
11 Rc1
A new move, but neither did 11 Qc2 cause Black any problems after 11 ... Qa5 12 Nfd2 bxc3 13 bxc3
Ba6 14 Rd1 Rfc8 15 Nc4 Qc7 16 Bf2 d5 in Huseinov-Budnikov, Azov 1991.
11 ... bxc3 12 bxc3 Ba6 13 Nbd2 Rb8 14 Bf1 Qd7
He might have considered the line 14 ... f5!? 15 e5 dxe5 16 fxe5 Qc7 17 Bxc5 Nxe5.
15 Nb3 e5 16 f5!
A correct and brave decision. If White plays passively, Dolmatov will get the initiative on both sides.
16 ... gxf5 17 Bh3 Qe8
Accurate defence. On 17 ... Qc7 18 exf5 e4 19 Ng5 Bxd3 20 Qh5 h6 21 f6 White develops a mighty
initiative, e.g. 21 ... hxg5 22 Bxg5!, or 21 ... Bxf6 22 Ne6! Be2 23 Qxh6 fxe6 24 Bxe6+ Rf7 25 Qxf6
winning.
18 exf5 e4 19 Ng5 (Diagram 15)

Diagram 15
A critical moment in the game

19 ... Nd5!
19 ... exd3? would have been very dangerous, viz 20 Qh5 h6 21 Ne4 (21 f6 Nd5 holds out) 21 ... Nd5
22 Bxh6 Qxe4 (or 22 ... Bxh6 23 Qxh6 Qxe4 24 Re1) 23 Re1! Qc4 24 Bxg7 Kxg7 25 Qg5+ Kh8 26 f6
Nxf6 27 Qxf6+ Kg8 28 Qg5+ Kh8 29 Qh6+ Kg8 30 Bf5 and White is winning. Instead Sergei Dolmatov
gives a pawn for active play.
20 Nxe4
Not 20 Qh5? as 20 ... Nf6 seals everything up in Black’s favour.
20 ... Qxe4! 21 dxe4 Bxe2 22 exd5 Ne5 23 Bg2
23 Kf2 Bc4, or 23 Bf1 Bg4 or 23 Bf4 Rfe8 are all nice continuations for Black.
23 ... Rfe8 24 Bf4 Rb6! 25 Bg5 h6 26 Bh4 Nd7 27 h3 h5 28 Bf1?
A mistake in time trouble, but any case Black had more than enough compensation for the pawn.
28 ... Bxf1 29 Kxf1 Re5 30 Bd8
Or 30 g4 hxg4 31 hxg4 Re4.
30 ... Rxf5+ 31 Kg2 Rb7 32 Rc2!? Nf6
Not falling into the trap. After 32 ... Rxd5? the situation would be changed completely: 33 Na5! Rb5
34 a4 Rb4 35 Nc6 Re4? (35 ... Rd3) 36 Ne7+ etc.
33 Na5?
White’s last chance was 33 Bxf6 Bxf6 34 Rf1 and
he is only slightly worse.
33 ... Nxd5!
White’s pawns fall like leaves, and the final part of the game doesn’t deserve any comment.
