2

First published in 2007 by Gloucester Publishers Limited, Northburgh House,

10 Northburgh Street, London EC1V 0AT

Copyright © 2007 Richard Palliser

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

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

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

British Library Cataloguing-in-Publication Data

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

ISBN Kindle: 978-1-78194-068-6

ISBN epub: 978-1-78194-069-3

Distributed in North America by National Book Network,

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

Distributed in Europe by Central Books Ltd.,

99 Wallis Road, London E9 5LN. Ph 44(0)845 458 9911.

All other sales enquiries should be directed to Everyman Chess,

Northburgh House, 10 Northburgh Street, London EC1V 0AT
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.

3
Contents
Bibliography
Introduction

Part One – The Colle System

1 Black Plays ...c5 and ...Nc6

2 The Main Line with 9...Qc7
3 Black Plays ...c5 and ...Nbd7
4 Other Black Set-ups after 3...e6

Part Two – The Anti-Colle

5 Black’s Third Move Alternatives

6 Black Delays ...Nf6

Index of Complete Games

4
Bibliography
A Killer Chess Opening Repertoire, Aaron Summerscale (Everyman 1998)
Chess Openings for Black, Explained, Lev Alburt, Roman Dzindzichashvili & Eugene Perelshteyn (CIRC 2005)
Colle Plays The Colle System, Adam Harvey (Chess Enterprises 2002)
Dealing with d4 Deviations, John Cox (Everyman 2005)
How to Beat 1 d4, James Rizzitano (Gambit 2005)
How to Succeed in the Queen Pawn Openings, John Watson & Eric Schiller (Trafford 2006)
Ideas Behind Modern Chess Openings: Black, Gary Lane (Batsford 2005)
Meeting 1 d4, Jacob Aagaard & Esben Lund (Everyman 2002)
Play 1 d4!, Richard Palliser (Batsford 2003)
The Chess Advantage in Black and White, Larry Kaufman (Random House 2004)
The Chigorin Defence According to Morozevich, Alexander Morozevich & Vladimir Barsky (New in Chess 2007)
The Colle System, Dimitrij Oleinikov (ChessBase CD 2003)
The Colle-Koltanowski System, Valerij Bronznik (Schachverlag Kania 2004)
The Dogs of War: Practical Play in The Colle System, Terese Hatch & David W. Hatch (2nd edition, Pawn Promotions 2005)
The Ultimate Colle, Gary Lane (Batsford 2001)
Winning With The Colle System, Ken Smith & John Hall (2nd edition, Chess Digest 1990)

5
Introduction
In amateur play, I don’t think there is a more popular and useful opening (from White’s perspective) than the
Colle System. It’s quick and easy to learn, safe, and contains quite a bit of sting.
Jeremy Silman (acclaimed author of How to Reassess Your Chess)

1 d4 d5 2 Nf3 Nf6 3 e3 (Diagram 1)

Diagram 1 (B)
The Colle System

brings about the Colle System, or should I say systems? After 3...e6 4 Bd3, White has the choice between two simple set-ups which both
offer attacking chances on the kingside without being particularly theoretical.

The Colle System

The player who specializes in the white side of the Queen’s Gambit Declined (1 d4 d5 2 Nf3 Nf6 3 c4 e6 4 Nc3) will have to make some
important decisions over the course of the next few moves, depending on whether Black plays an early ...Bb4, heads for the sharp and
theoretical Semi-Slav with 4...c6 or employs one of the many Classical defences. Compare that with the Colle: 1 d4 d5 2 Nf3 Nf6 3 e3 e6 4
Bd3 c5 5 c3 and now no matter how Black develops his king’s bishop and queenside, White can play 6 Nbd2, 7 0-0 and then look to break with
e4 (Diagram 2).

6
 Diagram 2 (B)
The main line Colle

The e4-break is the key to White’s play. Should Black exchange on e4, then White will recapture with his
knight after which the c1-bishop is free to move and White has easy development. But should Black ignore
the pawn on e4, he may find that it quickly advances to e5, driving away the f6-knight and heralding the
beginning of a kingside attack.
This system dates back to the Nineteenth Century, but came to be named after by far its leading protagonist in the 1920s, the Belgian
Edgard Colle (1897-1932). Following Colle’s untimely death, his favourite system was adopted by his countryman, George Koltanowski
(1903-2000); a player chiefly remembered today outside the United States for his vast blindfold simultaneous exhibitions. Undoubtedly
Koltanowski did much to popularize the opening in America (after arriving there during the Second World War), and it is nowadays sometimes
even referred to as the ‘Colle-Koltanowski System’.

The Zukertort Attack

Also known as ‘the Colle-Zukertort’, this system shares much in common with the traditional Colle. White will begin as before with 1 d4, 2
Nf3, 3 e3 and 4 Bd3, but rather than support d4 with c3 and look to break with e4, he instead prefers to fianchetto his queen’s bishop: 5 b3 and
then 0-0, Bb2 and Nbd2 (Diagram 3)

7
 Diagram 3 (B)
A typical Zukertort

completes White’s easy-to-employ set-up.

This is also a pretty reasonable option for White, but do please note that we will only concentrate on the pure Colle in this work, not the
Zukertort set-up.

The Kingside Attack

Throughout this book we will witness a number of white kingside attacks and typical attacking motifs. Probably the most important of these is
the attack on h7 and the Colle is an opening which allows the white clubplayer not just to dream about making a Greek Gift sacrifice, but of
actually carrying one out.
NOTE: A ‘Greek Gift’ sacrifice occurs when White sacrifices his light-squared bishop on h7,
intending to mate after 1 Bxh7+ Kxh7 2 Ng5+ Kg8 3 Qh5.
(Diagram 4)

Diagram 4 (W)
8
 White to play and win

shows the Greek Gift in its purest form (of course, in an actual game Black would have many more pieces
and the queenside wouldn’t be so empty).
White wins with 1 Bxh7+! Kxh7 (Black could decline the bishop with 1...Kh8, but he has then lost a vital
defensive pawn; White will probably retreat the bishop before continuing his attack with Ng5 and Qh5+, or
Qg4-h4 after ...g6) 2 Ng5+ (Diagram 5) and now Black is defenceless:

Diagram 5 (B)
The Greek Gift is decisive

a) 2...Kg8 3 Qh5 threatens mate on h7 and in this particular instance White’s fastest win after 3...Rd8 is 4
Ba3, cutting off the Black king’s escape. That motif sometimes crops up in practice too, but more usually the
presence of Black’s other pieces prevents his king from escaping past e7 and White wins more slowly with 4
Qxf7+ Kh8 5 Qh5+ Kg8 6 Qh7+ Kf8 7 Qh8+ Kg7 8 Qxg7+ and mate follows. Note that even if f7 was
defended, such as by a black rook on a7, White’s attack is still usually decisive following 4 Qh7+ Kf8 5
Qh8+ Kf8 6 Qxg7 and then either f7 falls or a nasty check occurs on the h4-d8 diagonal. Of course, in our
particular and contrived case, the alert reader will no doubt have spotted that 4 Qh7+ Kf8 5 Qh8+ Kf8 6
Qxg7?? fails to 6...Rd1 mate; a most embarrassing turnaround which reminds us to always be alert to our
opponent’s threats.
b) 2...Kh6 usually fails to a lethal discovered check (on e6) or even a discovered and double check (on f7),
which may well net a queen on d8 or c7, if not actually help White force mate.
NOTE: White’s dark-squared bishop is often a silent but vital piece on the c1-h6 diagonal in any
Greek Gift. Without it Black may be able to defend with 1 Bxh7+ Kxh7 2 Ng5+ Kh6 (depending on
whether he has a defence to 3 Qg4, such as 3...Rh8), and White must also make certain that 2 Ng5+
doesn’t just drop the knight to a black queen on d8.
c) 2...Kg6 is usually the most challenging defence. In our position White’s quickest mate begins with 3
Qd3+, but in practice he is more likely to be successful with 3 Qg4 (when Black will probably have to try
3...f5) or with 3 h4, intending 4 h5+.

Of course, any Greek Gift is reliant upon any f6-knight having been driven away from its defensive duties,
usually with e4-e5, White’s ideal Colle advance. When preparing to sacrifice the light-squared bishop it is
9
always wise to view the whole board and to consider the individual position of each black piece.
WARNING: Turn back to Diagram 4 and place a black dark-squared bishop on e7. 1 Bxh7+??
Kxh7 2 Ng5+ now fails to 2...Bxg5, and such a bishop on e7 is normally sufficient to thwart any Greek
Gift. White must also be aware of Black defending back along the b1-h7 diagonal after 2...Kg8 3 Qh5+
with either his queen or bishop (Diagram 6).

Diagram 6 (B)
3...Qd3 or 3...Bd3 covers h7

Stronger readers may well be wondering if they’ve opened a primer on attacking chess, rather than a book
on the Colle, but the Colle does abound like no other opening with chances at lower levels for White to
execute a textbook bishop sacrifice on h7. In practice White will often attack but without carrying out a
bishop sacrifice on h7. He may threaten one and thereby induce a weakening ...h6 or ...g6 advance, or White
might just force one of those moves by building up against h7, such as with Ng5 and Qh5, assuming no f6-
knight, or with one Bc2, Qd3 and then either Ng5 or Bg5 (Diagram 7).

10
 Diagram 7 (W)
Bg5 or Ng5 will increase the pressure

Throughout this work we will see a number of different attacking mechanisms for White, but in just about
every case the defence of h7 should be at the forefront of Black’s mind. Fortunately for white club players, it
frequently isn’t. The Greek Gift sacrifice, for instance, may be as old as the hills, but even players who should
know better have fallen for it, possibly having been lulled into a false sense of security by White’s apparently
simple (yet deadly!) early development. Let’s look first at a Victorian classic:

D.Baird-J.Hanham (see Diagram 8)

New York 1893

This position is a little different from the standard Greek Gift affair since there isn’t a white knight on f3.
That does not, however, preclude the sacrifice, especially when, as here, f7 is also vulnerable.

11
 Diagram 8 (W)
Where are Black’s kingside defenders?

12 Bxh7+!
Already this is pretty much decisive. Should Black capture the bishop with 12...Kxh7 then White reveals
another important Greek Gift motif, namely a rook lift: 13 Qh5+ Kg8 14 Qxf7+ Kh8 15 Re3! and mate
follows (15...Nf6 16 Rh3+ Nh7 17 Ng6mate).
In the game Hanham exploited the vacant f8-square with 12...Kf8, but that wasn’t enough to save him: 13
Ndf3 (correctly introducing more pieces into the attack, rather than getting involved in the flashy 13 Nxf7)
13...f6? 14 Ng6+ Kf7 15 Nh8+! Kf8 (15...Rxh8 16 Qxe6+ Kf8 17 Qe8mate was White’s point) 16 Nh4 and
Black was crushed.
Hanham wasn’t a particularly weak player (and he is chiefly remembered for the variation of the Philidor
which bears his name), so how could he overlook White’s sacrifice? It may help if we consider that his last
move (Diagram 9)

12
 Diagram 9 (B)
The e5-knight is very strong

was 11...Nf6-d7.
Undoubtedly Hanham was rather concerned about White’s dominant knight on e5 and wanted to exchange
it off or to chase it away with a well-timed ...f6. His problem was that an immediate exchange with
11...Nxe5?? would have cost him a piece after 12 dxe5 and 11...Bxe5 was also undesirable; not only does
Black cede bishop for knight, but after 12 dxe5 Nd7 13 Nf3 (and not 13 Bxh7+? due to 13...Kxh7 14 Qh5+
Kg8 when Black defends after both 15 Nf3 Nf8 and 15 Re3 Ncxe5 16 Rh3 Ng6), White has a useful space
advantage and can build up almost at his leisure on the kingside.
Sacrifices on h7 aren’t always played with the aim of mate. Instead they can be used as a device to pick up
a loose piece or pawn, often down the d-file or on e8 (with Black’s f-pawn missing or able to be won with
check). Here’s a simple and recent example:

13
 Diagram 10 (W)
Bxh7+ can be used to win material

J.Curtis-T.Headlong (see Diagram 10)

British League 2006

24 Bxh7+! Kxh7 25 Qh4+ Kg8 26 Bxe7 (Diagram 11)

Diagram 11 (B)
White has netted a useful pawn

and White had won a pawn, while also increasing his kingside attacking chances.

The Attacking Stonewall Set-up

In the Colle White will often play for an early Ne5, especially after a premature ...cxd4; exd4 exchange. From e5 the knight towers over the
black camp and is not at all easy to chase away. White will then look to attack on the kingside, but may well first strengthen his grip on e5 with
f2-f4. The resultant ‘Stonewall’ structure is both easy to handle and difficult for Black to defend against. For example:

S.Reshevsky-A.Kevitz (see Diagram 12)

US Championship, New York 1936

14
 Diagram 12 (W)
The e5-knight is very powerful

Black’s handling of the opening has not been especially accurate: his premature exchange on d4 (still quite
a typical mistake at club level), has allowed White a fine square for his knight on e5. Reshevsky now
increased his grip over that key square.
12 f4!
Setting up a cramping Stonewall formation. We will see much more of this set-up in Chapters Three and
Four, and this is a particularly good version since Black lacks some essential counterplay. Note too how he
cannot capture on e5: 12...Nxe5?? 13 fxe5 forks and wins a piece, while 12...Bxe5? 13 fxe5 merely opens
lines for White’s attack.
12...Rfe8 13 g4!?
We are taught not to advance pawns in front of our king, but when the opponent has no counterplay it can
be safe to do so. Reshevsky wants to drive the f6-knight away and begin a strong attack. In such positions
White can also consider 13 Ndf3 heading for g5, but here that would allow Black to awkwardly interfere with
White’s build-up after 13...Ne4!.
13...Nf8 14 g5 N6d7 15 Nf1!
Unravelling and heading for h5. At some point Black must break out with ...f6 or ...f5 (White should
capture en passant in either case) which is positionally desirable for him, but in this case only helps White’s
attack.
15...Re7 16 Ng3 f6
Otherwise Nh5 would have increased the pressure in any case, and then one effective follow-up is Qg4-h4
and Re3-h3.
NOTE: Swinging at least one major piece across to the h-file is an integral part of many Stonewall
attacks.
17 gxf6 gxf6 18 Nxd7 Rxd7 19 Nh5 Rf7 20 f5! (Diagram 13)

15
 Diagram 13 (B)
White has a strong attack

Opening a path for the c1-bishop. White enjoys a dangerous attack and Reshevsky went on to win.

The Disparaging Views of Strong Players

What we have seen so far sounds pretty ideal: White develops his pieces in straightforward manner and then has good attacking chances on the
kingside, especially with a Bxh7+ sacrifice. The first part of that is definitely true, although I’m not quite sure I fully agree with the famous
author C.J.S. Purdy when he wrote: ‘A player who specializes in the Colle System needs to spend only about a tenth of the time studying the
openings (for White) that he would otherwise have to.’ Perhaps that is so for when White gets to play the Colle after 1 d4 d5 2 Nf3 Nf6 3 e3 e6,
but Black can obviously also deviate earlier as we will see. As for White’s attacking chances, these are real and frequent at club level, although
to hear some strong players speak the Colle is just a bad opening. What exactly has caused such views?

1. White-to-play-and-win Manuals
The Colle is an excellent opening for those new to the game or keen to avoid masses of theory: White’s development system is very easy to
employ and he will struggle to blunder early on. Unsurprisingly a number of books and articles have been written to cater for such an audience.
They do contain some good ideas, but all too often exaggerate White’s chances and downplay or even outright ignore some of Black’s better
defences.
Let’s say Joe Bloggs crushes Sam Smith in a Colle miniature in the U1400 tournament at their local club.
A few weeks later the two meet again in the Minor section of a weekend congress. This time Sam improves, a
complex but roughly balanced middlegame ensues and the game is heading for a draw until a late blunder
decides. Upon returning home Joe consults his Colle source; a work which has always impressed him since it
contains a whole white opening repertoire, unlike the many specialist 1 e4 manuals also on his bookshelf.
Unfortunately for Joe, Sam’s natural-looking 10th move isn’t anywhere to be found and he was sure that this
line was very good for White. Disheartened he turns on his computer, but even the might of the German chess
engine can’t find a way for White to gain a large advantage or even to launch a strong attack on move 11.
Even worse, Joe’s book hasn’t explained all the key ideas for both sides in the line in question. At some stage
in every game we are ‘taken out of book’; in the Colle that shouldn’t be a large worry since the key ideas for
both sides are usually fairly easy to understand.
The Colle is an easy opening to learn and with knowledge of a little theory and some typical motifs, one
can be pretty successful with it up to a certain level. However, if Black also understands the Colle pretty well
and knows some theory, unfortunately White has no objective advantage. But think about it: can White really
force an advantage against the Najdorf or Nimzo-Indian, assuming best play from both sides? He can’t, and
such reasons shouldn’t put White off the Colle.

16
 In short, the key challenge for your author is to instruct the reader in the main ideas of the Colle, while at
the same time not neglecting Black’s best defences and ways for White to then play the resulting positions.

2. Grandmasters Don’t Play the Colle

Usually some sort of exception is made for Yusupov (a man who once reached a Candidates’ semi-final no
less), but one will often hear that few or even no grandmasters specialize in the Colle. Such an argument most
certainly has its merits. Anand, Svidler and Ponomariov might have employed either the Colle or the
Zukertort, but that was only on the odd occasion: respectively when facing a computer, when very young and
a mere 2480, and against a much weaker opponent in a rapidplay. Quite simply at 2600+ level, the likely
absence of strategical middlegame mistakes in the Colle, aligned to good theoretical knowledge and defensive
technique, means that White’s winning chances are pretty slim in comparison with the sharper 1 e4 and 1 d4
main lines in vogue at the highest level.
So the Colle might not be a good opening at the likes of Linares, but at slightly less exalted levels, the Colle and Zukertort are both found;
Grandmaster Kovacevic has employed the opening against all-comers for more than twenty years. Going back to the 1920s and 1930s, we find
a number of Colles in the leading events of the time, and by no means only when Colle himself was playing: Alekhine (who even employed the
Colle against both Bogoljubow and Euwe in World Championship matches), Capablanca, Euwe and Rubinstein all made good use of the
opening. Quite simply the World’s elite in those days were not yet certain that 3 c4 (after 1 d4 d5 2 Nf3 Nf6) was a superior move to 3 e3.
Chess understanding has greatly increased since the thirties, but the elite of those days were still somewhat superior to your average clubplayer
nowadays. If they found the Colle a good opening to employ, and their opponents often struggled against it, there can be no reason for the
amateur player to reject it as ‘not strong enough’ today.

3. Strong Players Don’t Actually Like facing the Colle

A final reason why some strong players may be a little disparaging about the Colle is because they, like many of their lower-rated club
colleagues, don’t actually like facing it, chiefly for one of two related reasons: (i) White’s opening is fairly untheoretical but very sound, and
Black needs to be both alert and quite patient when facing it; and (ii) it can be hard to defeat weaker opposition when they employ the Colle. As
Watson and Schiller put it, ‘Black must create his own imbalances if he wants to play ambitiously’, but of course such an approach doesn’t suit
everyone and may play into White’s hands.

What Lines are Included?

As we’ve already mentioned this is a work on the Colle System, not the Colle-Zukertort. It would, though, be a little criminal to only consider
the main line of the Colle which is usually characterized as beginning 1 d4 d5 2 Nf3 Nf6 3 e3 e6. Those who employ the Colle must also be
ready for some other tricky options from Black on his second and third moves; these will be examined in Chapters Five and Six.
The Colle player must also be aware that while this work will help them to combat 1 d4 d5, it does not
feature 1 d4 Nf6. Many Colle fans like playing 2 Nf3 when it is possible to meet 2...e6 with a Colle approach:
3 e3 b6 (3...d5 returns to our main subject matter) 4 Bd3 Bb7 5 0-0 c5, and now 6 c4!? leads to a tricky line of
the Queen’s Indian, but staying in Colle vein and preparing the traditional e3-e4 push with 6 Nbd2 (Diagram
14)

17
 Diagram 14 (B)
The Colle Queen’s Indian

is also quite playable. However, I’m afraid that I cannot recommend meeting 1 d4 Nf6 2 Nf3 g6 with 3 e3
(White should instead consider 3 Bg5, 3 Bf4 and 3 Nc3, as well as 3 c4), since (i) on d3 White’s bishop will
now bite on granite and (ii) Black has easy play with ...d6 and ...e5 (see Diagram 15).

Diagram 15 (W)
An ineffective Colle set-up

Selection of Games
This is a Starting Out guide (with all its usual features, including notes, tips and warnings as we’ve already seen), not a comprehensive
theoretical tome on the Colle. There is some theory on the opening (and clearly the stronger club player may wish to try and remember a little
more of it than their weaker counterpart), and indeed one or two positions are quite critical, but it is an opening in which a knowledge of where
the pieces go and of typical strategical motifs is generally more important than any specific knowledge of theory. As a result, the illustrative
games have usually been chosen with the aim of highlighting common types of middlegame and typical plans arising.

18
 The reader will also find that the illustrative games are generally between fairly strong players. Not only
did I not want to rehash the classics for those who already employ the opening (although a few were still too
instructive to leave out; likewise some games by lower-rated players), but also by studying some of the better
games available we are more likely to appreciate both White’s main plans and especially Black’s best
defences. Only the player comfortable with their opening even when Black plays one of the better lines
against it, can be said to be on the road towards slowly mastering that opening.
Some readers may well want to know more about what’s going to happen at lower club levels than in an
international game featuring a grandmaster or two. As a concession to any such readers, I have at times
overused the word ‘typical’; for instance, ‘a typical trap’ refers not to a trap a grandmaster is likely to fall for,
but rather one which is fairly common at club level. I do not, however, propose to give white win after white
win in low-level games. Should anyone having read this work still want to study a collection of lower-level
Colle games, they are referred either to ChessBase’s Mega or Bigdatabases (collections of over 3.5 million
games; unsurprisingly they now contain a lot of games by untitled and even internationally unrated players),
or to Hatch and Hatch’s creatively-titled The Dogs of War: Practical Play in The Colle System; a collection
of games played by all levels of player from 900-rating to GM strength.
Finally, it must be said that the selection of lines covered and illustrative games supplied was not influenced only by Colle theory and
typical Colle motifs. We live in an information age, and of late a number of repertoire works having sprung up offering Black advice in
handling the Colle amongst other openings. On the grounds that the reader’s opponents may well turn to such sources in a bid to avoid being
wiped out by the Colle, we will pay particular attention to the lines they propose.

How Best to Use This Book

Some readers may wish to systematically work through every page and note, but for those new to the Colle I would recommend the following:
play through the main line which begins each section, only stopping to examine any sub-variations which catch the eye or contain a fair amount
of explanatory text, and then study the illustrative games at the end of each section. After that you should be ready to begin using the Colle in
your own games, but after each outing with the opening, please look up the opening moves and study that particular line in a little more depth.
This way your Colle understanding will begin to build up, and without any information overload occurring in the process!
I should also mention that some readers may find the notes a little more theoretical than they expected at
times. Quite simply the Colle is not completely untheoretical and it would be most unfair on stronger readers
not to cover some critical and fairly sharp lines.
It only remains to thank my colleagues at Everyman Chess for all their help with this project, along with
Joachim Berndt, John Cox, Dylan Morgan, Jim Nicholson, Jeremy Silman, Kieran Smallbone and Norman
Stephenson for their kind assistance and thoughts, and to wish the reader all the best with his or her Colle
adventures.

Richard Palliser,
Harrogate,
March 2007

19
Chapter One
The Colle System: Black Plays ...c5 and ...Nc6
The Early Moves
1 d4 d5 2 Nf3 Nf6 3 e3 (Diagram 1)

Diagram 1 (B)
Black will often play 3...e6

3...e6
Black continues as he would in the Queen’s Gambit Declined, except that White hasn’t played c4. He
doesn’t have to block in his light-squared bishop (and the alternatives are covered in Chapter Five), but this is
by far Black’s most popular continuation, occurring more than three times as often as his main alternatives
according to ChessBase’s Big Database 2007 (our source for all future statistics).
4 Bd3 c5
As White hasn’t employed an early c4, Black naturally enough decides to play a sort of reversed Queen’s
Gambit himself. This is again the most popular option, although he can also delay ...c5 as we will see in
Chapter Four.
5 c3
A useful move and a key component in the Colle development scheme.
NOTE: 5 c3 both adds support to the d4-pawn and prepares a retreat to c2 for the light-squared
bishop should Black employ an early ...c4-advance.
5...Nc6 (Diagram 2)

20
 Diagram 2 (W)
A straightforward and common defence

The most aggressive square for the knight and by far the most common move seen in practice. Should
Black, though, want to fianchetto his queen’s bishop, he might well prefer not to block the bishop’s range
from b7 with 5...Nbd7, the subject of Chapter Three. Black can also develop his kingside before deciding
where to place his queen’s knight, but after 5...Be7 (or 5...Bd6) 6 Nbd2 0-0 7 0-0 play is likely to transpose to
either the lines of this chapter or the next (after 7...Nc6), or to those of Chapter Three (after 7...Nbd7).
Another option is 7...b6, reaching a position we will consider in Chapter Four in which 8 e4 isn’t White’s
only try; another very thematic move being 8 Ne5, followed by f4.
6 Nbd2 Bd6
Another natural move, both preventing Ne5 and should White continue too passively over the next couple
of moves, possibly preparing ...e5. Black can also play a little more cautiously with 6...Be7, but this doesn’t
make any difference should White be happy with the main line: 7 0-0 0-0 (Diagram 3)

21
 Diagram 3 (W)
Less aggressive than 6...Bd6

8 dxc5 leaves Black with nothing better than 8...Bxc5 9 e4 and we’re back in the main subject of this
chapter and the next. Of course, if White doesn’t want to play an early dxc5, then the bishop’s presence on e7
rather than on d6 makes a difference and I will endeavour to point out such differences as we go along.
At club level, Black may worry about having to lose a tempo with his dark-squared bishop (in the event of
...Bd6 or ...Be7; dxc5 Bxc5) and mistakenly decide to ease the central tension: 6...c4? is seen in our first
illustrative game and 6...cxd4?! in the second.
TIP: White should be happy to see Black picking up his c-pawn on move 6. No matter what he does
with it, the resulting positions tend to emphasize White’s attacking chances on the kingside and Black’s
potentially bad light-squared bishop.
A radical alternative to developing Black’s dark-squared bishop is 6...g5?!, an idea given some coverage
by Hatch and Hatch. Black is trying to play a reversed Shabalov Gambit (1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 Nc3 e6
5 e3 Nbd7 6 Qc2 Bd6 7 g4!?), but this tempo down version isn’t fully convincing: 7 0-0 (White can even
consider 7 Nxg5!? Rg8 8 Nxh7 Nxh7 9 Bxh7 Rxg2 10 Qf3 – Hatch and Hatch) 7...g4 (Diagram 4)

22
 Diagram 4 (W)
Ambitious play from Black

8 Ne5! (exploiting the absence of ...Qc7) 8...Nxe5 9 dxe5 Nd7 10 f4 c4 (White remains for choice after
10...gxf3!? 11 Nxf3 Be7 12 Qe2 and e4 – Black’s king is still some distance from finding a safe home on the
queenside) 11 Bc2 h5 (J.Markus-M.Steinbacher, Amsterdam 2004) 12 Qe2 and Black’s kingside advance is
beginning to look a little irrelevant with White ready to powerfully break with some combination of b3 and
e4.
7 0-0 (Diagram 5)

Diagram 5 (B)
White will break with e3-e4

WARNING: Advancing with e3-e4 is White’s main break in the Colle and one which he must strive
23
for. However, an immediate 7 e4 is premature and Black secures a comfortable game with 7...cxd4 8
cxd4 dxe4 9 Nxe4 Nxe4 10 Bxe4 Bb4+!.
In the Colle White can sometimes accept an isolated queen’s pawn (or IQP) in return for attacking
chances, but this is not one of those cases. The problem is that the side with the IQP needs to use his extra
space to attack, whereas the defender wants to trade pieces down to an ending in which the d-pawn becomes a
fatal weakness. Here after 11 Bd2 Bxd2+ 12 Qxd2 Bd7 at least one set of minor pieces too many have been
traded for White’s liking, and he has nothing better than a rapid d4-d5 to secure full equality.
7...0-0
Once again 7...c4 and 7...cxd4 are misguided approaches (the latter transposes to Game 2, while the former
is very similar to Game 1), but what if Black decides to try and seize the initiative with 7...e5? After all,
following 1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 Nc3 e6 5 e3 Nbd7 (this variation of the Semi-Slav is actually a
reversed Colle) 6 Bd3 Bd6, White’s best move is 7 e4 (or 7 0-0 0-0 8 e4). In our Colle position White is,
however, a crucial tempo ahead of Black in that Semi-Slav sequence which makes a vital difference: 8 dxe5
Nxe5 9 Nxe5 Bxe5 10 e4! (Diagram 6)

Diagram 6 (B)
White seizes the initiative

and Black is beaten to the bounce. White has good chances to seize a strong initiative here with 11 f4 and 12 e5 the immediate plan:
a) 10...Bxh2+?? 11 Kxh2 Ng4+ 12 Kg1 Qh4 fails to 13 Nf3 – remember this key defensive resource
should you ever be tempted to carry out a Greek Gift in the ...Nbd7 lines.
b) 10...0-0 11 f4 Bc7 12 e5 Ng4 13 Nf3 c4 14 Bb1 Bb6+ 15 Kh1!? (Diagram 7)

24
 Diagram 7 (B)
A white kingside attack is likely

left White with good attacking prospects (down the b1-h7 diagonal and with f4-f5) in R.Frosch-I.Ceschia,
Austrian Team Championship 2006. Note here that 15...Nf2+?? would be a howler due to the double attack
after 16 Rxf2 Bxf2 17 Qc2.
c) 10...Qc7!? is Black’s most combative defence, but White retains the advantage with the bold 11 exd5!.
Now 11...0-0 12 Nf3 Bg4 13 h3 is just good for White, so in T.Doeppner-E.Hermansson, German League
2002, Black tried 11...Bxh2+ 12 Kh1 0-0?! and White fell for the bluff, shying away from 13 g3! Bxg3 14
fxg3 Qxg3 15 Qf3 Qh4+ 16 Kg2 when there isn’t objectively enough for the piece. Admittedly the white king
would appreciate some pawn cover, but White’s pieces are quick to come to the kingside to defend and Black
might even live to regret half-opening the h-file.
d) 10...dxe4 is a safety-first approach, but one which can easily leave Black horribly tied up in knots after
11 Nxe4 Nxe4? (this time 11...Bxh2+?? 12 Kxh2 Ng4+ 13 Kg1 Qh4 fails to 14 Bf4, although 14 Bb5+ might
be even stronger; actually the best defence is the untried 11...0-0!, but White should be able to retain a little
nibble, such as with 12 Re1 Nxe4 13 Bxe4 h6! 14 Qxd8 Rxd8 15 Be3 Bd6 16 Rad1) 12 Bxe4 (Diagram 8)

25
 Diagram 8 (B)
Black is already struggling

12...Qxd1 13 Rxd1 0-0 14 Be3 when White’s bishops rule the board and it will cost Black a pawn to develop his c8-bishop.
NOTE: In this last sequence Black didn’t have to trade queens, but he should avoid 12...0-0 in view
of the typical Colle motif 13 Bxh7+! Kxh7 14 Qh5+, winning a clear pawn.
Finally, Black might prepare ...e5 with 7...Qc7, but after 8 dxc5 Bxc5 9 e4 he has nothing better than 9...0-
0, transposing to the main line and Chapter Two.
Returning to 7...0-0:
8 Qe2
Supporting the e4-push and avoiding multiple exchanges as can occur in the main line after 8 dxc5 Bxc5 9
e4 (Diagram 9)

Diagram 9 (B)

26
 The main line

9...dxe4 10 Nxe4 Nxe4 11 Bxe4 Qxd1 12 Rxd1 (not that White should at all fear this position as we will
see later in this chapter). 8 Qe2 is essentially a tricky transpositional move; White wants to transpose to
certain variations in the main line, while avoiding others.
It is also possible to play an immediate 8 e4 and this advance is much better timed than on the last move
since White has now castled. After 8...cxd4 9 cxd4 dxe4 10 Nxe4 (Diagram 10)

Diagram 10 (B)
White prefers an IQP set-up

an IQP situation has arisen and this should be a relatively comfortable version for Black. That said, a white
player experienced in such positions and/or knowing that his opponent doesn’t like to defend against the IQP
may find this to be a reasonable practical alternative to the main line. See Game 3 for details.
In the ...Nbd7 lines we will frequently see White preparing e4 with Re1, but here 8 Re1 is dubious. The
variation 8...e5! (8...Qc7 is also fine for Black) 9 dxc5 Bxc5 10 e4 Bg4 (Diagram 11)

27
 Diagram 11 (W)
Black has equalized easily

with active and easy development for Black demonstrates why: 11 h3 (with the queen on e2, White can try
11 exd5 Qxd5 12 Ne4, but here 12...Rad8 is clearly nothing for him) 11...Bh5! 12 exd5 Qxd5 13 Bc4 Qd8 14
b4 Bb6 15 Bb2 Qc7 16 Qb3?! e4! 17 Nd4 Qf4 left White in trouble in N.Amaral-R.Damaso, Portuguese
Championship 1993; White’s queenside expansion wasn’t especially convincing and Black took over the
initiative in straightforward style. White can also meet 8 Re1 e5 with 9 dxe5 Nxe5 10 Nxe5 Bxe5, but again
this fails to trouble Black after 11 Nf3 Bc7 or 11 e4 Bg4.
Returning to 8 Qe2:
8...e5! (Diagram 12)

Diagram 12 (W)
...Nc6 and ...Bd6 prepared 8...e5
28
 And why not? Just as against 8 Re1, this is the critical test of White’s move order.
NOTE: If Black is happy to meet the main line (8 dxc5 Bxc5 9 e4) with 9...Qc7 then here he may opt
for 8...Qc7 since White hasn’t anything better than 9 dxc5 Bxc5 10 e4 with a transposition.
That 8 Qe2 deserves some respect is shown by the variation 8...b6?! 9 e4!. The threat is 10 e5 and
9...dxe4?! 10 Nxe4 is already quite promising for White: the idea being to take advantage of the undefended
c6-knight to win a piece with 11 Nxf6+ and 12 Qe4. Meanwhile the d6-bishop is also vulnerable and Black is
in some trouble as was demonstrated by the fine game, M.Wildi-L.Mella, Zurich 1989: 10...Bb7 11 dxc5!
Bxc5 12 Nxf6+ (Diagram 13)

Diagram 13 (B)
12...Qxf6?? is a typical blunder

12...gxf6 (unfortunately for Black 12...Qxf6? 13 Bg5 runs his queen out of squares) 13 Rd1 Qc7 14 Bxh7+! (the presence of the pawn on
f6 and the strong alternative possibility of 14 Qe4 might well have deterred many from the classic sacrifice, but Wildi has seen deep into the
position) 14...Kxh7 15 Qe4+ (Diagram 14)

29
 Diagram 14 (B)
This deep Greek Gift is decisive

15...f5 (there was nothing better: for instance, White wins after 15...Kg7 16 Qg4+ Kh8 17 Bh6 Rg8 18 Qh4 or 16...Kh7 17 Ng5+! fxg5 18
Qh5+ Kg7 19 Qxg5+ Kh8 20 Rd3 – Bronznik) 16 Qh4+ Kg8 17 Qg5+ Kh7 18 Qh6+ Kg8 19 Ng5 f6 20 Nxe6 Qf7 21 Rd3! and the threat of
mate saw White regain his material with a large advantage. This variation might well tempt one to try 8 Qe2 in a rapid game, but unfortunately
for White 8...b6 is fine if Black’s bishop is back on e7 since 9 e4 doesn’t then threaten a piece-winning fork.
With the bishop on d6, a rare but interesting alternative is Neikirkh and Tsvetkov’s suggestion of
8...Nd7!?.
WARNING: No matter which of Black’s knights is on d7, White should usually refrain from
exchanging on c5 since ...Nxc5 enables Black to unravel with tempo against White’s key light-squared
bishop.
After 8...Nd7, play might continue 9 e4 cxd4 10 cxd4 Nb4 11 Bb1 (White really needs to keep this bishop
if he is to mount a successful attack) 11...b6 12 Re1 Ba6 13 Qe3 Rc8! (Bronznik) with good counterplay for
Black.
9 dxc5
Just as in the main line, this is really the only capture to consider. Instead 9 dxe5 Nxe5 10 Nxe5 Bxe5
leaves White the side looking to equalize, especially after 11 e4?! Qc7 12 h3 Re8 or 11 Nf3 Bc7 and Black
has aggressive options available in ...Qd6 and ...Bg4.
9...Bxc5 10 e4 (Diagram 15)

30
 Diagram 15 (B)
White counters in the centre

10...Be6
Preparing to recapture with the bishop in the event of an exchange on d5, and White doesn’t have a good
alternative way to develop his queenside. Black can also consider:
a) 10...Bg4 11 exd5 Qxd5 12 Ne4 (Diagram 16),

Diagram 16 (B)
Accurate play will equalize for Black

first played by Colle, is a variation with which White has enjoyed some practical success, although Black
should be fine after either 12...Nxe4 13 Bxe4 Qe6! or the more ambitious 12...Rad8!? 13 Nxf6+ gxf6 14 Be4
Qe6 15 h3 Bh5 16 Qc2 Bg6! (H.Polzer-R.Holmberg, correspondence 1999).
31
 NOTE: Both Polzer and Holmberg were rated over 2450 when this game was played and the
correspondence world (see, for example, Tim Harding’s Mega Corr 4 CD) can be a good source of ideas
for the studious player.
b) 10...dxe4 11 Nxe4 Be7 is fairly accommodating since White has been allowed to unravel. Nevertheless,
Black remains quite solid and obtained at least equality after 12 Rd1 Qc7 13 Nxf6+ (relying on piece play
with the more aggressive 13 Nfg5!? is probably a better try; White retains some awkward pressure after, for
example, 13...g6 14 Nxf6+ Bxf6 15 Be3 Ne7 16 Ne4 Bg7 17 Bc5! or 15...Bd7 16 Qf3!) 13...Bxf6 14 Bg5 Be7
15 Qe4 f5! 16 Qd5+ Kh8 17 Bxe7 Qxe7 and 18...e4 in E.Limp-G.Vescovi, Americana 2000.
c) 10...d4!? is a principled if double-edged advance, ruling out exd5 and seizing space in the centre. This
also appears quite playable for Black so long as he meets 11 Nb3 Bb6 12 cxd4 with the correct recapture
12...exd4!, intending to create some problems down the e-file and 13 e5 Re8 14 Bg5 Bg4 15 Bxf6
(W.Schraufl-H.Horther, Fuerth 2002) 15...gxf6 16 Qe4 f5 17 Qf4 Bxf3 18 Qxf3 Rxe5 shouldn’t give White
anything despite Black’s split kingside.
11 exd5
The only other option is to attack on the queenside with 11 b4?! Bb6 12 b5 Ne7, but this fails to convince:
13 Ba3 runs into 13...Ng6! and 13 Nxe5 Qc7 14 exd5 Nexd5 left White a little overextended in M.Tratar-
D.Sermek, Groningen 1993.
11...Bxd5 12 Ne4 Nxe4 13 Bxe4 (Diagram 17)

Diagram 17 (B)
Control of e4 is useful

White hopes to gain some pressure against e5 and then to either advance his queenside majority or to
attack on the kingside. However, the e5-pawn controls some useful squares and Black has good chances to
fully equalize. Possibly he should settle for the slightly meek 13...Bxe4 14 Qxe4 h6 15 b4 Bd6, as in
J.Krouzel-M.Rybak, Czech League 1996, although White could now try the critical 16 Nh4!?. In practice
Black has often preferred the active 13...f5!? 14 Bxd5+ Qxd5 15 Rd1 Qe4!, but after 16 Qb5! he must avoid
16...Bb6?! 17 Re1 Qc2 18 Be3 Bxe3?! 19 Re2! Qe4 20 Rxe3 as in the stem game F.Saemisch-S.Rosselli del
Turco, Baden-Baden 1925. Instead Bronznik has suggested that Black might continue in active vein with
16...Rad8!, although now I wonder about 17 Qb3+ Kh8 18 Re1! Qg4! (both 18...Qd3 19 Bg5 Rde8 20 Rad1
Qa6 21 Rd7 and 18...Qd5 19 Bg5 Qxb3 20 axb3 give White an edge) 19 Qxb7 (Diagram 18)

32
 Diagram 18 (B)
The struggle critically flares up

19...Qg6 20 b4 Bb6 when Black is well co-ordinated and has a dangerous central pawn duo, but White is a pawn up and may be able to beat
off the coming attack.

Theoretical Conclusion
8 dxc5 is the main line for the reason that it’s the most challenging move. With 8 Qe2, White aims to avoid some of Black’s 9th move options
in the main line, but these are probably not worth avoiding. The unbalanced positions after 8 Qe2 e5 give both sides opportunity to outplay a
weaker or inexperienced opponent, but objectively Black is fine and he can also consider the rare but promising 8...Nd7.

Illustrative Games

Game 1
E.Colle-Delvaux
Terneuzen 1929

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 (Diagram 19)

33
 Diagram 19 (B)
Facilitating Bc2 if needed

The correct move order. Instead 5 0-0 makes 5...c4!? a much more enticing prospect for Black and one
which has caught the eye of some strong players. A fairly recent example continued 6 Be2 b5! (Diagram 20)

Diagram 20 (W)
An interesting black option

7 Nbd2?! Bb7 8 Ne5 Nbd7 9 f4 Be7 10 Ndf3 Ne4 and White was already struggling, before he made
matters worse with the weakening 11 a4?! b4 in A.Adly-S.Rublevsky, FIDE World Championship, Tripoli
2004.
NOTE: The Stonewall set-up with Ne5 and f4 can lead to promising attacking possibilities for White,
but he must be aware that such a set-up is positionally committal: while White can never regain pawn
34
control of e4, Black can take control of the e5-square with a well-timed ...f6.
Adly lacked any real counterplay on the kingside and White does better to challenge on the queenside with
7 b3!, although Black should still be OK in an tough struggle; for example, 7...Bb7 8 a4 (8 bxc4 bxc4 9 Nc3
Nbd7 10 Rb1 Bc6 11 e4! dxe4 12 Nd2 Qa5! 13 Bb2 Nb6 was much more unbalanced but also about even in
the exciting clash G.Dizdar-M.Chandler, Jurmala 1983) 8...a6 9 axb5 axb5 10 Rxa8 Bxa8 11 bxc4 bxc4 12
Nc3 Nbd7 13 Bd2 Be7 14 Qa1 0-0 and neither side was able to get anywhere down the open queenside files
in G.Maróczy-A.Nimzowitsch, Gothenburg 1920.
5...Nc6 6 Nbd2 c4?
This advance is quite popular with rather inexperienced juniors. They usually envisage a queenside pawn
storm, completely forgetting that White also gets to both move and advance.
NOTE: Once White has played c3, ...c5-c4 is an instructive mistake. It prematurely releases the
central tension, is too slow and greatly strengthens the e4-break, since an exchange on e4 would leave
c4 rather weak.
7 Bc2
Of course, and this was the point of 5 c3.
7...Be7 8 0-0 b5
Continuing Black’s plan and he might as well do this. Otherwise after 8...0-0 9 e4 White threatens 10 e5
and Black lacks counterplay in the resulting French positions.
9 e4 (Diagram 21)

Diagram 21 (B)
White’s ideal central break

9...dxe4
Delvaux is understandably concerned about the impending attack, but I suspect Black has more chances to
resist after 9...0-0 10 e5 Nd7, although that is a relative statement! White will re-route his knight into the
attack via f1 and so in W.Hoffmann-K.Kreis, Bruchkoebel 1993, Black tried to counter in French style with
11 Re1 a5 (Black doesn’t need this preparatory move, but after 11...b4 Bronznik offers a cute and instructive
possible continuation: 12 Nf1 bxc3 13 bxc3 Qa5 14 Qd2! Ba6 15 Ng3 Rfb8 16 Nh5 Nf8? 17 Nxg7! Kxg7 18
Qh6+ Kg8 19 Bg5 and wins) 12 Nf1 f6, but after 13 exf6 Nxf6, as well as the game’s 14 Ne5, 14 Ng5 looks
pretty strong since the only way to defend e6 is the disgusting 14...Qd6 15 Qe2 Nd8.
10 Nxe4 0-0
35
 Harvey wonders whether ‘Black would do better to avoid castling’. I too suspect that he would, but after
10...Bb7 11 Qe2 it’s not so easy for Black to find a good plan. White, on the other hand, can try to crash
through on e6 or f7 after bringing a knight to g5, and can also weigh up the a4- and b3-breaks.
11 Qe2
Once again White targets h7; the threat being the fairly common tactic 12 Nxf6+ Bxf6 13 Qe4, winning a
piece.
11...Bb7 12 Nfg5! (Diagram 22)

Diagram 22 (B)
An omnipotent knight on g5

12...h6?
Black had to cover h7, but this weak move merely leads to further weaknesses. Seeing as Black shortly
plays ...g6 in any case, he was much better off with 12...g6 when there isn’t anything decisive, although
White retains a rather pleasant advantage after 13 Nxf6+ Bxf6 14 Ne4 Be7 15 Bh6 Re8 16 Rfe1.
13 Nxf6+ Bxf6 14 Qe4 g6
From here on Delvaux rather decides that he might as well be remembered as the loser of a fine miniature
than continue to grimly suffer after the marginally superior 14...Re8 15 Qh7+ Kf8 16 Ne4.
15 Nxe6! (Diagram 23)

36
 Diagram 23 (B)
A typical sacrificial breakthrough

Undoubtedly one of the main attractions for Colle to the opening which bears his name was the
opportunity to land such blows.
15...fxe6? 16 Qxg6+ Bg7 17 Qh7+ Kf7 18 Bg6+ Kf6 19 Bh5! (Diagram 24)

Diagram 24 (B)
There’s no defence

Quite possibly Delvaux overlooked this neat switchback when accepting the piece on e6. White’s threats
on the kingside now quickly decide:
19...Ne7 20 Bxh6 Rg8
Or 20...Bxh6 21 Qxh6+ Kf5 22 g4+! and the black king is driven forwards to its imminent demise.
37
 21 h4!
Keeping up the threats and Black now allows the game to terminate immediately. However, there hasn’t
been any good defence for several moves and here, for instance, 21...e5 22 Bxg7+ Rxg7 23 dxe5+ costs
Black a rook.
21...Bxh6? 22 Qf7 mate 1-0
A crushing victory from the man himself.

Points to Remember
1. White should meet 4 Bd3 c5 with 5 c3, not the slightly inaccurate 5 0-0.
2. ...c5-c4 is simply a badly misguided strategy – Black is too slow on the queenside.
3. White should react by aiming to force a favourable French with an e4-e5 advance. Black can avoid that with an exchange on e4 when
White should look for tactics not just against h7, but also e6 and f7.

Game 2
E.Gausel-K.Lie
Norwegian Championship, Asker 2000

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 cxd4?!

Once again the tension proves too much for Black to handle and he prematurely releases it.
7 exd4 (Diagram 25)

Diagram 25 (B)
White has the more active bishops

We now have the so-called ‘Carlsbad structure’ on the board. A structure which the reader may well have
seen arise in some other openings, and most notably the Exchange Variation of the Queen’s Gambit Declined.
This Colle version of the Carlsbad is pretty pleasant for White. Compare it with the QGD version: 1 d4 d5 2
c4 e6 3 Nc3 Nf6 4 cxd5 exd5 5 Bg5 Be7 6 e3 c6 7 Bd3 0-0 8 Qc2 Nbd7 9 Nf3 Re8 10 0-0 Nf8 (Diagram 26).

38
 Diagram 26 (W)
Our Colle structure in reverse

In that case White’s c1-bishop is actively developed outside the pawn chain and he can consider carrying
out a minority attack with b4-b5.
TIP: A ‘minority attack’ is the name given to a situation in which the side with just two pawns on a
flank advances them against the opponent’s three, aiming to create a weakness. In the Exchange QGD
version, for example, should White achieve b4-b5xc6 bxc6, he gains a useful outpost on c5, while Black
is left with a weak backwards pawn on c6.
In our Colle case not only is Black some way from arranging ...b5-b4, but he is also stuck with a bad light-
squared bishop, blunted by the e6-pawn. White on the other hand has free development, can develop his dark-
squared bishop much more easily than in the main line and enjoys a very handy outpost on e5.
7...Bd6 8 0-0 0-0
Black has also been known to panic here about White’s imminent kingside attack and play 8...Qc7 9 Re1
Bd7 10 Qe2 0-0-0?!, but this is a case of jumping from the frying pan into the fire. Not only does Black lack
counterplay, but White has the straightforward and very effective plan of pushing his queenside pawns; for
example, 11 Ne5 Be8 (or 11...Rdf8 12 Ndf3 Kb8 13 a4 Bc8 14 b4 Ne8 15 Nxc6+ Qxc6 16 Bd2 Nc7 17 b5
Qd7 18 b6! and Black was crushed in M.Golmayo de la Torriente-L.Prins, Sitges 1934) 12 Ndf3 (Diagram
27)

39
 Diagram 27 (B)
White will push his queenside pawns

12...Nd7 13 Nxc6 (Koltanowski’s idea of 13 Bd2 Nf8 14 c4!? is also quite dangerous) 13...Qxc6
(A.Brown-P.Thomas, Scotland 2004) 14 a4 Kb8 15 b4! Nf6 (15...Qxc3? 16 Bd2 Qc8 17 Rec1 Bc7 18 a5
merely opens lines for the attack) 16 Bd2 h5 17 b5 Qc7 18 a5 and White has a very strong and much the
faster attack.
NOTE: This game actually began via the move order 1 d4 Nf6 2 Nf3 c5 3 e3 e6 4 Bd3 d5 5 c3 Bd6 6
Nbd2 cxd4 7 exd4 Nc6 8 0-0 0-0, but I’ve standardized it for the reader’s convenience.
9 Re1 (Diagram 28)

Diagram 28 (B)
The e5-outpost beckons

40
 Preparing to advance the knight to e5, but Black resists...
9...Qc7 10 Qe2
...and so White tries again.
10...Nd7!
Such an undeveloping move may appear a little strange, but Lie correctly realizes that the fight for control
of e5 must be at the top of both sides’ priorities. Instead 10...Re8?! 11 Ne5 (Diagram 29)

Diagram 29 (B)
Hoping for 11...Nd7? 12 Bxh7+!

returns us to a position we considered in Baird-Hanham in the Introduction. We considered there both

11...Nd7? and 11...Bxe5, but Black might also try covering f7 with the prophylactic 11...Re7, as he did in
E.Colle-V.Soultanbeieff, Liege 1930. Objectively White should continue 12 f4, followed by Ndf3 and a
kingside attack, quite possibly involving either Ng5 or an advance of the g-pawn. However, Colle couldn’t
resist the lure of the sacrifice and preferred 12 Ndf3 Nd7 13 Ng5!? Nf8 14 Nxh7!? Nxh7 15 Bxh7+ Kxh7 16
Qh5+ Kg8 17 Re3 (Diagram 30)

41
 Diagram 30 (B)
How should Black defend?

and now before we condemn his defence, we should try putting ourselves in the no doubt stunned Soultanbeieff’s shoes. Cover up the
following variations and spend some time trying to work out exactly what’s going on (it’s a good calculation exercise). Undoubtedly White has
a strong attack and Black needs to be pretty careful:
a) 17...Re8? 18 Rh3 Kf8 was rather too simplistic a defence from the Russian. Colle now triumphed in
style with 19 Bg5 f6 20 Bxf6! (Diagram 31)

Diagram 31 (B)
A neat finish from Colle

and 1-0 since mate follows after 20...gxf6 21 Ng6+.

b) 17...g6?! also doesn’t help and allows White to target the f6-square: 18 Qh6 f5! (this goes unmentioned
by both Smith and Hall, and Harvey, but appears to be the best defence; the win is much more routine after
42
18...Re8 19 Ng4! f5 20 Rh3 Qg7 21 Nf6+! – Smith and Hall) 19 Rh3 Bxe5 20 dxe5 Qxe5 21 Qxg6+ Qg7 22
Qh5 (Diagram 32)

Diagram 32 (B)
Black must return material

22...Ne5 (22...Qh7 looks more natural, but is rebuffed by the machine’s fiendish 23 Rg3+ Kh8 24 Qf3 Qf7 25 Bg5 when Black is
defenceless; Black might consider too 22...f4!? 23 Bxf4 e5, but after 24 Rg3 exf4 25 Rxg7+ Kxg7 26 Qg5+ Kh7 27 Qxf4 White’s queen and
three connected passed pawns should gradually overcome Black’s rook and two pieces) 23 Rg3 Ng4 and White regains his piece, leaving him a
pawn ahead albeit in a slightly messy position after something like 24 h3 e5 25 Bh6 Qh7 26 hxg4 fxg4 27 f3 Bf5 28 fxg4 Bg6 29 Qg5 Re6.
c) 17...Bxe5!? 18 dxe5 Nxe5! (this obvious attempt to return the piece also goes unmentioned by previous
annotators, but appears forced: 18...Re8? is crushed by Smith and Hall’s 19 Rh3 Kf8 20 Bg5 Ne7 21 Bf6!,
while 18...g6? 19 Qh6 Re8 loses to the leisurely 20 Bd2 followed by 21 Rae1 and only then 22 Rh3 or 21...f5
22 Rg3) 19 Rh3 (White can also regain the piece with 19 Rxe5, but after 19...f6 20 Re3 e5 Black’s strong
centre at least compensates for his absent h-pawn) 19...f6 and White hasn’t anything better than perpetual
check.
d) 17...f6! 18 Rh3 Nxe5! (but not 18...fxe5?? 19 Qh8+ Kf7 20 Rf3+ Kg6 21 h4 and 22 h5mate) 19 Qh8+
Kf7 20 Rh7 Kg6! (Diagram 33)

43
 Diagram 33 (W)
White can only repeat

(Reinfeld) is the traditional defence given when White has nothing better than a repetition with 21 Rh4 Kf7
22 Rh7.
Not the most theoretically relevant of games, but one which shows how tough the pressure can be for
Black to handle when White goes all in on the kingside, as well as the attacking menace created by knights on
e5 and g5 (remember Colle had plenty of alternatives to sacrificing on move 14).
Returning to the superior defence, 10...Nd7:
11 c4! (Diagram 34)

Diagram 34 (B)
Changing tack with Ne5 unavailable

44
 Switching plans in a bid to exploit the congestion in Black’s camp. The Colle is an opening in which
White’s play is often thematic, but sometimes he must change his plan should Black try a radical or very
creative defence.
NOTE: Not for nothing did Koltanowski like to say: ‘You do not have to be orthodox about sticking
completely to the Colle line if your opponent varies.’
White can also continue in natural Colle style with 11 Nf1, but after 11...Re8! Black threatens to gain
counterplay with ...e5. White can prevent that with 12 Ne3?! (Tim Taylor’s 12 h3 e5 13 dxe5 Ndxe5 14 Bf4
is probably White’s best try for a small edge – again we can see White having to change plans; he can only
play for an attack when he controls the e5-square), but after 12...Nf6! he lacked a good continuation in
Y.Anokhin-A.Kalantar, Batumi 1961.
Gausel’s 11 c4 is designed to dissuade Black from ...e5. Admittedly 11...e5!? is still possible, but after 12
cxd5! Nxd4 13 Nxd4 exd4 14 Qh5!? (Bronznik), White has the initiative.
11...Nf6
Now that White has weakened the defence of d4, Black returns his knight to its most natural square. This
may appear a little slow, but Black doesn’t appear to be able to equalize in any case:
a) 11...b6 12 b3 Bb7 13 Bb2 Nb4?! (Black likes to make use of the b4-square in this variation, but
especially doing so here only plays into White’s hands) 14 Bb1 Rac8 15 a3 Nc6 16 Qd3! and White was a
little better due to his powerful queen and bishop battery in A.Perez de Miguel-A.Gerrikabeitia Uriarte,
Ourense 2006.
b) 11...dxc4 12 Nxc4 Bb4 13 Rd1! (correctly avoiding any exchanges) 13...Nf6 14 Bg5 Be7 15 Rac1
(Diagram 35)

Diagram 35 (B)
White has a grip on the position

gave White quite a good IQP set-up in G.Meszaros-S.Kovacevic, Harkany 2001; the c4-knight can
increase the pressure by heading to e5 or possibly even to e3 (supporting a d5-advance), while Bb1 and Qd3
should force a kingside weakness.
c) 11...Be7?! 12 b3 Bf6 13 Bb2 (13 cxd5!? exd5 14 Ba3 Rd8 15 Rac1 Qa5 16 Nb1! is also promising)
13...Qf4!? was a creative defence in R.Simon-T.Sarbok, German League 2001, but White’s superior
development saw him retain a comfortable edge after 14 g3! Qh6 15 Rad1.
12 b3 Bb4?!
45
 Such potential hanging pawn positions are never easy for Black if he is pretty inexperienced in them.
Whereas White usually intends to increase the pressure with Rad1 (or Rac1) and Ne5, Black has all manner
of possible piece configurations to employ. Lie decides to undertake active operations to counter White’s
harmonious Zukertort set-up, but they fail to fully convince. Instead White would have only been a little
better after 12...b6 13 Bb2 Bb7 14 Rad1 Rac8 15 Ne5 when Black remains pretty solid and can begin
counterplay with 15...Qe7, intending ...Rfd8 and ...Bb4.
13 Bb2 (Diagram 36)

Diagram 36 (B)
It’s Zukertortesque!

13...Qf4 14 Rad1 b6 15 a3 Bxd2!

Ceding the bishop pair might appear a little drastic, but Black hopes to undermine White’s defence of c4.
Furthermore, 15...Be7 could be met by either 16 b4 dxc4 17 Nxc4, when we again see a knight flexibly
placed on c4 and controlling a number of useful squares, or by 16 Ne5!?, hoping for 16...Nxd4? 17 Bxd4
Qxd4 18 Nc6 Qc5 19 cxd5, exploiting White’s superior development and winning material after 19...Nxd5?
20 Rc1 Qd6 21 Qe4 g6 22 Nxe7+ Nxe7 23 Qxa8 Qxd3 24 Qxa7.
16 Rxd2 Na5 17 Bc2 (Diagram 37)

46
 Diagram 37 (B)
...Ba6 is a useful option

17...dxc4
Perhaps an immediate 17...Ba6!? was superior since 18 Bd3 is now impossible. Possibly Lie was worried
by 18 Bc3, but 18...dxc4 19 Bxa5 bxa5 20 bxc4 Rfd8 21 Bd3 Qc7! doesn’t appear too bad for the second
player. White’s problem is that for the time being he cannot easily advance the hanging pawns. Indeed 22 c5?
would be a horrible move, ceding control of the d5-square and further weakening the d4-pawn.
NOTE: The term ‘hanging pawns’ refers to two pawns abreast with an open file on either side (i.e.
they aren’t supported by any other pawns).
The side with the hanging pawns can sometimes use them to powerfully break through in the centre (but
here 22 d5? exd5 23 cxd5 Bxd3 24 Rxd3 Qc5 costs White his d-pawn), or he may exploit the extra space they
grant him to launch a kingside attack. On the other hand, the side opposing the hanging pawns (Black in this
case) needs to prevent a powerful d4-d5 advance and aim to tie down White’s pieces to the defence of the
pawns.
18 bxc4 Bb7
Much more consistent was still 18...Ba6 when 19 Bd3 Rac8 20 Rc2 Rfd8 21 Ne5! reaches a critical
position in which we can see how useful White’s dark-squared bishop is in comparison to the note to Black’s
17th. 22 g3 will now drive Black’s queen offside and there are also nasty ideas of d5 and Nxf7 in the air; for
example, the ever-thorough Bronznik analyses 21...Nc6? 22 g3 (22 d5 also looks quite promising, but the text
is very strong) 22...Qh6 23 Nxf7! Kxf7 24 Qxe6+ Kf8 25 a4! with a huge attack.
19 Ne5 Rac8 (Diagram 38)

47
 Diagram 38 (W)
Spot a typical breakthrough!

20 Bc1?!
Rightly queried by Bronznik who points out that White had a much more thematic continuation available
in 20 d5! (the classic breakthrough in such positions, unfurling White’s powerful dark-squared bishop)
20...exd5 (20...Nxc4!? 21 Nxc4 Qxc4 22 Qxc4 Rxc4 23 dxe6 fxe6 24 Bb3 is also much better for White with
his bishop pair on an open board) 21 Rd4 Qh6 22 cxd5 Bxd5 (or 22...Nxd5 when White has a number of
tempting and strong continuations, including Bronznik’s 23 Bc1 Qf6 24 Qd3 g6 25 Nd7 Qc6 26 Rxd5!) 23
Ng4 (Diagram 39)

Diagram 39 (B)
All White’s pieces are attacking

48
 23...Nxg4 24 Rxg4 f6 25 Bc1 g5 26 h4 and Black’s kingside is unlikely to hold out for too much longer.
20...Nxc4??
In his desire to gain counterplay, Lie misses an important tactic. He has, though, never really looked at
home as Black in this game. Indeed aggressive players have often been known to struggle when facing the
Colle: Black needs to first patiently neutralize White’s aggressive ideas in the early middlegame; usually only
then can he look to undertake active operations himself.
Instead 20...Qh4 would have kept White’s advantage to a minimum; Black’s queen may become a tactical
liability on the kingside, but she can also be an effective defender. Now 21 Rdd1 is the sane continuation
instead of the rather messy 21 Rd3!? Nxc4 22 Nd7!? Nxd7!? (22...Rfd8 23 Rh3 Qxd4 24 Nxf6+ Qxf6 25
Bxh7+ Kf8 26 Bd3 is a little more comfortable for White due to his safer king) 23 Rh3 Qxh3 24 gxh3 Rfe8.
Black has good positional compensation for the queen here, but must play accurately to prevent White from
inflicting a kingside weakening (25 Qd3 is thus best met by 25...Nf8).
21 Nxc4 Rxc4 22 Qxc4 Ng4 23 g3 Qf3
The point of Black’s combination, but he had evidently missed one of White’s next two moves.
24 d5!
The ideal IQP advance; decisive in this case, albeit played for rather unusual reasons.
24...Bxd5 25 Be4 (Diagram 40)

Diagram 40 (B)
Defending and winning

25...Ne5 26 Bxd5 exd5 27 Qxd5 1-0

Points to Remember
1. An early exchange on d4 weakens Black’s control of the key e5-square.
2. White should then aim to quickly occupy e5 with a knight, intending to launch a kingside attack
featuring such moves as f4, Nd2-f3-g5 and possibly Bxh7+.
3. The best defence to the threat of 11 Ne5 is 10...Nd7 when White should probably change tack with 11
c4.
4. Hanging pawn positions require care and tactical alertness from both players. White needs to keep any such pawns defended, while
looking to either attack on the kingside or to carry out a thematic d4-d5 breakthrough.

Game 3
T.Straeter-J.Hladik

49
Pardubice (Game 3) 2006

1 d4 Nf6 2 Nf3 e6 3 e3 c5 4 Bd3 d5 5 c3 Nc6 6 Nbd2 Bd6

WARNING: White can only aim for an early IQP position when Black develops his bishop to d6:
6...Be7 7 0-0 0-0 (Diagram 41)

Diagram 41 (W)
The IQP approach isn’t possible

8 e4? simply costs a pawn after 8...dxe4 9 Nxe4 cxd4 10 cxd4 Nxd4.

Those determined for whatever reason to avoid the main line can consider meeting 6...Be7 7 0-0 0-0 with
either 8 Qe2 or the Stonewall approach, 8 Ne5. Objectively, though, neither is particularly testing against all
but fairly inexperienced opponents; for example, against the latter, the solid 8...Nd7 9 f4 f5! is fine for Black.
7 0-0 0-0 8 e4!?
Playing for an IQP position and a kingside attack. Surprisingly this option wasn’t considered by Watson
and Schiller in their anti-Colle coverage (as part of their How to Succeed in the Queen Pawn Openings), and
it continues to catch out some black players.
8...cxd4
The correct way for Black to begin to liquidate the centre. Instead 8...dxe4 9 Nxe4 Nxe4 10 Bxe4 cxd4 11
cxd4 (and not 11 Bxh7+? Kxh7 12 Ng5+ Kg6! when Black comfortably defends with ...f5) is a little better for
White since Black lacks any kingside defenders, but even worse for him is 9...cxd4?! 10 Nxf6+! (Diagram
42)

50
 Diagram 42 (B)
Black lacks a good recapture

10...gxf6 (10...Qxf6?? 11 Bg5 nets a whole queen) 11 Bh6 Re8 12 Nxd4 f5. Then 13 Qh5!? Nxd4 14 cxd4
gave White attacking prospects in P.Romanovsky-G.Rotlewi, St Petersburg 1909, but objectively he should
probably prefer something calmer having gained a structural advantage.
9 cxd4 dxe4
Before inflicting the IQP on White, Black can also bring his knight closer to the ideal blockading square on
d5 with 9...Nb4, as recommended by Rizzitano. However, he then overlooks after 10 Bb1 dxe4 11 Nxe4 Be7
(Romanovsky’s old suggestion of 11...Nbd5!? is rarely played but still seems quite reasonable; Black doesn’t
mind an exchange on d6 since he can then avoid being forced to play ...g6 after 12 Nxd6 Qxd6 13 Qd3 b6 14
Bg5), the possibility of 12 Nc3 transposing to our main game. White probably should keep pieces on in such
a manner and instead 12 Nxf6+ Bxf6 13 a3 Nd5 14 Qd3 g6 15 Bh6 Bg7 16 Bg5 Qd6 17 Re1 b6 18 Ne5 Bb7
(Diagram 43)

51
 Diagram 43 (W)
Model defence from Black

19 Qh3 Rac8 20 Be4 Rc7 saw him unable to generate anything against Black’s model defence in J.Cobb-
B.Lalic, British League 2000.

IQP positions tend to be quite divisive amongst players: some love to play with an IQP; others can’t think
of anything worse. The same applies to players facing an IQP and should Black not wish to do so, he might
prefer to detonate the centre with 9...e5!? (Diagram 44).

Diagram 44 (W)
A radical alternative to the IQP

Now 10 exd5 Nxd4 11 Nxd4 exd4 doesn’t lead anywhere for White, so he usually prefers a double
52
exchange on e5: 10 Nxe5 Nxe5 11 dxe5 Bxe5 12 exd5 Qxd5 13 Nf3 Bb8!? (13...h6?! 14 Nxe5 Qxe5 15 Re1
Qd5 16 Bc2 Qxd1 17 Bxd1 Be6 18 Bf3 Bd5 19 Bxd5 Nxd5 20 Bd2 gave White a very pleasant endgame
advantage with bishop against knight in J.Nogueiras-Y.Gonzalez, Santa Clara 2005, although Black
eventually held the draw after a gruelling 75-move defence) 14 Bg5 Bf5! (Black was seduced by 14...Bg4?!
15 Bxf6 Qh5!? in M.Ashley-G.Vescovi, Bermuda 1995, but was forced to play accurately to hold a pawn
down after 16 Bh4 Bxf3 17 Qxf3 Qxh4 18 g3 Qf6 19 Qxb7) 15 Bxf5 Qxf5 16 Qd4 Ng4! (16...Bc7!? is also
possible, relying on the bishop versus knight advantage to offset Black’s split kingside after an exchange on
f6, but the text is the simplest equalizer, threatening to capture on h2) 17 Be7 Re8 and if anyone needs to be
careful to ensure full equality, it’s now White; for example, 18 Rfe1 Bc7 19 h3? runs into 19...Nxf2!.
10 Nxe4 Be7!
Preventing the threat of 11 Nxf6+ and unfurling an attack on d4 which White parries by clearing the d3-
bishop’s path.
11 Nc3! (Diagram 45)

Diagram 45 (B)
An unbalanced structure

TIP: When playing with an IQP, always try to ensure that any exchanges force some sort of
concession from the opponent.
Here an exchange on f6 doesn’t and so Straeter retreats.
11...Nb4
Even if White plays passively, it takes Black a very long time to win the IQP. In the short term, more
important than the theoretical weakness of the pawn itself, is the fact that White has lost control of the d5-
square and Hladik exploits that in typical fashion.
12 Bb1
Setting up a battery on the long diagonal is probably the best try. Instead the instructive 12 Bc4 a6! 13 Bg5
b5 14 Bb3 Bb7 15 Ne5 Nbd5 16 Qe2 Nxc3! (transferring the weakness from d4 to c3 is a fairly common
ploy) 17 bxc3 Rc8 18 Rac1 Nd5 left Black slightly better in G.Werner-Y.Boidman, Hassloch 2006.
12...Bd7!?
An interesting way of bringing the bishop to the long diagonal without slightly weakening b5 and a6 as the
usual 12...b6 does. Play may then continue 13 Re1 Bb7 14 Ne5 (Diagram 46)

53
 Diagram 46 (B)
White wants to attack

and now:
a) 14...Nbd5?! is natural, but a little too compliant. J.Plaskett-N.Short, British League 2000 (a game which
incidentally arose from a French – IQP positions are full of such transpositions between different openings)
continued 15 Qd3 Rc8 16 Qh3! (stronger than an immediate 16 Bg5 since White may now be able to deflect
the f6-knight with Ng4, while the queen also supports tactics against e6 and f7) 16...Nxc3 17 bxc3 Qd5 18
Bd2 g6?! (Black should rarely make such an advance voluntarily) 19 Bc2 Qa5 and now Plaskett recommends
20 Bb3! (Diagram 47)

Diagram 47 (B)
A dangerous white set-up

54
 with advantage to White: he threatens various nasty sacrifices on the light squares; may also be able to
advance with Rad1, c4 and d5; and Black lacks counterplay for his slightly shaky king position. Black might
try to play for exchanges with 20...Bd5?!, but after 21 Qh6 a rook lift is threatened and 21...Ne4? fails to 22
Rxe4! Bxe4 23 Re1 Bd5 24 Re3! (Plaskett) with a winning attack.
b) 14...Nc6!? breaks with Black’s development scheme to awkwardly attack d4 and threaten an exchange
of knights.
c) 14...Rc8 also appears fine for Black; for example, 15 Bg5 Nbd5 16 Qd3 g6 17 Bh6 Re8 18 a3 Nxc3 19
bxc3 Qd5 20 f3 Nd7 was about equal in B.D.Smith-K.Sakai, correspondence 2002 (note that 21 Nxf7!? Kxf7
22 Ba2 isn’t decisive due to 22...Nc5).
13 a3 Nbd5 14 Qd3 Bc6 15 Re1 (Diagram 48)

Diagram 48 (B)
White will force a weakening ...g6

15...g6?!
A small inaccuracy as there was no need to hurry with this. Instead 15...Rc8 16 Bg5 g6 17 Bh6 Re8 would
have left Black a useful tempo up on the game. I suspect Straeter would have preferred 16 Ne5 when 16...Re8
(and not 16...Qb6?! 17 Nxd5! Bxd5? 18 Nd7) is a useful move to both cover e6 and prepare ...g6 and ...Bg7.
16 Bh6 Re8 17 Ne5 Bf8 18 Bg5!?
Keeping pieces on in a bid to maintain some pressure. Quite possibly it was, though, slightly preferable to
exchange on f8 when White can aim to exploit the weakened dark squares with Qd2-h6.
NOTE: An exchange on c6 with 18 Nxc6 bxc6 does not favour White, despite winning the bishop
pair and splitting Black’s queenside. The c6-pawn is not at all easy to attack and White will continue to
struggle to find a good role for his light-squared bishop, while Black may place White’s b-pawn under
some pressure.
18...Bg7 19 Ba2 Qd6 20 Qh3 (Diagram 49)

55
 Diagram 49 (B)
Black remains pretty solid...

White’s pieces now occupy fairly aggressive squares, but Black remains pretty solid...
20...Ne7?
...or rather he did. Hladik wants to meet 21 Qh4 with 21...Nf5, but unfortunately he underestimates the
potential power of White’s a2-bishop. Instead the position would have remained roughly equal after 20...Rac8
21 Qh4 Nxc3 22 bxc3 Nd5 23 Qg3 b5.
21 Nxf7! (Diagram 50)

Diagram 50 (B)
...but then falls for a typical blow

A standard IQP blow and already Black could pretty much resign since 21...Kxf7 22 Rxe6 is crushing.
56
 21...Qxd4 22 Qxe6 h5 23 Ne5+ Kh7 24 Bxf6 Ng8 25 Qf7 Nxf6 26 Qxg6+ Kh8 27 Nf7+ Kg8 28 Nh6+ Kh8 29 Qxg7+! 1-0

Points to Remember
1. Attacking players with some IQP experience may wish to consider 8 e4 as an alternative to 8 dxc5.
2. White can only play 8 e4 after 6...Bd6, and not against 6...Be7 when it just loses a pawn.
3. In any IQP position a Bb1/Bc2 and Qd3 battery is a useful way to force the weakening ...g6.
4. Especially when there’s a knight on e5, look out for sacrificial opportunities on f7 and e6.

White Plays 8 dxc5

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 (Diagram 51)

Diagram 51 (B)
White’s main continuation

White’s main choice, aiming to advance with e4-e5 and then to blow Black away on the kingside. Black’s
two main defences both address that plan: 9...dxe4 is the subject of our next section and the main line,
9...Qc7, that of the next chapter.
9...e5!?
From certain Colle sources, one would have thought that this was pretty much Black’s only alternative to
9...Qc7 and 9...dxe4. However, he actually has a whole raft of alternatives, many of which remain both quite
rare and rather unexplored:
a) 9...d4? 10 Nb3 wins the d-pawn, such as with 10...Bb6 11 e5 (Diagram 52)

57
 Diagram 52 (B)
White wins a pawn

11...Nd5 (and not, of course, 11...Ne8 12 cxd4 Bxd4?? 13 Nbxd4 Nxd4 14 Nxd4 Qxd4?? 15 Bxh7+) 12 Nbxd4.
b) 9...b6 ignores White’s ideas to continue Black’s development. Let’s see what happens if White simply
gets on with e4-e5: 10 e5 Nd7 11 Re1 and we reach the first of many French-type positions which we will
encounter. Black must, of course, avoid 11...Qc7?? 12 Bxh7+, but the thematic 11...f6 doesn’t appear too bad
for him; for example, 12 exf6 Nxf6 (12...Qxf6 13 Qe2 – and not the tempting 13 Bxh7+?! Kxh7 14 Rxe6
Bxf2+! 15 Kh1 Qf5 16 Rxc6 when White’s poorly co-ordinated pieces give Black dangerous compensation –
is probably a little better for White with the idea of 13...e5 14 b4! Be7 15 Bc2 and 16 Bb3) 13 Nb3 Bd6
(Diagram 53),

Diagram 53 (W)
Resembling a French Tarrasch

58
 although ...b6 isn’t especially useful once Black goes ...f6 and White is a little better after 14 Bg5 (with
Tarraschesque ideas of Bh4-g3, but an immediate 14 Nbd4 Nxd4 15 cxd4 is also a reasonable try for an edge)
14...Qc7 (and not 14...e5? 15 Bb5 – it’s never easy for Black to quickly advance his e-pawn in these
positions) 15 Bxf6! (15 Nbd4 Nxd4 16 cxd4 isn’t especially clear after 16...Ng4 or even 16...Ne4!?) 15...Rxf6
16 Nbd4 Bd7 17 Qe2 with a grip on the e5-square.
TIP: After 9...b6 White can immediately push with 10 e5, but he may in general find it helpful to
delay that advance until after 10 Qe2; a useful move which both covers f2 and helps to control the e5-
square in the event of an e5 f6; exf6 exchange.
After 9...b6, White might thus also opt for 10 Qe2 when E.Roblek-M.Vecek, correspondence 1997, for
example, continued 10...Bb7 (10...Qc7 is probably a better try, transposing to a position we’ll consider via
9...Qc7 10 Qe2 b6) 11 e5 Nd7 12 Nb3 Be7 13 Bf4 (Diagram 54)

Diagram 54 (B)
White has kingside chances

with a typical edge which turned into a useful attack after 13...Nc5 14 Nxc5! bxc5 15 b3 Qb6 16 Rfe1 a5
17 Ng5!.
c) 9...h6!? eliminates any notion of a Greek Gift after e5 and following 10 Qe2 Qc7 play has transposed to
the next chapter, as does 10...Bb6 11 Bc2!? Qc7.
WARNING: When he advances to e5, White must ensure that his pawn is well protected.
It usually is, but after 10...Bb6 11 e5? Nd7 12 Nb3 Bc7! 13 Bf4, 13...f6 is rather awkward, winning the e5-
pawn unless White tries the desperate and unconvincing 14 exf6 Bxf4 15 Qxe6+ Rf7 (A.Habib-E.Kengis,
Damascus 2003); for example, 16 Bg6 Nde5 17 Bxf7+ Nxf7 18 Qe2 Qxf6 and the two unopposed bishops far
outweigh White’s extra rook and pawn.
d) For 9...Be7 10 Qe2 Qc7 see the move order 9...Qc7 10 Qe2 Be7 in Chapter Two. Instead 10...b6 is very
similar to variation ‘b’ above. White can now close the centre immediately or try 11 Re1!?, aiming to deflect
Black’s light-squared bishop away from the defence of e6 so that Black finds it harder to counter e5 with
...f6; for example, 11...Bb7 12 e5 Nd7 13 Nb3 Re8 14 Nbd4 a6 15 Nxc6 Bxc6 16 Nd4 Bb7 17 Qg4 gave
White a pleasant edge in L.Fieschi-F.Grillo, Salsomaggiore Terme 2005. Here Black lacks counterplay
against White’s kingside build-up and will probably have to resort to ...f6 or ...f5 at some point, but then his
light-squared bishop is misplaced.
59
 e) 9...Bd6 10 Qe2 (Diagram 55)

Diagram 55 (B)
Pushing through e4-e5

10...Qc7 reaches another position we’ll deal with via 9...Qc7 – I’m afraid this section is full of
transpositions. Instead 10...Bc7!? (or 9...Bb6 10 Qe2 Bc7) is a radical plan to maintain control of the e5-
square. Since White can’t push e5 without losing his e-pawn, he should switch plans with Bronznik’s idea of
11 Rd1!?, preparing Nb3 and dissuading any notion of 11...e5?! in view of 12 exd5 Nxd5 13 Nc4.
f) 9...Re8 is another move which allows White to choose between waiting with 10 Qe2 and the immediate
10 e5. The former is Bronznik’s main subject, but the latter is also a reasonable try; for example, 10...Nd7 (or
10...Ng4!? 11 Qe2 f6 12 Nb3 Bb6 13 exf6 Nxf6 and now White should give serious thought to changing the
central structure with 14 c4!?; instead 14 Bf4? Bc7? 15 Ne5 Nd7? (Diagram 56)

Diagram 56 (W)
60
 What’s White’s best move?

16 Bxh7+! Kxh7 17 Qh5+ Kg8 18 Qf7+ Kh7 19 Nxc6 bxc6 20 Bxc7 was excellent for White in S.Marsh-
D.Henderson, Newcastle 1993, but 14...e5 would have been somewhat stronger) 11 Nb3 (or 11 Qe2!?,
intending to transpose after 11...Qc7 to a position we’ll consider via 9...Qc7 10 Qe2 Re8 11 e5 Nd7) 11...Be7
12 Re1 f6 (an essential break in this sort of position to both free Black’s position and break up White’s
kingside chances) 13 exf6 Bxf6 14 Qc2! (this is why 12 Re1 was preferred to 12 Qe2; Black is now a little
tangled and White slightly for choice) 14...Nf8 (and not 14...g6? 15 Bxg6! hxg6 16 Qxg6+ Bg7 17 Bh6 Re7
18 Bxg7 Rxg7 19 Qxe6+ Kh8 20 Qxd5 when White emerges with four pawns for the piece) 15 Nfd4 Bd7 16
Nxc6 Bxc6 17 Nd4 e5 18 Nxc6 bxc6 19 f3 Bg5 20 Bxg5 Qxg5 21 c4! gave White an edge in J.Lyon-
L.Cooper, correspondence 1989, but Black probably does better with 15...e5!?. Thus White might wish to
consider too 15 Rd1!?, aimed against ...e5 and intending to unravel after 15...Qe7 16 Bf1! with g3, Bg2 and
possibly Be3, continuing to restrain Black in the centre.
Returning to 9...e5 (Diagram 57):

Diagram 57 (W)
Black also seizes some space

10 exd5
This useful multipurpose move opens the d3-bishop’s path, while preparing to improve the d2-knight and
then to develop the dark-squared bishop. I believe it is also the most testing option at White’s disposal, but in
practice 10 Qe2 has actually been more popular, transposing to our main continuation after 8 Qe2 above
(8...e5 9 dxc5 Bxc5 10 e4).
10...Qxd5
One would be forgiven for thinking that Black’s active and fairly easy development (after either the text or
10...Nxd5 – the subject of Game 4) has already enabled him to equalize. A closer inspection reveals, though,
that Black’s pieces are on such squares as to permit White to drive them backwards as he develops with
tempo. Following either 11 Qc2 or 11 Qe2 White prepares Ne4 hitting the c5-bishop, after which one
harmonious option is to complete development with Bg5 and Rad1. Whether this is enough to give White a
definite edge is not yet totally clear, but certainly Black will need to find some accurate moves to remain on a
roughly even keel.
11 Qc2! (Diagram 58)
61
 Diagram 58 (B)
Eyeing up Black’s h-pawn

Once again 11 Qe2 is somewhat more common when 11...Bg4 12 Ne4 transposes to note ‘a’ to Black’s
10th move in ‘the early moves’ section. The text, taking aim at h7 and preparing 12 Ne4, was used by Colle
himself against no less a player than Rubinstein, but has so far only been repeated on eleven occasions.
Despite that, I believe it is the trickiest move for Black to face and not only if he again continues with
11...Bg4.
TIP: Colle theory is still developing and the opening remains a fairly fertile ground for those who
like to try and resurrect forgotten or discarded ideas. As well as the text, we’ve seen that 8 e4 has
recently obtained a new lease of life, but which variation will do so next?
11...Bg4?!
Black hopes that the attack on the f3-knight will prevent 12 Ne4, but if so then he’s in for a nasty surprise.
He has also tried:
a) 11...Re8? was seen in the stem game, E.Colle-A.Rubinstein, Budapest 1926, but this natural-looking
move fails to pay any attention to White’s plans. Colle in his usual fashion opted for the aggressive 12 Bc4!?
Qd7 13 Ng5, but as Bronznik points out there was nothing wrong with winning a clear pawn with 12 Ne4
(Diagram 59)

62
 Diagram 59 (B)
Black must lose a pawn

12...Nxe4 13 Bxe4 and h7 falls.

b) 11...h6 prevents White’s threat, but after 12 Ne4 Nxe4 (or 12...Be7 13 Rd1 Qa5 as in R.Frosch-H.Titz,
Austrian League 2003, and now not the slow 14 h3, but 14 Nxf6+ Bxf6 15 Qe2, intending 16 Qe4 and
retaining an edge even in the event of 15...Ne7 16 Nd2!) 13 Bxe4 Black’s queen is forced to a slightly
undesirable square and White can gain the advantage; for example, 13...Qd6 14 Re1 Re8 15 Be3! and White
had a nagging edge in S.Savova-G.Stanciu, Balatonfured 1987 (White has some pressure down the d-file and
against e5; in the game Black could find nothing better than ceding the c8-bishop with 15...Bxe3 16 Rxe3
Bg4 17 h3 Bxf3 18 Rxf3), or 13...Qe6 14 Re1 (Diagram 60)

Diagram 60 (B)
The pressure against e5 is awkward

63
 14...f6 (E.Michalski-M.Blasiak, Pyskovice 1991) 15 Bxh6! gxh6 16 Bxc6 (Bronznik) and White wins a
pawn since 16...Qxc6? loses to 17 Qg6+ Kh8 18 Qxh6+ Kg8 19 Qg6+ Kh8 20 Re4.
NOTE: The e4-square is an ideal home for White’s pieces in this line (from there the knight, light-
squared bishop and queen all often create threats), while Black may find that having his pawn on e5
isn’t always so useful (it weakens his control of the central light squares and to defend it usually
requires a small concession).
Even the ubiquitous Bronznik (the author of a fine and extremely detailed work on the Colle) has been
unable to find a way for Black to equalize against 11 Qc2. His best suggestion is 11...Be7 12 Ne4 h6, but
White remains slightly for choice after 13 Nxf6+ Bxf6 14 Rd1 Qe6 (14...Qa5 15 Qe2 transposes to our
improvement over Frosch-Titz, above) 15 Be3 and Be4, possibly followed by a doubling on the d-file or
some sort of queenside probe, like a4-a5.
12 Ne4! (Diagram 61)

Diagram 61 (B)
Logical and strong white play

A discovery of the Spanish Grandmaster, Marc Narciso Dublan. White doesn’t worry about the doubling
of his f-pawns as he’s going to win h7, and this is much stronger than the previously played 12 Ng5 h6 13
Nge4 Nxe4 14 Nxe4.
12...Bxf3 13 gxf3 Kh8?
Defending h7, but falling for the second part of White’s trap begun by 12 Ne4. This has actually occurred
in the only two games so far played with 12 Ne4 and I suspect White’s next move is quite easy to miss,
especially with Black’s attention focused on h7. Instead 13...Rfd8?! 14 Nxf6+ gxf6 15 Bxh7+ Kg7 16 Be4
leaves White a clear pawn ahead (and he is no slower than Black at getting a rook to the g-file), and even the
relatively best 13...Nxe4!? 14 Bxe4 Qe6 15 Bxh7+ Kh8 doesn’t appear to supply enough compensation after
16 Bf5 Qf6 17 Qe4; for example, the direct 17...g6 18 Bh3 Kg7 19 Qg4 Rh8 fails to 20 Bg5 Qf5 21 Qxf5
gxf5 22 Bxf5 Rh5 (or 22...Rag8 23 Kh1 Rh5 24 Be3 Bxe3 25 Bg4!) 23 h4 f6 24 Bg4 Rh7 25 Be3 Bxe3 26
fxe3 Rxh4 27 Kg2 with some advantage for White in the endgame.
14 c4! (Diagram 62) 14...Qd7?

64
 Diagram 62 (B)
Black loses material

As Bronznik points out, Black should try 14...Qd4 15 Be3 Nxe4! 16 Bxd4 Nxd4 17 Qa4 Nd2, although
White remains at least clearly better after 18 Be4!? Nxf1 19 Kxf1.
15 Nxc5 Qh3 16 Bf5!
White now defends with either 16...Qxf3 17 Qd3 (M.Narciso Dublan-L.Barczay, Budapest 2001) or 16...Qh5 17 Be3 Qxf3 18 Be4 Nxe4 19
Qxe4 (C.Drechsler-G.Milosevic, Zurich 2006), leaving Black a clear piece in arrears.

Theoretical Conclusion
Black’s alternatives to 9...Qc7 and 9...dxe4 aren’t too bad. Should White force through e4-e5 then he usually gains slightly the more
comfortable position and, should Black continue in a rather slow vein, good attacking chances. 9...e5 is a much more proactive handling of the
black pieces, but the rediscovery of 10 exd5 Qxd5 11 Qc2 appears promising.

Illustrative Games

Game 4
A.Benderac-Le Thanh Tu
Turin Women’s Olympiad 2006

1 d4 Nf6 2 Nf3 c5 3 e3 d5 4 c3 e6 5 Bd3 Nc6 6 0-0 Bd6 7 Nbd2 0-0 8 dxc5 Bxc5 9 e4 e5 10 exd5 Nxd5!? (Diagram 63)

65
 Diagram 63 (W)
The e4-square beckons White’s pieces

Black declines to develop her queen and then see it quickly attacked. Nevertheless, this alternative
recapture isn’t a simple equalizer: White will again gain a tempo on the c5-bishop and can target both e5 and
h7.
11 Ne4
Making good use of the e4-square and hinting at aggressive manoeuvres on the kingside. White has also
played both 11 Nb3 and 11 Nc4 intending 12 Be4, but against the latter, the active 11...Re8! 12 Be4 Be6 13
Ncxe5 Nxe5 14 Nxe5 Qc7 is a dangerous gambit, R.Schulz-S.Schoof, German League 1991.
11...Be7
Black naturally wants to keep her dark-squared bishop and this is the more prudent retreat, preparing to
assist in any kingside defence. Alternatively 11...Bb6 12 Bc4 (Diagram 64)

66
 Diagram 64 (B)
White has an edge

12...Be6?! 13 Nfg5! works out well for White, and even the superior 12...Nc7 leaves him with an edge
after 13 Qxd8 (13 Qc2!? also deserves consideration, intending 14 Neg5 and if 13...Bf5, 14 Bg5) 13...Rxd8
14 Neg5 Ne6 15 Nxe6 Bxe6 16 Bxe6 fxe6 17 Re1 (P.Heyn-W.Zeiske, correspondence 1993).
12 Ng3
Colle’s choice, aiming to dissuade Black from a swift advance of the f-pawn, but 12 Qe2!? (Diagram 65)

Diagram 65 (B)
White provokes ...f5

might be a better try. The main point is that 12...f5?! 13 Rd1! (Bronznik’s untested idea and a better move
order than the previously played 13 Bc4 when 13...b5!? gives Black good counterplay after both 14 Bb3 Na5
and 14 Bxb5 fxe4 15 Bxc6 exf3 16 Qxe5 Nb6 17 Bxa8 Nxa8) 13...fxe4? 14 Bxe4 Be6 15 c4 (Diagram 66)

67
 Diagram 66 (B)
Regaining the piece with interest

regains the piece with some advantage after 15...Nc3 16 bxc3 Qc7 17 Bd5. Finding a good improvement
for Black after 13 Rd1 is not so easy: 13...Be6? 14 Neg5 clearly isn’t one; 13...h6?! 14 Ng3 wins a pawn in
view of the neat tactic 14...e4 15 Nxe4! fxe4 16 Bxe4; and even Bronznik’s main line of 13...Kh8 14 Neg5!
Bxg5 (threatening 15 Bc4 and once again hoping for 14...e4?! 15 Nxe4!) 15 Bxg5 Qc7 16 Bb5 leaves Black
in serious trouble.
Almost certainly Black should prefer to keep his f-pawn firmly on f7 after 12 Qe2, and prophylactically
leaving the d-file with 12...Qc7 makes a lot of sense. Now 13 Ng3 angles for a favourable version of our main
game; White’s position is a little easier to play, but Black should be able to secure a reasonable game with
13...g6!? 14 Re1 (or 14 Rd1 Nf4!) 14...Re8, keeping White out of f5 and intending ...Be6 and/or ...Nf4.
12...f5!
Bravely, and correctly, taking up the challenge.
13 Bc4 Be6 (Diagram 67)

68
 Diagram 67 (W)
14 Qe2 contains a double threat

14 Re1
Just as at move 12, 14 Qe2!? looks like a more critical and better try for the advantage.

TIP: Both Qe2 and Re1 attack the e5-pawn, but only the former also threatens to increase the
pressure with Rd1.
After 14 Qe2, 14...e4?! (14...Qc7?! 15 Re1 e4? 16 Nxe4! was rather effective in A.Woods-F.Ashley,
correspondence 1964; but after the slightly artificial-looking 14...Qd6 it turns out that White doesn’t have
anything much; for instance, 15 Rd1 Rad8 16 Bb3 is repelled by Bronznik’s 16...Na5!? 17 Qxe5 Qxe5 18
Nxe5 Nxb3 19 axb3 f4 with good compensation) 15 Rd1 (Diagram 68)

Diagram 68 (B)
69
 Black can easily drop material

the multiple pins, allied to the threat of 16 Nxe4, are rather awkward for Black and he has to find 15...Qd6!
to avoid material loss (15...Qd7? is less effective due to 16 Ne1 Ne5 17 Bxd5 Bxd5 18 c4, but with the queen
defended on d6, 18...Nxc4 becomes possible).
14...e4 15 Ng5
Deviating from E.Colle-M.Euwe, Amsterdam 1928, in which 15 Bg5 Kh8!? (the future World Champion
allows White to retain a tiny pull in a more complex situation than that which would have arisen after
15...Bxg5 16 Nxg5 Qxg5, transposing to our main game) 16 Bxe7 Ncxe7 17 Nd4 Qb6 18 Qb3 Bg8 19 Rad1
(19 Qxb6!? axb6 20 Bxd5 Bxd5 21 b3 looks like a better try to retain an edge), and now 19...Qg6!?
(Diagram 69)

Diagram 69 (W)
Would you capture on f5?

has been roundly but wrongly criticized. Colle continued 20 Nc2, and the game was later drawn, passing over the tactical try 20 Ngxf5?!. I
suspect he was put off by the continuation 20...Rxf5! 21 Nxf5 Qxf5 22 Qxb7 Rf8 when Black has dangerous compensation with ideas of both
...e3, and ...Qg5 and ...Nf4; play around for a while with a human or silicon sparring partner and you’ll quickly get the idea.
15...Bxg5 16 Bxg5 Qxg5 17 Bxd5 Qg6
17...Qe7 was also enough for full equality.
18 Bxe6+
Surely 18 Qb3 Bxd5 19 Qxd5+ Qf7 20 Rad1 was a little more accurate, whereas the way White chooses to
exchange queens leaves her worse in the ending due to her inferior structure.
18...Qxe6 19 Qb3?! Qxb3 20 axb3 Rad8 21 Red1 g6 22 Kf1
One plan is to break up Black’s pawn phalanx before her king reaches e5 and it was quite possible to start
with 22 f3.
22...Kf7 (Diagram 70) 23 Ne2! a6

70
 Diagram 70 (W)
Unambitious play leaves White worse

23...Ne5 24 Nd4 a6 25 f3 would have eroded d3 as a permanent outpost, but the Vietnamese is in no hurry.
24 b4?!
Weakening c4 like this makes no sense to me. Better was 24 Nf4, heading for d5.
24...Rfe8 25 Rac1 Kf6 26 Nf4 Ke5 27 Ne2 Kf6
Curiously Black is happy with a draw when 27...Rd3 would have left White with a tricky defensive task.
Black wasn’t out-rated by that many points (2239 to 2293), but perhaps the repetition clinched a tight match
or she wasn’t enjoying the Olympiad’s ludicrously-fast FIDE time control.
28 Nf4 Ke5 29 Ne2 ½-½

Points to Remember
1. White should meet 9...e5 with 10 exd5, rather than 10 Qe2.
2. 10...Nxd5 is probably superior to 10...Qxd5, but Black must still play accurately to equalize, especially
should White try one of the two testing Qe2 approaches.
3. Piece control of e4 is a useful asset.
4. Should Black play ...f5, watch for tactics down the d- and e-files, and along the a2-g8 diagonal.
5. If exchanges are imminent, try to arrange them on your terms.

The Endgame
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 0-0 Bd6 7 Nbd2 0-0 8 dxc5 Bxc5 9 e4 dxe4
Ruling out e4-e5 once and for all. Technically speaking this exchange does not have to led to an endgame,
nor does it all that often in practice, but the following exchanges are often cited by some white players as a
reason for preferring 8 Re1 or 8 Qe2 to the main line. However, even those who were dreaming of a decisive
sacrifice on h7 should be happy if Black exchanges twice on e4 and then on d1 because the resulting position
is rather pleasant for White. Let’s see why:
10 Nxe4 (Diagram 71)

71
 Diagram 71 (B)
Black plays for exchanges

10...Nxe4
An important alternative is 10...Be7, keeping pieces on and reaching a structure which one would more
usually associate with a king pawn opening; for example, the Caro or French Rubinstein. Traditionally such a
position was often thought to slightly favour White due to his queenside majority. Nowadays, though, it is
realized that advancing the queenside pawns is no easy task, at least in the short term. Instead White should
look to use his extra space to generate pressure, down the central files and especially on the kingside.
White’s most popular response to 10...Be7 is 11 Qe2 as we will see in Game 5, but once again 11 Qc2!?
(Diagram 72)

Diagram 72 (B)
Taking aim at h7
72
 is a fairly rare but venomous alternative:
a) 11...h6?! 12 Nxf6+! (many other moves have been tried, but this is far the most dangerous) 12...Bxf6 13
Qe2 (Diagram 73)

Diagram 73 (B)
White again targets the h7-square

leaves Black sorely regretting ever touching his h-pawn. In P.Corrales Ibanez-A.Garcia Ruiz, Malaga
1999, White was quick to exploit his attacking opportunities: 13...Qc7 (perhaps Black already has to try the
radical and not exactly ideal 13...Qa5!?, with the idea of 14 Qe4 Qf5) 14 Qe4 Rd8 15 b3! Bxc3? 16 Qh7+ Kf8
17 Ba3+ Bb4 18 Bb2 f5? 19 Qh8+ Ke7 20 Qxg7+ Kd6 21 Be5+ 1-0.
b) 11...Nxe4 12 Bxe4 f5?! is an active attempt to avoid allowing White an attack as 12...h6 and 12...g6
would do. Here, though, it fails to convince in view of 13 Bxc6 bxc6 (A.Alekseeva-L.Fedoseeva, St
Petersburg 2001), and now Bronznik’s 14 Bf4 with a useful grip on the e5-square and potential pressure
against the e6-pawn.
c) 11...Nd5 12 Ng3! (forcing a kingside weakness after all) 12...h6 13 Re1 b6 14 Bd2 Bb7 15 Rad1 Qc7 16
Qc1! (Diagram 74)

73
 Diagram 74 (B)
Lining up a sacrifice on h6

gave White good attacking prospects in the classic game A.Romanovsky-B.Blumenfeld, St Petersburg
1905.
TIP: Whenever Black meets a threat to h7 with ...h6, White should always look out for sacrificial
opportunities on g7 and h6 in the resulting play.
Indeed Romanovsky-Blumenfeld continued 16...Bd6?! 17 Nh5 Ne5 18 Nxe5 Bxe5 19 Bxh6! with a
thematic and powerful attack.
Returning to 10...Nxe4:
11 Bxe4 (Diagram 75)

Diagram 75 (B)
Black shouldn’t rush into exchanges
74
 11...Qxd1?!
Continuing the hoovering policy. Should Black instead wait with, say, 11...h6, White has a pleasant edge
and attacking prospects whether he continues with 12 Bf4 or 12 Qe2, and after the latter one idea is
Pezerovic’s plan of b3, Bb2, a rook to the d-file and c4.
Black can also keep the queens on with 11...Qc7, but then the direct 12 Ng5 h6 13 Bh7+! Kh8 14 Bc2
(Diagram 76)

Diagram 76 (B)
The knight is immune

is rather cute and effective:

a) 14...e5? 15 Qh5 Qe7 16 Ne4! (Diagram 77)

Diagram 77 (B)
75
 Black lacks kingside defenders

demonstrates White’s attacking potential. The immediate threat is to capture on h6 and Black actually
lacks a good defence; for example, 16...Kg8 17 Bxh6! (17 Bg5!? first and only then 17...f6 18 Bxh6 gxh6 19
Nxc5 is also pretty effective) 17...gxh6 18 Qxh6 Bf5 (the best defence, whereas 18...f5?? loses immediately
to 19 Bb3+ Rf7 20 Nf6+) 19 Nf6+ Qxf6 20 Qxf6 Bxc2 21 Rfe1 and Black’s three minor pieces were
insufficient compensation for the queen in view of White’s ongoing attacking prospects in K.Gilg-W.Von
Holzhausen, Breslau 1925.
b) 14...Rd8? is another typical way in which Black has been known to underestimate White’s attack.
NOTE: White’s bishops on c1 and c2 might appear some way from the black kingside, but they are
both pointing straight at it. Indeed this is a classic case of the dangerous ‘raking bishop pair’.
After 14...Rd8? 15 Qh5! Kg8?! (but 15...Rf8 16 Nxf7+ Rxf7 17 Qxc5 is also rather grim for Black), White
wins with 16 Nxf7! Rd5 17 Nxh6+! (Diagram 78)

Diagram 78 (B)
Stripping away the defences

17...gxh6 18 Qe8+ Kg7 19 Bxh6+!, and the black king is hunted down, or more mundanely after 16...Qxf7
17 Bh7+ Kf8 18 Qxc5+ Qe7 19 Qe3 (Oleinikov).
c) 14...Be7 15 Re1 (Bronznik’s suggestion of 15 Qh5!? and if 15...Kg8, 16 Nf3 is also rather dangerous)
15...Qd8! (15...b6? 16 Qd3! g6 17 Qh3 once again sees the white pieces wreck havoc after 17...Kg7 18
Nxe6+! Bxe6 19 Qxh6+ Kg8 20 Bxg6 fxg6 21 Qxg6+ Kh8 22 Rxe6 or 17...h5 18 Nxf7+! Rxf7 19 Bxg6 Bd6
20 Qxh5+ Kg8 21 Re4, and in both cases Black is unlikely to survive long) 16 Qxd8 Bxd8 17 Ne4 saw Black
sensibly force the queens off, but White retained a small edge due to the weakness of the d6-square in
F.Rinaldi-W.Ottenweller, correspondence 1999.
d) 14...f5! is probably the best defence. The ever-creative Bronznik suggests that White can still sacrifice
with 15 Qh5 Kg8 16 Ne4 Be7 17 Bxh6?! (17 Ng3 should be preferred with an unclear tussle ahead), but for
once I find his line of 17...fxe4! 18 Bxe4 Qe5 19 Qg6 Qf6 20 Qh7+ Kf7 21 Be3 not wholly convincing after
just 21...Bd7 with the plan of fleeing with the black king.
12 Rxd1 (Diagram 79)

76
 Diagram 79 (B)
Compare the light-squared bishops

One of the attractions of the Colle to players like Euwe and Colle himself in the 1920s and 1930s was that
White often found himself with a 3v2 queenside majority. Nowadays, thanks to the efforts of many theorists
and writers, most notably John Watson, we appreciate much more Black’s central majority and ability to
sometimes even launch a queenside minority attack.
NOTE: Despite all our modern understanding, the text position is still somewhat better for White.
He has much the freer position (try talking to the c8-bishop and he’ll tell you!) and his queenside
pawns are quick to advance and with tempo.
12...f6
Rubinstein’s idea to mobilize Black’s central pawns may be a little slow, but is probably Black’s best
defence. Alternatively, the attempt to force further exchanges with 12...Rd8 is seen in Game 6, while 12...Ne7
13 Ne5!? (an interesting plan, although the more straightforward 13 Bf4 followed by Bd6 and a doubling of
rooks on the d-file is also pretty effective) 13...f6 14 Nc4 a5?! (this is probably too weakening, but 14...Nd5!?
15 Ne3 Nb6 16 b4 Be7 17 a4 is also better for White) 15 Be3 Bxe3 16 fxe3! Ra6 17 Rd6 Rxd6 18 Nxd6 Rd8
19 Nxc8 Rxc8 20 Rd1 gave White a rather favourable ending in A.Dergatschova Daus-J.Nill, German League
2001. Note the bishop’s superiority and also the way in which White’s rook will be quick to infiltrate against
Black’s queenside, such as with Rd6-b6-b5.
13 b4! (Diagram 80)

77
 Diagram 80 (B)
White has a useful initiative

Wasting no time in beginning queenside operations. Colle himself preferred 13 Be3 Bxe3 14 fxe3 and
White has also been quite successful with this approach, especially when Black doesn’t take control of d6
with 14...Nd8! 15 Rd2 Nf7.
WARNING: After 16 Rad1 g6, White’s control of the d-file is useful, but that on its own is not going
to win the game. He needs to advance his queenside pawns and/or to create further weaknesses before
Black unravels.
E.Colle-A.Rubinstein, Berlin 1926, was a good example of the dangers of drifting planlessly for White: 17
Bc2 a6 18 Ba4?! b5 19 Bb3 Ra7! 20 Nd4 Re7 and Black slowly took over the initiative. Going back, White
should prefer either Koltanowski’s suggestion of 18 Bb3 or 18 c4!?, intending Be4 and a rapid queenside
advance, although Black remains pretty solid in both cases after 18...b6.
13...Bb6
The bishop will be further targeted here, but neither is 13...Be7 sufficient to solve Black’s difficulties.
J.Roth-B.Foerster, Schwaebisch Gmuend 2003, continued 14 Be3 (Oleinikov’s suggestion of 14 a3 with the
idea of a clamping c4-c5 is also strong) 14...Nd8 15 Nd4 f5 16 Bf3 a6?! (preventing Nb5, but White is quick
to reveal the downside to this advance) 17 Nb3! Nf7 18 Na5 Rb8 19 Ba7 Ra8 20 Bb6 and Black remained
some way from being able to free his light-squared bishop.
14 a4 (Diagram 81)

78
 Diagram 81 (B)
Mobilizing the queenside majority

Undoubtedly White has some queenside initiative here and should be quite pleased with the outcome of the
opening. His immediate idea is shown by 14...e5? 15 a5 Bc7 16 a6! which significantly increased the pressure
in K.Nickl-H.Wohlfahrt, Austrian League 1994. Black can defend better, but his problems persist after
something like 14...a6 15 a5 Ba7 16 c4 e5 17 Bd5+ Kh8 (R.Rehm-T.Loesch, Waldshut 2002) when 18 Ba3 is
pretty strong, or 14...a5 15 b5 Ne7 16 Ba3 Re8 17 c4 (Diagram 82),

Diagram 82 (B)
Black remains badly cramped

intending Bd6 and c5.

79
Theoretical Conclusion
9...dxe4 is a reaction which one can expect from an opponent who finds themselves in the Colle main line without any theoretical knowledge
and worried about the e5-advance. This rather simple approach gives White good chances for an edge whether Black prepares to soak up some
kingside pressure after 10 Nxe4 Be7, or prefers to suffer in a queenless middlegame (it is often referred to as an endgame, but it’s really a little
too early – despite this section’s heading! – to use such a term).

Illustrative Games

Game 5
J.Pelikian-L.Coelho
Santos 2000

1 d4 d5 2 Nf3 Nf6 3 e3 c5 4 c3 e6 5 Bd3 Nc6 6 Nbd2 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 dxe4 10 Nxe4 Be7 11 Qe2 Qc7 (Diagram 83)

Diagram 83 (W)
There are several attacking plans

NOTE: With this particular move order, 11 Qc2!? is a good alternative, but the text reaches an
important position which also arises via 9...Qc7 10 Qe2 dxe4 11 Nxe4 Be7.
12 Nxf6+
White decides to force an immediate concession on the kingside, but in this particular position ...g6 is not
such a big concession since Black’s dark-squared bishop is ready to slot into g7. White might thus prefer:
a) 12 Nfg5!? g6 (and not 12...h6? 13 Nxf6+ Bxf6 14 Nh7, while 12...Nxe4?! 13 Qxe4 f5 14 Qc4 takes
immediate aim at the newly-created weakness on e6) 13 Qf3 (Diagram 84)

80
 Diagram 84 (B)
A weakening ...g6 has been induced

is an old idea of Tartakower’s, recently rediscovered. Black should respond not with 13...Nxe4?! 14 Bxe4
e5 15 Bd5 Nd8 16 Re1 which gave White awkward pressure in S.Tartakower-J.Molina, Argentina 1931, but
either the untested 13...Nd5 or Bronznik’s idea of 13...Nh5!?, intending 14 g4!? Ne5 15 Qh3 Nxg4! 16 Qxg4
f5 17 Qg2 fxe4 18 Bxe4 with a rather unclear position full of mutual weaknesses.
b) 12 Bg5 and now:
b1) 12...Nxe4 13 Qxe4 f5! (the correct approach since White isn’t in a position to fix e6 as a weakness;
instead 13...g6?! 14 Qh4! concedes some pressure) 14 Qh4 Bxg5 15 Nxg5 h6 16 Nf3 e5 (Diagram 85)

Diagram 85 (W)
Black’s central majority is dangerous

81
 sees Black mobilizing his centre to good effect. So too does 14 Qe2 e5 and even 14 Qc4!? (P.Shaw-
P.Large, Halifax (rapid) 2004) doesn’t promise White any advantage after 14...Bd6! and 15...Kh8 (Bronznik).
b2) Black doesn’t have to simplify, but should he prefer 12...Nd5!? then he must meet 13 Rad1 Rd8 14
Bb1!? Bd7 15 Bxe7 Ncxe7 16 Nfg5 with 16...h6!, and not 16...Nf5? which allowed Colle to carry out another
of his beautiful trademark attacks: 17 Qh5 Nh6 18 Rxd5! exd5 19 Nf6+! gxf6 20 Qxh6 Bf5 21 Bxf5 fxg5 22
Bxh7+ Kh8 23 Bg6+ Kg8 24 Qh7+ Kf8 25 Re1 and 1-0 in E.Colle-G.Van Doesburgh, Bussem 1926.
c) Waiting with 12 Re1!? also deserves attention and we will actually consider this position in the next
chapter (via the move order 9...Qc7 10 Qe2 Be7 11 Re1 dxe4 12 Nxe4).
12...Bxf6 (Diagram 86)

Diagram 86 (W)
Again White has options

13 Qe4
Setting up the typical Colle queen and bishop battery, but should the reader find this too simplistic they
may wish to explore, as well as White’s alternatives on the last turn, 13 Bg5!?; an idea which Smith and Hall
drew attention to. Following 13...Qe7 (13...Bxg5!? 14 Nxg5 h6 is an alternative defence; in F.Rinaldi-
S.Richkov, correspondence 1999, Black would have been close to full equality had he met 15 Bh7+ Kh8 16
Bc2 Kg8 17 Nf3 Rd8 18 Rad1 with 18...Bd7 19 Qe4 g6 and ...Kg7) 14 Qe4 g6, their idea was not 15 Qh4
Bxg5 16 Nxg5 f6 17 Ne4 e5 with equality (E.Colle-G.Maróczy, San Remo 1930), but 15 h4! e5 16 Qe3,
intending to recapture on g5 with a clamping hxg5 and retaining an edge, such as with 16...Bg4 17 Be4 Rfe8
18 Rfe1 (Bronznik) due to the pressure against e5.
NOTE: White must beware Black taking over the centre with ...e5 and ...f5, but ...e5 can sometimes
become a weakness and tie Black’s pieces down, particularly if the pawn can neither be easily defended
by ...f6 nor safely advanced.
13...g6 14 Bg5
14 Bh6 Bg7 15 Bxg7 Kxg7 is similar to the game: Black will complete his development with ...Bd7 and
...Rad8, and White lacks a good way to make any real progress on the kingside.
14...Bxg5 15 Nxg5 Ne5! (Diagram 87)

82
 Diagram 87 (W)
Black defends actively

Usefully attacking the bishop before driving away the knight from g5 with ...h6 or ...f6. Instead 15...h6 16
Nf3 e5 17 Qe3 Kg7 18 Rfe1 offers White chances of an edge, and even worse for Black is the panicky
15...Qe5?! 16 Qh4 Qg7 17 f4! h6 18 Nf3, losing control of e5 as occurred in a later game.
16 Bc2 h6 17 Nf3 Nxf3+ 18 Qxf3 Bd7
Completing Black’s development and leaving a draw already looking rather likely.
19 Rad1 Bc6 20 Qf6 Rfe8 21 Rd3 Qe7!
Neutralizing any remaining pressure and the way in which Coelho is happy to take his time opposing the
d-file (which Pelikian is unable to really do anything down), preferring to first eliminate any kingside danger,
is most impressive.
22 Qxe7 Rxe7 23 Rfd1 Kg7 24 f3 Rc8 25 Kf2 f5 (Diagram 88)

83
 Diagram 88 (W)
White has no advantage

Already it is obvious that Black’s central majority is at least as dangerous as anything White can generate
on the queenside.
26 Rd6 Rcc7! 27 g3 Rcd7 28 h4
This and White’s 30th just herald further exchanges, but otherwise Black gains space with ...g5 and then
brings his king to f6, trades all the rooks and begins to advance with ...e5.
28...Bd5 29 Rxd7 Rxd7 30 f4! Kf6 31 Ba4 Rg7 ½-½
And in view of the imminent ...g5, a draw was agreed. Not an ideal result for White considering his 300-
point rating advantage and Pelikian was never able to even begin to advance his queenside majority.