34 Re1 Rd7 35 Re8+ Kh7 36 Nc4 Nxc3 37 Ba5 Nb5 38 Rd2 Nd4 39 Bc3 d5 40 Ne3 Re5 41 Rxe5 Bxe5 42 Bxd4 Bxd4 43 Nf5
Bc3 44 Rd1 d4 45 Kf3 d3 46 Ne3 Kg6 47 Nc4 Kf5 48 Rc1 Bb4 49 a3 d2 50 Rd1 Bc3 51 Ne3+ Ke5 52 Ke2 Ke4 53 Rb1 Rd3 54
Nd1 Rxg3 55 Rb7 Rg2+ 56 Kf1 Rh2 57 Kg1 Re2 58 Nxc3+ Kd3 59 Kf1 Re1+ 0-1
Index of Complete Games
Adams-Dreev, Wijk aan Zee 1996
Adams-Epishin, Ter Apel 1992
Adams-Levitt, Dublin zonal 1993
Adams-Morozevich, Sarajevo 2000
Adams-Rozentalis, Belgrade 1999
Adams-Short, Sarajevo 1999
Almasi-Korchnoi, European Cup Final, Budapest 1996
Anand-Gurevich.M, Manila Interzonal 1990
Anand-Khalifman, Linares 2000
Anand-Timman, Dortmunder Schachtage 1999
Berg.E-Potkin.V, World Under 18 Championship, Oropesa del Mar 1999
Bezgodov-Sakaev, Moscow 1999
Chandler-Gurevich.M, Leningrad 1987
De Firmian-Kaidanov, US Championship 1999
Degraeve-Gurevich.M, Belfort 1997
Dvoirys-Dolmatov, USSR Championship, Odessa 1989
Fischer-Larsen, Candidates match, Denver 1971
Fischer-Rossolimo, US Championship 1965
Geller-Karpov, USSR Championship, Moscow 1976
Georgiev.Kir-Yusupov, Las Palmas 1993
Grabarczyk-B.Socko, Polish Championship, Plock 2000
Grischuk-Gulko, Esbjerg 2000
Grischuk-Korchnoi, Biel 2001
Gross-Blatny, Czech Championship 1998
Gurevich.I-Bareev, Biel 1993
Gurevich.M-Short, Manila Interzonal 1990
Hansen.L-Antonsen, Farum 1993
Hansen.S-Brynell, Gothenburg 1998
Hector-Barsov, Vikings GM, York 1999
Hodgson-Kupreichik, German Bundesliga 1998
Hracek-Socko.M, MK Cafe Cup, Koszalin 1999
Kalichkin-Gavrjushin, Russian Cup, Perm 1997
Karpov-Kuzmin, Leningrad Interzonal 1973
Karpov-Uhlmann, Madrid 1973
Kasparov-Anand, Reggio Emilia 1992
Kasparov-Bareev, World Cup of Rapid Chess, Group A, Cannes 2001
Kasparov-Gurevich.M, Bosna SuperGM, Sarajevo 2000
Kasparov-Korchnoi, Zürich rapid 2001
Kasparov-Ponomariov, Linares 2002
Kasparov-Shirov, Astana 2001
Kasparov-Shirov, Fujitsu-Siemens Giants, Frankfurt 2000
Kasparov-Short, Novgorod 1997
Khalifman-Timman, Reykjavik 1991
Kharlov-Beliavsky, Novosibirsk 1995
Korneev-Matamoros Franco, Elgoibar 1997
Kotronias-Anastasian, Yerevan 2000
Kudrin-Short, Monarch Assurance, Port Erin 1999
Lanka-Morozevich, Kishinev 1998
Lobzhanidze-Yagupov, St Petersburg 1997
Maus-Hübner, Lugano 1989
Morozevich-Dolmatov, Russian Cup 1997
Morozevich-Kiriakov, St Petersburg 1997
Movsesian-Gurevich.M, Sarajevo 2000
Mrdja-Gleizerov, Turin 2000
Nataf-Ulibin, Stockholm 1999
Nijboer-Gurevich.M, Escaldes 1998
Nijboer-Nikolic.P, Dutch Championship, Rotterdam 1997
Nunn-Nikolic.P, Belgrade 1991
Polgar.J-Korchnoi, Wijk aan Zee 2000
Polgar.J-Morozevich, Wijk aan Zee 2000
Polgar.J-Shirov, EuroTel Trophy, Prague 1999
Polgar.J-Shirov, EuroTel Trophy, Prague 1999
Reshevsky-Vaganian, Skopje 1976
Romanishin-Beliavsky, Belgrade 1993
Shirov-Akopian, Merida 2000
Shirov-Morozevich, Fujitsu Siemens Giants, Frankfurt 2000
Shirov-Timman, Groningen 1996
Shirov-Topalov, Sarajevo 2000
Short-Ivanchuk, Horgen 1995
Short-Kosten, Hastings 1988
Spassky-Korchnoi, Candidates final, Belgrade 1977
Sutovsky-Psakhis, Tel Aviv 1999
Sutovsky-Zifroni, Israel 2000
Svidler-Ivanchuk, Linares 1999
Timman-Andersson, Sigeman & Co, Malmoe 2000
Topalov-Bareev, Linares 1994
Topalov-Morozevich, Sarajevo 1999
Van der Wiel-Korchnoi, Euwe Memorial, Amsterdam 1991
Yermolinsky-Vaganian, Tilburg 1993
Yurtaev-Driamin, Russian Cup, Tomsk 1998
Zelcic-Zaja, Croatian Championship, Pula 1998