Points to Remember
1. After 9...dxe4 10 Nxe4 Be7, 11 Qc2 is a good alternative to the more usual 11 Qe2.
2. Transpositions are common in the Colle – be aware of the more important ones, such as the alternative
way in which this game could have begun.
3. Those seeking an advantage after 11 Qe2 Qc7 should closely explore the notes to White’s 12th
(variations ‘a’ and ‘c’ in that case) and 13th moves.
4. Sometimes ...g6 is not a particularly helpful ‘weakness’ to inflict when Black already has his dark-
squared bishop on the kingside.
5. Always try to force some concessions, no matter how small, when allowing Black to trade several pairs
of pieces early on.
6. Don’t overestimate White’s queenside majority! As Pelikian found out, it can easily be of no use at all against active play from Black.

Game 6
M.Haag-E.Doennebrink
German League 1997

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 dxe4 10 Nxe4 Nxe4 11 Bxe4 Qxd1?! 12 Rxd1 Rd8
As we’ve already mentioned, 12...f6 is a better try, but here we are going to see how White’s queenside
pressure can pan out in practice.
13 Bf4 (Diagram 89)

84
 Diagram 89 (B)
The bishop is strong on f4

Of course, White wants to retain control of the d-file, while the bishop hinders ...e5 and may even be able
to invade on d6 or c7.
13...f6
Keeping White’s pieces out of e5 is still the best defence. Instead 13...Be7?! (a position which also arises
from 12...Be7 13 Bf4 Rd8) 14 Rxd8+ Bxd8 15 Rd1 Be7 leaves Black badly struggling to free his queenside
pieces. One tempting idea is 16 b4!?, but White can also exploit Black’s woes with 16 Bd6 Kf8 17 Bxc6!
bxc6 18 Ne5 c5 19 Bxe7+ Kxe7 20 f4 (S.Marsh-D.Edmunds, Middlesbrough 2000). Exchanging on c6 may
appear at first a little strange, but just as in the similar and famous Fischer-Petrosian example, White is
swapping one useful positional advantage for an even better one. In this case, Black can’t easily move his
queenside pieces and White will invade with Nc6+ and Rd8, after which he can target the g- and h-pawns, as
well as the queenside isolanis.
14 b4! Bb6
In E.Colle-R.Calapso, Merano 1926, Black preferred to retain control of d6 for the time being with
14...Be7. After 15 b5 Rxd1+ 16 Rxd1 Nb8 White retained some advantage (17 Bd6 Kf7 18 Bxe7 Kxe7 19
Bxh7 wins a pawn, although White might well prefer something more subtle), although sadly our record of
the game is incomplete; even the indefatigable Harvey (author of a very well-researched and useful collection
of Edgard Colle’s own games) has been unable to unearth it.
15 Bd6! (Diagram 90)

Diagram 90 (B)
Can the d6-bishop be dislodged?

15...Nd4
Black tries to free his position through exchanges, but White has seen deeper into the position and realized
that even after this Black will struggle to develop his light-squared bishop. That couldn’t have been done
immediately since 15...Bd7?! 16 a4 Rac8 17 b5 Na5 18 Be7 costs Black material and 16...Rdc8 17 a5 Bc7 18
a6 is also rather unpleasant for him.
Perhaps, though, Black should have tried to reach a superior version of the game with 15...a5!? 16 b5 Nd4.
White doesn’t have to go along with that plan, however, and as well as the sensible 16 a3, I wonder about 16
85
Bc5!? Bxc5 17 bxc5, crippling White’s queenside pawns in return for being able to exchange a pair of rooks
and then pressurize b7.
16 Nxd4 Rxd6 17 Nb3
The simple approach, but as Bronznik points out, 17 Nb5!? Rxd1+ 18 Rxd1 and Nd6 was quite possibly
even stronger.
17...Rxd1+ 18 Rxd1 f5 19 Bf3 e5 20 Bd5+
Essential as White mustn’t allow Black to untangle with ...e4 and ...Be6. After Haag’s choice, White
retains a large advantage and I haven’t been able to find a good way to improve Black’s play over the next
few moves. White’s mobile queenside majority is a very useful asset indeed.
20...Kf8 21 a4 (Diagram 91)

Diagram 91 (B)
White enjoys strong pressure

21...a5?!
This makes matters even worse on the queenside since a5 can itself become a serious weakness should
White force through b5, c5 and b6. The machine’s suggestion of 21...Rb8 was probably more resilient,
although White retains some advantage with 22 c4 Ke7 23 c5 Bc7 24 b5 and the advancing pawns are
beginning to look rather dangerous.
22 b5 Ke7 23 c4! Be6
At last Black gets to develop his light-squared bishop, but it comes too late to really matter. Nevertheless,
there wasn’t really anything better; for example, 23...Bc7 24 c5 g5 25 b6 Bd8 is pretty disgusting for Black
with 26 Nd2 and Nc4 but one good way for White to begin to break down Black’s fragile fortress.
24 Bxe6 Kxe6 25 c5 Bc7 26 b6!
Wisely not allowing Black to blockade the queenside with 26...b6 27 c6 Bd6.
26...Bd8 27 Rd6+ Ke7 28 Kf1 e4 29 Ke2
With Black tied hand and foot, Haag realizes that the arrival of his king on b5 will force through a decisive
breakthrough with c6.
TIP: Never forget that the king is the most important piece in the endgame.
29...Ke8 30 Ke3 Be7 31 Rd5 g6 32 Kd4 Bh4 33 g3 Bf6+ 34 Kc4 Ke7 35 Kb5 (Diagram 92)

86
 Diagram 92 (B)
A decisive arrival of the white king

35...Bc3 36 c6 bxc6+ 37 Kxc6 1-0

There’s nothing to be done about the killer b-pawn.

Points to Remember
1. Don’t shy away from the early queen exchange with the slightly inaccurate 8 Re1 and 8 Qe2. The resulting
queenless middlegame is very pleasant for White.
2. Do not let Black easily move his bishop from c8 should he be foolish enough to wander into this ending.
3. Some combination of a queenside pawn advance and exploiting the weak d6-square should set White well
on the path to a large advantage.
4. Should Black become very passive, don’t forget to bring your king into play.

87
Chapter Two
The Colle System: The Main Line with 9...Qc7
Introduction
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 Qc7 (Diagram 1)

Diagram 1 (W)
Black prevents e4-e5

Black’s main continuation. We touched on a number of important themes in the last chapter, but now it’s
time to enter the main line of the Colle. Weaker club players may well encounter Black’s alternative
possibilities from the last chapter more often than 9...Qc7, but one should expect this move from either an
experienced or a booked-up opponent.
WARNING: Relative to many openings, the 9...Qc7 variation is not particularly theoretical, but it is
by Colle standards and this will be by far our most theoretical chapter.
The ideas behind 9...Qc7 are fairly obvious: Black wants to prevent e4-e5 and, as Rizzitano puts it so well,
‘Black’s goal is to contest the centre before White can initiate a kingside attack’. White, for his part, can still
hope to gain such an attack, but must be especially aware that forcing through e4-e5 by no means guarantees
him the advantage and may even on occasion play into Black’s hands.
The main move is now the natural 10 Qe2, but first we will consider an unbalancing alternative which
should particularly appeal to those who like to play against an isolated pawn:
10 exd5!? (Diagram 2)

88
 Diagram 2 (B)
White plays for a small edge

Once a favourite of Colle’s, this exchange was long considered fairly harmless, but has recently undergone
a small revival.
10...exd5
The traditional response, but by no means everyone likes to play with an IQP, as we discussed in Game 3.
Those who like to maintain a flexible structure, or players who are surprised by 10 exd5 and want to reply
solidly, are likely to prefer 10...Nxd5!?. After 11 Ne4 Be7 the position is very similar to the line 9...dxe4 10
Nxe4 Be7, except that Black has gained the moves ...Nd5 and ...Qc7. Those extra tempi may be enough for
him to equalize, but White should not give up on 10 exd5 just because of this (fairly rare) option. Indeed he
can point to Black’s slightly exposed kingside and hope to attack just as he often did in such positions in
Chapter One. Play might continue 12 Re1 (12 Qc2 is less impressive now that there isn’t an immediate threat
to win the h-pawn, while 12 Bg5 f6 13 Bh4 would have been fine for Black in K.Seiser-B.Steiner, Vienna
2004, had he continued with 13...Rd8 and ...e5) 12...Rd8 (12...Nf6?! 13 Nxf6+ Bxf6 was a rather retrograde
step in H.Schmitzer-R.Ophoff, Potsdam 1997, and now both the game’s 14 Bg5 and 14 Ng5!? g6 15 Qf3 are
promising) 13 Qc2!? (Diagram 3)

89
 Diagram 3 (B)
Again lining up against h7

13...b6 (13...Nf4 14 Bxf4 Qxf4 nets the bishop pair, but after 15 Rad1 White’s lead in development and potential kingside pressure
continues to set Black a few problems; for example, 15...e5 16 Ng3 g6 17 Re4 Qf6 18 Bb5 Bd7 19 Qe2 and White continues to build up against
e5) 14 Ng3 h6 15 Bd2 Bb7 16 Rad1, covering the potentially loose d3-bishop with a roughly equal position. Black might now try 16...Bd6,
aiming to use the f4-square and preventing any sacrificial ideas of Qc1 and Bxh6, and so the white queen might seek instead to join the game
via a4 and g4.
11 Nb3
WARNING: 11 h3? Bxh3! (Diagram 4)

Diagram 4 (W)
A nasty tactic to avoid

simply left Black a clear pawn up after 12 Ng5 Bd7 in A.Shahtahtinsky-B.Socko, Athens 2004, since

90
White couldn’t capture with 12 gxh3? in view of 12...Qg3+ 13 Kh1 Qxh3+ 14 Kg1 Ne4! with a crushing
attack.
11...Bb6 (Diagram 5)

Diagram 5 (W)
White needs to eliminate Black’s activity

The best retreat for the bishop and from b6 it prevents Be3 and is able to take part in an aggressive ...Qd6
and ...Bc7 regrouping. Furthermore, the alternative 11...Bd6 allows White to prevent ...Bg4 with 12 h3! after
which 12...Ne5 (this appears a little compliant, although otherwise White can hope to maintain control,
beginning with 13 Be3 and Rfe1) 13 Nxe5 Bxe5 14 Be3 Bd7 15 Bd4 (15 Re1 Rfe8 16 Qf3 may be a better
set-up; Nd4 and Bf5 is then one plan) 15...Rfe8 16 Qf3 Ne4 17 Rfe1 Bc6 18 Bxe5 Rxe5 19 Nd4 Rae8 20
Qf4! Nc5 21 Rxe5 Qxe5 22 Qxe5 Rxe5 23 Bf1 Re8 24 f3 saw White force some generally favourable
exchanges and retain an edge, although Black managed to hold the ending in M.Rygaard-L.Eklund, Skara
2002.
12 Bg5
NOTE: The position rather resembles the Tarrasch Variation of the French. The text indeed begins
a plan which is also common in that opening, namely to reduce Black’s attacking prospects by bringing
White’s bishop round to the h2-b8 diagonal.
White doesn’t have to begin with 12 Bg5, but once again he should avoid 12 h3?! Bxh3! 13 gxh3 Qg3+ 14
Kh1 Qxh3+ 15 Nh2 Ne5 16 Be2 Nf3! 17 Bf4 Nh4 18 Bf3 Nh5 when Black has a rather strong attack,
although a very thorough analysis from Rizzitano has shown that White should be able to draw if he defends
very accurately, beginning with 19 Be5!. We will see move of Black’s attacking ideas in Game 7 which
features Colle’s old choice, 12 Qc2.
As 12 Nbd4 Bg4 13 Be2?! Nxd4! 14 Nxd4 Bxe2 15 Qxe2 Bxd4 16 cxd4 already left Black slightly for
choice (his knight will eventually attack d4, whereas White’s bishop can never attack d5) in A.Lein-
V.Eingorn, Moscow 1989, White’s only other possibility is 12 Re1, but this also proved harmless in another
Eingorn game: 12...Bg4 13 Be3 Rfe8 14 Bxb6 Qxb6 15 Rxe8+ Rxe8 (Diagram 6)

91
 Diagram 6 (W)
Black has sufficient activity

and White had nothing better than 16 h3 Bh5 17 g4 Bg6 18 Bxg6 hxg6 when his overextended kingside at
the least balanced out the IQP in M.Drasko-V.Eingorn, Tallinn 1989.
Thus 12 Bg5 is not only White’s most logical move (as well as preparing Bg5-h4-g3, it also threatens to capture on f6), but also his best try
and we will return to it in Game 8.

Theoretical Conclusion
10 exd5!? is a good surprise weapon, but objectively it doesn’t trouble Black after either recapture on d5. The isolated queen’s pawn positions
are fairly unbalanced and while White must be careful, he can hope to outplay the opponent in them, particularly if he has some prior
experience of similar positions.

Illustrative Games

Game 7
E.Colle-I.Kashdan
Bled 1931

1 d4 d5 2 Nf3 Nf6 3 e3 c5 4 c3 e6 5 Nbd2 Nc6 6 Bd3 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 Qc7 10 exd5 exd5 11 Nb3 Bb6 12 Qc2 (Diagram 7)

92
 Diagram 7 (B)
Trying to avoid a ...Bg4 pin

Black correctly plays actively

White has the ideas of 13 Bf5 and 13 Bg5, now that ...Ne4 has been prevented, but 12 Qc2 does appear a
little slow.
12...Re8!?
Supporting ...Ne4 and deviating from Colle’s previous outing in this line: 12...Bg4 13 Nh4 Rae8 14 Nf5
Ne4 15 Nfd4 Bd7 16 Be3 f5 (Diagram 8)

Diagram 8 (W)

gave Black good counterplay in E.Colle-S.Tarrasch, Baden-Baden 1925. Perhaps Colle was ready to improve with 13 Nfd4 (or 13 Nbd4;

93
White lacks a convincing alternative with f3 attacked), rather than again waste time with his king’s knight, not that 13 Nfd4 is at all
problematic for Black after exchanges on d4 and c2.
Much more recently Black preferred to prevent Bg5 with 12...h6 and this also appears to be a playable
approach. E.El Gindy-S.Halkias, Linares 2003, continued 13 Re1 Bg4 14 Nfd4 Rae8 15 Be3 a6 and now 16
f3 and Bf2 would have retained a small edge, whereas El Gindy’s 16 Bf5?! ran into 16...Re5! when the
bishop had to scurry back to whence it came since 17 Bxg4? Nxg4 would have inflicted serious kingside
weaknesses.
13 Bg5
WARNING: The side facing the IQP should usually complete their development and prevent any
threats before looking to exchange pieces. Rushing to force exchanges may just help the opponent’s
initiative to grow.
Here, for example, 13 Bf5? Ne5! already leaves Black dangerously active. A knight is heading for g4 and
if 14 Nxe5?! Qxe5 15 Bxc8 Raxc8, Black threatens a decisive 16...Ng4 as well as ideas of ...Ne4 and ...d4.
13...Ne4 14 Rae1 Bf5 15 Be3!
Now White is able to exchange a pair of minor pieces, but his slightly misplaced rooks, allied to the strong
e4-knight, give Black sufficient compensation for the technical weakness on d5.
15...Bg6 16 Nh4!? (Diagram 9)

Diagram 9 (B)
Very provocative play by Colle

An extremely risky try when White had much simpler alternatives available, including Kmoch’s logical 16
Bxb6 Qxb6 17 Nbd4. One wonders if Colle underestimated the following sacrifice or if his legendary
calculatory powers were quick to work out that the attack wasn’t quite sound.
16...Nxf2!
Correctly taking up the gauntlet, although this was apparently based on a miscalculation.
17 Bxf2!
The only defence. As Harvey has shown, both 17 Kxf2?? Qf4+ 18 Nf3 Bxe3+ 19 Ke2 Bd2+ and 17
Bxg6?? Rxe3 18 Bxh7+ Kh8 19 Rxe3 Ng4 were hopeless for White. He also mentions 17 Bxb6? Qxb6 18
Bxg6 Rxe1 19 Bxh7+ Kh8 20 Rxe1 Nd3+ 21 Kf1 Nxe1 and now White can try to counterattack with 22
Qf5!?, but after 22...Qe3 23 Qh5 Qh6 Black remains an exchange ahead in the ending.

94
 17...Bxf2+ 18 Kxf2 (Diagram 10)

Diagram 10 (B)
A dangerous piece sacrifice

18...Qb6+!?
A tempting move and one often given without comment by previous annotators, but objectively Black
should probably have preferred 18...Ne5! when 19 Rxe5 (19 Bxg6 Ng4+ 20 Kf3 Nxh2+ 21 Kf2 hxg6 also
leaves White unable to avoid losing the exchange back) 19...Qxe5 20 Qd2! Bxd3 21 Qxd3 Qxh2 22 Nf3 Qh5
leads to a rather unclear position. Black has a rook and two pawns for the two knights, but probably isn’t
better since White is finally well co-ordinated once more and can play to target both d5 and the black
kingside.
19 Kg3
Once again the only move since 19 Kf3? would have allowed 19...Ne5+ with tempo, and 19 Nd4? Nxd4 20
cxd4 Qxd4+ wins one of White’s two loose pieces.
NOTE: As pointed out by John Nunn, loose pieces figure in many combinations. Always try to keep
your pieces co-ordinated and remember LPDO (‘loose pieces drop off’)!
19...Re3+ 20 Rxe3 Qxe3+ 21 Rf3 (Diagram 11)

95
 Diagram 11 (B)
Has Black enough compensation?

21...Qg5+?!
Mistakes often go together in pairs and this appears to be an example of that. At some point over the past
few moves Kashdan apparently realized that his original intention of 21...Bxd3 can be met by just 22 Qxd3; a
possibility which he had initially overlooked. Thus he prefers the check, but it was actually better to exchange
first on d3 and then go 22...Qg5+ 23 Kh3 Ne5 after which Black regains an exchange with 24 Rg3?! (White
should probably settle for a repetition after 24 Qd4 Nxf3 25 Nxf3 Qh6+ 26 Kg3 Qg6+ 27 Kh3 or here try 27
Kf2!?) 24...Nxd3 25 Rxg5 Nf2+ 26 Kg3 Ne4+ 27 Kg4 Nxg5 28 Kxg5 Re8 and the knights will do well to
contain the active rook in the ending. Nevertheless, the move played in the game, whilst not objectively
correct was still practically very dangerous – few people feel comfortable playing with their king on the
wrong side of its defensive pawn shield.

22 Kh3 Ne5 23 Rg3 Qh6 24 Bf5!

Once again the only move and a good one, but despite this Colle’s position remains tricky to handle,
especially as time trouble was quite possibly beginning to rear its ugly head.
24...Re8 25 Nd4 Bh5 (Diagram 12)

96
 Diagram 12 (W)
An unusual h-file line-up!

26 Qf2?
Trying to cover f2 and presumably Colle couldn’t find a defence to the threat of 26...g5. But there was one
and a rather good one at that in the active 26 Qa4! when White should be able to unravel and win with, for
example, Kmoch’s 26...Rd8 27 Qa5 Rd6 28 Qxa7 g5 29 Qb8+ Kg7 30 Bc8 (Diagram 13).

Diagram 13 (B)
White wins!

A most unusual position and one a little baffling to the human eye, but the threat of 31 Ndf5+ is decisive
since 30...Bg6 fails to 31 Rxg5! Qxg5 32 Qxd6.
26...g5 27 Bxh7+ Kf8
97
 Avoiding Colle’s trick of 27...Kxh7? 28 Qf5+ Bg6 29 Qxg5 when White survives on the kingside and is a
pawn ahead, but missing the best defence of 27...Kh8! with a crazy-looking line-up down the h-file! I think
Black is doing quite well here, although White can continue to resist after 28 Bf5.
28 Bf5
This game is full of amazing tactical possibilities which I hope the reader will take the time to fully explore
and enjoy. Here, for instance, White might have tried 28 Ne6+!? Qxe6+ 29 Nf5 Rc8 30 Qf4!? (the queen is
immune due to the mate on g8) 30...Bg6 31 Qxg5 Ke8 and it’s anyone’s guess what’s happening.
28...gxh4 29 Re3?

As pointed out by Oleinikov, White should have preferred 29 Qe3! when the game is likely to end in a
draw after 29...Qh8! 30 Rg5 f6 31 Ne6+ Rxe6 32 Rxh5 Qxh5 33 Bxe6.
29...Re7 30 Re1? (Diagram 14)

Diagram 14 (B)
Black has a neat deflecting sacrifice

The final and decisive mistake. Colle wants to avoid the threat of 30...Ng4, but misses that Black’s bishop
can also make good use of that square. 30 Qe1 Nd3 was hopeless, but White could have continued to defend
with the paradoxical 30 Qg1!, although Black remains for choice after 30...Qf4.
30...Bg4+! 31 Bxg4 Nd3!
A final neat tactic in this amazing game and Kashdan shortly emerges in an exchange-up endgame which
he converted without difficulty:
32 Qxh4 Nf4+ 33 Kg3 Qxh4+ 34 Kxh4 Nxg2+ 35 Kg5 Rxe1 36 h4 Ne3 37 Bf3 Nc4 38 Nf5 Rg1+ 39 Kf4 Rf1 40 Ne3 Rf2 41 Nd1 Rh2
42 h5 Nxb2 43 Ne3 Nd3+ 44 Kg3 Rxa2 45 Bxd5 Rb2 46 h6 Ne5 47 Kf4 Ng6+ 48 Ke4 b6 49 Nf5 Ne7 0-1

Points to Remember
1. From an objective viewpoint, 12 Qc2 is not White’s best option.
2. Try to neither develop too slowly nor always hurry with exchanges when faced with an IQP.
3. Always watch out for sacrifices in IQP positions on h2 and f2 (as in this game); as well as on f7 and e6
when White has the IQP.
4. Never be afraid of using your intuition when mounting a sacrificial attack. It’s rarely possible to see
everything. Kashdan’s play wasn’t perfect, but I suspect he sensed that his attack was always very dangerous
from a practical perspective.

98
 5. Enjoy this mind-boggling and very entertaining game!

Game 8
M.Rygaard-A.Lugovoi
European Club Cup, Chalkidiki 2002

1 d4 Nf6 2 Nf3 d5 3 e3 e6 4 Nbd2 c5 5 c3 Nc6 6 Bd3 Be7 7 0-0 0-0 8 dxc5 Bxc5 9 e4 Qc7 10 exd5 exd5 11 Nb3 Bb6 12 Bg5
Quite frankly, I’m surprised that this hasn’t received more than the three outings which I’ve been able to
locate. In view of the threat to take on f6, Black’s next is semi-forced. Bringing the knight to e4 might have
appeared to many Whites to only help Black, but matters are by no means so clear.
12...Ne4 13 Qc1!? (Diagram 15)

Diagram 15 (B)
Rare but interesting

A creative follow-up and probably also the best try for the advantage. White might instead continue his
Tarrasch-like plan with 13 Bh4, but here Black is fairly well placed to meet it and 13...Bg4 14 Be2 Rfe8 15
Nfd4 (White should avoid the course of L.Scalise-J.Suarez, Buenos Aires 2001: 15 Qxd5? Nxc3! 16 bxc3
Rxe2 and Black was already somewhat better) 15...Bxe2 16 Nxe2 Qe5 (Bronznik) leaves Black with
sufficient activity to offset the IQP.
13...Bg4
Natural, but this move is not the easy equalizer which it might at first appear. Before seeing why, we
should note that the main idea behind 13 Qc1 is that 13...Nxg5 14 Qxg5 Be6 15 Rfe1 (Diagram 16)

99
 Diagram 16 (B)
Black lacks activity

leaves White with a small but clear edge; the exchange on g5 has deprived Black of much of his activity
and it is now White who can increase the pressure, such as by doubling on the e-file and/or preparing Bf5.
NOTE: Two advantages of the IQP are the space it gains and the activity it gives to one’s pieces.
When playing with such a pawn, always try to remain active.
Probably Black’s best continuation is Watson and Schiller’s subtle suggestion of 13...Bf5!. White must
then deal with the threat to f2 and 14 Be3 isn’t particularly inspiring, although it may be the best try. More
consistent with our main game would appear to be 14 Bf4, but after 14...Qe7 15 Nfd4, the Americans point
out the big difference that Black can now play 15...Bxd4 16 Nxd4 Nxd4 17 cxd4 Qb4! when 18 f3 is ruled
out and White hasn’t any advantage.
14 Bf4 Qe7?!
Black’s last move was a little risky, but it can probably be justified by 14...Qd7!. The idea is that 15 Nfd4
Nxd4 16 Nxd4 Bxd4 17 cxd4 Rac8 is now playable: Black meets 18 Qb1 with 18...Bf5 and 18 Qe3 with
18...Rfe8 when White hasn’t anything better than allowing full equality after 19 Rfe1 Nc5 20 Qd2 Nxd3 21
Qxd3.
15 Nfd4! (Diagram 17)

100
 Diagram 17 (B)
Black will be driven backwards

Now Black’s advanced minor pieces on e4 and g4 begin to look a little misplaced in view of the threat of
f3. Surprisingly the position after 14...Qe7 was assessed as equal by Rizzitano, but it appears that small
inaccuracy actually grants White a pleasant edge.
15...Rac8
When White doesn’t have full piece control of d4, Black is always going to consider 15...Bxd4 16 Nxd4
Nxd4 17 cxd4, but here doing so is not sufficient to equalize. After 17...Rac8 18 Qe3 Black must retreat with
18...Be6, but then 19 Rfe1 Rfe8 20 f3 Nd6 21 Qf2 Qd8 22 g4 is one way to maintain a small but clear pull for
White. If Black can exchange the light-squared bishops in such a structure, he’s usually at least fine since the
knight can then attack d4, but here he can’t.
16 Qe3 Bd7 17 f3
Driving Black backwards and the following exchange of queens merely serves to highlight White’s edge.
Without the queens on, Black’s lacks sufficient counterplay to offset his weaknesses on d5 and d4.
17...Nf6 18 Qxe7 Nxe7 19 Bg5 (Diagram 18)

101
 Diagram 18 (B)
Inflicting further structural damage

19...Nf5!
There wasn’t a good square to move the f6-knight to and so Lugovoi allows a further weakening of his
structure in return for some much-needed counterplay and activity.
20 Rfe1 Rfe8 21 Kf1 Nd6 22 Bxf6 gxf6 23 Rxe8+!
Not only do exchanges help White in general terms, but removing all the rooks is also the best way of
countering the threat of ...Nc4.
23...Rxe8 24 Re1 Re5
Rygaard’s deep point would have been revealed by 24...Rxe1+ 25 Kxe1 Nc4 26 Bxc4! dxc4 27 Nd2 Bc7
28 g3 b5 29 Ne4 Bd8 (Black can’t really exchange either bishop and 29...Kg7? loses a pawn to 30 Nc5 Be8
31 Nde6+! fxe6 32 Nxe6+ Kf7 33 Nxc7) 30 Nc5 Be8 31 Ke2 (Bronznik) and White’s knight-pair fully
dominates Black’s bishops.
25 Nd2 Bc7 26 g3! b5 27 f4 Rxe1+ 28 Kxe1 a5
Good defence from Lugovoi, preparing to gain counterplay with a well-timed ...b4. Should White’s knight
leave d2, Black will try to exchange off the strong knight on d4. The resulting position is rather grim for him,
but this attempt to create a fortress is his best.
29 Nf1 Bb6 30 Ne3 Bxd4! 31 cxd4 Bc6 (Diagram 19)

102
 Diagram 19 (W)
How best to make progress?

32 Nf5?!
Rygaard is understandably tempted by the pure bishop ending in which Black’s bishop is badly tied down
defending d5. However, it is not then at all easy for White to infiltrate with his king and there was no need for
him to hurry. As indicated by Bronznik, White should have first fixed the queenside with 32 b4! (32...axb4 33
Nc2 regains the pawn) after which the king can be improved to f3. Only then need White decide whether and
when Nf5 is the best way to make progress. Another possible plan being to play f5 and Ne3-g2-f4. Black
meanwhile must be very careful about any ...Nc4 excursion; White may just ignore the knight, but he will
also be tempted by an exchange, blockading the resulting passed pawn with knight or king, and then using the
other piece to push forwards along with his own resulting passed pawn.
32...Nxf5 33 Bxf5 h6?
Grandmasters are usually much better than the rest of us at defending such dispiriting positions (it’s part of
the job I suppose), but Lugovoi’s initial play in the bishop ending is not the best. Black needs to avoid having
his bishop tied down to both b5 and d5, and so 33...b4! was correct.
TIP: Space is often a useful asset in an ending; always be on the look-out for ways to safely gain
some (i.e. in the process try not to overextend and/or to allow the opponent’s pieces to infiltrate), and
thereby deprive the opponent’s pieces of room in which to operate.
34 a3?!
The Swedish 2250-rated amateur misses the chance to really make his grandmaster opponent suffer with
34 b4! (Diagram 20).

103
 Diagram 20 (B)
Black’s bishop is very bad

Black cannot then exchange as White’s king will recapture and decisively reach c5, and so 34...a4 35 a3
Kg7 36 Ke2 Kg8 (the weakness of h6 leaves Black without an alternative to passing) 37 Kf3 Kg7 38 Kg4
Kg8 39 Kh5 Kg7 40 Bc8 Kh7 might well have occurred. Thus far, White’s progress has been pretty
straightforward, but how now to force zugzwang (i.e. reach a position in which Black cannot keep playing
just ...Kg7-h7-g7)?
As Bronznik has shown, White wins with a classic breakthrough on the kingside: 41 g4! (White can’t
simply manoeuvre his bishop: 41 Bf5+ Kg7 42 Bd3 is met by 42...Bd7) 41...Kg7 42 h4 (an immediate 42 g5
fxg5 43 fxg5 also does the trick: after 43...hxg5 44 Kxg5 the outside h-pawn is a decisive deflector) 42...Kh7
43 g5 (Diagram 21)

Diagram 21 (B)
104
 The winning breakthrough

43...hxg5 44 fxg5! fxg5 (or 44...Kg7 45 gxf6+ Kxf6 46 Kh6 and the h-pawn queens) 45 Kxg5 Kg7 46 Kf5
and White’s king infiltrates with decisive effect.
34...Kf8 35 Kf2?!
Time trouble may well have been affecting the proceedings. Lugovoi now finds the correct solution at the
third time of asking.
35...b4! 36 axb4 axb4 37 Bd3 Bd7 38 Kf3 Ke7 39 Bc2 Bb5 40 Bf5 Ba4! (Diagram 22)

Diagram 22 (W)
White’s king can’t infiltrate

41 Ke2
Black has managed to set up a fortress which 41 Kg4 no long breaks due to 41...Bd1+ 42 Kh4 Kf8 43 Bg4
Bc2 44 Kh5 Kg7 and there’s no way in on the kingside.
41...Bb5+ 42 Kd2 Kd6 43 Bg4 Ba4 44 Ke3 Bc2 45 Bh5 Ke6 46 Bg4+ Kd6 47 Be2 Ba4 48 Bf3 Bc2 49 Be2 Ba4 50 f5 Ke7 51 Kf4 Bc6 52
Kg4 Kf8 53 Bd1 Kg7 54 Kf4 Kf8 55 Bf3 Ke7 56 Bd1 Kf8 57 Bc2 Ke7 58 Bb3 Kf8 59 Kg4 Kg7 60 Kf3 Kf8 61 Ke3 Ke7 62 Kd2 Kd6 63
Bc2 Bd7 64 Kc1 Bb5 ½-½

Points to Remember
1. 12 Bg5 Ne4 13 Qc1 is an interesting way to set Black early problems.
2. Firm piece control of the square in front of an IQP can be a very useful asset. Rygaard’s knight on d4
fully dominated Black’s minor pieces in the early middlegame.
3. Exchanges generally favour the side playing against the IQP.
4. When you reach a clearly better ending, always give thought to how the exchange of any remaining pair of pieces affects both the kings
and any possibility that the defender may have of setting up a fortress.

White Plays 10 Qe2

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 Qc7 10 Qe2 (Diagram 23)

105
 Diagram 23 (B)
Supporting e4-e5

The main move, renewing the idea of e4-e5. Just as at move 9, Black now has a whole plethora of
possibilities. The two most important (the traditional main line, 10...Bd6, and the modern preference, 10...h6)
are the subject of the next two sections respectively, but first we have a number of other options to consider:
10...Re8?!
A not uncommon inaccuracy. Black prepares a retreat square for his bishop on f8 and especially hopes to
advance with ...e5. This was once the choice of Lasker, but it does rather underestimate White’s kingside
potential. The alternatives are:
a) 10...Ng4 is very similar to 10...Bd6 in that Black is trying to retain control of e5, and so we will examine
this option too in our next section.
b) 10...e5?! is a typical mistake: 11 exd5 Nxd5 12 Bxh7+! Kxh7 13 Qe4+ (Diagram 24)

106
 Diagram 24 (B)
A typical pawn-winning trick

13...g6 14 Qxd5 is a classic Colle sacrifice to net a pawn. After 14...Be7 (P.Harrison-J.Farr, Cheltenham
2004) 15 Nb3 Be6 16 Qb5 a6 17 Qe2 Kg7 (playing for ...Rh8, whereas 17...f5?! merely weakens the g5-
square) 18 Ng5! Bf5 19 f3 Black has some compensation, but not really enough due to White’s grip on the
e4-square and actively-placed knights.
c) 10...dxe4 11 Nxe4 Be7 transposes back to Game 5 of the last chapter (or after 12 Re1 to variation ‘e’
below). Black has also been known to try 11...Nxe4?! 12 Qxe4 f5, but this is rather committal. In P.Keres-
D.Adamson, Tartu 1935, the great Estonian was quick to pinpoint the weakness of e6: 13 Qc4! (Diagram 25)

Diagram 25 (B)
Black can’t get ...e5 in

13...Ne5?! 14 Nxe5 Qxe5 15 Bf4 Qd5 16 Rad1 Qxc4 17 Bxc4 a6 18 a4 a5 19 Rfe1 and White was in full
control of the position.
NOTE: ...f5 can be a useful way for Black to blunt any pressure down the b1-h7 diagonal, but he
needs to be able to follow it up with a quick ...e5. If he cannot, White may well be able to target the
backward e6-pawn and/or make good use of the e5-square.
d) 10...Ne5 is a more creative and possibly better way of simplifying; see Game 9.
e) 10...Be7 (covering the g5-square and hoping for 11 e5? Ng4 when the e5-pawn drops off) 11 Re1 dxe4
(11...Re8 12 e5 Nd7 13 Nb3 transposes to Game 10, and White is certainly ready with e5 after 11 Re1; for
example, 11...b6 12 e5 Nd7 13 Nb3 Nc5 14 Nxc5 bxc5 15 Bf4 left White slightly for choice due to his
kingside possibilities in E.Miolo-A.Yusupov, Indonesia 1983) 12 Nxe4 (Diagram 26)

107
 Diagram 26 (B)
Can Black equalize?

reaches a position which also occurs via 10...dxe4 11 Nxe4 Be7 12 Re1 (or even 9...dxe4 10 Nxe4 Be7 11
Qe2 Qc7 12 Re1), and now Black must decide how to complete his development:
e1) 12...b6 13 Nxf6+ Bxf6 14 Qe4 g6 15 Bf4 Qb7 16 Rad1 (Bronznik) leaves White fully in control.
e2) 12...Nxe4 13 Qxe4 f5!? (13...g6 14 Bh6 Rd8 15 Rad1 Bd7 16 h4! gave White the initiative in
L.Barredo-M.Vega Viejo, Santa Olaya 1998) 14 Qc4 Bd6 15 g3!? Bd7 16 Be3 b5 17 Qh4 was T.Voelker-
H.Strehlow, Berlin 2004, and now Black could have retained reasonable counterchances with 17...Ne5!.
e3) 12...e5 13 Ng3 Bg4?! 14 h3 Bxf3 15 Qxf3 favoured White in B.Reinhardt-D.Migl, German League
1992, but 13...Re8! is a clear improvement. Thus White might also wish to consider both 13 Nxf6+ Bxf6 14
Ng5 and 13 Nfg5!?; compare with the notes to Game 5.
f) 10...Bb6!? (Diagram 27)

108
 Diagram 27 (W)
Provocative black play

is a rather brave move, daring White to sacrifice on h7: 11 e5 (White should consider too developing with
11 b3!?; compare variation ‘g’ below) 11...Nd7 (11...Ng4 has been more common when 12 Bxh7+ Kxh7 13
Ng5+ Kg8 14 Qxg4 Qxe5 15 Qh5! Qf5 16 Ndf3 Qg6 17 Qxg6 fxg6 is similar to positions we will shortly
examine; here White is a little better after 18 Be3 Bc7 19 Rad1, C.Ahues-L.Engels, Bad Nauheim 1935) 12
Bxh7+!? (after 12 Re1 f6! the position is unclear, but it can be seen that both Black’s dark-squared bishop on
b6 and rook on f8 are fairly actively placed after this thematic break) 12...Kxh7 13 Ng5+ Kg6 (the only move,
but a good one) 14 Qd3+ f5 15 Nxe6 Qxe5 16 Nxf8+ Nxf8 (Diagram 28)

Diagram 28 (W)
A very unclear position

has occurred in two correspondence games and is rather unclear. Those who like unbalanced middlegames
should be happy to explore it as White, but saner options are available in both 11 b3 and 12 Re1.
g) 10...b6!? is very similar to the last variation. Again White can carry out his standard sacrifice on h7, but
Black believes that the resulting unclear positions should give him reasonable counterplay. To play 10...b6
requires good nerves and it remains fairly rare, but it did receive some much-needed coverage from the
comprehensive Bronznik, and so may be the future choice of some well-prepared opponents.
After 10...b6!?, a typically thorough piece of analysis from Bronznik runs 11 e5 (the untried 11 b3!? is
suggested by Bronznik and deserves serious consideration; after, say, 11...Bb7 12 Bb2 White is flexibly
placed to respond to whatever plan Black chooses, and has ideas of c4, as well as e5) 11...Ng4!? (11...Nd7 12
Bxh7+!? can be compared with line ‘f’ above, and White also has 12 Re1, intending 12...f6 13 b4! Be7 14
exf6 Nxf6 15 b5 Na5 16 Nb3 Nc4 17 Nbd4! e5 18 Nc6 with advantage or 15...Nd8 16 c4!) 12 Bxh7+! (the
only consistent follow-up to 11 e5 since that pawn is now hanging) 12...Kxh7 13 Ng5+ Kg8 14 Qxg4 Qxe5!
15 Qh5 (15 Ndf3 Qf6 16 Qh5?! runs into 16...Qh6!, as occurred in T.Parameswaran-P.Zarnicki, Yerevan
Olympiad 1996, but White should prefer the unclear and unexplored 16 Qh4 Qh6 17 Qa4 – Bronznik)
15...Qf5 (Diagram 29)

109
 Diagram 29 (W)
Black covers h7 in time

16 Ndf3 Ba6! 17 Rd1! Be2 (17...f6? 18 g4! Qc2 19 Rd2 reveals why the rook must go to d1, not e1) 18
Re1 Bxf3 19 Qxf3 Qxf3 20 Nxf3 leaving the position about even.
Going back, 12...Kh8?! 13 Bc2 Ncxe5?? 14 Nxe5 forced resignation in A.Huth-M.Meffert, Weilburg 1997
– it’s remarkable just how often weaker players decline the Greek Gift without fully considering its
acceptance.
TIP: Both sides’ must always try to calculate the consequences of any possible Greek Gift.
The experienced Colle player will, though, develop a feeling for when it does and doesn’t work, and
studying the many possible Bxh7+ sacrifices in this section should be of some help in acquiring that
experience.
Finally we return to 10...Re8; will the Greek Gift now work?
11 e5 (Diagram 30)

110
 Diagram 30 (B)
Greek Gifts are on the agenda

11...Nd7
Unlike in variations ‘f’ and ‘g’ above, 11...Ng4?! is rather asking for trouble with the rook on e8.
NOTE: Black’s king’s rook is generally better placed on f8 than e8 to support an essential ...f6-break
in any French-type position.
White has a pleasant choice after 11...Ng4:
a) 12 Bxh7+!? Kxh7 13 Ng5+ Kg8 14 Qxg4 Nxe5 (14...Qxe5!? 15 Ndf3 Qf5 16 Qh5 Qg6!? – else White
plays 17 Nh4 – 17 Qxg6 fxg6 is positionally horrible, but does defuse the attack and Black will get ...e5 in)
15 Qh5 (Diagram 31)

Diagram 31 (B)
White has a promising attack
111
 15...Ng6 16 Ndf3 with good kingside chances (especially after 16...Bd6 17 Qh7+ Kf8 18 h4!, the
continuation of E.Loiskandl-H.Glaser, Bad Wörishofen 1992).
b) 12 Nb3 Be7 13 Re1 f6 (pretty much forced in view of the threat of 14 h3) 14 exf6 Nxf6 (14...Bxf6?! 15
h3 Nge5 16 Nxe5 Qxe5 17 Bxh7+! was rather good for White in L.Belavenets-T.Fomina, Frunze 1975) 15
Ng5!? (Bronznik) with awkward pressure since White threatens to capture on h7 as well as e6.
12 Nb3! (Diagram 32)

Diagram 32 (B)
Wisely not hurrying with Re1

There’s no need to play 12 Re1 yet since Bf4 may be a better way of defending e5 should that pawn need
protection. 12 Re1 was, though, preferred in W.Winter-Em.Lasker, Nottingham 1936, which instructively
continued 12...f6! 13 Ng5?! (White gets carried away on the kingside; instead, as indicated by Alekhine, he
would have retained chances to gain the advantage with 13 b4! Bb6 14 exf6 Nxf6 15 b5 and c4) 13...Qxe5!
(13...fxg5!? was also possible since the f8-knight holds everything together after 14 Bxh7+ Kxh7 15 Qh5+
Kg8 16 Qxe8+ Nf8 17 Nf3 g4 18 Ng5 Rb8) 14 Bxh7+ Kf8 (Diagram 33)

112
 Diagram 33 (W)
Not an effective Greek Gift

and with the queens coming off, Black was better due to his potentially strong centre.
WARNING: Teeing up a Bxh7+ sacrifice is an important part of the Colle and especially this
variation. But White must be careful to ensure that the sacrifice at least gains him some attack,
especially when he has to lose his e5-bridgehead in the process: he does not want to have to retreat and
then watch as Black expands in the centre.
Returning to the more precise 12 Nb3 and now we will see the most common retreat, 12...Be7, covering
the g5-square and supporting ...f6, in Game 10. It is also instructive to see what happens after the alternative
retreats:
a) 12...Bf8 13 Bxh7+! (it is also possible to build up more slowly with Koltanowski’s suggestion of 13
Bf4, but the Greek Gift leads to a fascinating and dangerous attack – who could here resist playing it?)
13...Kxh7 14 Ng5+ Kg6 and now it at first appears that Black will escape after 15 Qd3+ f5 16 exf6 Kxf6, but
in M.Roelvaag-P.Fossan, Olso 2000, White found 15 f4! (Diagram 34),

113
 Diagram 34 (B)
A deep but promising Greek Gift

threatening both 16 Qd3+ f5 17 exf6+ Kxf6 18 f5! and especially 16 Qg4 f5 17 Nxe6+:
a1) In the game Fossan rather collapsed with 15...Ndxe5?! 16 fxe5 Qxe5 and now 17 Qd3+ was the
simplest way to win.
a2) Black had to find 15...f5! 16 Qf3 Nf6 (the best defence to the threat of 17 Qg3), and after 17 Qg3!
Nh5! (17...Ng4?! 18 h3 Qb6+ 19 Kh1 Be7 20 hxg4 Rh8+ 21 Nh3 fxg4 22 Be3 Qb5 23 Qxg4+ Kf7 24 Kg1
leaves White a pawn ahead) 18 Nxe6+ Nxg3 19 Nxc7 (thus far Bronznik) 19...Re7! 20 Nxa8 Nxf1 21 Kxf1
(21 a4!? b6 22 a5 Ba6 23 axb6 Be2 is another enjoyable position to analyse – Black is a piece ahead, but his
knight is trapped on f1, albeit a square from where it dominates White’s bishop!) 21...b6 22 Be3 Bb7 23
Nxb6 axb6 24 Bxb6 and with four healthy pawns for the piece, White is better.
b) 12...Bb6!? keeps the bishop as actively placed as possible.
WARNING: Of course, White cannot meet 12...Bb6 with a Greek Gift since the f8-square is vacant:
13 Bxh7+?? Kxh7 14 Ng5+ Kg8 15 Qh5 Nf8!; a motif we’ve seen before and a vital one to remember.
White does somewhat better with 13 Bf4 (Diagram 35)

114
 Diagram 35 (B)
White has a good French-type position

13...f6 (Black preferred the solid 13...Nf8 in A.Sherwood-M.Nemutlu, correspondence 2004, when White
should have exploited Black’s lack of counterplay to advance on the kingside with 14 h4!, intending to soften
Black’s dark squares with h5-h6) 14 Rae1! (a deep move; it’s useful to keep f2 covered and the king’s rook
may be well placed on f1 should exchanges follow on e5 and White want to play f4), and now:
b1) 14...Ncxe5? 15 Nxe5 Nxe5 16 Bxh7+! exploited the loose e8-rook in G.Koltanowski-J.O’Hanlon,
Dublin 1937.
b2) 14...fxe5?! 15 Nxe5 Ndxe5 16 Bxh7+! Kf8 17 Bxe5 Nxe5 18 Qxe5 Qxe5 19 Rxe5 also left White
much better in D.Bernstein-B.Hansen, correspondence 1995.
b3) 14...Nc5! (best) 15 Nxc5 (as Bronznik has shown, 15 Bxh7+? is insufficient after both 15...Kxh7 16
Ng5+ Kg6 17 Qg4 fxg5 18 Bxg5 Qf7! and 16 Nxc5 Bxc5 17 Ng5+ fxg5 18 Qh5+ Kg8 19 Qxe8+ Bf8)
15...Bxc5 16 Qd2! (Diagram 36)

115
 Diagram 36 (B)
Even ...f6 hasn’t really helped Black

and the threat to capture on f6 leaves White with a pleasant edge after 16...fxe5 17 Nxe5 Bd6 18 Qe2, while 16...f5? fails to 17 Bxf5!.

Theoretical Summary
There’s been a lot to cover in this section, but a close study of it should help the reader to better understand both the French-type positions
which can arise when White forces through e4-e5 and when Bxh7+ is an effective idea. Both 10...Bb6!? and 10...b6!? deserve further attention,
but Black’s other options, typically played by opponents unaware of any theory beyond 9...Qc7, are not especially fashionable and offer White
good chances of an edge.

Illustrative Games

Game 9
P.Harrison-D.Pritchard
Cheltenham 2004

This is very much an amateur battle, not a clash from an international event. White was rated 1600 and the experienced black player, veteran of
a number of British Championships and probably best remembered as the author of The Encyclopaedia of Chess Variants, about 1950. A 350-
point rating gap is never easy to overcome, but due partly to the thematic and straightforward nature of the Colle, White was able to:
1 d4 Nf6 2 Nf3 d5 3 e3 c5 4 c3 e6 5 Bd3 Nc6 6 Nbd2 Qc7 7 0-0 Bd6 8 Qe2 0-0 9 dxc5 Bxc5 10 e4 Ne5
This move can be characterized along with 10...dxe4: both moves aim to reduce White’s potential
attacking pressure (10...dxe4 by ruling out e4-e5; 10...Ne5 by exchanging a pair of knights and bringing the
black queen over to the kingside to defend). However, the early exchanges help White too by freeing his
game, and in both cases he still has good chances to mount a kingside attack.
11 Nxe5 Qxe5 (Diagram 37)

116
 Diagram 37 (W)
White can still force through e4-e5

12 Nf3
White’s main choice, but the alternatives are also well worth considering:
a) 12 exd5 is the positional approach which left White with a pleasant advantage after 12...Qxe2 13 Bxe2
Nxd5 (or 13...exd5 14 Nb3 Bb6 15 Bf4 Bf5 16 Rad1 and again Black lacked sufficient activity without the
queens for the IQP in E.Biaggi-P.Ricardi, Acasusso 1994) 14 Nb3 Be7 15 Bf3 a5 16 a4 (Diagram 38)

Diagram 38 (B)
Bxd5 follows with an edge

16...Bd7 17 Bxd5! exd5 18 Rd1 Bc6 19 Be3 Bd8 20 Nd4 Re8 21 Nb5! Ra6 22 Kf1 Bb6 23 Bd4 in
T.Bakre-Z.Mamedjarova, Abu Dhabi 2004.
117
 b) 12 Kh1!? frees the f-pawn and so Black might try 12...dxe4 (12...Bd7 13 f4 Qh5 14 Qxh5 Nxh5 15 e5
g6 16 Nb3 Be7 was C.Peter-H.Namyslo, Regensburg 1997, when 17 g4 Ng7 18 Be3 would have left White in
control of the position) 13 Nxe4 Nxe4 14 Qxe4 Qxe4 15 Bxe4, but once again the queenless middlegame is a
touch better for White. Black has a superior version of the ‘endgame’ from Chapter One, although the
problem of freeing the c8-bishop remains.
Play might continue 15...e5 (15...f5 16 Bf3 e5?! 17 Bd5+! keeps the bishop on the long diagonal and
favours White; Black might consider 15...Rd8!? 16 Bf4 f6, although it’s now far from ridiculous for White to
sacrifice a pawn to invade down the d-file with 17 Rad1!?) 16 f4 Re8!? (16...Bd6 was preferred in C.Rice-
A.Shah, Hastings 1995, and now 17 Rd1 Bc7 18 f5! is one route to maintain White’s advantage) 17 f5!?
(Diagram 39)

Diagram 39 (B)
The c8-bishop remains a problem

17...Rd8 18 b4 Bb6 19 a4 a5 20 b5 when Black remains passive but solid, and will hope to later show that
White has overextended. However, White’s superior light-squared bishop supplies at least an edge here, and
he will continue with Ra2-d2 or Ba3 and c4-c5.
12...Qh5
The main point of 10...Ne5; Black hopes that his queen will neutralize any attacking chances of White’s.
Instead 12...Qc7?! is rather inconsistent and 13 e5 (F.Birkholz-A.Heinrich, Schwabstedt 1994) 13...Nd7 14
Bxh7+! Kxh7 15 Ng5+ Kg6 16 Bf4! is pretty much business as usual for White. The threat is the slow Qf3-g3
and Black lacks a good defence: White will at least get to capture on e6 (after ...f5) with an ongoing and
powerful attack.
13 e5 (Diagram 40)

118
 Diagram 40 (B)
Bxh7+ has been ruled out

13...Nd7
Again, relatively best. Otherwise White gains a pleasant choice after 13...Ng4 14 h3 Nh6 between 15 Be3
(Bronznik), aiming for a good knight against bad bishop middlegame, and 15 Bxh7+!?; a sacrifice which
Black has twice rejected, leaving him a clear pawn in arrears! He should at least try 15...Kxh7 16 Ng5+ Kg6
when the position is a little messy, but White remains for choice after 17 g4 Qh4 18 Qd3+ Nf5 19 Kg2!.
14 Re1!?
Rare, but no means illogical. White has more usually preferred either 14 Bf4 or 14 Be3 b6 15 Bd4,
preparing to advance on the kingside after Qe3 and Nd2. P.Steen-W.Hohlfeld, German League 1981, for
example, continued 15...Bb7 16 b4! Be7 17 a4 a6 18 Qe3 Bd8 19 Nd2 (Diagram 41)

Diagram 41 (B)
119
 White’s ideal set-up

19...Re8 20 f4 f5 21 Rf3 and White was in full control of the position. Going back to 14 Be3, 14...Be7 has
also been seen, but again Black can easily land up without any counterplay. Thus in H.Rossetto-J.Bibiloni, La
Plata 1997, he tried 15 Bd4 Nc5 16 Bc2 b6 17 Rfe1 Bd7 18 Rad1 f5!? . The game was eventually drawn after
19 Nd2, but with all his pieces nicely centralized White should have realized that there was no need to allow
the position to close: 19 exf6! Bxf6 (or 19...gxf6 20 c4 with strong central pressure) 20 Bxf6 Rxf6 21 Rd4
(Bronznik) and White’s central bind leaves him somewhat better.
14...f6?!
Natural, but in view of White’s strong riposte, Black should probably have preferred 14...Be7. Then 15
Be3 drops the e-pawn (this is why 14 Be3 is White’s most precise option) and 15 Bf4 Nc5 16 Bc2 b6, while
still a little passive, is playable for Black since the f4-bishop only serves to hamper the plan of Qe3, Nd2 and
f4.
15 Be3! (Diagram 42)

Diagram 42 (B)
White plays for a positional advantage

15...fxe5
In keeping with his simplistic rabbit-bashing strategy (a concept familiar to all fans of Simon Webb’s
classic Chess for Tigers), Pritchard plays for further exchanges, but he had presumably underestimated how
hard White’s resulting grip would be to shake off. There were two obvious alternatives, although neither
suffices for equality:
a) 15...Bxe3 16 Qxe3 fxe5 17 Nxe5 Nxe5 18 Qxe5 Qxe5 19 Rxe5 leaves White with the better bishop and
structure. He might gradually advance on the kingside, as in our main game, or he could consider an
immediate assault on e6 with 19...Bd7 20 c4!?.
b) 15...Be7 keeps pieces on and may be relatively best, although White still has a pleasant enough edge
after 16 exf6 Nxf6 (or 16...gxf6 17 c4!; White must apply immediate pressure to Black’s centre like this after
...gxf6, before it rolls forwards with ...e5) 17 Bd4.
16 Bxc5 Nxc5 17 Nxe5 Qxe2 18 Rxe2 (Diagram 43)

120
 Diagram 43 (B)
e5 is an excellent outpost

The position White was aiming for with 15 Be3. The white knight dominates the black position, and Black
already faces a long and painful defence.

NOTE: ...f6 is a key break in these French-style positions, but Black rarely wants to lose control of
e5 after it.
18...Nxd3?
There was no need to hurry with or even play this. 18...b6 was a superior choice, giving White more to
think out (which might lead to a slip), such as whether Bc2 and b4 drives Black backwards or just allows
awkward counterplay against c3.
19 Nxd3 b6 20 Ne5 Ba6 21 Re3 Rae8 22 Rae1 Rf4 23 Nc6!
The knight was superbly placed on e5, but Harrison realizes that it will do just as good a job on d4, tying
down Black’s pieces to the defence of e6.
23...Bc8 24 Nd4 (Diagram 44)

121
 Diagram 44 (B)
Black faces a long and passive defence

24...Rf6 25 Re5 Bd7 26 g3

Now f4 will follow, while Black can only really sit and wait.
26...b5 27 f4 g6 28 Kg2 a5 29 Nc2 Rc8 30 a3! Kf7 31 Kf3 Rc4?
I suspect this was a blunder, but in some ways the following exchange benefits Pritchard; it at least gives
his bad bishop a little more scope. That said, one should always defend as tenaciously as possible, and here
Black should have forced White to time a gradual kingside advance before trying anything desperate.
32 Ne3 Re4!
The best try; 32...Ra4 33 Ng4 Rf5 34 Nh6+ costs Black an exchange, after which White follows up with
Rd1-d4.
33 Rxe4 dxe4+ 34 Kxe4 Ke7 35 Ng4 Bc6+ 36 Kd4 (Diagram 45)

122
 Diagram 45 (B)
White’s pieces head forwards

36...Rf8 37 Ne5
37 Kc5!? Bd5 38 Ne5 was more adventurous, but White wants to keep control and might well have been a
little low on time at this juncture.
37...Bd5 38 Rd1 Rc8 39 h4 Bb3 40 Rd2 Rc7 41 Ke3 Bd5 42 Kd4 Kd6 43 Ke3?!
There was no need to allow Black to improve his king; 43 h5! would have decisively opened a ‘second
front’ on the kingside.
TIP: Often when we have a large positional advantage, there’s no immediate way, as here, to exploit
it due to the opponent’s solid resistance. In such situations, opening a ‘second front’ will usually over-
stretch the defender’s resources.
Here, after 43 h5 gxh5 44 Rh2 (Diagram 46)

Diagram 46 (B)
A second front opens up

Black is in some trouble since 44...Rg7 45 Rxh5 Rxg3? 46 Rxh7 neatly traps Black’s king in mid-board;
Black must deal with the mate threat on d7, but after 46...Bc6 47 c4! he is helpless. He might thus prefer
44...Ke7 45 Rxh5 Kf8, but with his king back on the kingside, his rook must remain covering the c-file.
White wins by either bringing his knight to c5, cementing it with b4 and playing Ke5, or with a kingside
advance: g4-g5 and Ng4-f6.
43...Kc5 1-0
At this point the game was adjudicated in White’s favour. His last move has made the winning process
more difficult, but White should still win after 44 Rd4, intending to begin with b4+ and Kd2.

Points to Remember
1. After 10...Ne5 11 Nxe5 Qxe5 White has three promising options, and which one to employ depends on
personal taste.
2. Should the centre close and Black delay ...f6, remember the regrouping plan of Be3-d4, Qe3, Nd2 and
f4.
3. When positionally worse, at least make a fight of it. Pritchard made things too easy for White by both exchanging on d3 and falling for

123
White’s Ne3 trick.

Game 10
J.M.Cristia-J.Lynch
Buenos Aires 1931

1 Nf3 Nf6 2 d4 d5 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Qc7 7 0-0 Be7 8 Qe2 0-0 9 dxc5 Bxc5 10 e4 Re8?! 11 e5 Nd7 12 Nb3 Be7 (Diagram
47)

Diagram 47 (W)
Black is a little cramped

13 Re1
It doesn’t seem to make too much difference whether White begins with this or 13 Bf4 since the two are
likely to transpose; for example, 13 Bf4 Nf8 14 Re1 transposes to our main game. Instead 13...h6 14 Rfe1
Nc5 15 Nxc5 Bxc5 also gave White an edge in A.Hamed-E.Ahmed, Tanta 1997, and now two tempting
options are Bronznik’s 16 Qd2, threatening a nasty sacrifice on h6, and 16 b4!? Bf8 17 a3, followed by c4.
13...Nf8
This is a little passive, but Lynch wants to shore up his kingside. Alternatively:
a) 13...Nc5?! can just be met by 14 Bc2, but Bronznik has shown that White should really take up the
challenge: 14 Nxc5 Bxc5 (Diagram 48)

124
 Diagram 48 (W)
Spot White’s breakthrough!

15 Bxh7+! Kxh7 16 Ng5+ Kg8 (16...Kg6? 17 Qg4 f5 18 Qg3 Qb6 19 Be3 is a neat and decisive deflection,
intending 19...Bxe3? 20 Nxe6+ Kf7 21 Qxg7+! Kxe6 22 Qf6+ Kd7 23 Qd6mate) 17 Qh5 with a strong
attack; for example, 17...Nd8 18 Qh7+ Kf8 19 Ne4! Ke7 (19...dxe4 fails to 20 Bg5) 20 b4 Bb6 21 b5! (now
the bishop prepares to attack via a3) 21...Qc4 22 Qxg7 Kd7 23 Nf6+ Ke7 24 Bg5 and wins (Bronznik), or
17...Kf8!? 18 b4! Bb6 19 Qh8+ Ke7 20 Qxg7 Kd7 21 Nh7! with a very strong attack.
b) 13...f5?! 14 exf6 Nxf6 15 Ng5! (Diagram 49)

Diagram 49 (B)
e6 and h7 are en prise

won a pawn in A.Kushnir-V.Nedeljkovic, Rijeka 1963, in view of 15...e5 16 Nxh7! (another typical way to exploit an undefended e8-rook)

125
16...e4 17 Nxf6+ Bxf6 18 Bc2.
c) 13...b6 14 Bf4 Nc5 15 Nxc5 bxc5 16 Rac1 (a prophylactic 16 b3!? also deserves attention) 16...c4 17
Bb1 Rb8 18 Ng5 Bxg5 19 Bxg5 Ne7 20 h4! gave White a kingside initiative in A.Castell Feliu-D.Gomez
Andreu, Palma de Mallorca 2003; the immediate plan being to fatally weaken Black’s dark squares with h4-
h5-h6.
14 Bf4 Bd7 15 Qe3!?
Preventing ...f6 and preparing to swing the queen across to g3, although it was also possible to allow 15 h4
f6 16 h5 f5 when Black is rather passive and White could occupy d4, or play for g4 after 17 Kh1.
15...Rec8 16 h4 Qd8 17 h5! (Diagram 50)

Diagram 50 (B)
The h5-pawn cramps Black

This useful space-gaining idea is also seen in the French proper. The pawn may advance on to h6, although
Cristia is happy to leave it on h5, preventing ...Ng6.
17...Rc7
Already Black is in some trouble since he is rather low on counterplay. His main problem is that 17...f6?!
would have only made matters worse after 18 exf6 Bxf6 19 h6 g6 20 Nc5 and White’s pieces pour into
Black’s many weaknesses. Nonetheless, Lynch’s defence is far from impressive and he would have done
better to try 17...h6!?. This hopes to tempt White into an immediate sacrifice (which is a little premature) and
also enables Black to shore up his kingside defences with ...Nh7, especially in response to 18 Nh2.
18 Nh2 Be8 19 Qg3 (Diagram 51)

126
 Diagram 51 (B)
White can slowly build up

19 h6!? g6 20 Ng4 Nd7 21 Nd2, planning Nf3 and Bg5, was also rather good, but Cristia understandably
sees no need to hurry.
19...Kh8 20 Nf3 Rd7 21 Rad1 a5 22 a4!
Refusing to allow Black any counterplay whatsoever, and preparing to increase White’s grip by exploiting
the weakened b5-square.
22...Qb6 23 Bb5 Rc7?
A blunder in a poor position. Black had to sit tight, not that doing so was particularly pleasant, such as
with 23...Qd8 when I wonder if Cristia would finally have tried 24 h6; for example, 24...g6 25 Bg5! Rc8 26
Bxe7 Qxe7 27 Ng5 Rdc7 28 Rd3 and the pressure mounts.
24 Be3! (Diagram 52)

127
 Diagram 52 (B)
The black queen is trapped

Winning material and shortly the game.

24...d4 25 Nfxd4 Nxd4 26 Bxd4 Rc5 27 Nxc5 Bxc5 28 Bxe8 Rxe8 29 Qf4 Bxd4 30 Rxd4 Qxb2 31 Qxf7 Qxc3 32 Red1 Qc8 33 h6 gxh6
34 Rd7 Nxd7 35 Rxd7 1-0

Points to Remember
1. This was rather a meister-gegen-amateur crush. It does, though, drastically highlight the fact that Black needs to be able to break with ...f6 in
these French-style positions.
2. 10...Re8 is not a recommended defence, even when Black prefers 12...Bb6 (probably his best try) to this
game’s 12...Be7.
3. When the opponent is very passive, try to restrain any counterplay at the same time as improving your
own position: Cristia’s Qe3, h4-h5, Nh2 and Qg3 was an instructive and typical kingside build-up.

128
Restraining e5 With 10...Bd6
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 Qc7 10 Qe2 Bd6
(Diagram 53)

Diagram 53 (W)
Ruling out e4-e5 once more

Continuing to restrain e4-e5 and on ChessBase’s Big Database 2007 this is Black’s most popular option,
albeit one being rapidly caught by the 10...h6 of our next section. Alekhine once employed 10...Bd6 and it
was fairly popular in the eighties, but the main line is now considered to give White a small but clear edge.
A system closely related to 10...Bd6 in terms of both its immediate aim and the resulting positions is
10...Ng4.
WARNING: Move order is important after 10...Ng4: 11 h3 Nge5 12 Nxe5 Nxe5 13 exd5?! being an
inaccurate one, due not to 13...exd5 with a transposition to the position we’re after, but 13...Nxd3 14
Qxd3 Rd8! with advantage to Black.
In that sequence, White does better with 13 Bc2, although 13...d4!? 14 cxd4 Bxd4 15 Nb3 Bb6 16 Bf4 Bd7
17 Rac1 Rad8 18 Nd2 Qb8! 19 a4 Bc7 was fine for Black in B.Lajthajm-K.Georgiev, Budva 2003.
Thus the correct move order for White is 11 exd5! exd5 12 h3 Nge5 13 Nxe5 Nxe5 and now 14 Nf3
(Diagram 54)

129
 Diagram 54 (B)
The IQP leaves Black worse

gives White a small advantage. This position should be compared with our main (10...Bd6) line and just as
there Black lacks sufficient activity; for example, 14...Nxd3 (or 14...Nc6 15 Bc2!? – even against the IQP, the
good old queen and bishop battery can be effective – 15...Ne7 16 Be3 Ng6 17 Bxc5 Qxc5 18 Bxg6! hxg6 19
Rfd1 and White was doing well in J.Feller-H.Jonkman, Cappelle la Grande 1994) 15 Qxd3 Be6 16 Ng5!? (16
Be3 was a simpler approach) 16...g6 17 Nxe6 fxe6 18 Bh6 Rf7 19 Rae1 Re8 20 Re2 retained an edge in
J.Harmatosi-E.Winkelmann, Budapest 2005.
Going back, 12...Nf6 appears a little illogical, but does force White to be precise: 13 Nb3 Re8 14 Bxh7+!
(14 Qc2 Bb6! is Black’s idea when the threat of ...Bxh3 looms large and Black is at least fine after both 15
Bf5 Ne4 and 15 Nbd4 Nxd4 16 Nxd4 Bxd4 17 cxd4 Qxc2 18 Bxc2) 14...Kh8 (14...Kxh7? 15 Qc2+ Kg8 16
Nxc5 regains the piece with an extra pawn) 15 Qc2 (Diagram 55)

130
 Diagram 55 (B)
The tactics favour White

15...Bxf2+ (15...Bb6!? 16 Bf5 Ne4! 17 Bxc8 Raxc8 was the alternative; Black is quite actively placed, but
White should be able to gradually unravel after 18 Qd1!) 16 Qxf2 Nxh7 17 Nbd4 Nxd4 18 Nxd4 Be6 19 Bf4
and White held the edge in the high-profile encounter V.Bronznik-T.Luther, Arco 2005, although he was
shortly happy to acquiesce in a repetition.
Returning to 10...Bd6:
11 Re1 Ng4
Consistent with Black’s approach, whereas after 11...dxe4 12 Nxe4 I’m not too sure what the bishop’s
doing on d6. In J.Pelikian-L.Borges, Americana 2000, White was quick to gain a strong attack: 12...Nxe4 13
Qxe4 (Diagram 56)

Diagram 56 (B)
Forcing a typical kingside weakness

13...g6?! 14 Bh6 Re8 15 Qh4 and now Black panicked with 15...Ne5?, losing material to 16 Nxe5 Bxe5 17
Bb5; a good example of why one must rarely be ambitious when still not fully developed in the early
middlegame. Black does better with 13...f5, although this is a little better for White after either 14 Qc4 or 14
Qh4!? Ne5! (14...e5? 15 Bc4+ Kh8 16 Ng5 g6 17 Rd1! h5 18 Nf7+ was a typical method of levering open
Black’s position to win material in T.Heyl-E.Rendi, Budapest 2005) 15 Nxe5 Bxe5 16 g3 Bd7 17 Qc4! with
pressure against e6, D.Muse-A.Breier, Berlin 1995.
12 h3 Nge5 13 Nxe5!
Once again White’s best approach is to play against an IQP. It’s initially tempting to also consider keeping
pieces on with 13 Bc2, but that allows Black to unravel through exchanges: 13...h6 14 Nxe5 Nxe5 15 Nf3
Nxf3+ 16 Qxf3 dxe4! 17 Qxe4 f5 18 Qe2 Bd7 was fine for Black in K.Gilg-A.Alekhine, Kecskemet 1927,
since White lacked a good way to prevent ...e5 (19 Bb3 Kh8!).
13...Nxe5
After the rather less common 13...Bxe5, 14 exd5 exd5 15 Nf3 remains a good approach, although 14 Nf3
is also tempting, and if 14...d4, then 15 Nxe5 Qxe5 16 Bd2 b6 17 Rad1 Bb7 18 Qg4 (having harmoniously
centralized, White is ready to attack with his bishop pair) 18...Rfd8 19 Bf4 Qc5 20 Bh6, as in J.Curtis-
T.Headlong, British League 2006.
131
 14 exd5 exd5 15 Nf3 (Diagram 57)

Diagram 57 (B)
Black lacks sufficient activity

Koltanowski’s move, forcing further exchanges and leaving White a little better due to the weakness of d5.
We will see how this plays out after both 15...Nxf3+ (Game 11) and 15...Re8 (Game 12). Note that White is
also happy to see Black take the bishop pair, and 15...Nxd3 16 Qxd3 Qc4 17 Rd1 Qxd3 18 Rxd3 Rd8 19 Be3
Bf5 20 Rd2 Be4 21 Rad1 (d5 will fall should Black trade on f3) 21...Be7 22 Nd4! Rd7 23 f3 Bg6 24 Ne2!
(Diagram 58)

Diagram 58 (B)
Directly building up against d5

24...Rad8 25 c4 saw White thematically build up against d5 and then win that pawn in G.Koltanowski-
132
J.Almeda, Mollet 1935 (25...d4!? was now the best try, relying on the bishop pair or potential opposite-
coloured bishops to supply some drawing chances after 26 Rxd4 Rxd4 27 Rxd4 Rxd4 28 Bxd4 Bd3 29 Nf4
Bxc4 30 b3 Ba6 31 Bxa7).

Theoretical Conclusion
I’m slightly amazed that both 10...Bd6 and 10...Ng4 have been seen so much in practice. Should White know how to inflict an IQP, he can look
forward to a healthy advantage; not for nothing is 10...h6 the choice of many modern grandmasters.

Illustrative Games

Game 11
G.Koltanowski-M.Defosse
Belgium 1936

1 d4 Nf6 2 Nf3 d5 3 e3 e6 4 Bd3 Be7 5 Nbd2 0-0 6 0-0 c5 7 c3 Nc6 8 Qe2 Qc7 9 dxc5 Bxc5 10 e4 Bd6 11 Re1 Ng4 12 h3 Nge5 13 Nxe5
Nxe5 14 exd5 exd5 15 Nf3! Nxf3+ 16 Qxf3 Be6 (Diagram 59)

Diagram 59 (W)
Black is rather too passive

There was nothing better, but already it’s clear that Black faces a long defence.
17 Be3
Preparing to bring the bishop to d4, but a good alternative is 17 Qh5!? g6 18 Qh6 (an immediate 18 Qh4
Be7! 19 Bg5 fails to 19...Bxg5 20 Qxg5 Qb6! with counterplay against b2 and with ...d4) 18...Be5 19 Qh4!,
weakening Black’s dark squares.
TIP: An exchange of dark-squared bishops is favourable for White in such positions: it further
reduces Black’s prospects for counterplay and enables White to use the d4-square for his major pieces.
Here White threatens Bg5-f6 and so in B.Persitz-S.Brilla Banfalvi, correspondence 1985, Black tried
19...d4!? 20 cxd4 Bg7 21 Be3 Rad8 and now 22 Rac1!? Qd6 23 Bc4 was one way to retain the advantage.
17...Rad8?!
Instead 17...f6?! 18 Bxa7! Rxa7 19 Rxe6 cost Black a pawn in H.Deller-B.Abegg, Chexbres 2005, while
17...Rfd8 gave rise to an instructive game and one which mirrored Koltanowski’s efforts in I.Jensen-
J.Fernandez Fornes, correspondence 1992: 18 Bc2! Be5 19 Bb3 (tying Black down to the isolani) 19...Bf6 20
Rad1 Qc6 21 Bd4 Bxd4 22 Rxd4 (accomplishing part two of the plan; the desired trade of dark-squared

133
bishops) 22...Rd6 23 Re5 (Diagram 60)

Diagram 60 (B)
Swinging the rook into the attack

23...Rad8 and now with Black completely tied down to d5, White opened a second front in some style on
the kingside: 24 Rh5! g6? 25 Rxh7! Kxh7 26 Qf6 Bg4 27 Qxf7+ Kh6 28 Rxg4 and the attack was decisive.
Finally, Black’s best move is often said to be Koltanowski’s suggestion of 17...Be5, but 18 Rad1 Bf6 19
Bc2 Rad8 20 Bd4 Bxd4 21 Rxd4 still saw White logically increasing his advantage in G.Ziese-R.Roosaar,
correspondence 1992.
18 Bc2 b5 19 Bd4!
By keeping his bishop on d6, Defosse has avoided its exchange, but it’s not just d5 which White can target.
Koltanowski has also appreciated that he can even afford to ignore Black’s counterplay with ...b4.
19...Bc5 20 Rad1 b4 21 Be5 Bd6? (Diagram 61)

134
 Diagram 61 (W)
Can White sacrifice on h7?

Missing White’s dream sacrifice. 21...Qe7 was a better defence, albeit still somewhat better for White after
22 cxb4 (22 Bxh7+? Kxh7 23 Qh5+ Kg8 24 Bxg7 Kxg7 25 Rd3 fails to 25...Bxf2+! 26 Kxf2 Qf6+ 27 Rf3
Rh8) 22...Bxb4 and then either 23 Re3 or even 23 Bxg7!? Kxg7 24 Qg3+ Kh8 25 Qe5+ f6 26 Qxe6, with an
extra pawn in the ending.
After 21...Bd6, Black’s king lacks piece cover. How should White exploit this vulnerability?
22 Bxh7+! Kxh7 23 Qh5+ Kg8 24 Bxg7!
This classic double-bishop sacrifice is more usually associated with the Zukertort Attack than the Colle
proper, but here it’s just as effective as ever. There’s nothing that can be done about the coming rook lift.
24...Kxg7
Or 24...f6 25 Qh8+ Kf7 26 Bxf8 Bxf8 27 Qh5+! Ke7 28 Qh7+ Kd6 29 Rxe6+! and Black loses his queen.
25 Qg5+ Kh7 26 Rd4 (Diagram 62)

Diagram 62 (B)
Completing a beautiful combination

NOTE: Rook lifts can play a role in standard Greek Gifts and they are, of course, essential when
both bishops are sacrificed. Always consider carrying out a rook lift in any attacking position; the extra
firepower can be very useful as we also saw in Jensen-Fernandez above.
26...Bh2+ 27 Kh1 Qf4
Desperation.
28 Rxf4 Bxf4 29 Qxf4 Rg8 30 Re5 1-0
The second rook lift forces resignation. A beautiful attack from Koltanowski and this is rightly regarded as
a classic Colle game.

Points to Remember
1. This whole line for Black has been tested in a number of correspondence games, but simply condemns him
to a long and ultimately often unsuccessful defence.
2. Having inflicted an IQP, White should look to exchange the dark-squared bishops and to tie Black down
to the defence of d5 (Bd3-c2-b3 is often a useful manoeuvre).
135
 3. When Black is fully tied down, look to attack on the kingside. As ever in the Colle, h7 is an easy target.
4. Keep your fingers crossed that one day you will do even better than carry out a Greek Gift; managing instead à la Koltanowski to
sacrifice both bishops on the kingside.

Game 12
D.Root-S.Rachels
Los Angeles 1991

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Nbd2 c5 5 c3 Bd6 6 Bd3 Nc6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 Qc7 10 Qe2 Bd6 11 Re1 Ng4 12 h3 Nge5 13 Nxe5
Nxe5 14 exd5 exd5 15 Nf3 Re8
A pseudo-active attempt to gain counterplay.
16 Nxe5
Ensuring that White retains control of the d4-square. It’s also tempting to play 16 Bf4!? Nxf3+ (I haven’t
seen 16...Bg4!? mentioned anywhere, but this may have put some players off 16 Bf4; however, after 17 hxg4
Nxf3+ 18 Qxf3 Bxf4 19 Qxd5 Rxe1+ 20 Rxe1 Rd8 21 Qf3 the presence of opposite-coloured bishops
shouldn’t be too great a hindrance to White – they only become a problem if all the major pieces are
exchanged) 17 Qxf3 (Diagram 63)

Diagram 63 (B)
White wants to trade bishops

since that forces the dark-squared bishops off, but with 17...Bd7 18 Bxd6 (White sensibly preferred 18
Be3! in M.Narciso Dublan-E.Vovsha, Budapest 1999, and obtained a pleasant edge similar to our main game
after 18...Qc6 19 Re2 Bc5 20 Rae1 Bxe3 21 Rxe3) 18...Qxd6 19 Rad1 Ba4 Black develops some counterplay
(although White might be able to regroup after 20 Rc1 since 20...d4!? 21 Qxb7!? dxc3 22 Bc4 Qf4 23 Rxe8+
Rxe8 24 Rxc3 Re1+ 25 Bf1 appears to defend, albeit in a slightly hair-raising manner).
16...Rxe5 17 Be3 (Diagram 64)

136
 Diagram 64 (B)
Again White has a pleasant edge

17...Bc5?!
There’s no need for Black to do White’s work for him. He needs to retain the dark-squared bishops if
possible (White isn’t here threatening a mating attack à la Koltanowski-Defosse) and so should prefer
17...Bd7. After 18 Qd2 Rae8 White had nothing better than 19 Bd4 Rxe1+ 20 Rxe1 Rxe1+ 21 Qxe1 in
G.Maróczy-A.Olland, Utrecht 1922.
NOTE: Given a free choice of any ending against the IQP, White would prefer either a pure rook or
light-squared bishop ending. In both those cases, Black can become very passive and White’s king can
easily come to d4.
That suggests that White might be best off employing Narciso Dublan’s move order, although the position
Maróczy reached is still pleasant enough for White. Black must still avoid an exchange of the dark-squared
bishops and after 21...Be6 22 b4! b6 23 Qe3 Bf4 24 Qe2 Qd7 25 Bc2 (Diagram 65)

137
 Diagram 65 (B)
White’s pull persists

25...Bc7 26 Qe3 f6 27 Qd3 g6 28 Qe3 f5 29 f4, Olland probably wasn’t enjoyed White’s free squeeze; he
defended well for a long time, but eventually the pressure told and Black blundered on move 50.
18 Qd2 Bxe3 19 Rxe3 Bd7 20 Rxe5 Qxe5 21 Re1 Qd6 22 Bc2! (Diagram 66)

Diagram 66 (B)
Bb3 will tie Black down

Yet again a good ploy; the bishop heads for b3, but first Root will flick in Qd3, trying to force another
weakness.
22...Bc6 23 Qd3 Qg6!?
A radical and probably correct decision. Rachels decides that White can create fewer problems with the
138
queens off the board. Black’s position would certainly have been unpleasant after 23...g6 24 Bb3: c4 is
threatened, while another good plan is Qd4 and h4-h5, softening up the black kingside.
24 Qxg6
By analogy with our last note 24 Qd2!? Qd6 25 Qd4 deserved attention, continuing to keep both Black’s
major pieces tied down to defensive duties.
TIP: Trading pieces down is a standard way to realize a material or positional advantage, but
always try to make sure that each exchange is actually a favourable one.
24...hxg6 25 f4 Kf8 26 Kf2 Rd8 27 g3 a5 28 h4! (Diagram 67)

Diagram 67 (B)
Increasing White’s advantage

Before improving his king or rook, Root sensibly seizes some useful kingside space.
28...Rd7 29 Ke3!
Once again the exchange of pieces frees some of the pressure on Black’s position, but Root was probably
correct to allow the rooks to come off. White might first try with 29 Rd1 Ke7 30 g4 to force through the
useful g4-g5 (fixing f7 and/or g6 as a weakness), but Black is in time to develop awkward counterplay with
30...Kd6 31 Re1 Kc5, followed by either ...d4 or 32 Re5 Rd8! 33 Ke3 Rh8.
29...Re7+ 30 Kd2 Rxe1 31 Kxe1 Ke7 32 Kd2 Kd6 33 b4 b6 34 Ke3 Bd7 (Diagram 68)

139
 Diagram 68 (W)
White needs to be accurate

35 Kd4?!
A natural continuation, but I suspect that Root hadn’t foreseen the following deep idea which Rachels
comes up with. As pointed out by Oleinikov, White should have retained control of the g4-square with 35
Bd1!. This not only prevents ...Bg4, but also prepares to create the all-important second weakness on the
kingside with g4-g5; for example, 35...Bb5 (neither can Black easily oppose White’s plan: 35...f5?! 36 Kd4
Bb5 37 Bb3 Bc6 38 a3 – Oleinikov – forces zugzwang; Black must either then place yet further pawns on
light squares with 38...a4 or lose g6 after 38...Bb7 39 Ba4 and 40 Be8, since the pawn ending is also lost for
him) 36 g4 Bc4 37 a3 Bb5 38 g5 (Diagram 69)

Diagram 69 (B)
Fixing the black kingside
140
 38...Bd7 39 Kd4 Bf5 40 Bf3 Be6. It’s clear that White has made significant progress, although there may
still be a way for Black to draw. One possible continuation being 41 bxa5 bxa5 42 f5!? (42 Bg2 does force a
pawn move, but it’s not at all easy to break Black’s blockade after 42...f6 43 Bf3 fxg5 44 fxg5 Bf7) 42...Bxf5
43 Bxd5 f6 44 c4 Kc7 45 c5 and the bishop ending is very favourable. Note how White created another
weakness on a5, and how Black couldn’t earlier trade on b4 since cxb4 would have led to the creation of an
outside passed pawn.
35...Bg4! 36 Ba4 f6 37 Be8 g5!
White now wins a pawn on the kingside, but at the cost of having his remaining pawns doubled.
38 hxg5 fxg5 39 fxg5 Bf3 40 Bb5 Bg4 41 Bf1 Be6 42 g6!
Root continues to probe well...
42...Bd7 43 Bg2 Be6 44 Bf3 axb4 45 cxb4 Bf5 46 Bxd5 Bd7! (Diagram 70)

Diagram 70 (W)
Impressive defence from Rachels

...and Rachels continues to defend well. Instead neither 45...Bg8 (46 Bg4!) nor 46...Bxg6 47 a4 Be8 48 a5
bxa5 49 bxa5 Bb5 50 Bb7 was advisable for Black.
47 Bb3 b5!
The final important defensive move.
NOTE: White may be two pawns up, but he isn’t able to win due to the absence of two key
ingredients: an entry path for his king and an effective pawn break.
48 Bc2 Bc6 49 Kc3 Bd7 50 a4!
The last try, but Black is able to both keep White out on the queenside and prevent his king from running
over to win g7.
50...bxa4 51 Kb2 Bb5 52 Ka3 Ke7 53 Bxa4 Bc4 54 Kb2 Kd6 55 Kc3 Ba6 56 Kd4 Bf1 57 Be8 Ba6 58
Kc3 Ke7 59 Bf7 Bb5 60 Bc4 Bd7 61 Ba6 Kd6 62 Kc4 Bc6 63 Bb5 Bb7 64 Kb3 Kc7 65 Kc3 Kd6 66 Kb3
Kc7 67 Ka4 Kb6 ½-½
The kings continue to oppose each other.

Points to Remember
1. Rachels might have defended this ending brilliantly, but White most certainly shouldn’t mind the risk-free
141
grind which it gives him.
2. White wants to exchange the dark-squared bishops in the middlegame, but needs to retain control of the
key d4-square while doing so.
3. Always weigh up the pros and cons of any exchange; Root quite possibly didn’t need to rush to trade the
queens and could have forced some further weaknesses by keeping them on for a while.
4. One weakness is rarely enough to win an ending; always try to create a second one.

142
The Modern Defence: 10...h6
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Bd6 7 0-0 0-0 8 dxc5 Bxc5 9 e4 Qc7 10 Qe2 h6
(Diagram 71)

Diagram 71 (W)
A popular and prophylactic choice

As we’ve seen Black’s attempts to both reduce the pressure through exchanges (10...dxe4 and 10...Ne5) and to hold up 11 e5 (10...Bd6 and
10...Ng4) still leave him facing an uphill struggle to reach equality. Arguably his two best options which we’ve seen so far in this chapter are
the rather unexplored 10...b6 and 10...Bb6. 10...h6 has much in common with those two provocative moves, but with an important difference:
11 e5 Ng4 can no longer be met with 12 Bxh7+.

Who Plays 10...h6?

This tricky little move has received the support in recent years of a number of strong players, including Eingorn, Silman, Volokitin and Wells.
While it may be popular with grandmasters who’ve recently revised their anti-Colle repertoire, I would still expect it to be rather uncommon at
lower club levels. That said, the club player would do well to find a line they are happy with against 10...h6 because it has been recommended
in two recent books (Rizzitano’s How to Beat 1 d4 and Watson and Schiller’s How to Succeed in the Queen Pawn Openings).

Advice for the Studious Player

As often happens with a new discovery (10...h6 only became popular after the postal player Robert Reynolds published some analysis on it in
1986), Black has scored fairly well so far with 10...h6. Furthermore, theory is yet to agree on White’s best response; for example, Watson and
Schiller award 11 Bc2 an ‘!’, whereas Rizzitano relegates that possibility to his notes, making 11 e5 and 11 h3 his two main lines. As we will
see, White should probably choose between 11 Bc2 and 11 e5, but there is much still to be explored here for those who like to find new ideas in
their favourite openings.

An Important Transposition
Any white player wishing to further their understanding of this tricky variation would do to remember that the Meran Semi-Slav is actually a
reversed Colle.
NOTE: With 10...h6 Black is effectively playing, albeit a tempo down, the white side of a line of the
Anti-Meran, namely 1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 Nc3 e6 5 e3 Nbd7 6 Qc2 Bd6 7 Bd3 dxc4 8 Bxc4 0-0 9
0-0 Qe7 10 h3 e5.
Thus the Colle player, especially should they possess a good database or Semi-Slav tome, may be able to
find some new ideas in our Colle version by exploring that particular Anti-Meran line.
Now it’s time to return to the theory of 10...h6:
11 Bc2!? (Diagram 72)

143
 Diagram 72 (B)
Two can play prophylactically!

Just as in the Anti-Meran version (where ...Bc7 is a common move), White employs a useful little waiting
move and one which we will shortly better understand. It’s important to realize that 10...h6 is directed against
two specific moves: 11 e5 is sometimes denoted as ‘?!’ or even ‘?’ in view of 11...Ng4 12 Nb3 Bb6
(Diagram 73)

Diagram 73 (W)
e5 is vulnerable

when e5 is dropping off (there’s no Bxh7+ trick to save a pawn and 13 Bf4 met by 13...f6); and the old Smith and Hall suggestion of 11 h3
runs into 11...Nh5! (Diagram 74).

144
 Diagram 74 (W)
A nasty fork is imminent

We’ll see if matters are really so clear after 11 e5 in Game 13, but White should steer clear of 11 h3.

WARNING: 11 h3?! Nh5! 12 Nb3 (there’s nothing better; for instance, 12 Rd1 Nf4 13 Qf1 Nxd3 14
Qxd3 Rd8 favours Black) 12...Ng3 13 Qc2 dxe4! 14 Bxe4 Nxe4 15 Qxe4 Bd6 (Diagram 75)

Diagram 75 (W)
Black’s bishop pair is a useful asset

leaves White struggling to generate enough activity to combat Black’s advantage of the bishop pair.
Rizzitano devotes almost two whole pages of thorough analysis to the position after 15...Bd6; an analysis
which I have to agree with – White must avoid this line. Going back, White might have more chances to
145
equalize with 13 Qd1!?, as he managed after 13...Nxe4 14 Bxe4 dxe4 15 Nxc5 exf3 16 Qxf3 Ne5 17 Qe3 b6
18 Nd3 in D.Root-J.Silman, 3rd mathchgame, California 1990. However, this is hardly inspiring for White
and Black can also consider 13...Bb6!?.
Before returning to 11 Bc2, we should note that White has also tried to unbalance the position with:
a) 11 exd5 exd5 (Reynolds’ suggestion of 11...Nxd5 12 Ne4 Bb6 is also fine for Black) 12 Nb3 Bb6
reaches an IQP position in which ...h6 is pretty useful. Furthermore, ...Re8 is coming with tempo and in
R.Kholmov-V.Potkin, St Petersburg 1998, White was probably quite happy when Black liquidated down
towards a level ending with 13 Be3 Bg4 (13...Re8 14 Nbd4 Ng4!? is a more ambitious try) 14 Bxb6 Qxb6 15
Bc2 Rfe8 16 Qd3 a5 17 a4 Bxf3! 18 Qxf3 d4.
b) 11 b4 is the sort of queenside expansion Black sometimes employs in the Semi-Slav, but here it doesn’t
appear particularly promising: 11...Bd6 12 Bb2 (Diagram 76)

Diagram 76 (B)
An unbalanced situation

12...Ne5!? (or 12...e5 13 a3 Re8 14 c4! d4 15 c5 Bf8 which was unbalanced but roughly level in H.Ree-So.Polgar, Amsterdam 1995) 13
Nxe5 Bxe5 14 g3 Bd7 15 Rac1 Rad8!? (M.Maki Uuro-J.Puranen, Tampere 1996) is fine for Black in view of 16 f4 dxe4 17 Nxe4 Qb6+ 18 Rf2
Bc6! (Bronznik).
We are now in a position to better appreciate some of the ideas behind the subtle semi-waiting move 11
Bc2!?: White can then consider h3 since ...Nf4 no longer hits his bishop, and the path has been cleared to set
up a battery against h7 with Qd3.
11...Rd8
Black’s main continuation, but unsurprisingly with both sides jockeying for position, he has also tried a
number of alternatives:
a) 11...Bb6 is the other main option and was advocated by Rizzitano; we’ll see it in Game 14.
b) 11...dxe4 rather gives into the pressures of playing a waiting game and there’s no need for Black to
break the tension so early. However, this exchange may not be so bad and after 12 Nxe4 Nxe4 13 Qxe4 f5
(C.Escher-R.Gabriel, German League 2003), White is only a little better after the game’s 14 Qe2 or 14 Qc4!?
and if 14...Nb4!? (based on the vicious trick 15 cxb4?? Bxf2+! and White loses his queen), then 15 Bb3 Nd5
16 Qe2.
c) 11...Ng4?! doesn’t fit in with the whole ethos of 10...h6 and, just like after 10...Ng4, White had a
pleasant advantage following 12 h3 Nge5 13 exd5 exd5 14 Nxe5 Nxe5 15 Nf3 in J.Pelikian-I.Thompson,

146
Groningen 1997.
d) 11...Bd6 is another slightly misguided approach which one might face from an opponent not fully au fait
with 10...h6 (the sort who knows it’s a good move, but not why) or who hasn’t seen 11 Bc2 (a move
surprisingly not mentioned by NCO) before. J.Berndt-M.Becker, German League 2004, continued 12 Re1
dxe4 13 Nxe4 Nxe4 14 Qxe4 f5 15 Qh4 Bd7 (Diagram 77)

Diagram 77 (W)
e6 and h6 may prove vulnerable

and now Berndt wonders about 16 Bg5!? as well as his aggressive choice of 16 Bb3 Rae8 17 Bxh6!?
(keeping the bishop on the long diagonal with 17 a3!? was safer and hopes for 17...Kh8? 18 Bxh6! when
Black cannot recapture) 17...gxh6 18 Qxh6 when Black failed to find the best defence; 18...Rf7? (after either
Berndt’s 18...Re7 or 18...Nd8 19 Ng5!? Re7 20 Rad1 Bf4! 21 Nxe6! Nxe6 22 Rxe6 Bxe6 23 Bxe6+ Rxe6 24
Qxe6+ Qf7, White emerges with three pawns for the piece, but Black has sufficient activity with his bishop to
roughly hold the balance) should have been punished by 19 Ng5! Ne5 20 Nxe6 Bxe6 21 Bxe6 Rxe6 22 Qxe6
Kg7 23 Rad1 Rf6 24 Qe8 with some advantage for White.
e) 11...Re8 was met by 12 Kh1!? in J.Pelikian-A.Valle, Brasilia 1997; further prophylaxis against ...Nh5
after h3. The Brazilian IM is a connoisseur of 11 Bc2, but I feel that White can also continue more directly
with Oleinikov’s 12 e5 when 12...Ng4 13 Nb3 Bb6 14 Bf4 f6 15 Qd3! (Diagram 78)

147
 Diagram 78 (B)
Trying to force ...f5

should be compared with our main (11...Rd8) line.

f) 11...b6!? takes much of the sting out of Nb3. White might thus opt for a slower development and 12 b3
(Diagram 79)

Diagram 79 (B)
A manoeuvring game ensues

12...Rd8 13 Bb2 Bb7 14 e5 Nd7 (14...Ng4!? 15 Rae1 d4! is a more direct and probably better try; for
example, 16 cxd4 Nxd4 17 Nxd4 Bxd4 18 Bxd4 Qxc2 19 Qxg4 Qxd2 20 Be3 Qd5 21 Bxh6 Qxg2+ 22 Qxg2
Bxg2 23 Kxg2 gxh6 and the game continues to fizzle out) 15 Rfe1 a6 16 Rac1 Rac8 17 b4! Bf8 18 a4 Be7 19
Qe3 saw White outmanoeuvring his opponent to gain the edge in G.Hill-B.De Wolf, correspondence 2002.
148
Instead 12 e5!? is untested and can quickly lead to a rather unclear situation, such as after 12...Ng4 13 b4
Bxf2+!? 14 Rxf2 Nxf2 15 b5! Nb4! (Bronznik).
g) 11...a6!? (Diagram 80)

Diagram 80 (W)
Facilitating ...Ba7

is a common move in the Anti-Meran, but a fairly rare one in this reversed version. It enables Black’s
bishop to retreat to the slightly superior a7-square, and may be set for more outings having been
recommended by Watson and Schiller. A.Pablo Marin-M.Roeder, Barbera 1997, continued 12 e5 (White can
also wait with 12 h3!? when Black will probably still continue 12...Nh5) 12...Ng4 13 Nb3 Ba7 14 Bf4 f6!
(this position is similar to both our main line and variation ‘e’ above, except that here Black’s rook remains
well placed on f8) 15 Qd3! (White can’t defend e5 and so must make good use of the bishop’s retreat to c2)
15...f5 (Bronznik points out that 15...g5!? is also conceivable; then 16 exf6 gxf4 17 Qg6+ Kh8 18 Qxg4 Rxf6
is rather unclear) 16 h4 and now one reasonable option was 16...Nf6!? (this is also the reply to 16 h3) 17
Nbd4 Ne4 with a complex and roughly level position.
Returning to 11...Rd8:
12 e5
The traditional Colle advance, but opinion is divided over whether this is better than 12 h3!?. Which move
to prefer is a matter of taste, but 12 h3 may appeal to those with some acquaintance of the Anti-Meran. By
analogy with that Black might play 12...Bd6!?, but in practice he’s usually preferred the consistent 12...Nh5.
However, White can now reveal another important point behind 11 Bc2, namely 13 Qb5! (Diagram 81)

149
 Diagram 81 (B)
Another point behind 11 Bc2

and 13...Bb6 14 Nc4! dxc4 15 Qxh5 e5 16 Nh4 left him slightly for choice in M.Lazic-S.Sale, Ljubljana
1995.
12...Ng4!
Consistent with Black’s uncompromising approach. Otherwise after 12...Nd7 Black’s rook is misplaced on
d8 and White gains a free hand on the kingside with something like 13 Nb3 Be7 14 Re1 b6 15 Bf4.
13 Nb3 Bb6 14 Bf4 f6 15 Qd3! (Diagram 82)

Diagram 82 (B)
Deflecting the pressure from e5

We’ve already seen this important concept in notes ‘e’ and ‘g’ at move 11. Just as there this position is
150
extremely unexplored (it’s only arisen in one game to the best of my knowledge) and Black must tread
carefully. He may be able to maintain unclear equality, but White’s position is practically the easier to handle,
especially if has already explored it a little. Play might continue:
a) 15...f5! sensibly tries to blunt the diagonal and after 16 h3 Nf6 (but not 16...Ngxe5?? 17 Nxe5 Nxe5 18
Qg3 and this time the pin costs Black material; in this particular position 16...g5? 17 hxg4 gxf4 18 gxf5 Nxe5
19 Nxe5 Qxe5 20 fxe6 is also somewhat better for White) 17 Nbd4 Ne4 (Diagram 83)

Diagram 83 (W)
The black knight reaches a good square

the position is very similar to note ‘g’ above, and once again unclear.
b) 15...fxe5!? is risky but possible and after 16 Qh7+ Kf8 (note that were the rook on e8 here, as in
variation ‘e’, above, 17 Bg6 would be pretty strong) 17 Bg3 Nf6 18 Qg6 (Diagram 84)

Diagram 84 (B)
151
 Black cannot keep e5

White should regain his pawn with a small edge; the machine prefers to defend with 18...Bd7 19 Nbd4!
Nxd4 20 cxd4 Be8 21 Qd3 e4!?, but after 22 Bxc7 exd3 23 Bxd8 dxc2 24 Bxb6 axb6 25 Rfc1 Rc8 26 a4!
Black doesn’t have enough compensation with Ra2, b3 and Ne1 one plan.
c) 15...Ngxe5 is also rather risky for Black and after 16 Qh7+ Kf8 17 Nxe5 fxe5 (or 17...Nxe5 18 Bxe5 Qxe5 19 Rae1 Qg5 20 Bg6 e5 21
Kh1! with some initiative – Bronznik) 18 Be3 Qf7 (D.Berg-G.Waldschmidt, Ruhrgebiet 2005), the most precise continuation, as given by that
most excellent of Colle analysts, Bronznik, is 19 Rae1! Qg8 20 Qd3 with good play for the pawn in view of Black’s poor co-ordination; for
example, 20...Bd7 21 Bxb6 axb6 22 a3 Be8 23 Qe3! d4?! 24 Qg3 Bf7 25 f4 and Black’s royalty remain somewhat misplaced.

Theoretical Conclusion
It’s no surprise that 10...h6 is a current grandmaster favourite; it’s a challenging option! Many of the critical positions after it do, however,
remain badly unexplored and White should focus his attentions on either 11 Bc2 or 11 e5 Ng4 12 Nb3 Bb6 13 Nbd4.

Illustrative Games

Game 13
G.Danner-I.Novikov
Graz 1996

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nc6 6 Nbd2 Qc7 7 0-0 Be7 8 dxc5 Bxc5 9 e4 0-0 10 Qe2 h6 11 e5 (Diagram 85)

Diagram 85 (B)
A critical advance...

11...Ng4!
A key idea behind 10...h6. Instead 11...Nd7?! 12 Nb3 and 13 Bf4 is pleasant for White since Black then
finds it harder to arrange ...f6 having already played ...h6.
12 Nb3 Bb6 13 Bf4?!
Defending e5, but walking into a pin. Instead White really must enter the critical complications with 13
Nbd4! (Diagram 86).

152
 Diagram 86 (B)
...and an essential follow-up

I’ve only been able to find nine games with his sharp continuation and certainly much remains to be
discovered. Practice thus far has seen:
a) 13...f6 isn’t quite as effective as in our main game as 14 Nxc6 bxc6 15 Bf4 fxe5 16 Bxe5 Nxe5 17 Nxe5
(Diagram 87)

Diagram 87 (B)
White successfully controls e5

(Silman) gives White control of e5 and an edge. Instead 15...g5?! 16 Bg3 f5 17 Nd4! c5 18 Nb5 Qg7
(L.Pecot-C.Cranbourne, correspondence 1999) 19 c4! (Pecot’s suggested improvement) 19...h5 20 h3 Nh6 21
Qxh5 f4 22 Bh2 Bd7 is analysis from Cranbourne which I don’t find fully convincing after 23 Nd6! and
153
White is better although the position remains quite complex. However, 13...f6 is playable for Black should he
meet 15 Bf4 with 15...Bc5!; for example, 16 h3 Nxe5 17 Nxe5 fxe5 18 Bxe5 Bd6 19 Rae1 Bxe5 20 Qxe5
Qxe5 21 Rxe5 a5! when White controlled e5, but the half-open b-file gave Black sufficient counterplay in
D.Zakrevskaja-S.Maksimovic, Nova Gorica 1999.
b) 13...Ngxe5!? 14 Nxe5 Nxe5 (and not 14...Qxe5? 15 Nxc6! Qxe2 16 Ne7+ Kh8 17 Bxe2 Bd7 18 Bf3
when White will at least return the piece on d5 with some advantage) 15 Bf4 f6 16 Rae1 (Diagram 88)

Diagram 88 (B)
An unclear and critical position

is critical. Currently White is a pawn down, but he is nicely centralized and can hope to exploit Black’s
weakened kingside light squares. This position is in some need of further testing, but for now I’m inclined to
agree with Silman’s 1991 view (when he played his match with Root – see Game 22) that ‘White might have
a very small plus, but that Black should be able to equalize with careful play’:
b1) 16...Qf7 leads to a fairly forced sequence: 17 Bxe5 fxe5 18 Qxe5 Bc7 19 Bg6! Qe7 20 Qh5 (Diagram
89)

154
 Diagram 89 (B)
Can White gain a bind?

and now:
b11) 20...Bd7 21 Nf3!? (Silman’s idea; instead 21 g3 Rf6 22 Bd3 Qc5 23 f4 b5! was dynamically balanced
in J.Trapl-M.Orsag, Czech League 1998, and 21 Qxd5 Bxh2+! 22 Kh1! Bd6 23 Qh5 Qf6 – Rizzitano – is also
fine for Black) 21...Bb5 (Rizzitano’s 21...Qf6 22 Bd3 Be8 23 Qg4 Bf7 is a more solid and also fairly equal
way of handling the position) 22 Qxd5 exd5 23 Rxe7 Bxf1 24 Kxf1 is a reasonable exchange sacrifice, but
after 24...Bd6! 25 Rd7 Rfd8 26 Rxb7 Rab8 Black appears to be able to gain enough activity to hold a rough
balance.
b12) 20...Qf6 is also possible since 21 f4!? is met not by 21...Bxf4?! 22 g3 Be3+ 23 Kg2! (Silman), but by
21...Bb6 22 Rd1 Bd7 (Rizzitano) with again approximate equality in view of the coming ...Rac8-c4 and/or
...Be8.
b2) 16...Bd7!? 17 Qh5 Qc8 18 Bxe5 fxe5 19 Rxe5 Bc7 20 Re2 Rf6 21 Nf3! Qe8 22 Qxe8+ Rxe8 23 Ne5
(Silman) gives White his ideal edge. However, a much better try is 20...e5!:
b21) The American IM and leading writer could find nothing better than 21 Rxe5!? Bg4! (21...Bxe5 22
Qxe5 gives White two pawns and good play for the exchange since 22...Bc6? fails to 23 Nf5! and Black
cannot save the exchange; note here the weakness of g6 – a recurring problem for Black in these lines) 22
Qg6 Bxe5 23 Qh7+ Kf7 24 Qg6+ (Diagram 90)

155
 Diagram 90 (B)
Hoping for 24...Ke7?? 25 Re1

24...Kg8! and neither side can deviate from the repetition.

b22) White might also consider Oleinikov’s idea of 21 Qg6!? e4 22 Bc2 Be8! (or 22...Rf6 23 Qh5 g5 24
f3! Bb6 25 Bb3 Be6 26 h3! and the computer marginally likes Black, but many humans would shy away from
such an extended position) 23 Qe6+ when his 23...Bf7 (23...Qxe6 24 Nxe6 Bb5 also looks rather drawish) 24
Qxc8 Raxc8 25 f3 breaks up the black centre, but Black is able to liquidate with 25...exf3 26 Nxf3 (26 Rxf3
Bb6) 26...Rce8 27 Rfe1 Rxe2 28 Rxe2 Bb6+! when he has sufficient activity to hold the balance after 29 Nd4
(prudent; 29 Kh1?! Bh5 only invites trouble).
b3) 16...g5!? may also be playable. In L.Pecot-J.Barlow, correspondence 1999, White could find nothing
better than 17 Bxe5 (17 Bg3 Qg7) 17...fxe5 18 Qxe5 Qxe5 19 Rxe5 Rf6 20 Re2 Bd7 21 Nf3 Raf8 when
Black’s pressure down the f-file fully compensated for the hole on e5, and another repetition occurred with 22
Ne5 Bc8 23 Ng4 Rf4 24 Ne5 R4f6.
Having examined the critical 13 Nbd4 and several positions which the reader may wish to return to and
analyse, we return to 13 Bf4:
13...f6! (Diagram 91)

156
 Diagram 91 (W)
The pin is rather awkward

14 Rae1?!
It was better to angle for a transposition to our last note with 14 Nbd4!? when 14...Ncxe5 (advocated by
Rizzitano) 15 Nxe5 (and not 15 h3?! g5! 16 Bg3 Nxf3+ when White lacked a good recapture and was just a
clear pawn down in J.Callaway-R.Reynolds, correspondence 1982) 15...Nxe5 (or 15...fxe5 16 Bg3 Qf7 17
Nb5!? e4 18 Bxe4 Nxh2 19 Kxh2 dxe4 20 Nd6 and White most certainly isn’t worse) 16 Rae1 does indeed
transpose. After 14 Nbd4, Black also has 14...Ngxe5!? when 15 Nxe5 fxe5 16 Nxc6 exf4 17 Ne5 Bc5! gives
White some but probably not enough compensation for the pawn. He should, though, prefer this to 15 Kh1?!
Qf7 16 Nxe5 fxe5 17 Bxe5 Nxe5 18 Qxe5 Bc7 19 Qe3 e5 and Black’s centre rolled forwards in J.Papai-
K.Schreiber, correspondence 1999.
14...fxe5 15 Bg3
Forced since both 15 Nxe5? Nxf2! 16 Rxf2 Rxf4 (S.Ivanov) and 15 Bxe5? Ncxe5 16 Nxe5 Rxf2 are
excellent for Black.
15...Qf7 (Diagram 92)

157
 Diagram 92 (W)
Black has a strong centre

16 Nh4?!
The Austrian IM tries to drum up complications on the kingside, but this only really helps Black. However,
I’m at a loss to suggest what White should have played: matters are already quite bad.
WARNING: 10...h6 is one of Black’s best defences to the Colle. Do not neglect it in your
preparation!
16...Qh5?
Even grandmasters can err at times. Perhaps surprised by how good his position already is, Novikov fails
to find 16...e4! 17 Qxg4 (17 Bxe4? dxe4 18 Qxg4 g5 – Oleinikov – wins the trapped knight) 17...exd3 18
Qg6 e5 19 Qxf7+ Rxf7 20 Bxe5 Bg4 with excellent activity and a large advantage.
17 h3 Nf6 18 Ng6 Qxe2 19 Rxe2?
Now White again runs into trouble due to a dark-squared bishop and ...Nh5 combination. Much better was
19 Bxe2 Ne4! (19...Rf7 20 Nxe5 regains the pawn with the ideal bind on e5 and White retains the advantage
after 20...Ne4 21 Nxf7 Nxg3 22 Nd6) 20 Bh4 when Bd3 and c4 ideas are on the table. Here Black’s best is
probably to give up an exchange with 20...Rf4!, but this is far from clear after something like 21 c4!? and
Danner had to try it.
19...e4! 20 Bb5 Nh5 (Diagram 93)

158
 Diagram 93 (W)
This annoying motif once more

21 Bd6
This leads to a bad ending, but 21 Bh4 Rf7 would also have left White a pawn down for pretty much
nothing.
21...Rf6 22 Ne7+ Nxe7 23 Bxe7 Ng3!
Simplifying and preparing to slowly advance that extra central pawn.
24 Bxf6 gxf6 25 Nd4 Nxe2+ 26 Nxe2 e5 27 Rd1 Be6 28 g4 h5! 29 Ng3 hxg4 30 hxg4 e3 31 fxe3 Bxe3+
32 Kg2 Kf7!
Now Black’s rook combines well with his bishop pair to begin a strong attack.
33 Nf5 Bf4 34 Nd6+ Kg6 35 Nxb7? Rh8 36 Nc5 Bxg4 37 Rxd5 Rh2+ (Diagram 94)

Diagram 94 (W)
159
 The attack is too strong

38 Kg1 Rxb2 39 Nd3 Be3+ 40 Kf1 Be2+ 41 Ke1 Rxb5 42 Rxb5 Bxd3 43 Rb3 f5 44 c4 e4 45 Rc3 f4 0-1

Points to Remember
1. 10...h6 is tricky; just look how quickly Danner stumbled into a bad position.
2. 11 e5 may have been dismissed by a number of sources, but it’s quite playable so long as its followed up
by 11...Ng4 12 Nb3 Bb6 13 Nbd4!.
3. The position after 13 Nbd4 remains extremely unexplored and untested. Any reader who undertakes some investigation of their own here
should reap the reward over the board.

Game 14
S.Mamedyarov-A.Volokitin
Calvia Olympiad 2004

Watson and Schiller wonder how long it is since ‘a Colle system has appeared in a Super-Tournament between world-class players’. The two
young talents in this game weren’t quite members of the world’s elite when this game was played, but both were rated over 2650. Evidently
Mamedyarov doesn’t feel that the Colle is only for the use of clubplayers.
1 d4 Nf6 2 Nf3 e6 3 e3 c5 4 Bd3 Nc6 5 0-0 d5 6 c3 Bd6 7 dxc5 Bxc5 8 Nbd2 0-0 9 e4 Qc7 10 Qe2 h6 11
Bc2 Bb6 (Diagram 95)

Diagram 95 (W)
Prophylaxis against e4-e5

Borrowing an idea from the Anti-Meran and this is useful prophylaxis against Nb3. Thus White can no
longer really consider 12 e5 (12...Ng4 13 Nb3 Ngxe5 14 Bf4 Nxf3+ 15 Qxf3 e5 16 Qd3 supplies
compensation, but the superior 14...f6! is somewhat better for Black).
12 a4!?
Seizing some space and continuing the waiting game, but ultimately this rare idea doesn’t lead to any
advantage. Depending on his taste, White might prefer to play an IQP position with 12 exd5 exd5 13 Nb3
Bg4 14 Qd3, although this version doesn’t promise any advantage after 14...Rfe8 15 Nbd4 and now either
15...Ne4 or 15...Rad8 (K.Srivachirawat-I.Papaioannou, Calvia Olympiad 2004).
White’s main and probably best try, albeit still rather unexplored, is 12 Kh1!? (an immediate 12 h3 is also
possible, but after 12...Nh5 13 Rd1 Nf4 14 Qf1 Bd7 15 Nb3 dxe4! 16 Bxe4 Rad8 White’s passively-placed
queen meant that Black was fine in M.Roelvaag-L.Johannessen, Oslo 2000) 12...Bd7 13 h3 (Diagram 96)

160
 Diagram 96 (B)
White prepares e4-e5

and finally 14 e5 is a threat. Practice has now seen:

a) 13...dxe4 is a little simplistic and 14 Nxe4 Nxe4 15 Qxe4 f5 16 Qe2 (probably best, although 16 Qc4!?
to prepare Bf4 and target e6 also deserves attention) 16...Rae8 17 Bb3! Na5 18 Ne5 Nxb3 19 axb3 Bc6 20
Bf4 gave White a small edge due to his control of e5 in B.Ferrufino-J.Molina, Santa Cruz 2005.
TIP: Remember that whenever Black plays ...f5, White needs to both prevent ...e5 and try to occupy
the tempting e5-outpost.
b) 13...Ne5!? 14 Nxe5 Qxe5 should be compared with Game 9 and this might well be slightly problematic
for White. After 15 f4 Qh5! 16 Qxh5 Nxh5, we can see that the inclusion of Kh1 and h3 isn’t helping. Thus in
E.Borroni-A.Vouldis, correspondence 2003, 15 Nf3 Qh5 16 e5 was preferred, but Black now exploited his
extra tempi with 16...Ne4! (Diagram 97)

161
 Diagram 97 (W)
A promising black pawn sacrifice

17 Be3 (17 Bxe4 dxe4 18 Qxe4 is perhaps critical, but after 18...Bc6 19 Qf4 f6! Black’s raking bishops
and activity supply dangerous compensation) 17...f6 18 Bxb6 axb6 19 Rfe1 fxe5 20 Bxe4 dxe4 21 Nd2 Qxe2
22 Rxe2 e3! 23 fxe3 Bb5 24 c4 Bc6 which was about equal; Black’s active pieces compensating for his
weakened structure.
c) 13...Nh5 (regardless!) 14 Nb3 Nf4 (14...Bxf2?! 15 Qxf2 dxe4 16 Nfd2 Ng3+ 17 Kg1 Nxf1 18 Nxe4! –
Ottow – is also good for White, but 14...dxe4 15 Qxe4 f5! looks like a clear improvement and should be
acceptable enough for Black) 15 Bxf4 Qxf4 16 exd5 exd5 17 Qd3 g6 18 Qxd5 Rad8 19 Qd2 won a pawn for
insufficient compensation in P.Etchegaray-S.Estremera Panos, Zaragoza 1992.
Returning to 12 a4:
12...a6 13 g3? (Diagram 98)

Diagram 98 (B)
White shouldn’t weaken himself so

A slightly mysterious move, but perhaps even the young Azeri felt compromised by the stupidly fast FIDE
time control imposed on this Olympiad. Preventing ...Nf4 once and for all was probably the idea; doing so is
nice in theory, but in practice is too time consuming and weakening. As Finkel points out, White could still
have gained a reasonable, if not too special, position with 13 h3.
13...Bd7 14 Kg2
No longer was 14 exd5?! exd5 15 Nb3 really an option due to 15...Bh3.
TIP: Unless forced to move it, try not to voluntarily advance the g-pawn in front of a castled king. It
can create serious weaknesses, unless there’s a bishop on f1 (or f8) ready to complete a delayed
fianchetto.
14...Rad8 15 h3 e5
A logical advance, although Finkel’s suggestion of 15...dxe4 16 Nxe4 Nd5 17 Rd1 Ba7 also looks very
comfortable for Black; his knights have targets and good squares on the queenside.
16 exd5 Nxd5 17 Nc4 Qc8! 18 Rh1 Bc7 (Diagram 99)

162
 Diagram 99 (W)
Black prepares to roll in the centre

Now Volokitin is ready to advance his e- and f-pawns, and with some effect. White cannot afford to dally
and Mamedyarov is forced to reveal why he has gained such a good reputation as a strong and resourceful
player.
19 Qe4! Nf6 20 Qh4 Bf5 21 Bxf5 Qxf5 22 Ne3 Qe6 23 Qc4!
Correctly continuing to play for exchanges to ease the pressure on White’s position.
23...Nd5?!
Despite White finally having woken up and activated his queen, 23...e4! 24 Nd2 Qe7 (Finkel) would still
have left Black with some advantage; note how White continues to suffer from the weaknesses created by the
careless 13 g3.
24 Nxd5 Rxd5 25 Re1 Bb6 26 Be3 e4! (Diagram 100)

163
 Diagram 100 (W)
White must be careful

27 Ng5?
It’s easy to see why Mamedyarov didn’t like the influence of the pawn arriving on f3 with 27 Bxb6 exf3+
28 Kh2 Qd7 (28...Ne5 29 Qc7 also makes it hard for Black to coordinate), but after 29 Qe4 White defends.
The text is tempting, but merely succeeds in strengthening Black’s attack.
27...hxg5 28 Bxb6 Ne5 29 Qxe4 f5 30 Qc4 Rf6! 31 Qb3 f4 32 g4 f3+ 33 Kg1 Rf4! 34 Rxe5!
Good practical defence to keep the game live. Instead 34 Kf1? would have allowed 34...Kh7! (34...Kh8 35
Bd4 Rxg4! isn’t quite as good since White can fight on with 36 Qxd5! Qxd5 37 hxg4 Qc4+ 38 Kg1 Qe6 39
Rxe5 Qxg4+ 40 Kf1 Qg2+ 41 Ke1 Qh1+ 42 Kd2 Qxa1 43 Rxg5) 35 Bd4 (the only real way to prevent
35...Rxg4 is 35 Rxe5, but with 35...Rxe5 36 Qxe6 Rxe6 Black gains a somewhat superior version of the
game) 35...Rxg4! (Diagram 101)

Diagram 101 (W)

Destroying White’s defences

36 Qc2+ Kh8 37 hxg4 Qh6! and mate follows.

34...Rxe5?!
White’s bishop is pretty powerful after the text, so probably Black should have preferred Finkel’s
suggestion of 34...Rxg4+! 35 Kf1 Rxe5 36 Qxe6+ Rxe6 37 hxg4 Rxb6 38 Ra2 Rh6! 39 Ke1 Rh4 with a better
rook ending.
35 Qxe6+ Rxe6 36 Be3 Rfe4 37 Kh2 (Diagram 102)

164
 Diagram 102 (B)
White has enough for the exchange

37...Rxe3!
Perhaps a surprising decision, but definitely a good one. White’s bishop was very well stationed on e3 and
with at least one of f3 and g5 about to fall, Mamedyarov was set to land up with two pawns and good play for
the exchange.
38 fxe3 ½-½
Following 38...Rxe3 39 Kg3 f2+ 40 Kxf2 Rxh3 the draw becomes obvious; Black’s rook is too active and his
king too quick to centralize for White to do anything positive.

Points to Remember
1. Make sure you have a line ready for 10...h6. Even the 2660-rated Mamedyarov didn’t and quickly drifted
into a bad position.
2. Should something go wrong in the opening, don’t lose hope. As superbly demonstrated by the young Azeri,
sterling resistance can take one far in chess.

165
Chapter Three
The Colle System: Black Plays ...c5 and ...Nbd7
...Bd6 Set-ups
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nbd7 (Diagram 1)

Diagram 1 (W)
Black plans ...b6 or ...e5

We’ve now covered the main line of the Colle (according to ChessBase’s Big Database 2007, 5...Nc6 is
four times as popular as 5...Nbd7), but the text represents an important alternative. From d7, as opposed to c6,
the knight doesn’t place d4 under any pressure and it obstructs the development of Black’s light-squared
bishop, making a queenside fianchetto quite likely. There are, though, some benefits too from the knight’s
position on d7. One is that the knight supports its f6-colleague and may be able to rush to that square (thereby
keeping the important h7-square covered), especially after a double exchange on e4. Even more importantly,
from d7 the knight overprotects the c5-pawn.
WARNING: In these ...Nbd7 lines White should rarely capture on c5: ...Nxc5 comes with tempo
against the d3-bishop and increases Black’s control over the key e4-square.
6 Nbd2 Bd6
Continuing to cover the e5-square, just as in the ...Nc6 variation, both to prevent Ne5 and to prepare a
possible ...e5-break. The more modest 6...Be7 is, though, an important alternative and one discussed in the
last two sections of this chapter.
7 0-0 (Diagram 2)

166
 Diagram 2 (B)
Development as usual from White

7...0-0
As we saw in Chapter One, 7...e5?! 8 dxe5 Nxe5 9 Nxe5 Bxe5 10 e4! shows up ...e5 to be a premature
advance. Here it’s also possible to consider 8 e4!? when 8...dxe4 9 Nxe4 Nxe4 10 Bxe4 is a little better for
White, while 8...0-0 9 Re1! leads back to our main line below.
Black might also further his control of e5 with 7...Qc7 when 8 Re1 0-0 transposes to the ...Nbd7 and ...Bd6
main line. White can also consider 8 Qe2!? 0-0 (8...e5?! is rebuffed by 9 dxe5 Nxe5 10 Nxe5 Bxe5 11 f4!
Bd6 12 e4 – Smith and Hall) 9 e4 cxd4 10 cxd4 dxe4 11 Nxe4 when his own queen is ready to recapture on
e4, whereas Black’s is a little vulnerable on c7.
8 Re1! (Diagram 3)

Diagram 3 (B)
167
 The best way of preparing e3-e4

TIP: Remember that Black has played ...Nbd7 not ...Nc6!

That may sound obvious, but another important difference becomes apparent after the inaccurate 8 Qe2?!
(a move which is playable with the knight on c6). Black replies 8...e5! (Diagram 4)

Diagram 4 (W)
White’s queen is misplaced

when 9 e4?! is good for Black since 9...cxd4 10 cxd4 exd4 11 e5? runs into the pinning 11...Re8. Note too
that 9 dxc5?! is, of course, now met by the strong 9...Nxc5!, and 9 dxe5 Nxe5 10 Nxe5 Bxe5 reaches a
position in which, as I remarked in Chapter One, ‘White [is] the side looking to equalize’. An example: 11
Nf3 Bg4 12 h3 Bxf3 13 Qxf3 c4!? 14 Bc2 Qe7 15 Rd1 Rad8 16 a4 Rfe8 17 Qe2?! Bc7! 18 Bd2 h5 19 Be1
Qe5 and White remained hampered by his bad bishop in R.Weemaes-A.Goloshchapov, Le Touquet 2004; a
good example of how to punish passive white play.
8 Re1 is not White’s only option, and fans of the IQP can again prefer 8 e4. Now both 8...dxe4 9 Nxe4
Nxe4 10 Bxe4 cxd4 11 cxd4 and 8...cxd4 9 cxd4 dxe4 10 Nxe4 Nxe4 11 Bxe4 lead to an IQP position and the
subject of Game 15. White should also be aware that Black doesn’t have to race to exchange on d4, but
8...dxe4 9 Nxe4 Nxe4 10 Bxe4 (Diagram 5)

168
 Diagram 5 (B)
An IQP set-up beckons

10...Nf6 (note how the knight is better placed here than on c6) 11 Bg5 (Colle himself preferred 11 Bc2, but
this is well neutralized by 11...h6!; for example, 12 Qd3 Qc7 13 Be3 Rd8 and Black comfortably followed up
with ...b6 and ...Bb7 in E.Gonzalez Haro-A.Rodriguez Marquez, Malaga 1984) leaves Black, due to the pin
and slightly problematic c8-bishop, without a good alternative to 11...cxd4 12 cxd4 Be7 (12...h6 13 Bxf6
Qxf6 is a slightly risky alternative; Black will now struggle to develop his light-squared bishop and 14 Qe2
Rb8 15 Ne5! left him a little worse in C.Guimard-J.Rubinetti, Chacabuco 1968) 13 Bc2 and we’ve entered
Game 15 via yet another move order.
Returning to White’s main option, 8 Re1:
8...e5?!
A tempting advance, but it turns out that White can exploit the lack of pressure against d4, compared with
when the knight is on c6. Instead 8...Qc7, taking further control of e5, is Black’s main move and the subject
of our next section (where we’ll also discuss the related 8...Qb6!?). Black might also try:
a) 8...c4 is a little better than in Game 1 since Black’s knight is better placed on d7 than c6, but White
retains an edge after 9 Bc2 b5 10 e4 dxe4 (10...Bf4?! 11 e5 Ne8 12 Nf1 Bxc1 13 Qxc1 gives White a
typically good French and then 13...h6 14 Qf4 f5 15 exf6 Qxf6 16 Qg3 left the master of strategy struggling
against White’s e5-outpost in D.Przepiorka-A.Nimzowitsch, Frankfurt 1930) 11 Nxe4 Nxe4 12 Bxe4 Rb8 13
Qe2 (R.Piantoni-F.Cedriano, Bratto 2005); note that as well as keeping his kingside covered, Black must also
ensure that the desirable ...b4-advance is not rushed and thereby cost him c4.
b) 8...cxd4?! 9 exd4 is not, however, an improved version of Game 2. If anything it’s an inferior one since
Black has lost the ...Nf6-d7 resource. Colle now demonstrated how White should exploit his control of e5 and
spatial advantage in E.Colle-F.Schubert, Scarborough 1928: 9...Qc7 10 Qe2 Re8 (or 10...b6 11 Ne5 Bb7 12 f4
and we’ve reached Reshevsky-Kevitz in the Introduction) 11 Ne5 Nf8 12 Ndf3!? N6d7 13 Ng5! (Diagram 6)

169
 Diagram 6 (B)
A typical and strong attacking motif

(a strong attacking set-up unless the tactics after ...f6 don’t work for White...) 13...f6? (...and here they do;
Black had to grovel instead with 13...Re7) 14 Qh5! (threatening 15 Bxh7+ and ensuring both knights are
immune: 14...fxg5 15 Bxh7+ Nxh7 16 Qf7+ Kh8 17 Ng6 is mate, and 14...fxe5 15 Bxh7+ Nxh7 16 Qxh7+
Kf8 17 Qh8+ Ke7 18 Qxg7+ Kd8 19 Nf7+ also leads to the same outcome) 14...g6 15 Nxg6! (Diagram 7)

Diagram 7 (B)
The attack crashes through

15...fxg5 (or 15...hxg6 16 Bxg6 fxg5 17 Bf7+ Kg7 18 Bxe8 when Black can scarcely move and a decisive
rook lift is imminent) 16 Nxf8 Nf6 17 Qxg5+ Qg7 18 Nxh7! 1-0. White’s last ensured he remained three
pawns ahead since 18...Nxh7 19 Bxh7+ Kxh7 20 Qh5+ would have cost Black the exchange.
170
 TIP: Never forget that h7 is the source of many tactics in the Colle!

c) 8...Re8 9 e4 can transpose after 9...e5 to the note to Black’s 9th move in our main line, below. Also
occasionally seen is 9...dxe4 10 Nxe4 Nxe4 11 Bxe4! (11 Rxe4!? would be the move, as we will shortly see,
had Black played ...Qc7, rather than ...Re8; but here Black is better placed defensively and should be OK
after, for instance, 11...b6!?, although this remains untested) 11...cxd4 when White could just recapture on d4,
but Colle himself preferred something much more dramatic (I’m sure you can guess what!), as we’ll see in
Game 16.
In any case, Black doesn’t appear to be able to equalize after 11 Bxe4; for example, 11...Nf8 (Euwe’s
suggestion) 12 dxc5 Bxc5 13 Bf4 (Bronznik) supplies a pleasant edge, as did 11...Be7 12 Ne5! (simple and
strong; once again the ending is unpleasant for Black) 12...Nxe5 13 dxe5 Qxd1 14 Rxd1 (Diagram 8)

Diagram 8 (B)
White is much better

14...Rb8 15 Be3 b6 16 Rd2 Bb7 17 Bxb7 Rxb7 18 Rad1 Reb8 19 f4 and White’s space advantage
eventually prevailed in M.Franklin-G.Hughes, correspondence 1986.
d) 8...b6 is a non-theoretical, developing and pretty sensible approach, albeit one probably not quite
sufficient to equalize after 9 e4 dxe4 10 Nxe4 (Diagram 9):

171
 Diagram 9 (B)
Black doesn’t have to exchange on d4

d1) 10...Be7 11 Qe2 (11 Bf4 Bb7 12 Nd6 Qc7! 13 dxc5 Nxc5 14 Nxb7 Qxf4 15 Nxc5 Bxc5 16 Qe2 Rfd8
didn’t lead anywhere for White in A.Yegiazarian-A.Pashikian, Armenian Championship, Yerevan 2006, but
the 11 dxc5!? Nxc5 12 Nxc5 Bxc5 13 Bg5 Bb7 14 Qe2 Be7 15 Rad1 of D.Baudot-C.Pergericht, Cannes
1989, is a good alternative to retain the initiative) 11...cxd4 (11...Bb7 leaves White with the extra move Re1
compared with a line we will shortly examine, namely 6...Be7 7 0-0 0-0 8 Qe2 b6 9 e4 dxe4 10 Nxe4 Bb7; he
could now transpose with 12 Rd1 or try Bronznik’s idea of 12 Ng3!?) 12 Nxd4 Bb7 13 Bg5 Nxe4! (and not
13...Bxe4? 14 Bxe4 Nxe4 15 Nc6! and White won material in S.Marjanovic-P.Popovic, Novi Sad 1985; a
fiendish and useful trick to remember) 14 Bxe4 Bxe4 15 Bxe7 Qxe7 16 Qxe4 Nf6 (Bagirov); here Black has
equalized, and so White should probably prefer the rare 11 dxc5!?.
d2) Black might also try 10...Nxe4 11 Bxe4 Rb8 and now the natural 12 Bc2 Bb7 13 Qd3 g6 14 Bh6 was a
little better for White in C.Lakdawala-V.Akobian, Los Angeles 2001.
WARNING: Instead of 12 Bc2, 12 Bxh7+? might have tempted a few alert readers since 12...Kxh7 13
Ng5+ Kg8 14 Qh5 Nf6 isn’t the end of the story as 15 Qh4 threatens a decisive rook lift, but Black
defends and wins with 15...cxd4 16 cxd4 Qc7! 17 Re3 Qc2 (Oleinikov).
Returning to 8...e5:
9 e4! (Diagram 10)

172
 Diagram 10 (B)
Increasing the central tension

First played in E.Colle-M.Duchamp, Paris 1929, this is a powerful advance. White makes good use of the
e1-rook and, as Colle realized, the early liquidation of the centre is in his favour.
9...cxd4
An important alternative is 9...Re8, supporting e5 and preparing ...exd4 (which if played immediately loses
a piece to 10 e5). White must then begin to trade pawns and he retains the advantage with 10 exd5 cxd4 (or
10...exd4 11 cxd4 Rxe1+ 12 Qxe1 and Black is curiously a little short of a good move: 12...Nxd5?! runs into
13 Nc4 Be7 14 Bxh7+! Kxh7 15 Qe4+, and 12...cxd4 13 Ne4! Qb6 14 Bg5 Nxd5 15 Nxd6 Qxd6 16 Qe8+
Qf8 17 Re1 – Bronznik – also leaves Black struggling) 11 cxd4 exd4 12 Ne4 (Diagram 11)

Diagram 11 (B)
d4 is weaker than d5
173
 12...Ne5?! (12...Nxe4 13 Bxe4 improves, but also isn’t fully convincing for Black: 13...Nf6 14 Bg5 h6 15
Bxf6 Qxf6 was A.De Santis-P.Kiss, Budapest 1998, and now, as well as the game’s 16 Qxd4, 16 Bh7+!? Kf8
17 Rxe8+ Kxe8 18 Be4 was a good try to retain the advantage; likewise Bronznik’s 13...Ne5!? 14 Bf4!
Nxf3+ 15 Qxf3 Bxf4 16 Qxf4 favours White since he possesses far the superior d-pawn) 13 Bg5 Bg4
(A.Kotov-B.Levitas, Leningrad 1938) 14 Bb5! Bxf3 (or 14...Nxf3+?! 15 gxf3 Rxe4!? 16 Rxe4 Qb6 17 Bxf6
and White remains material ahead, as he does too after 14...Bb4?! 15 Qxd4! Bxe1? 16 Nxe5) 15 gxf3
(Bronznik) and White’s bishops are too powerful; for example, 15...Rf8 16 Nxd6 Qxd6 17 Bf4 Qb6 18 Bxe5
Qxb5 19 Qxd4 with a huge advantage.
Black can also support e5 with his queen, but this is no panacea: 9...Qc7 10 exd5 exd4 11 cxd4 cxd4
(11...Nxd5 is also well met by 12 Ne4) 12 Ne4 Nxd5 13 Nxd6 Qxd6 14 Re4! N7f6 15 Rxd4, for example,
gave White an edge with his bishop pair in J.Flieger-R.Kehl, Bad Kuehlungsborn 1964.
10 cxd4 (Diagram 12)

Diagram 12 (B)
White has the advantage

NOTE: This position can also arise from a line of the Slav!

The line in question being 1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 e3 e6 5 Nbd2 Nbd7 6 Bd3 Bd6 7 0-0 0-0 8 e4 e5 9
cxd5 cxd5 10 Re1.
10...dxe4
Not forced, but both 10...Qc7 11 exd5 exd4 and 10...Re8 11 exd5 exd4 transpose to positions we
considered in the notes to Black’s 9th.
11 Nxe4 Nxe4 12 Rxe4! (Diagram 13)

174
 Diagram 13 (B)
A powerful recapture on e4

The best and most aggressive continuation.

TIP: In the ...Nbd7 variation, the concept of an early rook lift is even more useful and effective than
those which we saw in the ...Nc6 lines.
Colle’s initial idea was instead 12 Bxe4 exd4 13 Qxd4, but Black can equalize with 13...Bc5! (improving
over 13...Qf6 14 Bg5! Qxd4 15 Nxd4 Nc5 16 Bd5! with the more active pieces and advantage in E.Colle-
G.Thomas, Paris 1929) 14 Qc3 Qb6 15 Be3 Bxe3 16 Rxe3 Nf6, as he did in K.Steijn-K.Engel,
correspondence 1986.
Returning to 12 Rxe4! and we can see that White has some initiative: e5 is attacked and Rh4 a strong
possibility. The most common continuation in practice has been 12...f5 13 Rh4!, which we’ll see in Game 17,
but also possible is 12...exd4 (or 12...Nf6, as in one game, when 13 Bg5! Bf5 14 Bxf6 gxf6 15 Re1 leaves
White with a clear structural advantage) 13 Rxd4 Bc5 14 Rd5! (a strong departure from the usual 14 Rh4, as
in this case Bronznik’s 14...Nf6 offers Black good defensive prospects) 14...Qb6 15 Qc2 g6 16 Bf4 (Diagram
14)

175
 Diagram 14 (B)
White is much the more active

and White’s far greater activity was obvious and gave him a clear advantage in R.Callergard-M.Unander, correspondence 1994.

Theoretical Conclusion
Against a ...Nbd7 and ...Bd6 set-up, White must strive to quickly open the centre with Re1 and e4. 8...e5 9 e4 only seems to play into his hands,
but neither do Black’s 8th-move alternatives fully solve his problems, although 8...b6 deserves more attention.

Illustrative Games

Game 15
V.Smyslov-O.Romanishin
Tbilisi 1976

1 d4 Nf6 2 Nf3 e6 3 e3 c5 4 Bd3 d5 5 c3 Nbd7 6 0-0 Bd6 7 Nbd2 0-0 8 e4 cxd4 9 cxd4 dxe4 10 Nxe4 Nxe4 11 Bxe4 (Diagram 15)

176
 Diagram 15 (B)
White accepts an IQP

11...h6!
Preventing White’s natural dark-squared bishop development to g5 makes a lot of sense. White will now
double on the b1-h7 diagonal, but it’s then not at all easy to undermine the f6-knight.
More popular has been 11...Nf6 when 12 Bg5 (or 12 Bc2 b6 13 Bg5 and now 13...Be7 is probably best,
transposing to 12 Bg5, whereas 13...Bb7 14 Qd3 g6 15 Bb3 Be7 16 Bh6 Re8 17 Ba4! Nd7 18 Qe3 a6 19 Qf4
saw Black coming under some pressure, although his defence remained solid – ...Bf8 is a key resource – in
P.Svidler-R.Gasimov, Leningrad 1990) 12...Be7 13 Bc2 b6 14 Qd3 (Diagram 16)

Diagram 16 (B)
A typical attacking battery

14...g6 15 Bh6 Re8 16 Ne5 Bb7 17 Rfe1 Qd6 18 a3 Rac8 was a typically unbalanced IQP position in
F.Izeta Txabarri-M.Illescas Cordoba, Los Yebenes 1990. The leading Spanish GM had here avoided any
early tactical blows, but the position remains quite complex and who to prefer is really a matter of taste.
Another option, and one which caused Edgard Colle some problems, is 11...Qb6 after which 12 Bc2 Nf6
13 Qd3 Bd7 (Diagram 17)

177
 Diagram 17 (W)
White mustn’t be too ambitious

14 Ne5? (White had to take time out for 14 a4 – Kagan) 14...Bb5 15 Qh3?! (continuing his plan, although
only losing a pawn with 15 Nc4 Qc7 was also very grim for White) 15...Qxd4! (but not 15...Bxf1?? 16 Nd7!
when White wins, as Colle was no doubt hoping for) 16 Bg5 Qxe5 saw White shortly have to resign in
E.Colle-E.Bogoljubow, Berlin 1926.
WARNING: The Colle allows one some beautiful tactical opportunities, but these usually only work
when one is in control of the position. Beware both playing too ambitiously too early and being seduced
by beautiful tricks, as Colle was here.
Colle’s improvement against 11...Qb6 was 12 Qe2 Nf6 13 Bd3 Bd7 after which 14 Bg5 (14 Rd1 is less
challenging and didn’t lead anywhere after 14...Ba4 15 b3 Bc6 16 Ne5 Nd5 in E.Colle-E.Bogoljubow,
Scarborough 1927; the exchanges and Black’s control of d5 balance out White’s ability to exchange either
bishop) 14...Be7 15 Ne5! is quite dangerous for Black. In the stem game, L.Majstorovic-I.Maricic,
correspondence 1954, he collapsed with 15...Rfd8? 16 Bxh7+! and White won a pawn since 16...Kxh7 fails to
17 Bxf6 Bxf6 18 Qh5+ Kg8 19 Qxf7+ Kh8 20 Nxd7. Black does better with 15...Bc6 16 Rac1 Rac8, although
after 17 Rc3! I still like White’s attacking prospects, even if this is objectively unclear.
Returning to Romanishin’s prophylactic 11...h6 (Diagram 18):

178
 Diagram 18 (W)
Black rules out Bg5

12 Bc2 b6 13 Qd3 Nf6 14 Re1 Bb7

Deviating from an earlier game, A.Konstantinopolsky-I.Veltmander, Tartu 1950, which quickly caught
fire: 14...Qc7!? 15 Bb3 (15 Ne5 now runs into 15...Ba6!) 15...Rd8 16 Bd2 Bb7 17 Rac1 Qb8 18 Bxh6!
(otherwise Black forces a favourable exchange with 18...Bf4) 18...Ng4! (Black is up to the task, whereas
18...gxh6? 19 Rxe6! would have given White a crushing attack; the rook is taboo due to the check on g6 and
19...Kg7 20 Rxf6! Kxf6 21 Qh7 is also extremely strong) 19 Ng5! Bxh2+ 20 Kf1 Ba6! (two can play at the
sacrificial game!) 21 Qxa6 Qf4 22 Qe2 Nxh6 23 Nh3 Qh4 24 Qe4 1/2-1/2.
15 Ne5 (Diagram 19)

Diagram 19 (B)
Unbalanced and roughly even
179
 15...Rc8
More recently 15...Bxe5!? 16 Rxe5 (16 dxe5 Qxd3 17 Bxd3 Rfd8 18 Be2 Nd5 is at least equal for Black
due to his strong centralized knight) 16...Rc8 17 Bf4 Ba6 18 Qd2 Ng4! 19 Ree1 Qh4 was also fine for Black,
C.Umetsubo-F.El Debs, Araraquara 2002.
NOTE: After 11...h6 by crucially playing actively, Black can thwart White’s attacking hopes and
maintain rough equality.
16 Bf4 Bb4 17 Re2
White would prefer to swing a rook, but 17 Re3?! Qd5! is awkward since 18 Rg3? runs into 18...Rxc2! 19
Qxc2 Nh5.
17...Qd5 18 f3!
Good defence from Smyslov who is now ready to begin to drive Black backwards with both Bb3, and a3
and b4. Before that can happen, the young Romanishin elects to simplify and soon gains a fairly easy draw
against the ex-World Champion.
18...Ba6!? 19 Qxa6 Qxd4+ 20 Be3 Qxe5 21 Bxh6! Qh5 22 Bf4 Bc5+ 23 Kh1 Nd5! (Diagram 20)

Diagram 20 (W)
Black has reasonable activity

After a fairly forced passage of play, Black is left with some useful central activity and White is forced
into further exchanges.
24 Bg3 Bd4 25 Be4 Qg5 26 Bxd5! exd5!? 27 Rae1 Rc1 28 Qd3 Bf6 29 f4 Rxe1+ 30 Rxe1 Qh5 ½-½
Black’s more active bishop fully compensates for his IQP.

Points to Remember
1. Just like against ...Nc6 and ...Bd6, so too is 8 e4 a reasonable option against ...Nbd7 and ...Bd6 for those
who enjoy attacking with the IQP.
2. The resulting IQP positions are unbalanced, but in general they are a little more comfortable for Black
than in Game 3 since his remaining knight can easily come to a good square (f6).
3. Objectively, 11...h6 is a slight theoretical worry for White, but such prophylaxis remains rare and
11...Nf6 is much more typical, especially at club level.
4. Should Black gain good activity (...Bb7 and ...Qd5 is often an effective set-up), White should be prepared to make exchanges to reduce

180
the pressure.

Game 16
E.Colle-J.O’Hanlon
Nice 1930

1 d4 d5 2 Nf3 Nf6 3 e3 c5 4 c3 e6 5 Bd3 Bd6 6 Nbd2 Nbd7 7 0-0 0-0 8 Re1 Re8 9 e4 dxe4 10 Nxe4 Nxe4 11 Bxe4 cxd4 (Diagram 21)

Diagram 21 (W)
Can h7 really be captured?

12 Bxh7+!
Over the years 12 Qxd4 has often been suggested as a better try, since a number of commentators have
considered the Greek Gift to be simply unsound or to lead to a draw at best. However, as was first shown by
Stefan Ottow in Kaissiber, Colle’s intuition was entirely correct. His sacrifice is at least extremely dangerous
and may well be decisive, although an extremely powerful man and machine team may still discover some
sort of defence in the fascinating notes to Black’s 13th.
12...Kxh7 13 Ng5+ (Diagram 22) 13...Kg6?

181
 Diagram 22 (B)
Black must be accurate

A surprising choice with hindsight, but I suspect that O’Hanlon overlooked White’s 15th. As most
annotators have agreed, 13...Kg8 would have been a better defence, but Black must still tread carefully after
14 Qh5:
a) It’s important to realize that f7 is attacked as well as h7, and 14...Nf6? loses to 15 Qxf7+ Kh8 16 Re4!,
overloading the f6-knight.
b) Neither can Black defend with 14...Ne5?! due to 15 Rxe5! Bxe5? 16 Qxf7+ Kh8 17 b3! (Ed.Lasker) and
the bishop arrives on a3 with crushing effect since 17...Bd6 allows 18 Qh5+ Kg8 19 Qh7+ Kf8 20 Qh8+ Ke7
21 Qxg7mate.
c) Harvey’s 14...Bxh2+!? runs into 15 Kh1! Qf6 16 Qh7+ Kf8 17 Ne4 Qf5! 18 Qh8+ Ke7 19 Qxh2 when
White retains a strong attack, such as with 19...d3 20 b3 Kd8 21 Nd6 Qxf2 22 Be3 Qb2 23 Nxf7+ Ke7 24
Ng5!.
d) Thus the only defence is 14...Qf6! 15 Qh7+ Kf8 16 Ne4 Qe5 17 f4! (ruling out a counter-sacrifice on
h2) 17...Qd5 18 c4 (Diagram 23)

182
 Diagram 23 (B)
White has promising compensation

and now Black has two main tries, but as shown by Ottow neither is sufficient. Here’s a quick summary of
the key points:
d1) 18...Qc6 19 Qh8+! (Nunn’s 19 f5!? Nf6 20 Nxf6 gxf6 21 Bg5! Ke7 22 Bxf6+ Kxf6 23 Qh4+ Kg7 24
f6+ Kg6 25 Re4 probably gives White no more than a draw after 25...Bxh2+! 26 Kxh2 e5 27 Rf1 Rh8 28
Qxh8 Qxe4 29 Qg7+ Kh5 30 Qxf7+ Qg6 – Oleinikov) 19...Ke7 20 Qxg7 Bb4!? 21 f5! Kd8 22 Bg5+ Kc7 23
Qxf7 Bxe1 24 Rxe1 Rf8 25 Qxe6 and, as shown by Ottow, White either regains his material with some
advantage or retains a strong attack; for example, 25...Qxe6 26 fxe6 Nb6 27 e7 Rh8 28 Bf4+ Kc6 29 Nf6!
(Bronznik) or 25...a5!? 26 Rd1 Kb8 27 Be7 Rh8 28 Rxd4 Ka7 29 Nc3.
d2) 18...Qa5 19 Bd2 Qc7 20 Qh8+ Ke7 21 Qxg7 (Diagram 24)

Diagram 24 (B)
183
 The attack grows in strength

21...Rf8 (or 21...d3 22 Ng5! and f7 falls leaving White doing well; 21...Bxf4 22 Ng5! Bxh2+ 23 Kh1 is
also pretty strong) 22 f5! Bxh2+ 23 Kf2! Kd8 24 g3 and, like Ottow and Bronznik, I haven’t been able to find
a defence for Black.
Of course, there are many other defensive tries for Black which there simply isn’t space to mention, but it
should be clear that White’s attack is very strong.
TIP: Should any reader find this particular sacrifice, or indeed any position in this book,
particularly fascinating, a good way to both explore it and improve one’s overall play would be to study
it in conjunction with a computer engine, like Fritz.
14 h4! (Diagram 25)

Diagram 25 (B)
A strong advance of the h-pawn

This had to be foreseen before sacrificing.

NOTE: As we’ve seen before, the h-pawn can be a key component in any Greek Gift when the black
king runs forwards.
14...Rh8?
Halting the h-pawn, but allowing a lovely and unexpected blow. A better defence was 14...f5, but White
retains a very strong attack after 15 h5+ Kf6 16 Qxd4+ Be5 17 Qh4 Qa5! (with both 18 Bf4 and 18 f4
threatened, this is again the best try; instead 17...Ke7 is routed by 18 Bf4 Bf6 19 Rxe6+ Kf8 20 h6!) 18 b4!?
(18 f4 is also good since 18...Bc7 permits 19 Nxe6+! Kf7 20 Ng5+ Kf8 21 Rxe8+ Kxe8 22 Qe1+ Kf8?! 23
Qe6 – Oleinikov) 18...Qd5 (or 18...Qa6 19 Rxe5! Nxe5 20 Nf3+ Kf7 21 Nxe5+ Kg8 22 h6 with a mating
attack) 19 Bb2 (Winants) and the threats of 20 c4 and 20 f4 are crushing.
15 Rxe6+!! (Diagram 26)

184
 Diagram 26 (B)
A crushing follow-up sacrifice

15...Nf6
Black would have been mated after 15...fxe6 16 Qd3+ and then either 16...Kf6 17 Qf3+ Bf4 18 Qxf4+ Ke7
19 Qf7+ Kd6 20 Qxe6+ Kc7 21 Bf4+ or 16...Kh5 17 g4+ Kxg4 18 Qf3+ Kxh4 19 Qh3mate.
16 h5+ Kh6
Alternatively, 16...Rxh5? removes a key flight square and permits 17 Qd3+ Kh6 18 Qh7mate, but
O’Hanlon is hardly going to escape after the text.
17 Rxd6
17 Qd3!, threatening to mate after 18 Nxf7+, was even better and more aesthetic, but Colle’s choice is
plenty good enough.
17...Qa5 18 Nxf7+ Kh7 19 Ng5+ Kg8 20 Qb3+ 1-0

Points to Remember
1. In this computer era, we can finally conclude that O’Hanlon’s 11...cxd4 was inaccurate and that Colle was correct to punish it with a Greek
Gift.
2. One useful pointer when weighing up a Greek Gift is to see if White can capture on f7 when the black
king retreats (being able to do so is often decisive).
3. Should the black king edge forwards to g6, as well as Qg4 and Qd3+ options, always remember to
consider too an advance of the h-pawn.
4. It’s often by no means possible to calculate everything when considering a sacrifice; you must instead try to learn to trust your intuition.

Game 17
R.Piantoni-F.Popa
Verona 2002

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 Nbd7 5 Nbd2 c5 6 c3 Bd6 7 0-0 0-0 8 Re1 e5?! 9 e4 cxd4 10 cxd4 dxe4 11 Nxe4 Nxe4 12 Rxe4 f5?
As we’ve seen 12...exd4 13 Rxd4 isn’t especially nice for Black, but the loosening, if tempting, text is
even worse.
13 Rh4! (Diagram 27)

185
 Diagram 27 (B)
A typical and effective rook swing

Superior to the also dangerous 13 Bc4+ Kh8 14 Rh4 since then 14...Be7! 15 Rh3 h6 (Bronznik) is a tough
defence.
13...Be7
Gaining a tempo to win control of the g5-square, but even this isn’t sufficient to save Black.
NOTE: 13 Rh4 tempts 13...e4, but that only plays into White’s hands: 14 Bc4+ Kh8 15 Ng5 h6 16
Qh5 is very strong; for example, 16...Qf6 17 Nf7+ Kh7 (or 17...Rxf7 18 Bg5!) 18 g4!? leaves Black
struggling.
White also has threats against e5 which 13...Qf6 does nothing to alleviate: 14 Bc4+ Kh8 15 dxe5 wins in
style after 15...Bxe5 (or the more mundane 15...Nxe5 16 Nxe5 Qxh4 17 Qxd6) 16 Rxh7+! Kxh7 17 Nxe5
(Diagram 28)

186
 Diagram 28 (B)
There’s no good defence

17...Qh4 18 g3 (Bronznik). Finally two prophylactic tries also fail to succeed: 13...h6 14 Bc4+ Kh8 15
Bxh6! gxh6 16 Rxh6+ Kg7 17 Rxd6 is crushing and 13...Kh8!? 14 Ng5 h6 15 Qh5 also strong, for example,
15...Qf6 16 Bc4 Be7 17 Rh3 f4 18 Nf7+ leaves Black with nothing better than to give up a clear exchange on
f7.
14 Rh3!
TIP: This line is extremely enjoyable for White to play; the threat now being 15 Bc4+ Kh8 16 Nxe5
Nxe5 17 Rxh7+ and mate on h5.
14...e4?!
This does nothing to prevent the knight’s planned advance, but after, say, 14...h6!?, 15 Qb3+ Kh8 16 Nxe5
Nxe5 17 dxe5 f4 18 Rf3 leaves White a pawn ahead.
15 Bc4+ Kh8 16 Ne5! (Diagram 29) 16...Qe8

Diagram 29 (B)
f7, g6 and h7 are vulnerable

Trying to cover the h5-square, but this runs into a cute interference tactic. However, in any case I think
Black was defenceless; for example, 16...Bh4 17 Qh5 Nxe5 18 dxe5 g5 19 Rxh4! gxh4 20 Bg5 and White’s
bishops wreck havoc.
17 Bf7! Rxf7 18 Ng6+ Kg8 19 Rxh7! (Diagram 30)

187
 Diagram 30 (B)
Finishing in style

Beautiful. Now the only way to avoid mate is 19...Rf6, but then Black loses his queen and the game.
19...Rf8 20 Qh5 1-0
Amazingly this exact game was later repeated in R.Piantoni-F.Berni, Bratto 2002; Signor Piantoni is a
lucky man!

Points to Remember
None really, bar to hope that one day you too will get to face 12...f5! Just remember that 13 Rh4! leads to a
very powerful attack, and enjoy working out all White’s neat tactical points at the board.

188
The ...Bd6 Main Line
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nbd7 6 Nbd2 Bd6 7 0-0 0-0 8 Re1 Qc7 (Diagram 31)

Diagram 31 (W)
Now 9...e5 is a threat

Black’s main move, but he can also deploy his queen with Keres’ 8...Qb6. This looks somewhat less
harmonious, but has been surprisingly popular in the correspondence world.
TIP: Leading correspondence games are an excellent source of opening ideas, since the players often
prepare their openings in meticulous detail.
With 8...Qb6, Black hopes that the increased pressure against d4 might deter e4, but White should not be
dissuaded! After 9 e4! dxe4 10 Nxe4 Nxe4 11 Rxe4 (Diagram 32)

Diagram 32 (B)
189
 Again the correct recapture

the position should be compared to the note to Black’s 9th move below. White gains a strong attack after
11...Nf6?! 12 Rh4! (12...g6 13 Ne5 Be7 14 Qf3 Nd7 15 Rxh7! was the powerful continuation of G.Sprott-
A.Nisbet, correspondence 1997), while 11...f5 12 Re2! leaves him a little better. Black might also exchange
on d4 after 11 Rxe4 (or prefer the move order 9...cxd4 10 cxd4 dxe4 11 Nxe4 Nxe4 12 Rxe4), but that still
leaves him facing the same problem: White has the initiative and Rh4 is threatened.
9 e4
The only move to consider; with ...e5 imminent, White can’t afford to dally.
9...cxd4
Black also plays 9...dxe4 10 Nxe4 Nxe4 11 Rxe4 and now 11...cxd4 12 cxd4 is a common way to return to
our main line. Instead 11...Nf6?! 12 Rh4 grants White a strong attack in just the same way as in our main line
below. Probably Black’s best option, both specifically here and in this whole ...Bd6 main line in general, is to
prefer 11...Re8!? (Diagram 33);

Diagram 33 (W)
Wisely vacating the f8-square

a useful prophylactic move endorsed by Timman. Black refrains from exchanging on d4 so that White
can’t gain a tempo with Rc1, and frees the f8-square for the use of his king, knight and bishop (the latter will
shore up h6 in the event an ...h6; Bxh6 sacrifice).
The position after 11...Re8 should be compared with our 19th illustrative game. Here, as there, 12 Rh4
(White preferred 12 Bc2!? in H.Ree-J.Timman, Amsterdam 2000, and now Bronznik devotes a lengthy
analysis to demonstrating that 12...b6! is fully viable for Black; instead the game saw 12...cxd4 13 cxd4 b6 14
Rh4 h6 and now Ree should have tried 15 Bxh6! gxh6 16 Qc1! Kf8 – and not 16...Bf8?? 17 Bh7+ Kxh7 18
Qxc7 – 17 d5! with dangerous compensation) 12...h6!? (12...Nf8?! is a less convincing defensive try; 13 Ng5
Be7 14 Qh5! gave White a strong attack in Ambrogi-P.Turturici, Rome 1973) 13 Bxh6? gxh6 14 Qd2 doesn’t
work due to 14...Bf8 15 Rxh6 Bg7, making good use of a key idea behind ...Re8. White must instead build up
more slowly and 13 Be3!? (Diagram 34)

190
 Diagram 34 (B)
White can’t rush on the kingside

looks like a good try. Now 13...e5 14 Qd2 Be7! 15 Re4 is untested but probably critical, since 13...b6? 14
Qd2 Bf8 15 Bf4! Qb7 (15...Qd8 16 Be4 wins an exchange) 16 Re1 Nf6 17 Be5! Qe7 18 Qf4 was excellent for
White in P.Berkell-T.Angqvist, Stockholm 1987.
10 cxd4 dxe4 11 Nxe4 (Diagram 35)

Diagram 35 (B)
A promising IQP position

White has accepted an IQP, but even those who usually prefer not to play with one shouldn’t be too
unhappy with this particular version: Black’s queen and dark-squared bishop are a little vulnerable, and
White can easily gain a strong attack.
191
 11...Nxe4
Black’s most popular move, but quite possibly he should prefer one of the alternatives:
a) 11...b6 12 Bg5 Bf4!? (trying to exchange some pieces: instead 12...Be7 13 Ne5!? Nxe5 14 dxe5 worked
out well for White in A.Preinfalk-B.Milic, Ljubljana 1945 – the d6-outpost is fairly useful; as did 12...Nxe4
13 Rxe4 Bb7 14 Rh4! f5! 15 Rc1 Qb8 16 Bc4 in M.Knezevic-G.Dizdar, Trencianske Teplice 1985) 13 Ne5
Bxg5 14 Nxg5 Bb7 15 Rc1 Qd6 was F.Moreno Ramo-E.Gilabert Mallol, correspondence 1990, and now the
startling 16 Qh5! would have either forced some kingside weaknesses or led to a slightly better ending after
16...Nxh5 17 Bxh7+ Kh8 18 Nexf7+ Rxf7 19 Nxf7+ Kxh7 20 Nxd6 Bd5 21 Rc7.
b) 11...Bf4!? aims for an improved version of our last note and is seen in Game 18.
c) 11...Nd5 brings a black knight to its ideal square, but this too doesn’t appear to be enough to equalize.
With 12 Nxd6! (breaking the rule about keeping pieces on when playing with the IQP, but this retains control
over e5 and is the only way to keep White’s attacking chances alive) 12...Qxd6 13 Ne5 (Diagram 36)

Diagram 36 (B)
Black is vulnerable on the dark squares

13...b6 (or 13...N7f6 14 Bg5 b6?! 15 Qf3! Rb8 – and not 15...Bb7?? 16 Bxf6! gxf6 17 Qg3+ Kh8 18
Nxf7+ and Black loses his queen – 16 Qh3 g6 17 Qh4 with a typical, direct and powerful kingside build-up in
T.Bennett-F.Vreugdenhil, Coulsdon 1999) 14 a3 Bb7 (G.Koltanowski-M.Feigins, Hastings 1936/37) 15 Qf3!
White retains the initiative; for example, 15...Rac8 16 Bh6!, hoping for the Qg3 and Nxf7+ trick, or 15...N7f6
16 Bg5 Rab8 17 Qh3 (Oleinikov) with a strong attack and the threat of 18 Ng4.
12 Rxe4!
NOTE: It cannot be emphasized enough how important it is to recapture with the rook in such
situations. Instead 12 Bxe4 Nf6 13 Bc2 Bd7 allows Black to develop and gives White far fewer attacking
chances than with his rook on h4.
12...Nf6?
A natural but not especially good move. There are a number of other typical inaccuracies which White
should be ready to punish:
a) One is 12...e5?! when 13 Rh4!? (13 dxe5 Nxe5 14 Nxe5 Bxe5 15 Qh5 was recommended by Bronznik,
but Black can continue to resist, albeit in a worse position, with 15...Qd6! 16 Rd4 g6 17 Rxd6 gxh5 18 Rd5
f6) 13...g6 (and not 13...h6? 14 Bxh6! when in V.Zivanovic-M.Feinig, Latschach 2006, White missed the
chance to punish 14...gxh6?? with 15 Rg4+! Kh8 16 Qd2 and there’s nothing to be done about mate) 14 Bh6
192
Re8 15 Rc1 Qb6 16 Qe2 leaves White with a strong initiative and Black struggling to free his position.
b) Also not recommended is 12...h6? since 13 Bxh6! (Diagram 37)

Diagram 37 (B)
The bishop cannot be touched

13...f5!? (the only real try as both 13...gxh6 14 Rg4+ Kh8 15 Qd2 and 13...Nf6 14 Rh4 gxh6? 15 Qd2 are
crushing) 14 Rxe6 gxh6 15 Rc1 Qb6 16 Rg6+ Kh7 (or 16...Kf7 17 Bxf5 Ke7 18 Qc2 Kd8 19 Rg7! with some
initiative) 17 Nh4! left White doing rather well in H.Zirwes-D.Appeldorn, correspondence 2003. Note that 18
Qh5 is White’s main threat, while 17...Nf6 fails to 18 Qd2.
c) Black can even run into trouble with 12...b6 as one future grandmaster discovered when he tried to
fianchetto: 13 Qe2 Bb7 14 Rh4 g6 15 Bd2 Rac8 16 Qe3 and White had the initiative in L.B.Hansen-
G.Vescovi, Copenhagen 1995, before finishing in style: 16...Rfd8? (16...Be7 17 Rh3 Bxf3 18 Qxf3 was
dangerous for Black, but he had to try this) 17 Bxg6! (Diagram 38)

193
 Diagram 38 (B)
Another typical breakthrough

17...fxg6?! 18 Qxe6+ Kg7 19 Rxh7+! 1-0 (mate follows after 19...Kxh7 20 Ng5+).
d) Thus Black might act prophylactically after 12 Rxe4, although White retains dangerous attacking
chances after both 12...Re8 and 12...Be7 as we’ll see in Game 19.
e) Finally we should note that Bronznik considered 12...Nb6!? to be Black’s best defence, but even this
allows White some initiative: 13 Rh4 h6 14 Ne5 (Black’s idea is revealed after 14 Bxh6!? gxh6 15 Qd2? f5!,
but White may actually be able to fall into the trap; certainly 15 Rc1! Qe7 16 Rg4+ Kh8 17 Qd2 f5 18 Qxh6+
Qh7 19 Qd2! Be7 20 Rc7 Bd7 21 Ne5 Rg8 22 Rxg8+ Rxg8 23 Rxb7 continued to swipe pawns and pose
serious problems for Black’s tangled pieces in A.Dror-R.Jaskula, correspondence 1999) 14...Bxe5?! 15 dxe5
Qxe5 16 Bxh6! f5 (16...gxh6? fails to 17 Rg4+ Kh8 18 Qd2 Qf6 – or 18...Qh5 19 Be2! – 19 Bg6! Qg7 20
Bb1 Qf6 21 Qd3) 17 Qh5 Nd7 18 Bg5 and White was much better in C.Goodman-J.Conde Ponderoso,
correspondence 1999.
Returning to 12...Nf6:
13 Rh4 (Diagram 39)

Diagram 39 (B)
The rook reaches its ideal square

13...Be7
Relatively the best defence since 13...h6? is again flattened by 14 Bxh6! and if 14...gxh6 then 15 Qd2,
13...g6 gives White a strong attack after 14 Bh6 Re8 15 Rc1 Qa5 16 Bg5! Nd5 17 Ne5 (M.Vujadinovic-
M.Crnic, correspondence 1999), and 13...Qe7? 14 Ne5! forced resignation in J.Zillmer-A.Detry,
correspondence; an early case of zugzwang, but 14...h6 allows the standard sacrifice and 14...g6 15 Bg5 Qd8
16 Qf3 Be7 17 Bh6 Re8 18 Bb5 picks up an exchange.
TIP: Whenever Black meets the pressure against h7 with ...h6, White should be on the look-out to
either immediately play Bxh6 or to try to tee up the sacrifice.
14 Bf4
White enjoys a powerful initiative which can quickly turn into a strong attack; for example, 14...Qa5 15
a3!? g6 16 Rc1 Nd5 17 Bg5! f6 18 Bh6 Rd8 19 Qc2 Qb6? 20 Bxg6! (L.Meeussen-A.Norris, correspondence
1995) or 14...Qd7? 15 Rc1 Bd8 16 Ne5 Qe7 17 Rc3! (Diagram 40)
194
 Diagram 40 (B)
Both white rooks are attacking!

17...Ne4 18 Rxh7! (D.Blair-R.Musicant, Parsippany 1999).

Theoretical Conclusion
Having received a fair amount of testing, we can conclude that the IQP position arising in the main line is both pretty dangerous for Black and
objectively a little better for White. If Black wishes to rehabilitate this variation, then postponing the exchange on d4 à la Timman is probably
his best try.

Illustrative Games

Game 18
E.Colle-S.Landau
Rotterdam 1931

1 d4 Nf6 2 Nf3 e6 3 e3 d5 4 Nbd2 c5 5 c3 Nbd7 6 Bd3 Bd6 7 0-0 0-0 8 Re1 Qc7 9 e4 cxd4 10 cxd4 dxe4 11 Nxe4 Bf4!? (Diagram 41)

195
 Diagram 41 (W)
Black plays for exchanges

12 Nxf6+
As ever, Colle chooses the most direct path, but quite possibly White should prefer one of two less
common approaches:
a) 12 Be3!? intends to recapture on e3 with the f-pawn and was endorsed by Smith and Hall. R.Crepeaux-
L.Prokes, The Hague 1928, continued 12...Nd5 13 Rc1 Qb6 14 Qd2! Bxe3 15 fxe3 N5f6?! 16 Nc5 Nxc5 17
Rxc5 Bd7 and now 18 Ne5 would have maintained White’s grip on the whole position, since 18...Rac8? runs
into 19 Bxh7+! and if 19...Kxh7?, then 20 Nxd7 Nxd7 21 Qc2+ wins the exchange.
b) 12 g3!? Bxc1 13 Rxc1 Qb6 14 Qc2! (Bronznik) leaves White in control of the c-file and, more
importantly, ready to launch on the kingside with 15 Neg5, or 14...Nxe4 15 Rxe4 and Rh4.
12...Nxf6 13 Ne5 Bxc1 14 Rxc1 Qd6 (Diagram 42)

196
 Diagram 42 (W)
White should swing the c-rook

15 Re3?!
Bearing in mind the problems White faces when Black opposes the c-file, it would have been better to
swing the c-rook over: 15 Rc3! Bd7 16 Bc2 Rac8 17 Rg3 Rc7 18 Qd2 g6! 19 Bb3 Bc6 20 Rh3 gave White a
superior version of our main game, albeit one still acceptable enough for Black in S.Djuric-R.Zelcic, Nova
Gorica 1997.
15...g6 16 Bb1 Bd7 17 h4!?
Typical Colle, but Landau, unlike some of his other opponents, refuses to crack under the kingside
pressure and methodically exchanges pieces.
17...Rac8 18 Rxc8 Rxc8 19 Rf3 Qe7 20 Qd2 Bc6! 21 Rh3 Be4 (Diagram 43)

Diagram 43 (W)
Logical and strong black play

In our next illustrative game, we will see Landau conducting a fine attack, but he was also a fine defensive
player. This manoeuvre was important not just for the sake of exchanging pieces, but to trade off White’s
dangerous bishop.
NOTE: Once ...g6 has been induced in an IQP position, White’s light-squared bishop can struggle to
enter the game, but Black ignores its potential at his peril: bishop sacrifices on e6 and g6 have won
many a game.
22 Bxe4 Nxe4 23 Qf4 f5 24 f3?!
This and White’s next see Colle going all in, but if ever there was an example that he could be too fervent
an attacker, this is it. Instead 24 Rb3, keeping the rook active, would have restricted Black to a slight edge at
best; it’s not easy for him to make progress once his kingside is compromised thus.
24...Nf6 25 h5?! Nxh5! 26 Qh6 Rc7 (Diagram 44)

197
 Diagram 44 (W)
There’s no way in on the kingside

27 Nxg6?
Presumably this had been planned some moves ago, but all it achieves is to greatly accelerate the end.
27...hxg6 28 Qxg6+ Ng7 29 Qh7+ Kf7 30 Qh6 Qf6 31 Qf4 Rc4 32 Qb8 Qxd4+ 33 Kh1 Qf4 0-1

Points to Remember
1. 11...Bf4 is a logical approach, best met by one of the slightly unorthodox 12 Be3 or 12 g3.
2. When playing with an IQP, try to keep as many pairs of pieces on as possible.
3. Should you begin an attack and realize it’s not going to work, unless fully committed and/or material down, see if you can redeploy your
pieces. Being able to put up stubborn resistance, unlike Colle here, is a very useful skill to acquire.

Game 19
S.Landau-E.Book
Kemeri 1937

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nbd7 6 0-0 Bd6 7 Nbd2 0-0 8 Re1 Qc7 9 e4 cxd4 10 cxd4 dxe4 11 Nxe4 Nxe4 12 Rxe4 Re8
Freeing the f8-square for defensive purposes. Another move directed against White’s attacking plans is
12...Be7!? when I wonder about both the untried 13 Qe2!? and even 13 Be3 Nf6 14 Rh4. The reason for
considering those untried approaches becomes clear after 13 Bf4 Qb6! 14 Qc2 Nf6 15 Bc7 (15 Re2 Bd7 16
Bc7 Qc6 is fine for Black too) 15...Qc6 (this is why the queen went to b6 rather than d8) 16 Qxc6 bxc6 17
Ree1 Bb7 18 Rac1 Rac8 19 Be5 c5 when Black has fully equalized (G.Traut-F.Sage, correspondence 2003).
13 Rh4 (Diagram 45)

198
 Diagram 45 (B)
For once ...h6 is a possible defence

13...Nf8?!
Covering h7, but this is a little passive. Black might consider 13...h6!? which sets a little trap.
WARNING: Always remember to calculate before sacrificing on h6! Here 14 Bxh6? gxh6 15 Qd2 is
rebuffed by 15...Bf8 16 Rxh6 Bg7 17 Rh3 Qd8 and Black was able to slowly unravel in S.Limberg-
P.Staller, Offenbach 2005.
White should thus prefer either 14 Be3 or Bronznik’s 14 Bc2!?, preparing Qd3 and keeping Bxh6 in
reserve.
14 Ng5! h6
Landau has been quick to strengthen his attack and already Book faces an unpleasant choice. The
alternative was 14...g6, but after 15 Qf3 f5 (J.Leuza Garcia-D.Merino Mejuto, Las Palmas 1964) 16 Bd2!,
followed by Rc1 and/or Bc3, White retains the upper hand.
15 Qh5 e5 16 Bd2!?
White didn’t have to sacrifice the pawn, but Landau’s move is very dangerous as White is now quick to
fully mobilize.
16...exd4 17 Rc1 Qe7 (Diagram 46)

199
 Diagram 46 (W)
White mustn’t get carried away

18 Ne4?
But this is too ambitious, albeit practically quite dangerous. Instead with 18 Nf3 White would have
retained a small pull and his kingside prospects without losing a pawn.
18...Ng6
Attacking the rook and blunting the b1-h7 diagonal, but Oleinikov’s 18...Be5!? was a reasonable
alternative, retaining the knight on f8 for defensive purposes. The Russian analyst then gives the line 19 Bg5
Qe6 20 f4 g6! 21 fxe5! gxh5 22 Nf6+ Kh8 23 Rxh5 and despite his extra queen, Black can’t avoid the
perpetual with 23...Qxe5 24 Rxh6+ Kg7 25 Nh5+ Kg8 26 Nf6+.
19 Bg5!?
Objectively 19 Nxd6 Qxd6 20 Re4 Bd7 21 Qc5 with some chances to hold was a better try, but it’s
important never to lose sight of the fact that chess is a practical game; something which Landau was fully
aware of.
19...Qe5
Of course, Black had to avoid 19...hxg5?? 20 Qh7+ Kf8 21 Qh8+! Nxh8 22 Rxh8mate, but some more
complex calculation lies ahead for Book.
20 f4 (Diagram 47)

200
 Diagram 47 (B)
Black must find some only moves

20...Qd5?
Badly underestimating White’s kingside threats, or missing one of them, and instead Black had to find the
active 20...Qa5!. Now 21 Nf6+!? (or 21 f5 Bxf5 22 Bd2 Rxe4! 23 Bxa5 Rxh4 24 Qxf5 Bxh2+ and Black
wins) 21...gxf6 22 Bxg6? fails to 22...Re1+. A much better try is 22 Rf1!?, but after 22...fxg5 23 Bxg6 Kf8!
Black appears to be able to run away and White’s attack gradually fades following 24 Bxf7 Re1 25 Bc4
Rxf1+ 26 Kxf1 Qf5 27 Qxh6+ Ke8.
21 Nf6+! gxf6 22 Bxg6
I wonder what Book missed around here; it’s not that hard to see that both White’s bishops are immune
due to his build-up on the h-file.
22...Bf8 23 Rc7
Straightforwardly increasing the pressure and already Black is pretty much defenceless.
23...Be6 24 Bxf6! Qxh5 25 Bxh5 Rec8
Now White wins a second pawn, but Landau had no doubt already seen that 25...Bg7? would have been
even worse due to 26 Bxf7+! Bxf7 27 Rg4.
26 Bxf7+! (Diagram 48)

201
 Diagram 48 (B)
Netting a key pawn

26...Kh7
Mate follows after 26...Bxf7 27 Rg4+, but two clear pawns down Book is quick to do the sensible thing
and shortly resigns.
27 Rxc8 Bxc8 28 Bxd4 Bf5 29 Rh5 1-0

Points to Remember
1. Don’t fall for the trap after 13...h6.
2. An attacking set-up with Rh4 and Qh5 is unsurprisingly quite dangerous.
3. We should aim to play the best moves, but practical considerations are also important to always
consider. The great Tal’s sacrifices were by no means always fully sound, but his opponents were rarely able
to find the correct defence in a complicated position and with the clock ticking; Book too found himself here
losing in such a scenario.

202
...Be7 Set-ups
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nbd7 6 Nbd2 Be7 (Diagram 49)

Diagram 49 (W)
Solid and sensible black play

In Chapter One, we saw that so long as White was happy to play the main line, it didn’t matter whether
Black began with ...Bd6 or ...Be7. Here, though, there are some important differences between the two bishop
moves. Placing the bishop on e7 is a slower and more passive option, but quite possibly also a prudent and
better one.
7 0-0 0-0 8 Qe2 (Diagram 50)

Diagram 50 (B)
White prepares e3-e4
203
 Of course, White can’t consider 8 dxc5?! (8...Nxc5!), and so the main plan is to prepare e4. With Black not
about to break with ...e5 and ...Re8 (unlike in the ...Bd6 variation), White can prepare e4 with his queen. This
is his most popular option and, as practice has shown, also his best way of opening the centre.
8 Re1 too might appear natural enough, but after 8...b6 9 e4 dxe4 10 Nxe4 Bb7 White has struggled
against Black’s solid and harmonized set-up. Note that sacrifices on f7 don’t work since Black is much better
developed than in related Caro-Kann positions; for example, 11 Neg5 h6 12 Nxf7? Rxf7 13 Bg6 Rf8 14 Qe2
Bxf3 15 Qxe6+ Kh8 left White struggling in T.Bardoux-G.Olarasu, French League 2002, since 16 gxf3 Ng8
retains the extra piece.
Colle himself preferred 8 e4 dxe4 9 Nxe4 Nxe4 10 Bxe4, but this is now also considered fairly harmless
for Black, as was shown by the 10...Nf6 11 Bc2 b6 (Diagram 51)

Diagram 51 (W)
Black has time to unravel

12 Bg5 Ba6! 13 Re1 Rc8 14 Qd2 cxd4 15 Qxd4 Qxd4 16 Nxd4 h6 17 Bh4 Rfd8 of G.Welling-J.Speelman,
Tilburg 1993. This is not a situation in which White’s queenside majority is a strength: the queenside pawns
are not easily able to advance, whereas Speelman was easily able to make further exchanges, centralize his
king and advance his central pawns. Going back, should any reader wish to re-examine these unfashionable
lines, they may wish to consider instead of 11 Bc2, the radical 11 Qe2!? Nxe4 12 Qxe4, parting with the
bishop pair but making it tricky for Black to free his c8-bishop. In this idea’s one and only test so far,
12...Qd5 13 Qxd5 exd5 14 Re1 Bd6 15 dxc5 Bxc5 16 Be3 Bxe3?! 17 Rxe3 gave White a pleasant anti-IQP
edge in J.Callaway-R.Katzl, correspondence 1974.
Finally, it must be said that White doesn’t have to play e4. He might instead prefer a Stonewall approach
and we will consider 8 Ne5 in our next section.
8...b6
With Black’s bishop deployed to e7, a queenside fianchetto is the only real way to make sense of his set-
up. He can begin too with 8...Qc7 when 9 e4 dxe4 10 Nxe4 b6 (but not 10...Nxe4?! 11 Qxe4 Nf6 12 Qh4
when White gains an attack and, for example, 12...g6 13 Bg5 Qd8 14 Rad1 saw the pressure quickly
increasing in A.Dridi-V.Sunthornpongsathorn, Manila Olympiad 1992) 11 Bg5 Bb7 12 Rfd1 transposes to
our main line. Just as there, bringing a rook to d1 is White’s best policy. Instead 12 Rfe1 (or 12 dxc5!? Nxc5
13 Bxf6 gxf6 14 Nxc5 Qxc5 15 Be4 Bxe4 16 Qxe4, as in V.Misanovic-A.Botsari, Thessaloniki Women’s
204
Olympiad 1988, but this type of unbalanced position, which also arises in the French Tarrasch, is fine for
Black after 16...f5 and ...Bf6) 12...Rad8 13 Rad1 Rfe8 14 Nxf6+ Bxf6 15 Bxf6 Nxf6 doesn’t lead anywhere
for White, and after 16 Bb5 Rf8 17 dxc5 Qxc5 18 Ba6 Ba8 19 Ne5 Rd5! Black gradually began to take over
the initiative in F.Bruno-M.Godena, Saint Vincent 2006.
9 e4 dxe4 10 Nxe4 Bb7 11 Rd1! (Diagram 52)

Diagram 52 (B)
White’s most testing set-up

NOTE: This is chiefly why 8 Qe2 is a superior try to 8 Re1. The rook takes up a more active role on
d1 and this is White’s most testing set-up.
11...Qc7!
Black’s main and best move, wisely vacating the queen from the d-file.
TIP: Queens rarely like to be on the same file as an opposition rook when there is a chance that the
file will open up (in this case after dxc5 or ...cxd4; Nxd4).
It is also quite typical, though, for Black to panic and immediately carry out exchanges, but such a policy
leads to a pleasant edge for White:
a) 11...cxd4 12 Nxd4 Nxe4 (or 12...Qc7 13 Nb5! Qc6 14 f3 Nd5 – otherwise a white knight lands on d6 –
15 c4! Nb4 16 Bb1 Rfd8 17 a3 Na6 18 Be3 Ndc5 19 Bc2 which saw White winning the manoeuvring battle
to leave Black’s pieces a little misplaced in C.Hoi-B.Sorensen, Copenhagen 1986) 13 Bxe4 Bxe4 14 Qxe4
Nc5 15 Qf3! Qc8 16 Be3 Bf6 17 Rd2 (Diagram 53)

205
 Diagram 53 (B)
White has the easier position

and White’s more active queen, ability to quickly double rooks and the weakness of d6 added up to a small
but useful plus in U.Kersten-O.Heinzel, German League 1991.
b) 11...Nxe4?! 12 Bxe4 Bxe4 13 Qxe4 is an even worse policy; Black may be able to draw the resulting
positions, but he is already playing for just two results and must also defend quite carefully to hold. Once
again the Danish grandmaster Carsten Hoi has demonstrated how White should carry out his free grind:
13...Nf6 14 Qe2 Qc7 15 Bg5 (15 dxc5!? Bxc5 16 Be3 Bxe3 17 Qxe3 Rfd8 18 Qe5! Qxe5 19 Nxe5 Kf8 20 c4
a5 21 b3 Ke8 22 f3 also leads to a clear endgame edge, although Black was able to hold with accurate
defence in M.Burgess-D.Johansen, London 1985) 15...Rfd8 16 dxc5! (correctly playing for exchanges; due to
the knight’s arrival on e5, White’s queenside pawns will be much easier to advance than Black’s central
majority) 16...Qxc5 17 Ne5 Rxd1+ 18 Rxd1 Rd8 19 Rxd8+ Bxd8 20 a4! (Diagram 54)

206
 Diagram 54 (B)
Another typical white edge

and White was somewhere between slightly and clearly better in C.Hoi-D.Holzapfel, Bad Mergentheim
1989; not only does White enjoy the more effective majority, but his minor pieces are the more active,
thereby making the defence all the tougher.

12 Bg5 (Diagram 55) 12...Rfe8

Diagram 55 (B)
The pin on f6 is a little awkward

Covering the e7-bishop in preparation for exchanges on e4, but this is far from forced. Instead 12...c4!? 13
Bc2 b5! releases the central tension, but Black’s control of the d5-square and minority attack compensate for
that. This position is dynamically about even and certainly 14 Nxf6+ Bxf6 15 Bxf6 Bxf3 16 Qxf3 Nxf6 17
Re1 Rab8 18 Re5 Rfd8 19 h4! Qb7 20 Qe2 was quite unbalanced in X.Lopez Sanchez-J.Lacasa Diaz, Catalan
Team Championship 1997.
13 dxc5
Once again White angles for a slightly better ending, but Black has a slightly better version of the positions
we considered in the notes to his 11th. Now 13...Bxe4 will be seen in Game 20. Instead 13...Nxe4?
(U.Groebe-M.Ulmer, Balingen 2006) drops a pawn to 14 cxb6!, since 14...Nxc3? fails to 15 bxc7 Nxe2+ 16
Bxe2 Bxf3 17 Bxf3 Bxg5 18 Bxa8 Rxa8 19 Rxd7, and 13...Nxc5 14 Nxc5 Qxc5 15 Ne5! Nd7? (Diagram 56)

207
 Diagram 56 (W)
White to play and win!

saw Black allow a textbook attack in A.Dubleumortier-U.Hasler, correspondence 1991: 16 Bxh7+! Kxh7 17 Qh5+ Kg8 18 Qxf7+ Kh8 19
Ng6+ Kh7 20 Bxe7 and White won.

Theoretical Conclusion
The ...Be7 approach looks like Black’s best way of handling the ...Nbd7 variation. With 8 Qe2 and 11 Rd1, White hopes for a risk-free
endgame edge, but Black can hold the balance with some accurate play, beginning with 11...Qc7. More aggressive Colle players may well thus
prefer to switch tack with the 8 Ne5 of our next section.

Illustrative Games

Game 20
I.Kashdan-H.Steiner
Pasadena 1932

1 d4 Nf6 2 Nf3 e6 3 e3 d5 4 Bd3 Be7 5 0-0 0-0 6 Qe2 Nbd7 7 Nbd2 c5 8 c3 b6 9 e4 dxe4 10 Nxe4 Bb7 11 Rd1 Qc7 12 Bg5 Rfe8 13 dxc5
(Diagram 57)

208
 Diagram 57 (B)
Black must be accurate

13...Bxe4!
The most precise for reasons we’ve seen. Black now liquidates three sets of minor pieces and draws close
to full equality.
14 Bxe4 Nxe4 15 Qxe4 Nxc5 16 Qc4 Bxg5 17 Nxg5 Qe7 18 Nf3
At first sight a draw looks imminent, but White has two small factors which offer him encouragement: (i)
his queenside majority is pretty mobile, especially should all the rooks come off, and (ii) whereas Black’s
knight is currently not causing any real difficulties, White can cause a few problems by bringing his knight to
d4 or e5.
18...Red8 19 Nd4 (Diagram 58)

Diagram 58 (B)
209
 Black still needs to be precise

19...Qh4?!
White isn’t especially troubled by the pin and Black’s queen turns out to be a little offside after this
pseudo-active move. Instead in H.Sadeghi-A.Johansen, correspondence 1994, Black unveiled a logical and
obvious improvement over some old analysis of Kashdan’s: 19...Rac8 20 b4!? Nb7! (instead Kashdan had
been hoping for 20...Na4? 21 Nc6! when White wins an exchange, while 20...e5!? 21 bxc5 exd4 22 cxd4
bxc5 23 dxc5 Rxd1+ 24 Rxd1 Rxc5 25 Qe4! also retained a small pull in Fuchs-Haack, correspondence 1988,
although now the accurate 25...Qf8! would have left Black very close to full equality) 21 Nc6 Qc7 22 b5 Re8!
(White can do nothing about the forthcoming exchange of knights) 23 Qa4 Nd8 24 Nxd8 Rexd8 25 Rxd8+
Rxd8 26 Rd1 Rxd1+ 27 Qxd1 g6 and Black held the draw without any trouble.
NOTE: White’s queenside majority can be a very useful factor in a pure queen ending, but only if he
can maintain flexibility amongst those pawns; here the b5-pawn had advanced too far ahead of its
compatriots.
20 Qe2 Rac8 21 a4!
And why not? The queenside pawns are White’s main trump and so he must put them to good use.
21...Qf6 22 b4 Nd7 23 Nb5 Nf8
To win such a position, one usually needs to force a number of small concessions. Forcing Black’s knight
back to a passive square is one, and already it’s clear that this isn’t going to be an easy draw for Steiner.
24 Qe3 (Diagram 59)

Diagram 59 (B)
White is a little better

24...a6?!
Another small inaccuracy. Black would have done better to keep White’s knight out with 24...Qe7. After
25 a5! Rxd1+ 26 Rxd1 Rd8 White might consider 27 Ra1!?, since Bronznik’s direct suggestion of 27 Rxd8
Qxd8 28 h3 (hoping for the instructive 28...a6?! 29 Qd4! Qxd4 30 Nxd4 b5 31 c4! with good prospects in the
knight ending) runs into 28...bxa5! 29 Qxa7 Qg5, threatening both the knight and perpetual (on c1 and f4),
while 30 Qb8?! a4! is a little risky only for White.
TIP: If defending a worse queen ending, always try to either target the enemy king, looking for
perpetual, or to create a passed pawn of your own; sometimes one passed pawn can at least compensate
210
for an extra two or three pawns of the opponent’s.
25 Nd6 Rb8?!
Allowing the dominant knight to remain on d6. 25...Rxc3? 26 Ne4 would have cost Black an exchange, but
he had to expel the knight with 25...Rc6 26 Ne4 Qe7 (Bronznik).
26 Rd2 Qe7 27 Rad1 Qc7 28 c4
Having doubled rooks, it’s now time for White to press ahead on the queenside. Kashdan’s play may
appear simple, but it’s also very instructive; not without good reason did Chernev include this game in his
famous The Most Instructive Games of Chess Ever Played.
28...Rd7 29 Qc3 h6 (Diagram 60)

Diagram 60 (W)
The d6-knight is a strong piece

30 Ne4
Both 30 h3!?, creating some useful luft, and 30 c5!?, wasting no time on the queenside, were good
alternatives, but Kashdan is content to simply continue to trade pieces.
30...Rxd2 31 Rxd2 Rc8
At first sight it looks like Black missed a good opportunity with 31...a5!? 32 b5 Rc8, but then, as pointed
out by Bronznik, 33 Nd6 Rd8 34 c5! bxc5 35 Nc4 Ra8 36 Rd6 would have been a strong pawn sacrifice,
gaining a passed pawn and leaving White in full control.
32 Nd6 Rd8 33 c5 (Diagram 61)

211
 Diagram 61 (B)
Black remains close to full equality

33...Qc6?
I can only imagine that both sides were short of time by this juncture. A much better defence was 33...bxc5
34 Qxc5 Qxc5 35 bxc5 Rd7 when Black should be holding; for example, 36 c6 Rc7 37 Rc2 e5! 38 Nf5 Ne6
(Oleinikov) and further white progress is impossible.
34 Qf3?
WARNING: Even the most positional-looking of positions contain tactics. Always be alert to them!
Inexplicably here both sides miss 34 Nf5! when the mate threat costs Black the exchange.
34...Qxa4?
Steiner’s flag must have been hanging, since the clear improvement 34...Qxf3 35 gxf3 bxc5 36 bxc5 e5 37
c6 Ne6 wasn’t that hard to find.
35 Qxf7+ Kh7 36 h3!
Taking care of any back-rank problems and threatening the vicious 37 Kh2 and 38 Ne4. Steiner avoids
that, but only at the cost of further material.
36...bxc5 37 bxc5 Rb8 38 Kh2 Ng6 39 Re2 Rf8 40 Qxe6 Qf4+ 41 g3 Qd4 42 Qe3! (Diagram 62)

212
 Diagram 62 (B)
The extra pawn should decide

42...Qxe3
There wasn’t anything better, but now White is a pawn ahead and his well-advanced c-pawn carries the
day.
43 fxe3 Ne5 44 Kg2 a5 45 Rf2?
I’m guessing that the gamescore must be incorrect around here; it’s hard to believe, despite the
shenanigans in the run up to move 40, that White would have blundered his c-pawn away so (45...Rxf2+ 46
Kxf2 Nd3+).
45...Rd8? 46 Rc2 Nc6 47 Nb5 Rd5 48 Nd4! Nxd4 49 exd4 Rxd4 50 c6 Rd8 51 Kf3 a4 52 Ke4 a3 53 Ke5 Rd3 54 c7 a2 55 Rxa2 Rc3 56
Kd6 Rd3+ 57 Kc6 Rc3+ 58 Kb7 Rb3+ 59 Ka8 Rc3 60 Ra7 1-0

Points to Remember
1. 11...Qc7 and 13...Bxe4 leaves Black very close to full equality.
2. However, White can still set a few problems; ones which Steiner certainly wasn’t able to solve as he fell
into a difficult position.
3. Always try to remain tactically alert!

Stonewalling
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5 5 c3 Nbd7 6 Nbd2
We will now consider White’s attempts to meet ...Be7 by setting up a Stonewall formation with Ne5 and
f4. I believe that the most accurate way of implementing such a set-up is with the move order in our main
line, but an immediate 6 Ne5 has also been tried. The problem with this comes when Black responds actively,
as indeed he must and often has done: 6...Nxe5! 7 dxe5 Nd7 (Diagram 63)

213
 Diagram 63 (W)
6 Ne5 is a little premature

8 f4 c4! and now, for example, 9 Bc2 Qh4+ 10 g3 Qh3 11 Nd2 f6 gave Black good counterplay in
P.Atanaskovic-B.Lajthajm, Leskovac 2002.
6...Be7
Black might also try to prevent Ne5 with 6...Qc7 when 7 0-0 Be7 leads to our next note. Observe too that
here White is again delighted to see a premature 7...e5?!: 8 dxe5 Nxe5 9 Nxe5 Qxe5 10 e4! and Black is
punished for breaking the ancient rule about not bringing out the queen too early. Here the e-file is opening
and Black is in some trouble; for example, 10...dxe4 11 Nxe4 Nxe4 12 Bxe4 Bd6 13 g3 Qf6 14 Bg5 Qxg5 15
Qxd6 Qe7 16 Qd3 with some initiative or 11...Be7 and now 12 Bb5+, displacing the black king, was but one
of several good moves available in H.Juehe-L.Kampmann, Altenbeken 1948.
7 0-0 0-0
Once again 7...Qc7 is an option, albeit not a troubling one for the white player happy to play the main line
of our last section (8 Qe2 0-0 9 e4 dxe4 10 Nxe4 b6 11 Rd1 Bb7 transposes). Fortunately for fans of the
independent 7...0-0 8 Ne5, White has a good alternative available in 8 e4 dxe4 9 Nxe4 (Diagram 64)

214
 Diagram 64 (B)
...Qc7 isn’t that useful a move

with an improved version of the e4-break of our last section. Here White hasn’t played Qe2, while Black
has committed his queen to c7 a little too early. Now:
a) 9...cxd4 10 cxd4 0-0 11 Bg5 gives White some initiative in an IQP position; after ...b6, Black would
much rather his queen was still on d8.
b) 9...0-0 10 Nxc5!? Bxc5 (or 10...Nxc5 11 dxc5 Bxc5 12 Qa4!; an idea of Bronznik’s, swinging the queen
into the attack) 11 dxc5 Nxc5 12 Bc2 e5 when White enjoys the bishop pair and could have retained a small
advantage with 13 Bg5!? in E.El Gindy-E.Magerramov, Dubai 1999.
c) 9...Nxe4 10 Bxe4 0-0 (10...Nf6 11 Bc2 can transpose after 11...0-0; instead 11...cxd4 12 Qxd4!? 0-0 13
Bf4 Qd7 14 Qe5 Qd5 15 Rfd1 Qxe5 16 Nxe5 gave White an edge due to his superior bishops in K.Gawehns-
K.Schiffer, German League 1998) 11 Bc2 b6 12 Qd3! Nf6 13 Bg5 (Diagram 65)

215
 Diagram 65 (B)
A standard attacking set-up

and 13...g6 was forced in W.Estrada Degrandi-I.De los Rios, Asuncion 1960, leaving White with good
attacking prospects after 14 Rfe1 Bb7 15 Ne5.
8 Ne5 (Diagram 66)

Diagram 66 (B)
White hopes to follow up with f4

8...Nxe5
Just as at move 6, Black’s main option is to immediately exchange on e5 before White can support his
knight with f4. An important alternative is 8...Qc7 9 f4 when White’s Stonewall set-up is complete and he
will be quick to attack on the kingside. Indeed attack is the name of the game for White; should he dally, his
e5-knight may change from being a rotund piece striking fear into the opposition camp to a pudgewibble one,
watching haplessly as Black breaks through on the queenside.
TIP: To break down Black’s kingside defences, White will often need to involve his g-pawn in any
attack as well as to swing a major piece across to h3.
Play usually continues 9...b6 10 Qf3 Bb7 11 g4 (Diagram 67)

216
 Diagram 67 (B)
Black must be careful, not panic

(but not 11 Qh3?! Ne4, blunting White’s play and freeing Black’s f-pawn; an important resource to
remember and avoid) and now:
a) 11...h6 appears rather weakening, but isn’t so easy to refute. White needs to prepare g5 (12 g5 hxg5 13
fxg5 Nxe5! 14 dxe5 Nh7 looks rather risky, but e5 is falling and after 15 Bxh7+ Kxh7 16 Qh5+ Kg8 17
Rxf7!? Rxf7 18 g6 Bd8 19 Qh7+ Kf8 20 Qh8+ Ke7 21 gxf7 Kxf7 Black calmly defends – White usually
needs to introduce some of his queenside pieces into any attack) and so should begin with either 12 Kh1 or
the untried 12 Qg2!?.
b) 11...Nxe5 is not objectively bad, but can be seen as a sign of panic (albeit less of one than the pawn-
losing 11...Ne4?). C.Flear-M.Sadler, Barnsdale 1989, continued 12 fxe5 Nd7 13 Qh3 g6 14 Nf3 Qd8 and now
I like Bronznik’s suggestion of the direct 15 Qh6!? (Diagram 68),

217
 Diagram 68 (B)
The kingside pressure mounts

intending to follow up with either h4 or e4 and Ng5. Note too that Black cannot defend here with 15...f6?
due to 16 Bxg6! hxg6 17 Qxg6+ Kh8 18 Nh4 and the knight arrives in the attack with deadly effect.
c) 11...Rad8!? is a brave move, tempting White forwards on the kingside, but also not a bad one at all as
we will see in Game 21.
d) 11...Ne8!? is another defensive resource which one can expect from a strong opponent. Black focuses
both his knights on the weakened e4-square and also frees his f-pawn to blunt White’s kingside advance: for
example, 12 g5 (or 12 Qh3 Ndf6 and Black’s next move is 13...Ne4 unless White tries 13 f5?! exf5 14 Rxf5,
as he did in G.Minarelli-P.Votruba, Forli 1989, but now 14...Bc8! 15 g5 Nh5! would have repelled the attack)
12...f5! 13 gxf6! (the best try; White can’t easily get anywhere if he allows the position to become locked)
13...Ndxf6 14 Qh3 (V.Kozomara-S.Popov, Sombor 1966) 14...Qd6 15 Ndf3 Ne4! (Bronznik) with an unclear
and roughly level position.
Returning to 8...Nxe5:
9 dxe5 Nd7 10 f4 f5 (Diagram 69)

Diagram 69 (W)
Black tries to close the position

Trying to block the position, but it may well be the case that the slightly suspicious-looking 10...g6!? is
actually a better option. E.Mohamed-U.Andersson, FIDE World Championship, Las Vegas 1999, continued
11 Qg4!? b6?! (an immediate 11...f5 was more accurate, transposing to the game in the case of 12 Qh3 b6 13
Nf3, while 12 exf6 Nxf6 13 Qg3 Nh5 is also fine for Black) 12 Nf3 (missing his chance for the logical 12 e4!
and now 12...Bb7 13 exd5 exd5 14 Nf3 or 12...dxe4 13 Bxe4 Rb8 14 Bc2 Bb7 15 Ne4, with an edge in both
cases – Tsesarsky) 12...f5 13 Qh3 a5 14 Bc2 Rf7 (Diagram 70)

218
 Diagram 70 (W)
Blocked and about equal

and a tough manoeuvring struggle ensued in which White should probably have first improved his dark-
squared bishop with the important manoeuvre 15 Bd2-e1-h4.
11 exf6!
White has tried a host of alternatives, but they are all inferior.
WARNING: When Black arranges ...f5, White needs to be ready to capture en passant. Otherwise
Black gains a comfortable position, and a long and rather dull manoeuvring game ensues.
11...Nxf6
Best, whereas both 11...Bxf6 12 e4 c4 13 Bc2 g6 14 e5 Be7 15 Nf3 Rf7 16 Qe1! (D.Dinev-M.Jokic,
Belgrade 2006), and 11...Rxf6 12 e4 Nf8 13 Nf3 Ng6 14 e5 Rf8 15 Qc2 (G.Breyer-G.Maróczy, Berlin 1920)
are a little more comfortable for White.
12 e4 (Diagram 71)

219
 Diagram 71 (B)
White forces open the position

12...c4
Preparing counterplay with ...Nc5 like this is probably the best option. Instead 12...Qb6 13 c4! dxe4 14
Nxe4 Nxe4 15 Bxe4 is a little better for White, H.Cording-D.Holzapfel, German League 1991.
After 12...c4 we reach a critical position and one in which I feel White is probably a little better. However,
he needs to play accurately to prove that. Theory has suggested that 13 Be2 is the way to begin, but we will
see in Game 22 that matters remain quite unresolved after that. Thus attention should also be given to 13
Bc2!? and now 13...e5! (before White plays the e5-advance himself) 14 exd5 Bc5+ 15 Kh1 Ng4 16 Ne4
(Diagram 72)

Diagram 72 (B)
A sharp situation
220
 was seen in the heavyweight encounter, E.Bareev-V.Tukmakov, Tilburg (rapid) 1994. Tukmakov ran into trouble after 16...Qh4?! 17 h3
Ne3 18 Bxe3 Bxe3 19 Qe1!, but Black does better with 16...Nxh2! 17 Nxc5 Nxf1 18 Qxf1 Qxd5 (Silman). This position is pretty open and
unclear, but I prefer White due to his bishop pair after 19 b4! when 19...Bf5 20 Bd1 prepares to bring the bishop to f3, and now 20...a5 21 Bf3
e4 22 Be2 b5 23 a4! axb4 24 Be3 bxc3 25 Rd1 Qf7 26 axb5 further opens the position to White’s advantage.

Theoretical Conclusion
8 Ne5 contains a fair amount of practical sting. In response Black is probably best advised to remain calm (with either 8...Qc7 or 8...Nxe5 9
dxe5 Nd7 10 f4 g6), but doing so is not so easy unless one has good nerves and/or is well prepared.

Illustrative Games

Game 21
K.Berg-N.De Firmian
Danish League 1999

1 d4 Nf6 2 Nf3 e6 3 e3 b6 4 Bd3 Bb7 5 0-0 d5 6 Ne5 Nbd7 7 f4 Be7 8 Nd2 0-0 9 Qf3 c5 10 c3 Qc7 11 g4 Rad8!? (Diagram 73)

Diagram 73 (W)
Black wants to open the d-file!

12 g5
Continuing on the kingside, but White might do better to prefer 12 Nxd7!? Nxd7 13 Qh3. The position is
then unbalanced and about equal after either 13...Nf6-e4 (Bronznik) or even 13...g6 14 g5 Rfe8 15 Nf3 a6 16
Qg4 Bc8 17 Ne5 Bf8! 18 Rf3 Bg7 19 Qh4 f5! 20 gxf6 Bxf6 21 Qg4 Nxe5 22 fxe5 Bg7 and Black held his
own on the kingside in G.Gara-L.Zsinka, Hungarian League 1994.
12...Nxe5 13 fxe5
One of the key ideas behind 11...Rad8 becomes clear after 13 dxe5 Ne4! 14 Nxe4 dxe4 15 Bxe4 Bxe4 16
Qxe4 c4! (Diagram 74)

221
 Diagram 74 (W)
Black controls the d-file

with the superior bishop, control of the d-file and good compensation.
13...Ne4 14 Nxe4
There wasn’t anything better, but now we see another point behind 11...Rd8; g5 is falling.
14...dxe4 15 Bxe4 Bxe4 16 Qxe4 Bxg5 17 Rf2
Trying to mount some sort of attack down the half-open files, but while White mobilizes his queenside,
Black is able to challenge on the kingside.
17...Qe7 18 Bd2 f5! (Diagram 75)

Diagram 75 (W)
Black’s f-pawn has an important role

222
 19 exf6 gxf6!?
19...Bxf6 20 Raf1 Qd6 and ...e5 was also fine for Black, but De Firmian is keen to unbalance the position
against his lower-rated opponent.
20 Kh1 Kh8 21 Rg1 f5 22 Qf3 Bf6 23 Rfg2
Berg continues to play sensibly and, with all the rooks set to come off, a draw begins to look likely.
23...Rg8 24 Rxg8+ Rxg8 25 Rxg8+ Kxg8 26 Be1! (Diagram 76)

Diagram 76 (B)
White can hold the ending

NOTE: Here White is only in time to activate his bishop to help him draw, but Bc1-d2-e1-h4/g3 is an
important feature of many Stonewall positions; the bishop’s arrival has strengthened many an attack.
26...Qd7 27 b3 cxd4 28 cxd4 b5!
Placing his pawns safely on light squares as Black possibly hopes that White will be hampered by his
central dark-squared pawns. However, Berg’s queen is active enough and once she is swapped, his king is
quick enough to centralize.
29 Bb4 Kf7 30 Kg2 a6 31 Bc5 Qd5 32 Qxd5 exd5 33 Kf3 Bg5 34 Bd6 Ke6 35 Bf4 Be7 36 Bb8 Bb4 37 Kf4 Be7 38 Kf3 Bd8 39 Kf4 Bh4
40 Kf3 Kf6 41 Bc7 Kg6 42 Bd6 Bg5 ½-½

Points to Remember
None really, bar that White needs to try and come up with something against 11...Rd8. This game might have seemed rather dull, but is quite
typical of the manoeuvring struggle which can occur when Black actively fights back in the centre against White’s aggressive set-up.

Game 22
D.Root-J.Silman
1st matchgame, California 1990

1 d4 Nf6 2 Nf3 e6 3 Nbd2 d5 4 e3 c5 5 c3 Nbd7 6 Bd3 Be7 7 0-0 0-0 8 Ne5 Nxe5 9 dxe5 Nd7 10 f4 f5 11 exf6 Nxf6 12 e4 c4 13 Be2
(Diagram 77)

223
 Diagram 77 (B)
Black will use the c5-square

At first sight this may appear a less attractive square than c2 for the bishop, but from e2 it keeps an eye on
c4, thereby hindering any ...d4-plans, and will often be redeployed to g4.
13...b5
Black later tried 13...Bc5+!? 14 Kh1 Be3 and perhaps this is his best option against 13 Be2. Now I wonder
if White should try 15 exd5!? exd5 16 Ne4, angling for pressure against d5 after something like 16...Bxc1 17
Nxf6+ Rxf6 18 Rxc1 Be6 19 Qd2, although 16...Nxe4 is one way for Black to reach a reasonable position.
Instead 15 e5 Bxd2! 16 Bxd2 Ne4 17 Be3 b6 18 Qd4 Bb7 left Black very solidly placed and was shortly
drawn in J.Schmall-H.Schmidt, Baden 1992.
14 e5 Nd7
The knight heads for c5, but even when it gets there I feel that White retains an edge, partly due to his
having a powerful knight of his own on d4. Thus this is a good moment for Black to look for an
improvement; two ideas being 14...Qb6+ 15 Kh1 and now either 15...Nd7 16 Nf3 b4 (Tsesarsky) 17 Nd4 Nc5
18 Be3 bxc3 19 bxc3 Qa6 (Diagram 78),

224
 Diagram 78 (W)
Queenside counterplay

with counterplay down the b-file and via the a4-square, or the more ambitious 15...Ne4 16 Nxe4 dxe4 17
a4 Rd8 18 Qc2 (O.Stjazhkina-J.Jackova, Frydek Mistek 1996) 18...Rd3!?, trying to offer the exchange in
return for a protected passed pawn and the light squares.
15 Nf3 Nc5 16 Nd4 (Diagram 79)

Diagram 79 (B)
The d4-knight is powerfully placed

16...Bd7 17 Bg4
White later deviated with 17 Be3 in K.Gawehns-G.Treppner, German League 1996, when a critical
continuation is 17...Qc8!? (shoring up e6 and c5 à la Silman; instead 17...a5 18 Bg4 Ra6 19 f5! with
225
advantage was the game) 18 f5!? Qc7 (but not 18...exf5?! due to Bronznik’s 19 Bf3 Kh8 20 Bxd5 Rb8 21 b3!
and White is better) 19 Bf4 (or 19 fxe6 Rxf1+ 20 Qxf1 Nxe6 21 Nf5 Bf8 and the e6-knight holds things
together for Black while e5 remains weak) 19...Qb6 20 Kh1 Rad8 with an unclear position, and one really in
need of more analysis than I can give it.
17...Qc8!
Probably the best square and certainly Black had to avoid 17...Qb6 18 Be3 b4? 19 cxb4 Qxb4 20 Nxe6!
(Silman) when the vulnerable d5-pawn costs him material.
18 f5! (Diagram 80)

Diagram 80 (B)
f4-f5 is White’s ideal break

18...Kh8 19 Bf3
Both sides have played pretty well so far, but this backward move doesn’t fully convince. Root was correct
to avoid 19 f6? gxf6 20 Bh6 fxe5! 21 Bxf8 Bxf8 22 Nf3 Nd3 when Black has more than enough for the
exchange, but he should have preferred 19 fxe6 Nxe6 (Bronznik’s 19...Rxf1+!? might be a better try, but even
so I slightly prefer White after 20 Kxf1 Nxe6 21 Bxe6 Bxe6 22 Kg1 Bg4 23 Qe1) 20 Rxf8+ Bxf8 21 Bxe6
Bxe6 22 Be3 (Silman) with an edge due to his strong knight and passed e-pawn which rather tie down
Black’s forces.
NOTE: This last variation, as well as the notes to Black’s 13th and 14th moves, shows just how much
remains to be discovered in this variation.
Indeed this line reminds me somewhat of the 10...h6 of Chapter Two; both variations are very critical,
currently quite viable for Black and in need of much more exploration and testing.
19...Nd3 20 Qe2 Bc5! (Diagram 81)

226
 Diagram 81 (W)
Black remains very solid

Now White loses his dominant knight and with it any real hope of the advantage.
21 fxe6 Bxd4+ 22 cxd4 Bxe6 23 b3 Qc7 24 bxc4 Nxc1
24...bxc4!? 25 Ba3 Qb6 26 Qe3 Rf7 was also acceptable enough for Black, but in the first game of this
match, and facing his opponent’s pet opening, Silman was happy to trade down towards a solid draw.
25 Raxc1 bxc4 26 Qe3 Qa5 27 Rf2 Rab8 28 Rcf1 Kg8 29 h3 Qb4 (Diagram 82)

Diagram 82 (W)
Black holds the balance

30 Kh2
The sharper 30 Bg4!? Bxg4 31 hxg4 Rxf2 32 Rxf2 c3 33 Rc2 Rc8 34 e6 Qb1+ 35 Rc1 c2 36 e7 Re8 would
227
also have led to a draw.
30...Rbc8 31 Bd1 ½-½

Points to Remember
1. As shown in the notes to Black’s 13th and 14th moves, 13 Be2 might not be as promising for White as
theory has suggested.
2. White players would do well to examine both the lines of this game and 13 Bc2 to look for improvements.
3. By bringing his knight to d4 and pushing through f5, Root gained a small initiative, but he needed to try
and retain his knight on that dominant d4-square.

228
Chapter Four
The Colle System: Other Black Set-ups after 3...e6
An Immediate Fianchetto
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3
It’s now time to turn our attention away from the main lines to Black’s other approaches. Employing an early
queenside fianchetto may be somewhat less popular than both the ...c5 and ...Nc6, and the ...c5 and ...Nbd7
variations, but it is also quite a logical plan for Black, focussing on the key e4-square. Black can either
fianchetto after ...c5 and castling (the subject of our next section), or do so immediately, as recommended in
Aagaard and Lund’s Meeting 1 d4.
4...b6 (Diagram 1)

Diagram 1 (W)
Black wants to control e4

5 0-0
It’s also interesting for White to begin with 5 Nbd2 Bb7 6 Ne5 (Diagram 2);

229
 Diagram 2 (B)
Cunningly delaying castling

a set-up which was quite popular in the early Twentieth Century. Play is similar to our main line, and may
even transpose, but White doesn’t have to castle: for example, 6...Bd6 7 f4 0-0 (or 7...Nbd7 8 Qf3 c5 9 c3
Qc7 10 g4!? cxd4 11 exd4 Bxe5? 12 fxe5 Nxe5?! 13 dxe5 d4 14 Qe2 Bxh1 15 exf6 dxc3 16 Ne4 and White
was much better in S.Tartakower-W.Winter, London 1932, but Black can improve such as with 11...h6!?,
intending 12 h4 h5! 13 g5 Ng4) 8 Qf3!? Nfd7 (Black’s defence can again be strengthened; for example, the
more active 8...c5 9 c3 Ba6 10 Bc2 Ne8!? 11 Qh3 g6 12 Ndf3 f6! 13 Qxe6+ Kg7 14 Qh3 Qe7 was a creative
defence in S.Simonenko-M.Sharbaf, Hyderabad 2005, and the position now became very unclear after 15
g4!?) 9 Qh3! f5 10 Ndf3 Nf6 11 Bd2 Ba6 12 Rg1 (Diagram 3)

Diagram 3 (B)
Direct but strong
230
 12...Bxd3 13 cxd3 Qe8 14 Ke2 Nbd7 15 Qh4 c5 16 Bc3 cxd4 17 Bxd4 Nc5 18 g4! and White had a strong
attack in M.Sultan Khan-H.Matisons, Prague 1931.
NOTE: From these instructive games we can see that Black’s dark-squared bishop is often
misplaced on d6 against a Stonewall set-up. Black doesn’t really want to play ...Bxe5 and he can’t easily
free his position with ...Nxe5 since fxe5 wins a piece, unlike when the bishop is on e7.
5...Bb7 6 Ne5!? (Diagram 4)

Diagram 4 (B)
Heading for a Stonewall set-up

A favourite move of Yusupov’s and one which should appeal to readers who liked the final section of our
last chapter. Furthermore, this option goes unmentioned by Aagaard and Lund. Instead they chiefly focus on
both 6 c4 (leading to an unusual type of Queen’s Gambit Accepted should Black capture which is actually a
sideline of the Queen’s Indian) and 6 Nbd2 c5 (6...Be7 7 Qe2 Ne4!? is also possible; then 8 Ne5 0-0 9 f3
Nxd2 10 Bxd2 Nd7 11 f4 f5 12 g4! was a little better for White in R.Chiaretti-A.Sardo, Bergamo 2002, but
Black might consider 10...f6!?) 7 c3 Nbd7 when 8 Qe2 (or 8 Ne5 g6!? 9 f4 Bg7 10 Ndf3 0-0 11 Bd2 Ne4!
and Black was fine in S.Tejero Lamana-D.Garcia Palermo, Zaragoza 1993) 8...Be7 9 e4 dxe4 10 Nxe4 0-0 11
Rd1 Qc7 transposes to the variation 4...c5 5 c3 Nbd7 6 Nbd2 Be7 7 0-0 0-0 8 Qe2 b6 9 e4 dxe4 10 Nxe4 Bb7
11 Rd1 Qc7; see Chapter Three.
6...Bd6
Black can’t invade the e4-square just yet: 6...Ne4 is, of course, met by 7 f3. However, while d6 might be
the most popular square for Black’s dark-squared bishop, as we have seen, it is not necessarily the best one.
Instead 6...Be7 7 Nd2 0-0 8 f4 c5 (only now that White has played f4 does 8...Ne4!? become an option; see
Game 23) 9 c3 Nbd7 10 Qf3 Qc7 takes us back to a position considered after 4...c5 5 c3 Nbd7 6 Nbd2 Be7 7
0-0 0-0 8 Ne5 Qc7 9 f4 b6 10 Qf3 Bb7 in the last chapter.
Yusupov’s opponents have also tried 6...Nbd7 7 f4 g6!? (Diagram 5),

231
 Diagram 5 (W)
Black tries to shore up his kingside

exploiting the delay in moving the f8-bishop to strengthen Black’s kingside with a fianchetto.
TIP: Remember Koltanowski’s advice: ‘You do not have to be orthodox about sticking completely to
the Colle line if your opponent varies.’
Thus after 7...g6, White shouldn’t continue with 8 c3 and 9 Nbd2, but should prefer 8 b3!? Bg7 9 Nd2 c5
10 Bb2 0-0 11 Qf3! and by switching to a Zukertort system, White was slightly for choice in A.Yusupov-
V.Anand, Linares 1991. It’s not that easy for Black to gain counterplay here, while White will attack on the
kingside as indeed Yusupov did: 11...Ne8! 12 Qh3 (preventing the desired ...f6 for the time being) 12...Nd6
13 Ndf3 Re8 14 Ng5 Nf8 15 dxc5! bxc5 16 Rad1!? f6! 17 Nxh7! Nxh7 18 Nxg6 (Diagram 6)

Diagram 6 (B)
A dangerous breakthrough
232
 with good compensation and a rook lift imminent. Both sides have played very well thus far, but now
Anand had to find 18...f5! 19 Bxg7 Kxg7 when 20 Rf3 Nf6 21 Rg3 only leads to perpetual check after
21...Nfe4 22 Bxe4 Nxe4 23 Ne5+ Nxg3 24 Qxg3+ Kf8 25 Ng6+ (Yusupov).
7 Nd2
White might also consider switching to a Zukertort-type set-up with 7 b3!?, as Yusupov has also been
known to do on occasion.
7...0-0 8 f4 c5
Continuing in classical vein. Yusupov has also faced (this variation is evidently quite popular for Black at
grandmaster level) 8...Ne4 when 9 c4!? Nd7 10 cxd5 exd5?! 11 Nxe4! dxe4 12 Bc4 Nf6 13 Qb3 exploited the
absence of early pressure against d4, as occurs after ...c5, and left White better in A.Yusupov-M.Drasko,
Sarajevo 1984. However, Black can improve with 9...f6!? when Yusupov feels that 10 cxd5 Nxd2 11 Qh5!
f5! 12 Bxd2 Bxe5! 13 fxe5 Qxd5 14 Rf3 is unclear, but White might prefer to regroup with 14 Qe2.
9 Qf3!? (Diagram 7)

Diagram 7 (B)
Preventing ...Ne4

Tactically preventing ...c4 (9...c4? 10 Bxc4!) and fighting for control of the e4-square. Instead 9 c3 Nbd7
(9...Ba6!? along the lines of our main line might be better; then 10 Qe2!? Bxd3 11 Qxd3 Nbd7 12 Kh1 Rc8
13 a4 Qe7 14 b3 cxd4 15 exd4 Rc7?! – missing his chance for 15...Ne4! – 16 Bb2 Rfc8 17 Rae1 Qd8 18 Qh3
was an interesting build-up from White in G.Bogdanovich-K.Pilgaard, Denmark 1995, before Black
collapsed with 18...g6? 19 Nxf7!) 10 Qf3 Re8 11 g4 Nf8 12 g5 N6d7 13 Qh5 Qe7 14 Rf2 f5! 15 gxf6 Nxf6
16 Qh3 Ne4 17 Rg2 Bxe5 18 fxe5 saw Black defending well and was about equal in Z.Kozul-B.Jobava,
European Championship, Ohrid 2001.
9...Qc8!
Covering the b7-bishop and also preventing any ideas of 10 b3 and a fianchetto.
10 c3 Ba6 11 Bxa6 Nxa6 12 g4 Ne8
NOTE: As is so often the case in these Stonewall positions, ...Nf6-e8 or ...Nf6-d7 is the key to a
successful defence; the knight can be redeployed to d6 or later return to f6, and crucially Black’s f-
pawn is free to advance.
13 Qg2 (Diagram 8)
233
 Diagram 8 (B)
An unbalanced situation

This unbalanced position was reached in A.Yusupov-Z.Almasi, German League 1997; an unclear struggle lies ahead.

Theoretical Conclusion
4...b6 and 5...Bb7 is a simple way for Black to play and one popular with several grandmasters. The most challenging and unbalancing
response is Yusupov’s favourite 6 Ne5, although the resulting Stonewall positions don’t promise White any theoretical advantage.

Illustrative Games

Game 23
A.Yusupov-P.Zarnicki
Minneapolis 2005

1 d4 Nf6 2 Nf3 e6 3 e3 b6 4 Bd3 Bb7 5 0-0 d5 6 Ne5 Be7 7 Nd2 0-0 8 f4 Ne4!? (Diagram 9)

234
 Diagram 9 (W)
Black leaps into the hole

9 c4
In view of the simplification which follows, one might wish to criticize this or to at least advocate a
reasonable alternative, but I’m not sure there is one! Certainly playing c4 whenever Black delays ...c5 makes
a lot of sense and the alternatives are all pretty unimpressive: for example, 9 Nxe4 dxe4 10 Be2 f6 11 Ng4
Qd7 12 b3 Rd8 13 Bb2 a5!? 14 c4 a4 15 Qc2 Nc6! 16 Rfd1 Nb4 17 Qd2 c5 gave Black good counterplay in
L.Filatov-R.Ziatdinov, Washington 1997.
9...Nd7 10 cxd5 Nxd2
The correct choice; 10...exd5?! 11 Nxe4 dxe4 12 Bc4 would have favoured White along the lines of the
previously discussed Yusupov-Drasko.
11 Bxd2 Nxe5 12 fxe5 Qxd5! (Diagram 10)

Diagram 10 (W)
Black roughly equalizes

Exploiting another benefit to having the bishop on e7 as opposed to d6.

13 Qg4
This aggressive deployment doesn’t really lead anywhere, but neither does 13 Qe2 f5!? (13...c5 is also
possible) 14 exf6 Rxf6 15 Rxf6 Bxf6 16 Rf1 Rf8 when White can’t easily arrange e4 and Black is fine.
13...c5 14 Rf3 f5!
NOTE: Yet again we can see that the secret to playing these Stonewall positions as Black is to
respond actively. Fortunately for the club player, at his level passive defence is instead quite typical.
15 exf6 Rxf6 (Diagram 11)

235
 Diagram 11 (W)
Black has correctly broken with ...f6

16 Rg3
Or 16 Rxf6 Bxf6 17 Rf1 Rf8 with full equality.
16...g6 17 Bc3 cxd4 18 Bxd4 e5 19 Qe2 Kg7 20 Bc4
Neither side has to go in for the forthcoming repetition, but then neither too has a good reason to deviate.
Yusupov probably considered that 20 Bc3 Rd8 was fine for Black, and that it was better to make a draw and
look for an improvement for next time.
20...Qe4
Black might have tried 20...Qd6 21 Bc3, but I feel that this displays a few signs of drifting, especially in
the case of 21...Raf8?! 22 Rg5! Rf5 23 Rxf5 Rxf5 24 Rd1 Qc6 25 Bb5 Qc5 26 Rd7 when White has seized
the initiative.
21 Bd3 Qd5 (Diagram 12)

236
 Diagram 12 (W)
A repetition looms

22 Bc4 Qe4 23 Bd3 ½-½

Points to Remember
None really. Against grandmaster opposition, the Stonewall is much less dangerous than at lower levels. Zarnicki’s 6...Be7 and 8...Ne4 looks
like another reasonable way for Black to play.

Black Plays ...c5 and ...b6

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 c5
Here we’ll examine what happens when Black doesn’t immediately follow up ...c5 by developing his queen’s knight, but instead prefers first a
queenside fianchetto. Note that our main line also comes about via the move order 4...Be7 5 0-0 0-0 6 Nbd2 c5 7 c3 b6.
5 c3 Be7
In Chapters One, Two and Three we saw Black often fianchettoing having first castled, but what about if
he develops his queen’s knight and then tries to fianchetto before castling? One such approach is 5...Nbd7 6
Nbd2 Bd6 7 0-0 b6?!, but this doesn’t convince since White is ready to immediately open the centre: 8 e4
(Diagram 13)

237
 Diagram 13 (B)
Exploiting a premature ...b6

8...cxd4 (or 8...dxe4 9 Nxe4 Nxe4 10 Bxe4 Rb8 11 dxc5 and now 11...Nxc5 12 Bc6+ displaces the black
king, but even worse is 11...Bxc5? 12 Bf4 and White won the exchange in E.Cohn-J.Mieses, Vienna 1908) 9
cxd4 dxe4 10 Nxe4 Be7 (10...Nxe4 11 Bxe4 Rb8 12 Bg5 f6 13 Bh4 also favoured White in J.Sadzio-
M.Bauer, Duisburg 2006) 11 Ne5! (Diagram 14)

Diagram 14 (B)
The a4-e8 diagonal spells trouble

and Black’s vulnerable queenside light squares cause him no end of trouble; for example, 11...Bb7 12
Nxf6+ Bxf6 13 Bb5 Bxe5 14 dxe5 Bd5 15 Qg4 forced 15...Kf8 (15...0-0? 16 Bh6) and was excellent for
White in R.Zysk-G.Danner, Budapest 1987.
238
 Black does better to fianchetto with his king’s bishop less exposed on e7, namely 5...Nbd7 6 Nbd2 Be7 7
0-0 b6 (Diagram 15).

Diagram 15 (W)
The Colle is full of transpositions

Now 8 e4 (Stonewallers can also consider 8 Ne5!?) 8...dxe4 9 Nxe4 Bb7 leaves White with nothing better
than 10 Qe2 0-0 11 Rd1 and a transposition to the ...Be7 main line of Chapter Three.
Finally, Black might even follow up 5...Nc6 6 Nbd2 Be7 7 0-0 with 7...b6. Now 8 e4 dxe4 9 Nxe4 cxd4!
10 Nxf6+ Bxf6 11 Be4 Bb7 12 Nxd4 Qd7 or even 12...Bxd4!? 13 cxd4 0-0 is fine for Black. 8 Qe2 0-0 is also
fully acceptable for him as we mentioned in Chapter One, and so White might consider 8 Ne5!? Nxe5 9 dxe5
Nd7 10 f4 (Diagram 16),

Diagram 16 (B)
Kamsky takes the Stonewall route
239
 as he did in G.Kamsky-E.Prie, Paris 1990, and now Black should probably employ the unclear 10...c4! 11
Bc2 Nc5 12 Nf3 g6 as indeed Prie later did.
6 Nbd2 0-0
Remaining flexible, but in practice both 6...Nc6 and 6...Nbd7 are seen somewhat more often, transposing
to Black’s two main defences to the Colle.
7 0-0 b6 (Diagram 17)

Diagram 17 (W)
White has three options

This was Black’s last chance to transpose to one of the main lines, but electing to fianchetto is also a pretty
reasonable option (and was recommended in the recent Chess Openings for Black, Explained). White must
now respond by taking into account how he likes to meet the 4...c5 5 c3 Nbd7 6 Nbd2 Be7 variation:
a) 8 e4 dxe4 9 Nxe4 Nbd7 10 Qe2 Bb7 (Diagram 18)

240
 Diagram 18 (W)
Yet another transposition

transposes into one of the main lines of Chapter Three. Black might also consider both 8...Ba6!?, as played
by Taimanov, and 8...cxd4. After the former, 9 Bxa6 Nxa6 10 Qe2 Nb8 11 e5 Nfd7 12 Re1 Nc6 13 Nf1 led to
an unclear French-type position in E.Raitza-J.Seils, Rostock 2005; here Black can’t really break with ...f6,
while the absence of the light-squared bishops weakens White’s attacking potential.
b) 8 Ne5 Bb7 (8...Ba6!? is again possible as an independent try; 9 Nc6 Qc7 10 Nxe7+ Qxe7 11 Bxa6 Nxa6
12 Qe2 Qb7 13 b3 Rfc8 didn’t really see Black missing his dark-squared bishop in D.Stroppa-I.Zaragatski,
Gross Gerau 2000) 9 f4 Nbd7 (9...Ne4!? as briefly recommended by Alburt, Dzindzichashvili and
Perelshteyn is another option) 10 Qf3 Qc7 takes us into the other ...Be7 main line of Chapter Three.
c) 8 Qe2 Bb7 (or 8...Nbd7 9 e4 dxe4 10 Nxe4 Bb7) 9 e4 dxe4 10 Nxe4 Nbd7 also takes us back into the realm of Chapter Three. Black
might consider too 8...Nc6, transposing to an acceptable enough position for him and one which also arises via 4...c5 5 c3 Nc6 6 Nbd2 Be7 7 0-
0 0-0 8 Qe2 b6.

Theoretical Conclusion
This has been a short section. It has, though, reminded us just how much remains to be discovered in the Colle (such as the ...Ba6 options
mentioned in the notes above), and shown that playing 4...c5 and then 7...b6 often only transposes to the ...Nbd7 and ...Be7 variation of Chapter
Three.
Black Responds Passively
1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 Be7
It’s now time to have a look at what happens when Black plays either quite slowly or rather planlessly in
the opening; not an atypical situation at club level. Such lines are usually associated with Black playing
neither an early ...c5 nor an early queenside fianchetto. However, it’s also important to know how to react
when Black plays something unexpected like 4...c5 5 c3 Nc6 6 Nbd2 Be7 7 0-0 and now not 7...0-0, the main
line of Chapters One and Two, but 7...Bd7?!.
NOTE: As a system-based opening the Colle is ideal for meeting such misguided moves. White either
continues to complete his initial development or has to decide between a line-opening e4 plan and
heading for a Ne5 and f4 set-up.
In this case with 7...Bd7, White has already carried out his usual Colle development and it’s time to act.
There’s no-one better to show us how to continue here than the man himself. In E.Colle-C.Piccardt,
Amsterdam 1931, White carried out his usual central break and 8 dxc5 Bxc5 9 e4 (Diagram 19)

241
 Diagram 19 (B)
Getting on with White’s main plan

9...dxe4 10 Nxe4 Nxe4 11 Bxe4 Be7 12 Qc2 h6 13 Bf4 0-0 14 Rad1 left him with a clear edge; Black’s
queen lacks a good square.
A line which cannot really be considered passive, but is also quite rare is 4...Nbd7 5 Nbd2 Bd6 and now 6
0-0 0-0 7 e4 (and not 7 c3?! e5 when Black at least equalizes) 7...dxe4 8 Nxe4 Nxe4 9 Bxe4 should be
compared with our main line below. Black’s bishop is a little more active on d6 than e7, but any pin down the
h4-d8 diagonal will be more awkward to deal with; for example, 9...Nf6 10 Bd3 b6 11 Bg5 Bb7 12 Qe2 Be7
13 Rad1 Re8 14 c4! (Diagram 20)

Diagram 20 (B)
White enjoys a spatial advantage

242
 with an edge, A.Rubinstein-A.Conde, Hastings 1922/23.
White can also meet 4...Nbd7 5 Nbd2 Bd6 with 6 e4, but it’s worth being aware that 6 0-0 e5?! is nothing
to fear; on the contrary Black’s attempt to play a Colle set-up himself with a quick e-pawn break merely helps
White: 7 e4! (Diagram 21)

Diagram 21 (B)
Punishing a premature ...e5

7...dxe4 (7...exd4? 8 e5 Nxe5 9 Nxe5 Bxe5 10 Re1 wins a piece) 8 Nxe4 Nxe4 9 Bxe4 0-0?! (or 9...exd4
10 Qxd4 Qf6 11 Bg5! Qxd4 12 Nxd4 – Bronznik – with a healthy lead in development) 10 dxe5 Nxe5 11
Nxe5 Bxe5 12 Bxh7+! Kxh7 13 Qh5+ Kg8 14 Qxe5 and White’s standard combination netted a pawn in
G.Koltanowski-M.Golmayo de la Torriente, Sitges 1934.
Black has also been known to employ the slightly baffling 4...a6?!. Once again it’s time to hand over to
Colle in Game 24 for a subsequent model handling of the opening.
5 0-0 0-0 6 Nbd2 Nbd7
Sometimes Black even plays 6...c6?!, but it’s a little too late for him to adopt a Semi-Slav set-up. White
can continue with 7 e4, although there’s also nothing wrong with 7 c3 and now 7...Nbd7 8 e4 dxe4 9 Nxe4
Nxe4 10 Bxe4 Nf6 11 Bc2 c5 12 Bg5 (O.Kriz-L.Chmelik, Slovakian League 1997) is similar to both
positions we’ve already examined and certain lines of the French Rubinstein, except that Black has lost a
tempo (playing ...c6, then ...c5). That helps to give White a pleasant edge; Re1 and Ne5 being one good plan.
7 e4 (Diagram 22)

243
 Diagram 22 (B)
Exploiting the absence of ...c5

There’s no real reason to delay the key Colle break, but White’s most popular move in practice has
actually been 7 c3 which has the advantage of remaining fully within Colle waters. Then 7...c6 8 e4 dxe4 9
Nxe4 Nxe4 10 Bxe4 Nf6 enables White to build a queen and bishop battery with 11 Bc2 and Qd3, but Black
should prefer 7...c5, transposing to Chapter Three.
TIP: Remember that there’s no need to hurry with c3 until Black has played ...c5, increasing the
pressure on d4 and possibly threatening ...c4.
Fans of a Stonewall approach should also note that 7 Ne5 is possible, albeit not particularly accurate.
Nonetheless, I can’t resist showing the entertaining scrap, W.Lombardy-A.Mestel, Lone Pine 1978: 7...Nxe5
8 dxe5 Nd7 9 f4 g6 10 b3 f6! 11 exf6 Bxf6 12 Rb1 e5 13 Ba3 e4?! 14 Bxf8 Qxf8 15 Be2 Qc5 16 Kf2 g5?! 17
b4! Qb6? 18 Nxe4!! dxe4 19 Qd5+ Kg7 20 fxg5 Be5 21 Bc4 Kg6 22 Qg8+ Kh5 23 Be2+ Kh4 24 g3+ 1-0; a
fine punishment of Black’s lack of queenside development and overambitious, if impressively creative play.
7...dxe4
Otherwise e5 follows and 7...h6?! 8 e5 Nh7 9 c4 already left Black a little cramped and worse in
A.Alekhine-S.Vorkapich, Los Angeles (simul) 1932.
8 Nxe4 Nxe4 9 Bxe4 Nf6 10 Bd3 (Diagram 23)

244
 Diagram 23 (B)
Black is solid but a touch passive

10...c5
Black’s thematic break. Alternatively, 10...b6 11 Qe2 Bb7 12 Rd1 Qc8 was A.Molet-B.Gobeaut, La Fere
2006, and now White retained a small edge with the instructive 13 c4!, intending both a queenside fianchetto
and to meet 13...c5?! with 14 d5.
11 dxc5 Bxc5 12 Bg5
We’ve actually transposed to a position which can also occur from the French Rubinstein (namely 1 e4 e6
2 d4 d5 3 Nc3 dxe4 4 Nxe4 Nd7 5 Nf3 Ngf6 6 Nxf6+ Nxf6 7 Bd3 c5 8 dxc5 Bxc5 9 0-0 0-0 10 Bg5), and
theory now considers Black’s best chance of equalizing to be either 12...b6 or 12...Qc7.
12...Be7?! 13 Qe2
We’ve been following E.Colle-V.Berger, Hastings 1928/29, in which White’s freer development gave him
the advantage. Now Black had to avoid 13...b6?? (I trust you spotted 14 Bxf6 Bxf6 15 Qe4!), but Berger was
quick to fall for another common trick: 13...Qc7 14 Rad1 Rd8 15 Ne5 Bd7? (Diagram 24)

245
 Diagram 24 (W)
Spot White’s typical breakthrough!

16 Bxh7+! Kxh7 17 Bxf6 Bxf6? (he had to play 17...Be8) 18 Qh5+ Kg8 19 Qxf7+ and 1-0 since 19...Kh8 20 Rd3 decides.

Theoretical Conclusion
Players used to defending passive positions may meet the Colle in this way, but there’s no need for Black not to play something a little more
active! White retains an edge by castling and carrying out the standard e4-break, although he may wish to insert first a prophylactic c3.

Illustrative Games

Game 24
E.Colle-J.Aguilera
Barcelona 1929

1 d4 d5 2 Nf3 Nf6 3 e3 e6 4 Bd3 a6?! 5 Nbd2 Nbd7

I suppose 5...b5 might be more consistent when 6 e4 Bb7 7 e5 Nfd7 8 0-0 c5 9 c3 leads to a sort of St
George Defence (1...a6) with White well set to attack on the kingside.
6 0-0 Bd6 7 e4 (Diagram 25)

246
 Diagram 25 (B)
Standard and strong white play

Having competed White’s kingside development, it’s time for part two of the usual Colle plan.
7...dxe4 8 Nxe4 0-0
Perhaps Black was better off with 8...Nxe4 9 Bxe4 Nf6 10 Bd3 (Oleinikov), but this is still nice for White;
...a6 has turned out to be a complete waste of time.
9 Bg5 Be7 10 Qe2 Nd5!?
Trying to free his position through exchanges. Aguilera is right to avoid 10...b6? 11 Nxf6+ Bxf6 12 Qe4,
when White wins material, and 10...Nxe4 11 Qxe4, forcing a kingside weakness, was also undesirable for
him.
11 c4! Nb4 12 Bb1 Bxg5!
Good defence; Black had to take action on the kingside before White got in 13 a3 Nc6 14 Qc2.
13 Nfxg5 h6 14 f4?!!? (Diagram 26)

247
 Diagram 26 (B)
A very creative idea

A move which reveals much about Colle’s approach to the game: it’s dangerous and extremely creative,
but not fully sound. Objectively White should have settled for a rather pleasant advantage after 14 Nf3 b6 15
a3 Nc6 16 Rd1.
14...hxg5 15 fxg5 f5 16 Qh5!
Undoubtedly it was this possibility which attracted Colle to 14 f4; Black must now defend extremely
carefully.
16...fxe4
Correct. Aguilera wisely avoids 16...Nf6 17 gxf6 Qe8 18 Qh4! when White is much better since 18...fxe4?
fails to 19 Bxe4 Rxf6 20 Qh7+ Kf8 21 Bg6 (Harvey). Here it’s also worth noting the cute line 17...Qxd4+? 18
Kh1 gxf6 19 Qg6+ Kh8 and now 20 Bd3!! prepares a decisive rook lift.
17 Bxe4 Nf6?
Presumably a sign of panic at lines such as 17...Rxf1+ 18 Rxf1 when mate follows, but, like Oleinikov,
I’m surprised that Black didn’t try 17...Rf5! (Diagram 27),

248
 Diagram 27 (W)
The correct defence

blocking the b1-h7 diagonal. Indeed this option would have revealed Colle’s attack to be unsound. Now 18
Rxf5? exf5 19 Bxf5 loses to 19...Nf8, but perhaps Aguilera missed that after 18 Bxf5 exf5 19 Rxf5!
(renewing White’s kingside threats, since now 19...Nf8 no longer works due to 20 Qf7+ and 21 Qxf8), Black
has 19...Ne5! (19...Nf6 may also be possible, but allows White at least a draw after 20 Rxf6 gxf6 21 Qg6+
Kf8, and perhaps even more with 22 Qh6+ Kf7 23 Qh7+ Ke6 24 Re1+!? Kd6 25 a3) 20 Rf4 (20 dxe5 Bxf5
21 g6 Bxg6 simply leaves Black a piece up) 20...Bg4! (Oleinikov) and White is struggling. He has to play 21
Qh4 Ng6 22 Qxg4 Nxf4 23 Qxf4 with three pawns for the piece, but here that’s insufficient compensation
since Black is quick to activate his major pieces with 23...Qe7 and ...Rf8.
18 gxf6 Qxd4+ 19 Kh1 (Diagram 28)

Diagram 28 (B)
249
 Black is defenceless

Now it’s all over; stripped of piece and pawn cover Black’s king can’t survive for long.
19...Qxe4 20 f7+ Rxf7 21 Qxf7+ Kh8 22 Rf4 1-0

Points to Remember
1. Should Black play a little pointlessly in the opening, stick to the usual plan: arrange e4 and then look for
attacking chances on the kingside.
2. Even slightly unsound piece sacrifices can be tricky for Black to defend against when he lacks kingside
defenders and/or is lagging in development.

250
Chapter Five
The Anti-Colle: Black’s Third Move Alternatives
Introduction
1 d4 d5 2 Nf3 Nf6 3 e3 (Diagram 1)

Diagram 1 (B)
Black doesn’t have to play 3...e6

By far Black’s most popular choice at this juncture is 3...e6, but as we have seen, it is by no means as easy as many believe for him to
equalize after it. Thus many strong players prefer to avoid 3...e6 and it is now time to consider those alternatives. The French GM Eric Prie is
even on record as describing 3...e6 as a ‘ridiculous move’; he feels that Black shouldn’t voluntarily block in his light-squared bishop and such a
viewpoint certainly has its logic. Nevertheless, Black doesn’t simply equalize just by preferring 3...Bg4 or 3...Bf5; he still needs to follow up
accurately.

Black Plays 3...Bg4

1 d4 d5 2 Nf3 Nf6 3 e3 Bg4 (Diagram 2)

251
 Diagram 2 (W)
A pinning approach

An active deployment of Black’s light-squared bishop and a move which received some coverage in
Watson and Schiller’s recent How to Succeed in the Queen Pawn Openings.
WARNING: In these anti-Colle lines it is vital that White pays attention to Black’s set-up. He cannot
develop routinely as he does against 3...e6.
Thus 4 Bd3?! is not recommended and after 4...Nbd7 (playing for ...e5 is the simplest way to equalize,
although 4...c5 5 c3 Nc6 6 0-0 e6 7 Nbd2 Bd6 with a reversed Queen’s Gambit Declined was also fine for
Black in M.Chodera-S.Tartakower, Bardejov 1926) 5 0-0 (or 5 h3 Bxf3 6 Qxf3 e5 7 dxe5 Nxe5 8 Qf5 Bd6 –
Watson and Schiller – with rapid and easy development for Black) 5...e5 (Diagram 3)

Diagram 3 (W)
Black equalizes
252
 6 dxe5 Nxe5 7 Be2 Bxf3! 8 Bxf3 Qd7 9 b3 Nxf3+ 10 Qxf3 Bd6 11 Bb2 Ng4! Black was already slightly
for preference in J.Feys-M.Sadler, Ostend 1992.
Likewise 4 Be2 is not active enough: 4...e6 5 0-0 (or 5 Ne5 Bxe2 6 Qxe2 Nbd7 7 f4?! c5 8 c3 Bd6 9 0-0 0-
0 10 Nd2 Rc8 and Black had an edge in S.Perich-P.Niccoli, Arco 2005; White badly misses his light-squared
bishop in such a Stonewall set-up) 5...c5 (Diagram 4)

Diagram 4 (W)
A reversed QGD

and White is again playing a reversed QGD. Such an approach will not suit many Colle players and this
isn’t a reversed position in which the extra tempo is really of much use; for example, 6 b3 (or 6 Nbd2 Nc6 7
b3 cxd4 and now the liquidating 8 Nxd4 is correct; instead 8 exd4 Qa5!, threatening ...Qc3 and intending 9
Bb2 Ba3, was already rather pleasant for Black in J.Trapl-J.Stocek, Tatranske Zruby 2003) 6...Nc6 7 Bb2
cxd4 8 Nxd4 Bxe2 9 Qxe2 Nxd4 10 Bxd4 Be7 11 Nd2 0-0 12 c4 and ½-½ in W.Arencibia-A.Rodriguez
Cespedes, Matanzas 1997. White has no advantage here and he must play much more actively.
NOTE: Black doesn’t have an easy ride to equality in the main lines of the Colle, and we should try
too not to allow him one when he avoids 3...e6.
4 h3!?
Wasting no time in putting the question to Black’s bishop. I quite like such an approach, but White’s main
move and another good option is 4 c4. Now:
a) 4...c6 is Black’s most solid choice, taking play into a line of the Slav. We’ll return to this position and
show that White shouldn’t mind the transposition in Game 25.
b) 4...Bxf3 5 Qxf3 (Lasker once preferred 5 gxf3!? e6 6 Qb3 which is also possible, although the active
6...Nc6!?, preparing to gambit the b-pawn, was a slightly awkward response in B.Westermann-R.Reichert,
Willingen 2001) 5...e6 6 Nc3 c6 should be compared with Game 25. White can transpose with 7 Bd3 Nbd7 8
0-0 Bd6 (Diagram 5)

253
 Diagram 5 (W)
Black employs a solid Slav formation

9 Rd1 0-0 10 h3 to a line considered in that illustrative game, but may well be able to put his extra tempo
to better use; for example, 9 e4 dxe4 10 Nxe4 Nxe4 11 Bxe4 Nf6 12 Bc2 (W.Buchanan-E.Rutherford,
Scottish Championship, Aberdeen 2001) when Watson and Schiller propose 12...Qe7, but now 13 Bg5 should
retain an edge due to the bishop pair.
c) 4...dxc4 5 Bxc4 e6 transposes to yet another opening, namely a sideline of the Queen’s Gambit
Accepted.
NOTE: In the Anti-Colle, White must be prepared to transpose to lines of the Slav and QGA. Doing
so is often his only way to fight for the advantage, while few of the transpositions are to especially
theoretical or critical positions.
Here this sideline of the QGA has long been considered to offer White good chances of an edge. The plan
is the straightforward 0-0, h3 and e4; i.e. 6 Nc3 Nbd7 7 0-0 Bd6 8 h3 Bh5 and now the critical continuation is
9 e4 e5 10 g4! (readers interested in the details should consult NCO, my Play 1 d4! or a specialist work on the
QGA). That line is promising and not actually that complicated, but the reader may prefer the simpler 9 Be2
0-0 10 e4 e5 11 dxe5 Nxe5 12 Nd4 (Diagram 6)

254
 Diagram 6 (B)
There’s plenty of play left

when a tough positional struggle lies ahead; for example, 12...Bc5 13 Nb3 Qxd1 14 Bxd1 Bb6 15 a4 Bxd1
16 Rxd1 a5 17 Bg5 c6 and the position remained dynamically balanced in A.Antunes-A.Miles, Andorra 1995.
d) 4...e6 5 Qb3 (exploiting the bishop’s early development to target b7) 5...Bxf3!? (beginning a critical
pawn sacrifice; instead 5...Qc8 is a little passive and 6 Nc3 c6 7 Ne5 Bf5 8 f3! leaves White better after both
8...h6 9 cxd5 exd5 10 e4! Be6 11 exd5 Nxd5 12 Nxd5 Bxd5 13 Bc4, V.Pirc-J.Foltys, Podebrady 1936, and
8...h5 9 cxd5 Nxd5?! 10 e4 Nxc3 11 bxc3 Bh7 12 Rb1! b6 13 Ba3, J.Redpath-Du.Walker, Scottish League
2005) 6 Qxb7!? (bravely taking up the challenge; 6 gxf3 Nbd7 7 Qxb7 transposes, but White can also
consider 7 Nc3, followed by Bd2 and possibly f4) 6...Nbd7 7 gxf3 (Diagram 7)

Diagram 7 (B)
Critical, if greedy
255
 7...dxc4! (taking play into a line of the QGA which was popular in the sixties and seventies; instead
7...Rb8 8 Qxa7 Bb4+ 9 Nc3 0-0 10 Qa4 c5 11 dxc5 Nxc5 12 Qc2 probably doesn’t give Black quite enough
for his two pawns, A.Iglesias-F.Perez, Havana 1970) 8 Bxc4 c5 9 dxc5 Bxc5 and Black has open lines, a
better structure and slightly superior development, which adds up to sufficient compensation. Who to prefer
here is really a matter of taste (Korchnoi, for example, would probably take White), but I suspect that many
readers may prefer the calmer 4 h3. Play might continue 10 Nc3 0-0 11 f4 Nb6! 12 Be2 Nfd5 13 0-0?! Qh4
and with ...f5 imminent, White came under kingside pressure in M.Quinteros-A.Miles, Amsterdam 1977.
Returning to 4 h3:
4...Bh5
Maintaining the pin, but as Watson and Schiller point out, a more prudent choice is 4...Bxf3!? 5 Qxf3
(White might also consider 5 gxf3!? and after 5...Nc6, I wonder about an ambitious 6 f4 and a3 before
arranging c4) 5...e6 (5...c6 6 c4 and 5...Nbd7 6 c4 e6 7 Nc3 c6 both transpose to Game 25) and now, as Prie
points out, one option is to play a solid reversed Moscow Semi-Slav with 6 Nd2 c5 7 c3. However, I prefer
the active 6 c4 and now:
a) 6...c6 has been Black’s most popular choice when 7 Nc3 transposes to Game 25.
b) 6...Bb4+ 7 Bd2 Bxd2+ 8 Nxd2 0-0 9 cxd5 Nxd5 10 Bc4 is a little better for White.
c) 6...c5!? 7 cxd5 Qxd5 (actively trying to avoid 7...exd5 8 Bb5+! Nc6 9 0-0 a6 10 Bxc6+ bxc6 11 b3 with
an edge for White against the coming hanging pawns) 8 Bb5+ Nc6 (Diagram 8)

Diagram 8 (W)
Very active play from Black

9 Qxd5 exd5!? 10 dxc5 a6 11 Bxc6+! bxc6 12 b3 Bxc5 13 Bb2 Ne4 is an enterprising idea of Prie’s,
designed to exploit Black’s slightly faster development. Indeed this looks like an excellent piece of analysis
from the French GM; after 14 Ke2 Bb4! 15 Rc1 Kd7 16 Nc3 Bxc3 17 Bxc3 Nxc3+ 18 Rxc3 a5! Black is very
close to full equality.
5 g4!
Little known, but vigorous and very much in the right spirit of how White should handle the Anti-Colle in
general.
5...Bg6 6 Ne5 (Diagram 9)

256
 Diagram 9 (B)
Preparing to advance the h-pawn

This was the main idea behind White’s last two moves. At first it appears that he is simply gaining the
bishop pair and some useful space, which wouldn’t be a bad deal at all.
TIP: White actually has a more fiendish plan in mind: to target the g6-bishop with a rapid advance
of the h-pawn.
6...Nbd7?!
A typical mistake. Black cannot afford to continue developing naturally like this, and must challenge the
dominant e5-knight with 6...Nfd7!; see Game 26. Instead 6...Be4? 7 f3 Bg6 8 h4! (Diagram 10)

Diagram 10 (B)
Black is in trouble

257
 8...h5 9 Nxg6 fxg6 10 Bd3 only helped White in G.Bogdanovich-K.Gutsche, Leutersdorf 2004, and now
Black decided he had nothing better than to give up the g6-pawn since 10...Kf7 11 g5 Ng8 12 c4 e6 13 Nc3
would also have been pretty bad for him. Likewise 6...e6?! 7 h4 should be compared with our main line;
again White is doing rather well after 7...h6 8 Nxg6 fxg6 9 Bd3 (J.Kivimaki-J.Pitkanen, Finnish League
1996).
7 h4! h6
Slightly less weakening than the similar 7...h5, while we should note that 7...Nxe5? 8 dxe5 Nd7 9 h5 Be4
10 f3 costs Black a piece and is the point which underpins White’s whole concept.
8 Nxg6 fxg6 9 Bd3 Kf7 (Diagram 11)

Diagram 11 (W)
White’s ideal position after 4 h3

Already we can see that the opening has not been a success for Black. White may now play quite slowly,
aiming to make good use of his bishops and extra space (as the calm 10 h5 g5 11 c4! e6 12 Bg6+ Kg8 13
cxd5 exd5 14 f3, retaining some advantage, did in M.Dietrich-O.Hach, Vienna 2006), but it may be even
better to continue aggressively with Prie’s 10 Rg1!?. The French GM and ChessPublishing writer now offers
10...e5 11 g5 hxg5 12 hxg5 e4 13 gxf6 exd3 14 Qg4 Nxf6 15 Qxg6+ Kg8 16 cxd3 and 10...g5!? 11 hxg5
hxg5 12 Nd2 e5 13 Nf3 e4 14 Nxg5+ Kg8 (and not 14...Kg6? 15 Qf3! (Diagram 12),

258
 Diagram 12 (B)
White wins

neatly exploiting the pin to force mate) 15 Bf1 followed by Nh3, with some advantage in both cases.

Theoretical Conclusion
3...Bg4 is not the easy equalizer it’s often portrayed as. White must, though, challenge with either 4 h3 or 4 c4, and both can lead to the same
line of the Slav which is slightly more pleasant for White. Indeed that is a good reason to prefer 4 h3 over 4 c4: the latter allows Black to
induce complications with 4...e6 5 Qb3 Bxf3 6 gxf3 Nbd7, whereas 4 h3 Bh5 5 g4! is promising for White.

Illustrative Games

Game 25
K.Georgiev-I.Sokolov
Sarajevo 2002

1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 e3 Bg4 5 h3

Once again we see White hurrying to challenge the g4-bishop, thereby forcing Black to make a committal
early decision: to exchange or not to exchange. I’ve also enjoyed some success with 5 Qb3, partly because
5...Qb6 6 Ne5 (note how neither side wishes to swap queens: White because it would ease any pressure on the
black position; Black because axb3 opens the a-file and may also give White a dangerous option of c5 and
b4-b5) 6...Be6 (Black should prefer 6...Bf5) 7 Nc3 Nbd7 (Diagram 13)

259
 Diagram 13 (W)
Has Black equalized?

was given as equal in NCO and thus caught the attention of a few black players. However, White retains a
small pull as shown by, for example, 8 Nxd7 Bxd7 9 Na4! Qxb3!? 10 axb3 0-0-0 (R.Palliser-A.Ledger,
British League 2003) 11 Nc5 b6 12 Nd3.
TIP: Never trust every assessment and piece of analysis you read, including in this book! It’s always
best to verify things with one’s own eyes; analysing a position may lead to the discovery of
improvements and will definitely further your understanding.
5...Bxf3
Solid and probably best. Instead 5...Bh5 6 Nc3 e6 7 g4! Bg6 8 Ne5 (Diagram 14)

Diagram 14 (B)
Chasing down the bishop
260
 wins the bishop after all, while gaining some useful space in the process. Once again Black must also be
careful that his bishop isn’t targeted; for example, 8...Nbd7?! 9 h4! dxc4 10 Nxc4 h5 11 g5 Ne4 12 Bg2 Nxc3
13 bxc3 Qc7 14 a4! Rd8 15 0-0 Be7 16 e4 left White better right across the board in M.Hebden-A.Ledger,
Cork 2005.
6 Qxf3 e6 7 Nc3 Nbd7 8 Bd3 Bd6 9 0-0 (Diagram 15)

Diagram 15 (B)
A tough struggle lies ahead

White has developed harmoniously thus far, and now it’s useful to realize that 9...e5? simply lose a pawn
to 10 cxd5 cxd5 11 Nxd5, while 9...dxc4 10 Bxc4 e5 11 Ne4! favours White. Black must prepare his
counterplay more slowly.
9...0-0 10 Bd2
Continuing to develop, but there’s no need to move the bishop just yet. Instead 10 Rd1 looks like a better
try for the advantage and now:
a) 10...Qe7 11 e4!? e5?! (natural but wrong; instead 11...dxe4 12 Nxe4 opens up White’s bishop pair, but
11...dxc4! 12 Bxc4 e5 13 d5 Nb6 14 Bb3 cxd5 15 Nxd5 left Black very close to full equality in E.Bareev-
B.Gelfand, Khanty Mansyisk 2005) 12 c5! (improving over the less effective 12 exd5 exd4 13 Ne4 cxd5 14
cxd5 Nxe4 15 Qxe4 Qxe4 16 Bxe4 Rfe8 of X.Parmentier-N.Pert, Montpellier 2003) 12...Bb8 (the best try;
12...Bc7 13 exd5 exd4 14 d6 is very good for White) 13 exd5 exd4 14 d6 (Diagram 16)

261
 Diagram 16 (B)
The complications favour White

14...Qe5 (14...Qe6!? improves, but 15 Na4 Ne5 16 Qg3 still resolves itself slightly in White’s favour) 15
Na4! b5 16 b4! (revealing White’s main idea; 16 Qxc6 bxa4 17 Qxa8 Nxc5 18 Qf3 Qxd6 isn’t so clear)
16...bxa4 17 Qxc6 and with 17...Qd5 well met by 18 Bb5, White wins the rook in the corner on his terms and
with a large advantage; an excellent discovery by John Cox.
b) 10...Re8 11 Bd2 (without the queen on e7, 11 e4?! is no longer effective: 11...e5! 12 exd5 exd4 13 Ne4
Nxe4 14 Bxe4 c5! gave Black good play on the dark squares in U.Capo Vidal-N.Pert, Turin Olympiad 2006)
11...Qe7 12 Bf1 Rad8 13 Rac1 h6 14 a3 a6 15 Re1 Bb8 16 g3 dxc4! 17 Bxc4 e5 (Diagram 17)

Diagram 17 (W)
...e5 is Black’s ideal break

262
 and the position was fairly level in A.Yusupov-R.Dautov, German League 2002.
10...Qe7 11 b3!?
A useful move directed against ...dxc4 and there’s no need to rush. Indeed Georgiev is happy to provoke
...e5, while Sokolov shortly falls into a bad position.
11...Rfe8 (Diagram 18)

Diagram 18 (W)
White must avoid 12 e4? e5!

12 Qd1
WARNING: If he’s going to break with e3-e4 in this line, White must get his timing right. Here, for
example, 12 e4? backfires after 12...e5! and White lacks a good move since 13 c5 exd4 14 cxd6 Qxd6
regains the piece with some advantage.
12...a6?!
Possibly intending ...b5, but this creates weaknesses and fails to convince. I assume that Sokolov didn’t
like the look of the natural 12...e5 13 cxd5 cxd5 14 Nb5, but this doesn’t seem too bad at all for Black after
14...Bb8!.
13 c5! Bc7 14 f4!
Clamping down on Black’s pawn breaks (which were ...e5, ...c5 and ...b5), and now Black is in some
trouble. He can’t afford to passively sit and wait as White builds up on one flank or the other, and so Sokolov
already finds himself having to employ a slightly speculative piece sacrifice.
14...Rad8
Sokolov’s Informant notes mention the possibility of 14...Ba5!? 15 Qf3 Bxc3 16 Bxc3 Ne4! 17 Bxe4 dxe4
18 Qxe4 f5 19 Qf3 Nf6. He assesses this as slightly favouring White, but I feel that it was Black’s best. In
return for a pawn, he gains good knight against bad bishop and White will have to be patient to build up on
either flank.
15 Qf3 (Diagram 19)

263
 Diagram 19 (B)
White has a serious clamp

15...e5!? 16 fxe5 Nxe5 17 dxe5 Qxe5 18 g3 Ba5!

Continuing to play actively and pose problems. Indeed Black cannot afford to lose the initiative as would
have been the case after 18...d4? 19 Ne2 dxe3 20 Bc3 (Sokolov).
19 Nb1! Bc7
Alternatively, 19...Bxd2 20 Nxd2 Qc3 21 Qe2 Qxc5 22 Rae1 fully co-ordinates White’s pieces, while
19...Qxa1? 20 Bxa5 Rd7? 21 Bc3 Qxa2 22 Bxf6 leads to a crushing attack.
20 Nc3
Missing his chance with 20 Na3! (Diagram 20)

Diagram 20 (B)
Complicated but good for White
264
 when the knight is en route to the strong d4-square and White is clearly happy to lose g3 to get the queens
off. The tactical justification is 20...Ne4 (or 20...Re7 21 Qg2!) 21 Bxe4 dxe4 22 Qxf7+ Kh8 23 Be1! Qxa1 24
Nc4 (Sokolov) and Black remains in serious trouble, despite having won a clear rook. The threat is 25 Ba5
and the black queen can’t retreat along the a1-h8 diagonal, while 24...Qxa2 fails to 25 Bc3 Rg8 26 Qxc7. Of
course, going from being a piece up to an exchange down, even in return for such an initiative, isn’t
psychologically an easy thing to do, but this was by far White’s best way of utilizing his advantage.
TIP: When we have a material or positional advantage, it is often tempting to calculate as little as
possible. Keeping things simple can work well, but it is often the case that working out the tactics leads
to a simpler conversion.
20...Ba5 21 Rac1?!
It still wasn’t too late for 21 Nb1 Bc7 22 Na3!.
21...d4! 22 Nb1 Bxd2 23 Nxd2 dxe3 24 Nc4 Qxc5 25 Be2 Qe7 (Diagram 21)

Diagram 21 (W)
Black has dangerous compensation

Perhaps by sacrificing two pawns to get the dark-squared bishops off, Georgiev thought that he would
quell Black’s activity and could then round up the e-pawn. If so, this was a surprising misjudgement from
such a strong grandmaster; the e3-pawn ties down White’s pieces and is far from easy to win. Sokolov now
feels that the prophylactic 26 a4!? was White’s last chance to retain a small advantage.
26 Rc3?! Nd5 27 Rd3 b5!
Ever since falling into a passive position Sokolov has played in an impressively practical and active
manner. Unsurprisingly he continues in such vein.
28 Na5 Qc7 29 Nxc6!?
Imaginatively returning the piece to create problems with a strong pin down the d-file. Objectively,
though, 29 b4! Nxb4 30 Rxd8 Rxd8 31 a3 Nc2 32 Nxc6 Rd6 33 Rd1 was probably a better way for White to
maintain a rough balance.
29...Qxc6 30 Qxf7+ Kh8 31 Rfd1 Re5 32 h4 h6 33 a4 bxa4 34 bxa4 Qc5 35 Qf3 a5 36 Rd4 (Diagram
22)

265
 Diagram 22 (B)
White hopes to force perpetual

36...Qc2?!
It wasn’t at all easy for Black to make progress due to the pin, but this allows Georgiev to force an
immediate draw with a tactic which he finds at the second time of asking.
37 g4?! Qc5 38 Kh1 Qe7?! 39 Rxd5!
Seizing his opportunity; there’s nothing Black can now do about the coming perpetual.
39...Rexd5 40 Rxd5 Rxd5 41 Qxd5 Qxh4+ 42 Kg2 Qf2+ 43 Kh3 Qxe2 44 Qa8+ Kh7 45 Qe4+ ½-½

Points to Remember
1. This is a very solid line for Black, but White can hope for a small pull with both 10 Rd1 and 10 Bd2.
2. White must time any e4-break very carefully; likewise Black must make sure that his pawn breaks aren’t
eliminated, as occurred in the game.
3. When a piece ahead for some activity, don’t be afraid to sacrifice to retain and transform the advantage.
4. When struggling always consider the active approach. Sokolov’s piece sacrifice wasn’t fully sound, but it still caused plenty of trouble
for his 2650+ rated opponent.

Game 26
G.Bogdanovich-D.Ortmann
Ditzingen 2002

1 d4 Nf6 2 Nf3 d5 3 e3 Bg4 4 h3 Bh5 5 g4 Bg6 6 Ne5 Nfd7 (Diagram 23)

266
 Diagram 23 (W)
Black’s correct response

7 Nxg6
WARNING: After the accurate 6...Nfd7!, White must content himself with winning the bishop pair,
not carry on with 7 h4?! since 7...Nxe5 8 dxe5 h5! now leaves him a little overextended.
7...hxg6 8 c4
White’s natural break, but he doesn’t have to hurry with this. Instead 8 Bg2 c6 (8...e6 9 c4 c6 10 Qe2 transposes) 9 Qe2 e6 (9...e5?! 10 e4!
opens the centre to White’s advantage in thematic Colle style) 10 c4!? (10 Nd2 Nf6 11 c3 Nbd7 left Black solidly placed in S.Mannion-J.Shaw,
Scottish Championship, Hamilton 2004; c4 rather than e4 is a more challenging pawn break to employ) 10...dxc4 (otherwise White develops
with Nc3, Bd2 and possibly f4, while Black is a little cramped) 11 Qxc4 Nb6 12 Qb3 N8d7 13 Nc3 is slightly better for White. Note that
there’s no need to hurry to castle short in this position; indeed should Black go short himself, White may well prefer to go long and push the h-
pawn.
8...dxc4!? 9 Bxc4
Against Black’s active approach (8...c6 remained the solid alternative), I also quite like 9 Qf3!? c6 (or
9...Nc6 10 Bxc4 e6 11 Nc3 and if 11...Nb6, then 12 Bb5!? is an option) 10 Bxc4 e6 11 Bd2 (Diagram 24)

267
 Diagram 24 (B)
White’s extra space is useful

when White remains quite flexibly placed. In J.Harmatosi-T.Johansen, Balatonbereny 1997, Black now
wrongly hurried to exchange queens: 11...Qf6?! 12 Be2 Qxf3 13 Bxf3 Na6 14 Nc3 Nb4 15 Kf1 0-0-0 16 Kg2
Be7 17 Rad1 Nf8 18 a3 Nd5 19 b4! f5 20 b5 and White had instructively increased his advantage.
9...Nb6
Neither did 9...e6 10 Nc3 Be7 11 Qf3 Nc6 12 Bd2 Bh4 13 Ne4 0-0 14 g5! Bxg5 15 h4 Bh6 16 0-0-0,
threatening to both gain a strong attack and entomb Black’s bishop, convince for Black in O.Rodriguez
Vargas-A.Miles Las Palmas 1978.
10 Bb3 e6 11 Qf3!
The queen is well placed on this aggressive square, pressurizing b7 and f7, and supporting White’s
kingside pawns.
11...c6?!
A little passive and inconsistent with Black’s 8th. Instead Prie points out that 11...Nc6! 12 Bd2 Qd7 13
Nc3 (raising the stakes with 13 a4!? a5 14 Kf1 and then Nc3-b5 is also possible) 13...0-0-0 14 0-0-0 Na5 15
Bc2 Nac4 was a much better try, restricting White to just a tiny edge.
12 Nc3 N8d7 13 Bd2 a5 14 a3!
Keeping the bishop on the a2-g8 diagonal, both to prevent ...Nc4 and to support a later advance of White’s
central majority.
14...Nf6 15 0-0-0 a4 16 Ba2 Nfd5?!
This doesn’t help matters, but Black was already in some trouble with his king lacking a safe home and e4
imminent; for example, 16...Qd7 17 e4 and due to 17...Qxd4? 18 Be3 winning a piece, White has time to
follow up with Kb1, Bf4 and h4-h5 with a very pleasant position indeed.
17 e4 Nxc3 18 Bxc3 Qg5+ 19 Kb1 Qb5 20 Ka1 Bd6?! 21 d5! (Diagram 25)

268
 Diagram 25 (B)
Blasting open the position

Thematically blowing open the centre.

TIP: When you have the bishop pair, try to open lines for it. When attacking in harness, the two
bishops can be a deadly asset.
21...cxd5 22 Bxg7 Rh7 23 Bf6 dxe4 24 Qxe4 Nc8?!
Making matters easy, but after 24...Nd7 25 Rxd6 Nxf6 26 Rxe6+! fxe6 27 Qxe6+ (Prie), White regains a
piece while maintaining a crushing attack; for example, 27...Re7 28 Qxf6 Ra6 29 Bf7+! Rxf7 30 Re1+ and
mate follows.
25 Bxe6! (Diagram 26)

Diagram 26 (B)
A decisive sacrifice
269
 The Russian IM continues in true textbook style to demonstrate how to attack a king stranded in the centre.
25...fxe6 26 Qxg6+ Rf7 27 Qg8+ 1-0
A slightly premature resignation perhaps, but after 27 Qg8+ Rf8 28 Qxe6+ Be7 29 Rhe1 Rf7 (or 29...Qc5
30 Qd7+ Kf7 31 Bxe7 Nxe7 32 Rd5) 30 Rd8+! Kxd8 31 Qxf7 (Prie) it becomes apparent that Black cannot
maintain his extra piece and is helpless against the attack.

Points to Remember
1. After 6 Ne5, Black must find 6...Nfd7.
2. Nonetheless, White gains a slight advantage with his bishop pair and extra space, and should break with
c4.
3. There’s no need for White to hurry to castle; the king may turn out to be best placed on b1, g1 or even
f1.
4. White may be able to target Black’s king; either with h4-h5 or, should Black not castle, by advancing his central majority.
Black Plays 3...Bf5
1 d4 d5 2 Nf3 Nf6 3 e3 Bf5 (Diagram 27)

Diagram 27 (W)
A Slav beckons

A solid approach and one recommended in John Cox’s fine repertoire work, Dealing with d4 Deviations.
4 Bd3
Neither 4 b3 nor 4 Be2 lead anywhere for White; it could be said that he is playing against the London
System, albeit with an extra tempo, and so needs to employ one of Black’s better defences there.
NOTE: Rather than drift into passivity, White must place Black under some pressure in the centre
which usually involves playing an early c4.
More flexible and probably more promising than our main line is the immediate 4 c4 when Black should
transpose to a Slav:
a) 4...e6?! is a typical inaccuracy from players who haven’t bothered to prepare a proper response to the
Colle. White should immediately pinpoint the mistake with 5 Qb3! (Diagram 28)

270
 Diagram 28 (B)
Black begins to regret ...Bf5

and now:
a1) 5...b6 is rather weakening and after 6 Nc3 Be7 7 Ne5! Black struggles to develop his queenside since
...Nbd7 is met by Nc6; for example, 7...0-0 8 Be2 Ne4 9 cxd5 Nxc3 10 bxc3 exd5 and now an immediate 11
c4 is one good move, while 11 0-0 f6?! 12 Nd3 Bg6 13 Nf4 Bf7 14 c4 saw Black continuing to struggle on
the light squares in G.Maróczy-D.Marotti, London 1922.
a2) 5...Nc6!? is an active try, but just as in the 2...Bf5 variation (see the next chapter), 6 c5! (and not 6
Qxb7? Nb4!) is a good response; for example, 6...Rb8 (6...Qc8 is also well met by 7 Bb5) 7 Bb5 (Diagram
29)

Diagram 29 (B)
White has some pressure
271
 7...Be7?! 8 Ne5! 0-0 9 Bxc6 bxc6 10 Nxc6 Rxb3 11 Nxd8 Rb8 12 Nc6 left White a clear pawn ahead in
A.Rubinstein-B.Blumenfeld, St Petersburg 1905.
a3) 5...Qc8 6 Nc3 c6 is the solid approach when 7 Nh4! Bg4 8 h3 Bh5 9 g4 Bg6 10 Nxg6 hxg6 11 Bg2
gives White a small edge due to his bishop pair and extra space.
b) 4...c6 reaches a Slav. Readers should really consult a specialist work on that opening, but it is possible
to play this position without knowing much more than a few key principles. Here’s a quick round-up of two
different approaches:
b1) 5 Nc3 e6 6 Nh4! (the only real try for the advantage; instead 6 Bd3 transposes to our main line, while 6
Qb3 Qb6 is fine for Black who meets 7 cxd5 with 7...Nxd5) 6...Bg6 (6...Be4 7 f3 Bg6 is similar and after 8
Nxg6 hxg6 9 a3!? Nbd7 10 g3 Be7 11 f4 play has actually transposed to Game 27; at one time Black
preferred 6...Bg4, but White’s bishop pair and extra space after 7 Qb3 Qb6 8 h3 Bh5 9 g4 Bg6 10 Nxg6 hxg6
11 Bg2 is considered to give him a little something) 7 Nxg6 hxg6 (Diagram 30)

Diagram 30 (W)
A fashionable position

sees White bag the bishop pair, but Black remains very solid. This line has always had its grandmaster
supporters, but recently the white cause was injected with a number of new ideas in the Elista World
Championship match; see Game 27.
b2) 5 cxd5 cxd5 6 Qb3!? is Mark Hebden’s tricky favourite and a line which I endorsed in Play 1 d4!. It is
still a good try at club level since White can easily emerge with a positional advantage; for example, 6...Qc7
(5 Qb3 can be met by 5...Qb6 since Black then has 6 cxd5 Nxd5!, but here he has already recaptured with the
pawn; thus 6...Qb6 is no longer so easy for him and 7 Qxb6 axb6 8 Nc3 e6 9 Bb5+ Nfd7 10 Ne5 Bd6 11
Nxd7 Nxd7 12 Bd2 gave White a small but clear edge in M.Hebden-C.Crouch, British League 1998) 7 Bd2
Nc6 8 Bb5 e6 9 0-0! (much more testing than the immediate and more common 9 Bb4) 9...Bd6 10 Bb4
(Diagram 31)

272
 Diagram 31 (B)
White has an edge

10...0-0 11 Bxc6 bxc6 12 Nbd2 when the weakness of c6, allied to Black’s non-too-useful light-squared
bishop, gives White a nibble. Unfortunately there is, however, currently a theoretical problem with this line,
as Dealing with d4 Deviations observes, in the shape of the rare 7...e6 8 Bb5+ Nbd7! when 9 0-0 a6 10
Bxd7+ Nxd7 11 Ne5 was R.Palliser-J.Houska, British League 2003, and now 11...Bc2 was one way to
equalize.
Returning to 4 Bd3:
4...Bxd3
Black can also aim to unbalance the position with 4...e6!?. White might then continue 5 c4 when play
should transpose after any exchange on d3. Instead in Game 28 we’ll consider the alternative structure arising
after 5 Bxf5 exf5.
5 Qxd3 c6 (Diagram 32)

273
 Diagram 32 (W)
White must break with c4 or e4

Preventing any check on b5 and now play usually once more transposes to a Slav.
6 c4
6 0-0 e6 7 Nbd2 Nbd7 8 e4 dxe4 9 Nxe4 Be7 10 Nxf6+ Bxf6 11 Bf4 0-0 was pretty even in L.Vadasz-
E.Postny, Budapest 2000, before White was outplayed following 12 c4 c5!? 13 dxc5 (Black’s thematic
equalizing break; this is often delayed for a while, but here the point is that 13 d5 exd5 14 cxd5 Nb6 hits both
d5 and b2) 13...Nxc5 14 Qe3?! Qb6 15 Rad1 Na4! 16 b3 Nc3 and Black seized control of the d-file.
6...e6 7 Nc3 Nbd7 8 0-0 Bb4
Both sides have developed rather naturally so far. Black’s last aims to take control of the e4-square, but he
can also be less adventurous with 8...Be7: for example, 9 b3 0-0 10 Bb2 a6!? (hinting at ...b5 and so forcing
White into action in the centre) 11 e4 dxe4 12 Nxe4 Nxe4 13 Qxe4 Qa5! (the black queen is often well placed
here in such positions; White now has troubling forming a positive plan) 14 a4 Rad8 15 Rfe1 Bf6 (Diagram
33)

274
 Diagram 33 (W)
Black is very solid

16 Rad1 Rfe8 17 h3 Qc7 18 Qc2 Nf8 19 g3 c5 20 d5 and with exchanges imminent, a draw was agreed in
V.Malakhov-A.Dreev, Solin 2006.
9 Bd2
Setting a little trap (9...0-0? 10 Nxd5!). Instead 9 a3 Bxc3 10 Qxc3 0-0 is fine for Black who follows up
with ...Ne4 and/or ...b5.
9...a5! 10 a3 Be7!?
And Black in turn hopes for a fiendish little trick, rather than simply exchange on e4.
11 b3 0-0 12 Rfd1! (Diagram 34)

Diagram 34 (B)
It’s about even
275
 WARNING: White must avoid 12 e4? Nc5! when 13 dxc5 dxe4 14 Qxd8 Rfxd8 regains the piece,
leaving Black a pawn ahead.
After 12 Rfd1 the position is pretty level, although both sides can hope to patiently outmanoeuvre a weaker opponent. One game continued
12...Re8 13 e4 dxe4 14 Nxe4 Nxe4 15 Qxe4 Bf8 16 Bc3 Qc7 17 h3 b5! and Black was fine in G.Kallai-R.Dautov, French League 2004.

Theoretical Conclusion
3...Bf5 is a very solid approach. White is unlikely to get anywhere with 4 Bd3 and so 4 c4, taking play into a Slav, is his most challenging
option, albeit one in which Black still remains rather solid and hard to break down.
A Tricky Transposition: 3...c6
A related variation to 3...Bf5 is 1 d4 d5 2 Nf3 Nf6 3 e3 c6!? (Diagram 35).

Diagram 35 (W)
Black again plays for a Slav

Now 4 Bd3 Bg4 is pretty comfortable for Black and so White must really take up the challenge with 4 c4.
In this Slav position, Black will often play either 4...Bf5, which we’ve just examined, or 4...Bg4, a move
which we saw in Game 25. However, he also has:
a) 4...g6 is the solid Schlechter system; White has good chances of gaining an edge with straightforward
positional play after 5 Nc3 Bg7 6 Bd3 0-0 7 0-0.
b) 4...a6!? is currently quite trendy. White can now develop in Colle-related style with 5 Bd3, followed by
Nbd2 and/or Qb3, or consider one of the main lines: either the positional 5 Nc3 b5 6 b3 or there the trendy 6
c5!?.
c) 4...e6 hopes to reach a Semi-Slav after 5 Nc3. One good independent option is 5 Nbd2!? when 5...Nbd7
(5...c5, often leading to an IQP situation, is probably more critical) 6 Bd3 (6 b3 and Bb2 is another reasonable
option, clamping down on the e5-square in Zukertort fashion) 6...Bd6 7 0-0 0-0 8 e4 e5 9 cxd5 cxd5 10 Re1!
(Diagram 36)

276
 Diagram 36 (B)
Have I not seen you before?

transposes to a position we considered at the beginning of Chapter Three!

Illustrative Games

Game 27
V.Topalov-V.Kramnik
World Championship (Game 9), Elista 2006

1 d4 d5 2 c4 c6 3 Nf3 Nf6 4 e3 Bf5 5 Nc3 e6 6 Nh4 Bg6 7 Nxg6 hxg6 8 a3!?

Topalov’s legendary preparation came up with a new concept in this much-tested position, namely to
remain flexible, being ready to either clamp down in the centre (with f4 and c5) or to quickly advance on the
queenside. White begins with a3 to prevent Black from trading off his second bishop. Indeed when Kramnik
adopted this line as White in the final matchgame proper, that’s exactly what occurred: 8 g3 Nbd7 9 Bd2
Bb4! (9...Be7 was more common before the match) 10 Qb3 Bxc3 11 Bxc3 Ne4 (Diagram 37)

277
 Diagram 37 (W)
The b-pawn is poisoned

12 Bg2 (12 Qxb7?! Rb8 13 Qxc6?! Rxb2!! is excellent for Black since 14 Bxb2?? allows mate beginning
with 14...Qa5+ – Notkin) 12...Nxc3 13 Qxc3 f5 and Black was solidly placed in V.Kramnik-V.Topalov,
World Championship (Game 12), Elista 2006.
With his next white in the match, Topalov opted for an immediate queenside advance: 8 Rb1!? Nbd7 9 c5
a5 10 a3 e5 11 b4 axb4 12 axb4 Qc7 13 f4!? exf4 14 exf4 and White was perhaps slightly for choice in
V.Topalov-V.Kramnik, World Championship (Game 11), Elista 2006.
8...Nbd7 9 g3
This had been played before, but not with Topalov’s follow-up (which was a novelty) in mind.
Considering that, White might also give consideration to beginning with 9 f4!? when 9...a6 10 Bd2 c5!? 11
cxd5 exd5 12 Qf3 Nb6 13 Bd3 cxd4 14 exd4 Bd6 15 0-0 Kf8! was roughly balanced in R.Palliser-S.Ansell,
British Rapidplay Championship, Halifax 2006.
9...Be7
Black has also tried 9...a6 when the thematic 10 c5! e5 11 Bg2 Be7 12 0-0 Kf8 13 b4 Kg8 14 f4! e4 15 a4
gave White some queenside initiative and the advantage in V.Milov-M.Tissir, Gibraltar 2007.
10 f4! (Diagram 38)

278
 Diagram 38 (B)
Seizing some useful space

10...dxc4
Much remains to be discovered in this complex variation; for example, 10...g5!? 11 fxg5 Nh5 12 g6 f5 13
g4!? Ng3 14 hxg3 Rxh1 15 gxf5 Nf6 (Notkin) gives White fair play for the exchange, but is far from clear.
11 Bxc4 0-0?!
Too slow. Kramnik’s relative lack of Slav experience appeared to trouble him at certain points in this
match, and here he needed to play more actively with either 11...Nb6 and ...c5, or even an immediate
11...c5!?.
12 e4
Taking over the centre and preparing to meet ...c5 with e5, thereby ensuring White of the advantage; Black
is a little too passive.
12...b5 13 Be2 b4 14 axb4 Bxb4 15 Bf3 (Diagram 39)

279
 Diagram 39 (B)
Black must get in ...c5

15...Qb6?
Black must gain counterplay before White consolidates his positional advantages, but after this his central
pawn breaks fail to impress. Much better was 15...c5! and if 16 e5, then 16...cxd4 17 Qxd4 Nd5 18 Bxd5 Bc5
19 Qd3 exd5 (Golubev) and Black gains counterplay with White beginning to look a little overextended.
16 0-0 e5
Now 16...c5 is powerfully met by 17 e5 Bxc3 18 bxc3 Nd5 19 Qa4 with some advantage as the position
begins to open up for White’s bishops.
17 Be3 Rad8 18 Na4 Qb8 19 Qc2 (Diagram 40)

Diagram 40 (B)
Black badly lacks counterplay
280
 Pinpointing one of Black’s many weaknesses and now Kramnik is forced to exchange in the centre,
leaving him facing a rather grim defence.
19...exf4 20 Bxf4 Qb7 21 Rad1 Rfe8 22 Bg5 Be7 23 Kh1!
There’s no need to hurry and so White tucks his king away; a useful move in case the g1-a7 diagonal
should later open.
TIP: When you have a number of positional advantages and the opponent lacks counterplay, there is
often no need to hurry. Grandmasters frequently improve as many of their own pieces as possible
before embarking on progressive action.
23...Nh7 24 Be3!? Bg5 25 Bg1! (Diagram 41)

Diagram 41 (B)
There’s no hurry for White

Of course, White wants to keep the bishops on. An exchange would not only help to free some of Black’s
cramp, but would also lessen White’s advantage.
NOTE: Two bishops work well together by complementing each other; if one is exchanged off, the
other’s influence is often reduced.
25...Nhf8 26 h4 Be7 27 e5!
The pressure mounts. Kramnik is now reduced to grovelling around on his back two ranks; unsurprisingly
such a defence is unlikely to hold out for long.
27...Nb8 28 Nc3 Bb4 29 Qg2 Qc8 30 Rc1 Bxc3 31 bxc3 Ne6 32 Bg4! Qc7 33 Rcd1 Nd7 34 Qa2 Nb6 35
Rf3 Nf8?
Collapsing, but after something like 35...Nd5 36 Rdf1 Rd7 37 Qh2!, h5 would have followed with some
effect.
36 Rdf1 Re7 37 Be3! Nh7
Forced to cover g5, but now White crashes through.
38 Rxf7! (Diagram 42)

281
 Diagram 42 (B)
Crashing through

38...Nd5
The game’s up: 38...Rxf7 39 Rxf7 Qxf7 40 Be6 costs Black his queen.
39 R7f3 1-0

Points to Remember
1. Much remains to be discovered in this line. It’s thus a good choice for any reader who doesn’t mind sometimes following grandmaster
fashion.
2. An early f4 can cause Black serious trouble gaining enough central counterplay.
3. When the opponent is passive, don’t hurry: Topalov built up calmly and most impressively.
4. Treasure the bishop pair, a large asset in many open and semi-open positions.

Game 28
A.Alekhine-M.Euwe
World Championship (Game 17), Eindhoven 1935

1 d4 d5 2 Nf3 Nf6 3 e3 Bf5 4 Bd3 e6 5 Bxf5 exf5 (Diagram 43)

282
 Diagram 43 (W)
Are the doubled pawns good or bad?

This line was relatively popular in the thirties, but is now rarely seen. Quite simply the early tests
demonstrated that Black’s grip on the e4-square, as well as his potential pressure down the half-open e-file,
fully compensates for the doubled pawns.
6 Qd3 Qc8 7 b3
White needs to try and open the centre if he is to prove that Black’s doubled pawns are a weakness; d5 can
only now be defended by Black’s c-pawn, no longer also his e-pawn. Alekhine had previously preferred the
slightly slower 7 0-0 Bd6 8 b3 0-0 9 c4 c6 10 Nc3 and now 10...b6?! 11 cxd5 Nxd5 12 Nxd5 cxd5 13 Bb2 g6
14 Rfc1 Qa6 15 Qc3 Rd8 16 Ne5 gave White the advantage in A.Alekhine-B.Kostic, Bled 1931, but as
practice later demonstrated the much more solid 10...Na6 11 a3 Nc7 is a clear improvement.
White might thus begin with an immediate 7 c4, but in a rare recent test Black developed sufficient
counterplay following 7...dxc4! 8 Qxc4 Bd6 9 Nc3 c6 10 b4 0-0 11 b5 Nbd7 12 Bb2 Nb6 (V.Vorotnikov-
M.Kobalija, Moscow 1999).
7...Na6! (Diagram 44)

283
 Diagram 44 (W)
The knight is well placed on the rim

Euwe was a highly respected theoretician and this is exactly how Black should handle the opening.
NOTE: As we’ve seen, White would like to open the c-file and to seize a queenside initiative. The text
prepares to recapture on d5 with the queen’s knight (from c7 or b4), thereby keeping lines closed.
8 0-0 Be7 9 c4 0-0 10 Nc3
Of course, 10 cxd5 would have been met by 10...Nb4 and ...Nbxd5.
10...c6 11 Bb2 Ne4 (Diagram 45)

Diagram 45 (W)
Black is comfortable

Only now does Black occupy the hole on e4. White can hardly exchange the knight off and already it is
284
clear that Black has quite an easy game.
12 Rfc1
White would like to continue 12 cxd5 Nb4 13 Qe2, but 13...Nxc3! 14 Bxc3 Nxd5 still sees Black carrying
out his ideal recapture on d5. He may then regroup with ...Bd6, ...Qe6 and ...Re8, after which ...Qe4 or ...f4
may follow.
12...Rd8 13 Qe2 Qe6 14 a3
Threatening to exchange on d5 and also hoping to seize the initiative on the queenside. However, pushing
the queenside pawns is quite a slow plan and Euwe is in time to gain counterplay in the centre.
14...Nc7 15 c5!? Re8 16 b4 f4! 17 exf4 Nxc3 18 Qxe6 Nxe6 19 Rxc3 Nxf4
If anyone must be careful here it’s White since he wouldn’t want to play a pure bishop ending, but
Alekhine manages to keep Black’s knight out and to overprotect d4.
20 Rb3! a6 21 g3 Ne6 (Diagram 46)

Diagram 46 (W)
It’s roughly level

22 a4 Bf6 23 Rd1 ½-½

The position remains rather equal: Black hasn’t managed to achieve anything down the e-file, while White is ready to further liquidate if
necessary with b5.

Points to Remember
None really. The exchange on f5 isn’t troubling for Black so long as he knows the ...Na6 defence.
Black Plays 3...g6
1 d4 d5 2 Nf3 Nf6 3 e3 g6 (Diagram 47)

285
 Diagram 47 (W)
Blunting any bishop on d3

A tricky option. By fianchettoing Black not only has ideas of later breaking with ...e5, but takes much of
the sting out of Bd3. White now has a number of set-ups to choose between.
4 Bd3
Staying in true Colle territory, but it is also possible to switch to a Zukertort approach with 4 b3 to contest
the long diagonal. Now 4...Bg7 5 Bb2 0-0 6 Be2 was recommended by Summerscale and resembles a
reversed closed Catalan. That is a solid choice for Black, but here it’s hard to make any real use of the extra
tempo; for example, 6...c5 (Black’s main option, but there’s also nothing much wrong with the solid 6...c6 7
0-0 Bg4 8 Nbd2 Nbd7 when 9 c4 leads to a tough manoeuvring game) 7 0-0 Nc6 (7...cxd4 8 exd4 Nc6 9
Nbd2 transposes; instead 9 Re1 Ne4 10 c3?! e5! gave Black the initiative in T.Mamedjarova-G.Ginsburg,
Lausanne 2005) 8 Nbd2 cxd4 9 exd4 (Diagram 48)

286
 Diagram 48 (B)
A manoeuvring struggle ensues

9...Bf5 (9...Ne4!? 10 c4 Bf5 is also quite critical and about equal) 10 Nh4?! (rather planless; a quick c4, as
indeed White shortly plays, is the only way to challenge Black) 10...Be6 11 Nhf3 Rc8 12 Rc1 Nb4 (the more
active 12...Ne4!? 13 c4 Qa5 was also possible) 13 c4 Bf5 14 Ne5 Nd7! was unbalanced and left Black
slightly for preference in L.Lauer-J.Lautier, Paris 2000.
Once again I feel that White should at least consider taking play away from the Colle with 4 c4 Bg7 5 Nc3
(Diagram 49) which actually transposes to a Grünfeld!

Diagram 49 (B)
Does Black play the Grünfeld?

NOTE: Watson and Schiller, in keeping with modern theory’s view, describe this variation as
‘harmless’. They don’t give any further moves, but just how many clubplayers play, or are familiar
with, both 1...d5 and the Grünfeld?
Following 5...0-0 I believe that White has two interesting ways to pose problems:
a) 6 b4!? prevents Black’s standard ...c5-break, which is the reply to 6 Be2, and has recently received
attention from grandmasters Akobian, Alexandrov, Bruzon and Rustemov. Black now has a number of
possible approaches including the sharp and critical 6...b6. For fairly up-to-date coverage, the reader is
referred to a two part survey by Panczyk and Ilczuk in New In Chess Yearbook (volumes 78 and 80).
b) 6 cxd5 Nxd5 7 Bc4 is the Keres variation. Just as against 6 b4, Black can equalize with accurate play,
but White can pose a number of practical difficulties as Timothy Taylor demonstrates in his recent Beating
the King’s Indian and Grünfeld.
I feel that 4 c4 and then one of these two sub-variations of the Grünfeld is White’s most challenging option
against 3...g6, but by no means every reader will want to have to consult another source. Whether to do so
really depends on:
i. How often you face 3...g6 (probably a more common move at higher levels).
ii. How comfortable you are against 3...g6 with either 4 Bd3 or 4 b3.
If your answers are ‘often’ and ‘not very’ then why not give 4 c4 a go? It should certainly come as an
unpleasant surprise to many opponents.
Returning to 4 Bd3:
287
 4...Bg7 5 0-0 0-0 6 Nbd2 (Diagram 50)

Diagram 50 (B)
The standard Colle deployment

Continuing in true Colle style.

WARNING: Against the solid kingside fianchetto, a Stonewall set-up no longer convinces.
One example being 6 Ne5?! c5 7 b3 (7 c3 is perhaps more consistent, but the thematic 7...Qb6 8 b3 Nc6 9
f4 Nd7 10 Kh1 f6! 11 Nxc6 Qxc6 was slightly better for Black in E.Piankov-V.Bukal, Halle 2004) 7...Nfd7 8
f4?!, as in J.Stockham-R.Palliser, Yarm 2000, and now the common trick 8...cxd4 9 exd4 Qb6! 10 c3 Nxe5
11 fxe5 Bxe5 nets a pawn.
6...c5
Black’s most popular move. He has scored quite well with this, but it’s by no means his only playable
option. Indeed both 6...b6 and 6...Nc6!? also deserve serious consideration; for example, after the latter, 7 c3
(or 7 c4 Re8 8 cxd5 Nxd5 9 Nc4 Nb6! 10 Bd2?! e5 with excellent Grünfeld-style counterplay in P.Doostkam-
A.Mallahi, Teheran 2001) 7...e5 (Diagram 51)

288
 Diagram 51 (W)
Black will regain his pawn

8 Nxe5 Nxe5 9 dxe5 Ng4 sees Black regain his pawn with easy equality.
7 c3 Nbd7
Overprotecting c5. Instead 7...Nc6 is rarely met by the critical 8 dxc5!?, but it probably should be. This
pawn grab has been condemned by theory ever since the game E.Colle-E.Bogoljubow, Bled 1931, but it is by
no means such a bad option: 8...Nd7 (or 8...e5 9 e4! – the ideal Colle break – 9...Qe7 10 b4 Rd8 11 Qe2 Nh5
12 Nb3 and now in K.Karlsson-A.Ziegler, Jonkoping 1987, even the critical 12...Nf4!? wouldn’t have given
Black enough compensation after 13 Bxf4 exf4 14 b5 Ne5 15 exd5 Rxd5 16 Bc4) 9 Nb3 a5 10 a4 Nde5 and
now Colle’s 11 Be2 was perhaps a little passive. White might also consider 11 Nxe5 Nxe5 12 Bb5!?
(Oleinikov) when it will take him a while to untangle (Nd4, b3 and Ba3 is one plan), but it isn’t clear that
Black has sufficient compensation since 12...Qc7 leaves the d-pawn hanging.
8 e4 (Diagram 52)

289
 Diagram 52 (B)
Less effective than after 3...e6

Continuing in typical Colle fashion. White often adopts such a plan, but prefers first to insert 8 Re1. For
that option see Game 29 which demonstrates that both sides can try to win the unbalanced but equal positions
that result from the e4-advance.
Smith and Hall advocated the idea of 8 b4!?, challenging Black on the queenside. This is an interesting
way to set Black different problems to e4 followed by central liquidation, but I’m afraid that it also doesn’t
suffice for the advantage. Play might continue 8...b6 (probably best; 8...c4?! 9 Bc2 followed by e4 is one of
White’s ideas) 9 bxc5 (or 9 Qa4!? cxb4 10 cxb4 Bb7 11 Bb2 a5 12 Rfc1 Ne4 with an unclear position,
L.Dubois-E.Daillet, Clichy 2000) 9...bxc5 10 Ba3 Qc7 (Diagram 53)

Diagram 53 (W)
An interesting white approach
290
 and now Kovacevic’s suggestion of 11 c4!? deserves attention, whereas 11 Qa4 Bb7 12 Rab1 Rfb8 13
Rfc1 c4! 14 Be2 Bc6 15 Qc2 Nb6 gave Black good play in V.Kovacevic-G.Sax, Zagreb 1985.
8...dxe4
Black might also consider 8...cxd4!?, angling for 9 Nxd4?! Nc5 or 9 cxd4 dxe4 10 Nxe4, with a
transposition to our next note. White must thus take up the gauntlet with 9 e5! when 9...Nh5 (Watson and
Schiller prefer the ambitious 9...dxc3!? 10 exf6 Bxf6 11 bxc3 Bxc3 with three good pawns and a strong
centre for the piece) 10 cxd4 Nf4 11 Bc2 f6! is untested, critical and unclear.
9 Nxe4 cxd4 10 Nxd4
The best recapture. Unlike in Chapter Three, White can’t really consider 10 cxd4?! since 10...Nxe4 11
Bxe4 Nf6 12 Bc2 Qb6 leads to a rather unimpressive IQP position; White hasn’t any attacking chances and
d4 is vulnerable.
10...Nxe4 11 Bxe4 Nf6 (Diagram 54)

Diagram 54 (W)
White hasn’t any advantage

The position is about equal; White’s slightly more active pieces compensate for Black’s extra central pawn. E.Colle-M.Monticelli, San
Remo 1930, for example, continued 12 Bf3 e5! 13 Nb5! Qb6 (13...Qa5!? was perhaps more accurate) 14 a4 Rd8?! and now White missed 15
a5! Rxd1 16 axb6 Rxf1+ 17 Kxf1 a6 18 Nd6 with advantage.

Theoretical Conclusion
3...g6 is a pretty reasonable black option. How to respond depends on taste: some readers may prefer 4 Bd3, others 4 b3. Neither promises any
advantage, but then 4 c4!? doesn’t either, although it is probably the most challenging option.

Illustrative Games

Game 29
N.Divinsky-W.Holowach
Canadian Championship, Saskatoon 1945

1 d4 d5 2 Nf3 Nf6 3 e3 c5 4 c3 g6 5 Bd3 Bg7 6 Nbd2 Nbd7 7 0-0 0-0 8 Re1 Re8
Supporting ...e5. Black can do so too with 8...Qc7 when 9 e4 dxe4 10 Nxe4 Nxe4 11 Bxe4 cxd4 12 Nxd4
(Diagram 55)

291
 Diagram 55 (B)
Unbalanced and about equal

leads to another roughly balanced position. Both sides must though aim to stay active; for example,
12...Nc5 13 Bc2 Rd8 14 Be3 (14 Qf3!?) 14...Ne6! 15 Nxe6 Bxe6 16 Qf3 Bd5 17 Be4 e6 18 Bxd5 Rxd5 19
Rac1?! b5! gave Black a useful minority attack in M.Puckett-N.Siegel, Washington DC 1996.
8...b6 is another approach when White should possibly prefer 9 a4!? to 9 e4 cxd4 10 cxd4 (10 Nxd4? Nc5
has left White struggling in a few games and is one to avoid) 10...dxe4 11 Nxe4 Bb7 with a comfortable IQP
position for Black.
9 e4 dxe4 10 Nxe4 cxd4 11 Nxd4 e5 12 Nf3 (Diagram 56)

Diagram 56 (B)
How adventurous will Black be?

292
 12...Qc7
12...Nxe4!? 13 Bxe4 Qc7 and ...Nc5 was another reasonable option. The position is about level, but both
sides can certainly play for a win.
13 Bg5 h6 14 Bh4!?
Raising the stakes...
14...Nxe4 15 Bxe4 f5
...and Black is up for the challenge. 15...Nc5 remained a safe alternative, but Holowach was probably right
to be provoked.
16 Bd5+ Kf8!
The best square; Black had to avoid 16...Kh8 17 Bf7 Rf8 18 Bxg6, and 16...Kh7 17 Qa4!? g5? 18 Bxg5!
hxg5 19 Nxg5+ Kg6 20 Qh4 (Oleinikov) would also have been rather dangerous for him.
17 Bg3! (Diagram 57)

Diagram 57 (B)
White is ready to sacrifice...

17...f4?
Tempting, but too greedy. Now Black’s king position becomes a cause for concern and he should have
preferred 17...g5! when 18 h3 would have been rather grim for White. I assume, though, that Divinsky would
have preferred 18 Qe2 f4 19 Bxf4! gxf4 20 Nh4 when 20...Re7 21 Ng6+ Ke8 22 Nxe7 Kxe7 23 Rad1 Nb6 is
rather unclear; Black’s king remains exposed, but his minor pieces are much better placed to help with the
defence than in the game.
18 Nh4!
Threatening mate in one.
WARNING: Always be very careful when entering a position in which your king lacks a flight
square.
18...Qd6 19 Bb3
Good enough, but more precise was 19 Bxf4! when the attack is very strong and White’s kingside pieces
immune; for example, 19...exf4? 20 Rxe8+ Kxe8 21 Bf7+ Ke7 22 Nxg6+ (Oleinikov) and Black loses his
queen.
19...Qf6 (Diagram 58)

293
 Diagram 58 (W)
...and now has a direct win

20 Qg4?
Undoubtedly White’s decision to entomb his bishop required a lot of calculation and perhaps Divinsky was
already low on time. His choice of queen move lets Black right back into the game, whereas 20 Qd3! would
have begun a decisive attack; Oleinikov supplies the key line: 20...e4 (20...g5? now loses to 21 Ng6+ and
neither does 20...Re6 21 Bxe6 Qxe6 22 Nxg6+ Kg8 23 Nxf4 help Black) 21 Rxe4 Rxe4 22 Qxe4 Ne5 (to
cover g6) 23 Bxf4 Qxh4 24 Qd5 and Black is helpless since 24...Qxf4 25 Qd8 is mate.
NOTE: Colour complexes play a large role in many attacks. Here White is extremely strong on the
light squares and should continue attacking on them.
20...g5 21 Qh5 Re6! 22 Bxe6 gxh4??
A blunder. Instead 22...Qxe6 23 Ng6+ Kg8 24 Nxf4 gxf4 25 Bxf4 b6 26 Rad1 Bb7 would have led to a
very unclear situation; White would then have had to stay active and so doubling on the d-file is probably his
best.
23 Bxd7 Bxd7 24 Bxh4 Qe6 25 Qe2 Bc6 26 f3 Qg6 27 Rad1 Kg8 28 Rd8+! Rxd8 29 Bxd8 Qe8 30 Bc7 Bb5 31 c4 Qc6 32 Bxe5 Bxc4 33
Qe4 Bd5 34 Qxf4 Bf8 35 Qf6 Qxf6 36 Bxf6 Bxa2 37 Bd4 a6 38 Kf2 Bd5 39 Rd1 Bc6 40 Be3 Kf7 41 f4 1-0

Points to Remember
1. An ideal advance for Black in these lines is ...e5, but it can lead to weaknesses should ...f5 be mistimed as a follow-up.
2. These positions aren’t particularly interesting, but one can certainly hope to outplay a weaker opponent
in them. Furthermore, Divinsky was still able to make the game rather exciting with some fine creativity.
3. Always be especially careful when your king is chased out to f1 (or f8) and can’t return to g1 (or g8) due to a raking enemy bishop.

Black Plays 3...c5

1 d4 d5 2 Nf3 Nf6 3 e3 c5 (Diagram 59)

294
 Diagram 59 (W)
A reversed Queen’s Gambit

Another active and tricky response. In order to fight for the advantage White is pretty much forced to
contest a reversed variation of the Queen’s Gambit, the only option being which one.
4 dxc5!?
Aiming to exploit the extra tempo in a reversed Queen’s Gambit Accepted is probably the most
challenging course, and has endorsed by both Bronznik and Summerscale. Instead 4 Bd3 Nc6 (4...c4!? also
deserves consideration) 5 c3 Bg4 6 0-0 e6 gives Black easy and good development; for example 7 Nbd2 Bd6
8 h3 Bh5 9 Qe2 0-0 (Diagram 60)

Diagram 60 (W)
An improved main line for Black

295
 10 dxc5 Bxc5 11 e4 Ne5! gave Black a superior version of the main line in T.Gauss-L.Van Wely,
Kuppenheim (blitz) 2004.
NOTE: Playing a standard Colle set-up fails to convince when Black develops his light-squared
bishop outside the pawn chain.
Likewise 4 b3 cxd4! (sensibly ruling out any notion of dxc5) 5 exd4 Nc6 6 Bb2 Bg4 is fine for Black.
A.Savage-L.Kaufman, Virginia Beach 2004, for example, continued 7 Be2 Bxf3!? 8 Bxf3 and now Kaufman
opted for 8...Qb6, but in his subsequent repertoire work The Chess Advantage in Black and White, he
preferred 8...g6 9 0-0 Bg7 10 c4 0-0 with a solid game for Black and good pressure against d4.
In practice White’s most popular move has actually been 4 c3 (White can also play a reversed QGD: 4
Nbd2 cxd4 5 exd4 is a reversed Exchange Variation) when 4...e6 is by some margin Black’s most common
response. That transposes back to our main lines, but Black can be more clever with 4...Qc7!?. This equates
to 1 d4 d5 2 c4 e6 3 Nf3 c6 4 Qc2 Nf6 with an extra tempo, but unfortunately that variation is generally
considered to give White an edge. The extra tempo here enables White to equalize in a reversed version of the
main lines (generally involving a quick dxc5), but no more than that against accurate black play.
After 4...Qc7 5 Nbd2 Nbd7 (or 4...Nbd7 5 Nbd2 Qc7), Colle himself used to like 6 Qa4!? (Diagram 61).

Diagram 61 (B)
A favourite of Colle’s

This too doesn’t supply any objective advantage, although it does pose some practical difficulties. Black
might well reply 6...e6, but in practice he has usually opted for 6...g6 and then 7 c4 was E.Colle-
A.Rubinstein, Rotterdam 1931. Now 7...cxd4! 8 Nxd4 Bg7 9 cxd5 Nxd5 would have been fine for Black
since 10 Bc4 N5b6 11 Bxf7+? Kxf7 12 Qb3+ Ke8 13 Ne6 Qe5 14 Nf3 (Smith and Hall) fails to the neat
14...Qxe6!! 15 Qxe6 Nc5 (Oleinikov), trapping White’s queen and leaving Black a piece ahead.
It’s also worth noting that Black can meet 4 c3 Nbd7 5 Nbd2 with 5...g6 (when 6 Bd3 Bg7 7 0-0 0-0
transposes to the 3...g6 variation). These lines might seem quite equal and as such a little depressing to the
reader, but in practice things are not so bad unless one is playing a grandmaster! Not that many club players
will be familiar with this anti-Abrahams line with White (i.e. 1 d4 d5 2 c4 e6 3 Nf3 c6 4 Qc2), and so
shouldn’t be too comfortable when playing it as Black. White players may wish to examine it and look at
ways to use the extra tempo, especially with 4 c3 Qc7 5 dxc5 followed by a quick b4.
Capturing on c5 isn’t the only way to pose problems: 4 c4 is a tricky little transpositional move. After
4...e6 5 Nc3 play has transposed to the Semi-Tarrasch, 4...dxc4 5 Bxc4 e6 is the Queen’s Gambit Accepted
296
and 4...cxd5 5 exd4 Nc6 6 Nc3 leads to the Panov-Botvinnik Attack!
TIP: The Colle is full of transpositions. Be aware of their possibilities so that you can both make
good use of them and not fall for any of your opponents’ move order tricks.
Returning to 4 dxc5 (Diagram 62):

Diagram 62 (B)
The reversed QGA

4...e6
Preparing to recapture the pawn in natural Queen’s Gambit style, but there are two important alternatives:
a) 4...Nc6?! is quite a typical mistake because after 5 c3! the extra tempo becomes pretty useful: 5 ..a5 (or
5...e5 when it might be most accurate to begin with 6 Bb5 since 6 b4 e4 7 Nd4 Ng4!? 8 Bb5 Qh4 9 g3 Qh6 10
Nd2 Be7 11 Qe2 0-0 12 Bb2 Nge5 13 f4 a6! was far from clear in D.Root-D.Campora, Lone Pine 1981;
Black’s play was though rather ambitious here and White should be able to improve, perhaps with 8 Be2!?,
intending 8...Qh4 9 g3 Qh3 10 Nb5) 6 Bb5 e6 7 b4! (Diagram 63)

297
 Diagram 63 (B)
And now a reversed Abrahams!

and White appears to have engineered a favourable reversed position with an extra tempo.
NOTE: White is now playing the sharp and dangerous Abrahams (or Noteboom to continentals)
variation, namely 1 d4 d5 2 c4 e6 3 Nc3 c6 4 Nf3 dxc4!? 5 e3 b5 6 a4 Bb4.
In our reversed position, play might continue 7 ..Bd7 8 a4 axb4 9 Bxc6 Bxc6 10 cxb4 b6 11 Bb2 bxc5 12
b5 Bb7 13 Nbd2 Bd6 14 0-0 0-0 with a reversed main line; Black hopes that his centre and bishop pair will
compensate for White’s dangerous queenside majority. In the standard Abrahams this line is considered very
unclear (most grandmasters now prefer 1 d4 d5 2 c4 e6 3 Nc3 c6 4 e4 – the Marshall gambit), but the extra
tempo may well give White some objective advantage; for example, 15 Qc2 Re8 (G.Drogou-L.Geffroy, Issy
les Moulineaux 2003) 16 Rfe1 (Diagram 64)

Diagram 64 (B)
298
 White arranges ...e5; e4

16...e5 17 e4! (a typical Abrahams motif; indeed White must usually arrange to meet ...e5 with e4) 17...dxe4 (17...d4 18 Nc4 blockades
Black’s central pawns and leaves White’s own majority free to advance; note too that Ra3 may well become useful) 18 Ng5 and White is better
(Summerscale).
b) 4...Qa5+ 5 Nbd2 and now:
b1) 5...Qxc5 6 a3 g6 (the most common move; practice has also seen 6...Nc6 7 b4 Qd6 8 Bb2 Bg4 9 c4! e6
10 Qa4 and Black faced some awkward pressure before collapsing with 10...dxc4? 11 Nxc4 Qd5? 12 Nb6 and
1-0 in A.Pyhala-S.Pitkanen, Lahti 1997) 7 c4 Bg7 8 b4 Qc7 9 Bb2 0-0 10 Rc1 (Diagram 65)

Diagram 65 (B)
Easy development for White

gives White free development and a small pull, although Black does remain quite solid; for example,
10...dxc4 (or 10...a5 11 cxd5 Qd8 12 Qb3 axb4 13 axb4 Qxd5 14 Bc4 Qd8 15 0-0 Nbd7? 16 Ng5! (Diagram
66)

299
 Diagram 66 (B)
White seizes the initiative

and Black was in trouble, since 16...e6 17 Bxe6! costs a pawn, in V.Kovacevic-B.Abramovic, Zenica
1986) 11 Bxc4 Qb6 12 Qb3! (the best square for the white queen in this line; as we’ve seen, f7 may become
vulnerable) 12...Nc6 13 0-0 Ne8 14 Bxg7 Nxg7 15 Rfd1 and Black never quite managed to equalize in
L.Breivik-E.Rodriguez Guerrero, Linares 2005.
b2) 5...Nc6 6 a3 Qxc5 transposes to variation ‘b1’ above. Instead 6...Bg4 7 Be2 Qxc5 8 b4 Qb6 was seen
in C.Guimard-R.Wade, Barcelona 1946, when one way to gain a small edge is Bronznik’s 9 Bb2 e6 10 c4
Be7 11 b5!? followed by exchanging on d5.
b3) 5...Bg4 can just be met by 6 Be2, but in J.Rogers-J.Cox, British League 2003, White preferred the
much more creative 6 b4!? Qxb4 7 Rb1 Qxc5 8 Rxb7 Qc8 9 Rb1 and after 9...e6 10 h3 Bxf3 11 Nxf3 Be7 12
Bb5+! Nbd7 13 0-0 0-0 14 Bb2 enjoyed good activity, albeit in a still rather unclear position.
5 b4!?
Continuing in provocative reversed-Abrahams style. White can also force an IQP position with 5 c4 Bxc5
6 Nc3 0-0 7 cxd5 exd5, but theory considers this version to be fairly comfortable for Black (for one thing,
he’s not lost a tempo with his dark-squared bishop); for example, 8 Be2 Nc6 (Diagram 67)

300
 Diagram 67 (W)
A reasonable IQP for Black

9 0-0 Re8 10 b3 a6 11 Bb2 Ba7 (as pointed out by Watson and Schiller, 11...Bg4 is another reasonable
option) 12 Qd3 Be6 13 Rad1 Qd7 14 Rfe1 Rad8 and Black is well-centralized in this unclear position,
W.Paschall-G.Kaidanov, Stratton Mountain 1999.
5 ..a5
A typical Queen’s Gambit motif and an essential move if Black wishes to recover his pawn.
6 c3 axb4
Black’s main approach. Instead 6...Nc6 7 Bb5 transposes to note ‘a’ to Black’s 4th move, but also possible
is 6...b6 when 7 a4!? (an alternative to the usual 7 Bb5+ approach) 7 ..bxc5 8 b5 Bd6 9 c4 0-0 10 Bb2 Nbd7
11 Nbd2 Bb7 12 Bd3 Qe7 13 Qc2 was rather unclear, but perhaps a little better for White in A.Summerscale-
J.Plaskett, Edinburgh 1999.
7 cxb4 b6 (Diagram 68)

301
 Diagram 68 (W)
White enjoys a queenside majority

8 Bb5+
Exchanges generally help Black in the Abrahams proper, so here White is keen to exchange pieces,
although it’s also possible to again consider 8 a4!? bxc5 9 b5 Bd6 10 Bb2 (M.Sorsa-A.Koponen, Finnish
League 2000).
8 ..Bd7 9 Bxd7+ Nbxd7
A natural recapture, but Watson and Schiller prefer 9...Nfxd7!?. Now the critical 10 a4!? remains untested,
but was analysed by Prie on ChessPublishing. He gave 10...bxc5 11 b5 Qf6!? (an alternative is 11...Be7!? 12
0-0 Bf6 13 Ra2 0-0 – Watson and Schiller) 12 Ra2 Qg6 13 Bb2 Qxg2 14 Rg1 Qh3 15 a5 ‘with interesting
compensation for the pawn’. Black should probably respond with the critical 15...c4!? and this is yet another
Colle position in need of some testing. Indeed this whole variation with 5 b4 remains rather undiscovered.
10 a4! bxc5 11 b5 (Diagram 69)

302
 Diagram 69 (B)
A rather unbalanced position

Compared with the reversed Abrahams proper (see note ‘a’ to Black’s 4th move), Black has failed to gain
the bishop pair, but his knight does help to fight for control of the e5-square and may also restrain White’s
queenside passers from c5. Objectively this exciting position is rather unclear, but a white player who studies
it a little should score quite well, especially if Black is a non-1 d4 player. Some examples:
a) 11...c4!? 12 0-0 Nc5 13 Bb2 Bd6 14 Nbd2 0-0 15 Qc2 Re8 was J.Redpath-P.Motwani, Scottish
Championship, Stirling 2002, when 16 Ne5 would have left matters rather unclear and double-edged.
b) 11...Bd6 (a more popular choice) 12 Bb2 0-0 13 Nbd2 Qc7 (instead 13...Bc7 14 0-0 e5?! 15 e4! Nxe4
16 Nxe4 dxe4 17 Nd2 e3?! 18 fxe3 Qh4 19 e4 Nf6 20 Qe2 turned out rather well for White in A.Vitor-
I.Fancsy, Matinhos 1994; Black does better with an immediate 13...e5!? when 14 e4 c4 15 0-0 Qc7 16 exd5
e4 was far from clear in C.Broeker-G.De Lorenzo, correspondence 2002) 14 0-0 c4 15 Qc2 Rfc8 (Kaufman
and his silicon assistants suggest 15 ..Rfb8!? to try and restrain the white pawns) 16 Bc3 e5 17 h3 Nc5 18
Rfb1 Nfd7 19 a5 (Diagram 70)

303
 Diagram 70 (B)
The complex struggle continues

was once again extremely unclear in F.Christenson-Y.Yakovich, Bergen 2002.

Theoretical Conclusion
3...c5 is a challenging response to the Colle and one against which White has often replied too timidly. Solid
players may be happy with 4 c3 and 5 Nbd2, but the fascinating 4 dxc5!? looks like the best way to pose
Black problems.

304
Chapter Six
The Anti-Colle: Black Delays ...Nf6
2...Bf5: Baltic-style
1 d4 d5 2 Nf3
We now come to some variations which are technically outside the world of the Colle. However, as the Colle
player must be prepared to meet them, it would have been a little criminal not to include some coverage of
these alternatives to 2...Nf6.
2...Bf5 (Diagram 1)

Diagram 1 (W)
An early bishop deployment

The Baltic Defence. It should also be mentioned that Black has a few minor options at this juncture:
a) 2...Bg4 is a reversed Trompowsky, but that is about the best that can be said about it. One direct and
easy way to the advantage is 3 Ne5! Bh5 (or 3...Bf5 4 c4 e6 5 Qb3 and b7 is vulnerable; compare with our
2...Bf5 main line below) 4 Qd3!? (threatening both 5 Qb5+ and the nasty 5 Qh3; 4 c4 f6 5 g4!? is a good
alternative) 4...Qc8 (the only way to prevent both threats: 4...Nd7? 5 Qb5 costs Black a pawn and 4...c6?! 5
Qh3 Nf6 6 g4! Bg6 7 Nxg6 fxg6 8 g5 is promising for White with his excellent light-square control) 5 c4
(Diagram 2)

305
 Diagram 2 (B)
The c4-break causes problems

5...f6 6 Nf3 (more recently 6 g4!? fxe5 7 gxh5 e4? 8 Qb3 c6 9 Bh3 Qc7 10 cxd5 cxd5 11 Nc3 won a clear
pawn in E.Solozhenkin-F.Rodkin, St Petersburg 2004; note how the early piece exchange opens lines for
White’s light-squared pieces and leaves Black struggling to develop) 6...e6 7 Nc3 Bg6 8 Qd1 c6 9 e3 Bd6 10
Bd2 Ne7 11 Rc1 (preventing ...e5 due to the threat to exchange on d5; ...f6 has created a clear weakness and
White enjoys a pleasant edge) 11...Nd7 12 Nh4! f5?! 13 g4! and White opened the kingside with some effect
in W.Steinitz-M.Chigorin, World Championship (Game 2), Havana 1889.
b) 2...g6?! 3 c4 Nf6 (3...dxc4 can even be met by 4 e4; Black should not concede the centre like this) 4
cxd5 Nxd5 5 e4 Nf6 6 Nc3 is a superior version of the Grünfeld for White: Black lacks his usual counterplay
against White’s centre (with ...c5 after ...Nxc3; bxc3).
c) 2...c6 is a much better option for Black. White can now take the game immediately into a Slav with 3 c4
or prefer 3 e3. After the latter, 3...e6 reaches a position we’ll discuss in the final section of this chapter, while
3...Bf5 4 c4 takes play back into our last chapter (4 Nh4!? is an offbeat alternative for those who like such
things).
3 c4! (Diagram 3)

306
 Diagram 3 (B)
Again c4 is the best approach

Just as was often the case in the last chapter, immediately placing Black’s centre under some pressure is by
far White’s best approach.
3...e6
Keeping matters within the Baltic. Instead 3...dxc4 transposes to a rather rare form of the Queen’s Gambit
Accepted (Black’s bishop would prefer to be on b7 or g4 in that opening) and 4 Nc3 Nf6 5 e3 e6 6 Bxc4
Nbd7 7 0-0 Be7 8 Qe2 is a clear edge for White: forcing through e4 is the plan (thus 8...Ne4 is best met by 9
Nd2!).
Instead 3...c6?!, transferring play into a bad line of the Slav, is quite a typical mistake and one that it’s
useful to know how to punish: 4 cxd5 cxd5 (4...Bxb1 5 Rxb1 Qxd5 6 a3 concedes the bishop pair for no real
compensation, but might nevertheless be best) 5 Qb3! (Diagram 4)

307
 Diagram 4 (B)
Black has fallen for a trap

leaves both b7 and d5 rather vulnerable. Here the retrograde 5...Bc8 might actually be best, since 5...Qc7 6
Nc3 e6 7 Bf4! Qb6 8 Qxb6 axb6 9 e3 Nf6 10 Bb5+ Nbd7 11 Ke2 Be7 12 Rhc1 (Sadler) leaves White with a
large positional advantage; he may well be able to infiltrate via c7.
TIP: As we’ve seen already and will see again, whenever Black develops his light-squared bishop
very early in the game and White has played c4, always consider Qb3; b7 is often not at all easy to
defend.
4 Qb3!
Wasting no time in attacking b7 and now Black is forced to play actively.
4...Nc6! 5 c5!? (Diagram 5)

Diagram 5 (B)
Ruling out ...Nb4

Blocking the a3-f8 diagonal, thereby preparing to capture on b7 and cause Black problems on the light
squares.
WARNING: White should avoid the hasty 5 Qxb7?! when 5...Nb4 6 Ne5!? (hoping for 6...Nc2?! 7
Kd1 Nxa1?? 8 Qb5+ Ke7 9 Nc6+) offers Black a pleasant choice between 6...f6!? and forcing a draw
with 6...Rb8 7 Qxa7 Ra8 8 Qb7 Rb8, etc.
However, a good and calm alternative is 5 Bd2, hoping for a small positional edge and, for example,
5...Rb8 6 Nc3 Nf6 7 cxd5 exd5 8 Bg5 gave White exactly that in G.Gutman-R.Londyn, Pardubice 2004;
Black’s queen’s knight is misplaced and 8...Nb4 met by 9 Rc1.
5...Rb8
Black should never weaken himself with 5...b6? in this type of position. With his bishop outside the pawn
chain, he will suffer on the light squares. In this particular case punishment is swift: 6 e4! dxe4 7 Bb5
(Diagram 6),

308
 Diagram 6 (B)
Black loses material

winning material (Watson and Schiller).

6 Nc3
The best move order. Baltic players generally love to fight in complex and unusual positions, and so White
should avoid 6 Bf4?! f6! 7 e3 g5 8 Bg3 Nge7 when Black had easy play in E.Bareev-J.Speelman, Hastings
1992/93.
6...e5?!
Critical and a move which at first appears scary, but ultimately fails to convince. However, if Black has to
resort to 6...Nf6 then 7 Bf4 gives White a bind on the position and an edge (see Game 30).
7 e4! (Diagram 7)

Diagram 7 (B)
309
 Opening the centre favours White

A useful move to remember; White blows open the centre to exploit his slightly freer development.
7...exd4
Black preferred 7...Nxd4!? 8 Nxd4 exd4 9 Qa4+ c6 10 Qxd4 Bxe4 in M.Taha-Z.Rahman, Kuala Lumpur
1993. One idea here is 11 Bf4 Rc8 12 0-0-0! with a strong initiative for the pawn and 11 Qe5+ Be7 12 Nxe4
dxe4 13 Qxg7 Bf6 14 Qg3 was also quite promising for White in the game.
8 exf5 dxc3 9 Bb5!
Sacrificing a pawn to continue to develop as rapidly as possible.
9...Bxc5 10 0-0 Nf6
Against 10...Nge7!? Portisch intended 11 Bxc6+ bxc6 12 Qxc3 Qd6 13 Qxg7 when White’s safer king
position supplies some advantage.
11 Qxc3 Qd6 12 Bf4! Qxf4 13 Qxc5 Qd6 14 Rfe1+ (Diagram 8)

Diagram 8 (B)
Black hasn’t managed to castle

We’ve been following L.Portisch-P.Keres, Wijk aan Zee 1969; a model example of energetically exploiting a lead in development. White is
doing rather well here: 14...Ne4 (14...Kd8 15 Qe3 leaves Black facing continued problems co-ordinating his pieces) 15 Qxd6 cxd6 16 Rad1
regained the pawn in the game.

Theoretical Conclusion
The Baltic Defence is fairly rare these days; one reason being the awkward pressure created by 4 Qb3 and 5 c5. White should be fairly happy to
see any of the lines considered in this section.

Illustrative Games

Game 30
H.Olafsson-M.Knezevic
Neskaupsstadur 1984

1 Nf3 d5 2 d4 Bf5 3 c4 e6 4 Qb3 Nc6 5 c5 Rb8 6 Nc3 Nf6 7 Bf4 (Diagram 9)

310
 Diagram 9 (B)
Can Black create counterplay?

7...a6
Ruling out any notion of Nb5, not that it was an immediate threat.
8 e3 Be7 9 Be2
The pawn structure rather determines both sides’ strategy. White will look to advance his queenside
pawns, while Black must aim to develop some counterplay on the kingside. Knezevic’s next begins to do just
that.
9...Ne4! 10 Nxe4
WARNING: White must be a little careful about any advance of the black g-pawn. Here 10 0-0?
would be mistaken because of 10...g5! 11 Bg3 g4 and the fork on d2 nets the exchange.
10...Bxe4 11 0-0
In view of Black’s options on his next turn, 11 Qa4!? may well be a more accurate choice. White has ideas
of Ne5, and 11...0-0 12 0-0 Bf6 13 Rfd1! Ra8 14 b4 gave him the upper hand in M.Krallmann-F.Drill, Kassel
1994.
11...0-0
A little submissive. Instead Olafsson has shown that 11...g5 12 Bg3 h5 isn’t as scary as it may at first
appear; White gains the advantage with 13 Nd2! h4 14 Nxe4 dxe4 15 Be5 since 15...f6 16 Qxe6 Qd7 (and not
16...fxe5?? 17 Qg6+ Kf8 18 f3 with a decisive attack) 17 Qxd7+ Kxd7 18 Bg4+ Kd8 19 Bxc7+ Kxc7 20 d5
(Pyshkin) regains the piece with a slight advantage.
Black should, though, probably prefer 11...Bxf3!? 12 Bxf3 g5 13 Bg3 h5 when 14 h3 f5 15 Qc3 Kd7 16
Be2 h4 17 Bh2 (Diagram 10)

311
 Diagram 10 (B)
Consistent play from Black

17...Qg8 18 f3 Bf6 was rather unbalanced in J.Kristinsson-A.Pyshkin, correspondence 1989, although I

feel that White remains slightly for choice after 19 Kh1.
12 Qc3
As Olafsson points out, 12 Ne5 Nxe5 13 Bxe5 was a simpler way to both restrict any counterplay and gain
an edge.
12...Bf6 13 Rfd1 Qe7 14 Ne5! Bxe5?!
Black had to play 14...Nxe5 15 dxe5 Bg5, followed by exchanging bishops and ...f6; now White gains a
grip on the position.
15 dxe5 h6 16 b4 (Diagram 11)

Diagram 11 (B)
312
 White advances on the queenside

16...f6?
An understandable bid for counterplay with a4 and b5 imminent, but one which just makes matters worse.
17 exf6 Rxf6
Not an ideal recapture, but both 17...gxf6 18 Bxh6 and 17...Qxf6 18 Qxf6 Rxf6 19 Bxc7 would have cost a
pawn.
18 Bg3 Rbf8 19 a4 Qf7
This fails to trouble White, but 19...e5?! 20 b5 would have won the e5-pawn and even the relatively best
19...a5!? allows White to retain some advantage with 20 bxa5! (but not 20 b5? Nb4, bringing the knight to a
fine square and halting White on the queenside) 20...Ra8 21 a6 bxa6 22 a5! (Olafsson), fixing a6 as a serious
weakness.
20 f3 Bf5 21 b5 axb5 22 axb5 Nd8 23 c6! (Diagram 12)

Diagram 12 (B)
Black is in trouble

Now c7 becomes rather vulnerable and Black must also watch out for tactics against his back rank...
23...bxc6 24 bxc6 Rg6 25 e4! dxe4? 26 Ra8! 1-0
A piece goes west in view of 26...Qe7 27 Rd7 Qg5 28 f4.

Points to Remember
1. The c5-pawn gives White a useful spatial advantage and one which he should exploit by advancing his b-
pawn in the early middlegame.
2. Black needs to stir up counterplay on the kingside, but White shouldn’t be overly worried about a ...g5; Bg3 h5 advance; Nd2 is often a
useful resource.

2...Nc6: Chigorin-style
1 d4 d5 2 Nf3 Nc6 (Diagram 13)

313
 Diagram 13 (W)
White doesn’t have to play 3 c4

Strictly speaking the Chigorin only arises after 2 c4 Nc6, but Chigorin players often meet 2 Nf3 with their
favourite queenside development and, of course, White can transpose should he wish. Indeed, I feel that the
fairly simple line 3 c4 Bg4 4 Nc3 e6 5 Bg5 (or 5 Bf4) isn’t too hard to learn or play, and recommend it to any
reader wishing to remain consistent by pushing their c-pawn whenever Black avoids the main lines of the
Colle! White can also continue in offbeat and solid vein with 3 Bf4, but it’s also possible to stay with a Colle
approach, at least for a move:
3 e3
Relatively unusual, but by no means a bad choice. Indeed in this millennium it’s been employed by 2600+
grandmasters Graf, Kharlov, Lputian, Nikolic and Volkov. Furthermore, it’s not mentioned as far as I can see
in Gary Lane’s black opening repertoire, Ideas Behind Modern Chess Openings: Black, which many club
Chigorin players may well use. That said, the man himself does supply some coverage in his inspiring The
Chigorin Defence According to Morozevich.
3...Bg4
Standard Chigorin development and the best move. 3...Nf6 might not look like much of a mistake, but it is
a small one and hasn’t been played by anyone highly rated. Black’s problem is that after 4 c4! (Diagram 14),

314
 Diagram 14 (B)
Black lacks a good move

he can’t defend d5 with ...e6 since that would block in his light-squared bishop (and leave his knight a little
misplaced on c6). Neither is 4...dxc4 5 Bxc4 a particularly good version of the QGA for him (5...Bg4 6
Qb3!), which leaves 4...Bg4 5 cxd5 Bxf3 (5...Qxd5 is the other standard Chigorin response to an early capture
on d5, but here after 6 Nc3 Black can’t pin with ...Bb4 and 6...Bxf3 7 gxf3 transposes to 5...Bxf3) 6 gxf3!?
(doubling White’s pawns, but taking control of e4; 6 Qxf3 also looks pretty sensible when Black will
probably have to aim for a quick ...e5 to try and combat White’s bishop pair) 6...Qxd5 7 Nc3 and White is
slightly for preference after both 7...Qd7 8 Bb5 and 7...Qa5 8 Bd2; after the latter, note the Scandinavian-type
trap 8...e5 9 Bb5 exd4?? 10 Nd5!.
4 Bb5!? (Diagram 15)

Diagram 15 (B)
315
 A promising independent try

An interesting attempt to fight for control of the key e5-square.

NOTE: Should Black develop slowly his queen’s knight can become misplaced on c6, obstructing his
c-pawn. Thus 2...Nc6 is usually associated with quick activity and, where possible, an early ...e7-e5
break.
White’s most popular choice in practice has been 4 Be2, but should he be after quite a simple set-up, he
may do better to begin with 4 Nbd2 and now:
a) 4...e6 5 c4 Bb4 (5...Nge7!? 6 a3 g6 7 b4 Bg7 8 Bb2 0-0 9 Qb3 Bxf3 10 Nxf3 a5 was much more
unbalanced in A.Graf-M.Breutigam, German Championship, Saarbruecken 2002, and now 11 cxd5 should
have been met by 11...Nxd5! 12 b5 a4) 6 a3 Bxd2+ 7 Bxd2 Nf6 8 Be2 0-0 9 0-0 Ne4 10 Be1 a5 11 h3 Bh5 12
Rc1 gave White a small but clear edge (the two bishops and queenside potential) in P.Nikolic-M.Peek,
Amsterdam 2004.
b) 4...e5!? 5 h3 Bxf3 (5...Bh5!? 6 g4 Bg6 7 Nxe5 Nxe5 8 dxe5 Qe7 – Martin – is a reasonable try to
unbalance the position) 6 Nxf3 e4 7 Nd2 (Diagram 16)

Diagram 16 (B)
A reversed French

7...Nf6 8 c4 (S.Volkov-M.Peek, Port Erin 2006) gives White a reasonable reversed French with the bishop
pair and d5 collapsing.
4...e6
Continuing to develop. Instead 4...Qd7 5 h3 is a little tricky for Black, as is 4...a6 5 Bxc6+ bxc6 6 c4 when
Qa4 is on the agenda; for example, 6...Bxf3 (a concession, but 6...dxc4 too is hardly ideal for Black;
N.Sulava-S.Peric, Nice 2002, continued 7 Nc3 c5 8 h3 cxd4 9 hxg4!? dxc3 10 Qa4+ Qd7 11 Qxc4 cxb2 12
Bxb2 with some initiative for the pawn, and even the simpler 9 exd4 is promising) 7 Qxf3 e6 8 0-0 Nf6 9 b3
Be7 10 Ba3! Bxa3 11 Nxa3 Qe7 12 Nc2 0-0 13 Rfc1 slightly favoured White due to his better structure and
queenside prospects in S.Lputian-S.Brynell, European Championship, Silivri 2003.
5 Nbd2
Bronstein’s choice. White has a number of other moves to consider in this rather unexplored position: 5 c4
and 5 h3 being two, while I quite like the consistent 5 Bxc6+ bxc6 6 c4 (Diagram 17) and now:

316
 Diagram 17 (B)
Black’s structure is a mess

a) 6...dxc4?! is again rather unaesthetic, but Black shouldn’t mind about such things if he can gain active
counterplay. Here, though, I’m not sure he can: 7 Qa4 Bxf3 8 gxf3 Qd5 9 Nd2 Nf6 10 e4 Qb5 11 Qxc4 Qxc4
12 Nxc4 c5 13 Be3 Nd7 14 Ke2 favoured White who could easily begin to target Black’s weakened
queenside in C.Wisnewski-H.Hebbinghaus, Kiel 2006.
b) 6...Nf6?! 7 Qa4! Bxf3 8 Qxc6+ Nd7 9 gxf3 Rb8 10 cxd5 Rb6 11 Qc4 wasn’t a very convincing gambit
in N.Sulava-C.Marzolo, French League 2002.
c) 6...Bxf3! is the prudent choice, leaving Black a tempo up on Lputian-Brynell. White retained a small
pull with 7 Qxf3 (7 gxf3 c5! 8 cxd5 exd5 9 Qc2 Qd6 10 dxc5 Qxc5 11 Qa4+ c6 is much less clear, N.Sulava-
S.Brynell, European Team Championship, Plovdiv 2003) 7...Bb4+ 8 Bd2 Bxd2+ 9 Nxd2 Ne7 10 0-0
(Diagram 18)

317
 Diagram 18 (B)
Simple play from White

10...0-0 11 Rfc1 a5 12 Rc2 in V.Gaprindashvili-A.Hamad, Adana 2006, but Black may be able to improve
with something like 8...Qe7.
5...Bd6
A modern player probably doesn’t mind obstructing the f8-bishop for the time being with 5...Nge7!? and
this is perhaps a slightly more accurate continuation (for a start Bxc6+ loses its sting). J.Barle-B.Kovacevic,
Ljubljana 1998, continued 6 h3 Bxf3 7 Nxf3 a6 8 Bd3 Nf5! 9 c3! Nd6 10 0-0 Be7 11 Qc2 f5 12 c4 and a
tough struggle lay ahead.
6 c4 (Diagram 19)

Diagram 19 (B)
Now Qa4 is possible

6...dxc4 7 Bxc6+ bxc6 8 Qa4

The key motif.
8...Ne7 9 Qxc4
9 Nxc4!? Bxf3 10 gxf3 also deserves attention; White can then capture on d6 at some point if needs be and
would like to advance with e4 and Be3.
9...0-0 10 0-0 Rb8 11 a3 Bh5 12 b4 (Diagram 20)

318
 Diagram 20 (B)
Black is a little cramped

White’s superior structure gives him a small advantage in this complex position, D.Bronstein-L.Aronson, USSR Championship, Moscow
1957.

Theoretical Conclusion
We’ve only scratched the surface here, but 3 e3 looks like a reasonable way of getting Chigorin players away from their standard positions. All
of 4 Nbd2, 4 Bb5 e6 5 Nbd2 and 4 Bb5 e6 5 Bxc6+ deserve further testing. Indeed for now the onus is really on Black to show that he has
sufficient counterplay.

Black Plays 2...c5

1 d4 d5 2 Nf3 c5 (Diagram 21)

Diagram 21 (W)
An active black approach

319
 An active approach; with White having rejected 2 c4, Black decides to play the Queen’s Gambit himself.
White must again react vigorously.
3 dxc5!?
Just as with 2...Nf6 3 e3 c5 4 dxc5 in the last chapter, this is the most challenging response. Indeed play
will often transpose to that line and positions we considered in Chapter Five.
3...e6
Played to regain the pawn as quickly as possible, but Black has three important alternatives:
a) 3...Nc6?! 4 c3! is once again a promising way to hold on to the pawn; for example, 4...g6 (or 4...a5 5 e4!
dxe4 6 Qxd8+ Kxd8 7 Ng5 Ke8 8 Nxe4 f5 9 Ng5 h6 10 Nf3 Nf6 11 Na3! e5 12 Nb5 and White was better in
J.Augustin-P.Bakalar, Brno 1990) 5 Nbd2 Nf6 6 b4 Bg7 7 e3 0-0 8 Bb2 (having held on to c5, White
develops logically; only once Black advances in the centre does he need to consider more active options)
8...e5 9 b5!? e4 10 Nd4! Ne5 11 c4 (Diagram 22)

Diagram 22 (B)
Complex but good for White

11...Bg4 was M.Kopylov-M.Bach, Hamburg 1999, and now White could have gained a clear advantage
with the forcing 12 Qc2! Rc8 13 h3 Bd7 14 cxd5 Nxd5 15 Qxe4 (Tsesarsky).
b) 3...Qa5+ 4 Nbd2 Nf6 5 a3 Qxc5 enables White with 6 e3 to transpose to note ‘b1’ to Black’s 4th move
in the 3...c5 section of our last chapter or he might prefer an immediate 6 b4!?.
c) 3...Nf6 4 e3 is another transposition back to Chapter Five, although White can also consider 4 c4 e6 5
a3!? (playing in reversed QGA fashion; instead 5 cxd5 Bxc5! gives Black a reasonable IQP position since 6
dxe6?? loses to 6...Bxf2+! which is certainly a tactic to avoid) 5...Bxc5 (5...a5?! 6 Nc3 Bxc5 7 Bg5 0-0 8 e3
dxc4 9 Bxc4 gave White a clear edge in V.Burmakin-A.Pakhomov, St Petersburg 2000; b5 is a serious
weakness) 6 b4 Be7 (6...Bb6 7 e3 0-0 8 Bb2 a5!? might be a more critical try, P.Klain-V.Ivanov, Bankia
1990) 7 e3 0-0 8 Bb2 b6 9 Nbd2 Bb7 10 Be2 when his position is slightly the easier to play, Pham Duy Kien-
Nguyen Tuan Loc, Can Tho 2003.
4 e4!? (Diagram 23)

320
 Diagram 23 (B)
An IQP position is likely

An independent approach and one not considered by Watson and Schiller when advocating 2...c5 for
Black; a slightly surprising omission considering that 4 e4 has been played by Lasker, Capablanca and
Seirawan. Instead 4 c4 Bxc5 5 e3 Nf6 transposes to a position we’ve already considered via the move order
2...Nf6 3 e3 c5 4 dxc5 e6 5 c4 Bxc5.
4...Bxc5
Black should avoid 4...dxe4?! 5 Qxd8+ Kxd8 6 Ng5 when White retains some advantage, although he can
also play 4...Nf6. Now 5 exd5 Qxd5 6 Bd3 should be compared with a French Tarrasch (White isn’t
committed to Nbd2), or White can transpose to our main line with 5 Nc3 Bxc5 6 Bb5+ Nc6 7 exd5 exd5 8 0-
0 0-0.
5 exd5 exd5 6 Bb5+
White might also consider 6 Bd3!? Nf6 7 0-0 0-0 8 Bg5 (J.Granda Zuniga-E.Vladimirov, Cienfuegos
1986) with similar play to the IQP positions we considered in earlier chapters.
6...Nc6 7 0-0 Nf6
Black can also develop in French Tarrasch style with 7...Nge7. Indeed 8 Nbd2 leaves him a tempo up on a
line of that opening (Black has played ...Bf8xc5 in one move, rather than ...Bf8-d6xc5), although White still
ground out a win following 8...0-0 9 Nb3 Bd6 10 c3 Bg4 11 Be2 Re8 12 Nfd4 in P.Nikolic-P.Leko, Horgen
1994.
8 Nc3 0-0 9 Bg5 (Diagram 24)

321
 Diagram 24 (B)
A typically unbalanced position

White has some pressure against the IQP, but objectively Black shouldn’t be worse here. He should
probably continue with Kouatly’s suggestion of 9...Be6, although 9...Bg4 (9...Be7?! 10 Bxf6! Bxf6 11 Qxd5
Bxc3 12 Qxd8 Rxd8 13 bxc3 Na5 14 Rfe1 didn’t give Black enough for his pawn in E.Lasker-S.Tarrasch,
World Championship (Game 15), Munich 1908) 10 h3 Be6 11 Qd2 h6 12 Bh4 a6 13 Bxc6!? bxc6 14 Rfe1
Be7 15 Nd4 Bd7 16 Bxf6 Bxf6 17 Na4 Qc7 didn’t really give White much in Y.Seirawan-M.Dlugy, US
Championship, Cambridge Springs 1988.
Theoretical Conclusion
This little section has mainly been about transpositions to the 2...Nf6 3 e3 c5 4 dxc5 variation. After 2...c5 3 dxc5 e6, 4 e4 is an interesting
alternative to 4 c4 depending on which type of IQP positions one prefers.

Black Plays 2...e6

1 d4 d5 2 Nf3 e6 (Diagram 25)

322
 Diagram 25 (W)
Black might be after a main line...

This can be played with the aim of transposing to the main lines of the Colle, but Black can also follow up
with something more aggressive.
3 e3 c5
An important alternative is 3...c6 4 Bd3 f5 (Diagram 26),

Diagram 26 (W)
...but might prefer a Stonewall

adopting one of our favourite plans, the Stonewall. As we’ve seen in Chapters Three and Four, the side
facing a Stonewall set-up must not sit by and watch as a strong attack is mounted with ...Bd6, ...Nf6-e4, ...0-0
and ...Rf6-h6. Instead White must fight for the initiative with 5 c4 Nf6 6 0-0 (6 Nc3 Bd6 7 h3 0-0 8 g4!? is an
323
aggressive alternative which should appeal to those who like an uncompromising approach) 6...Bd6 and now
7 b3! (Diagram 27).

Diagram 27 (B)
Preparing to trade bishops

NOTE: White strives for an exchange of dark-squared bishops in this line of the Dutch, not just to
trade off his bad bishop, but especially to weaken both Black’s control of e5 and any later kingside
attack.
An ideal situation for White arises after 7...0-0 8 Ba3 Bxa3 9 Nxa3 when 9...Nbd7?! is a further mistake:
10 cxd5! cxd5 (Black would like to recapture with his e-pawn, but that would drop f5; a useful concept to
remember) 11 Rc1 a6 12 Qc2 Ne4 13 Qc7 and White is in full control. Black does better with 7...Qe7 when 8
Bb2 0-0 9 Qc1 is one way to force an exchange of bishops, but we will focus on another (8 a4!?) in Game 31.
4 c3
Should he be expecting a Stonewall then White might prefer 4 Bd3, intending to transpose to our last note
with 4...f5 5 c4, but Black can prefer 4...c4!? 5 Be2 b5 which is rather unclear.
4...Nc6
Now 4...c4?! would be a typical mistake. White replies 5 b3!, either gaining a useful central majority in the
case of 5...cxb3 6 axb3 or a strong queenside initiative in the event of 5...b5 6 a4! Qa5 7 Bb2, continuing to
undermine Black’s overextended queenside.
5 Bd3 Bd6
Having played ...c5, Black rarely follows up with 5...f5, but this may not be so bad. However, Black’s
centre is less stable than when he has a pawn on c6 and White should exploit that by still aiming for c4; for
example, 6 b3 Nf6 7 0-0 Bd6 8 Bb2 0-0 9 c4! cxd4 10 exd4 b6 11 Nc3 was slightly better for White in
R.Monje Peinado-J.Mindeguia Guruceaga, Vlissingen 2006; Black would prefer his f-pawn back on f7.
6 Nbd2 Nf6 (Diagram 28)

324
 Diagram 28 (W)
We’re back in the main line

Play has transposed back to Chapter One.

Theoretical Conclusion
Black players with some experience of the Dutch or just those looking to hack may like to employ a Stonewall set-up, but White gains a small
edge by responding with c4 and b3, seeking to trade the dark-squared bishops.

Illustrative Games

Game 31
V.Korchnoi-A.Yusupov
Montpellier Candidates 1985

1 Nf3 d5 2 c4 c6 3 e3 e6 4 d4 f5 5 Bd3 Nf6 6 0-0 Bd6 7 b3 Qe7 8 a4!? (Diagram 29)

325
 Diagram 29 (B)
Insisting on an exchange of bishops

8...0-0
Black can also halt White’s queenside expansion with 8...a5 and this may be a better plan. Now 9 Ba3
Bxa3 10 Nxa3 0-0 11 Qb1!? Ne4 12 Nc2 Na6 was M.Ulibin-P.Tregubov, Berlin 1996 – two great Stonewall
experts in opposition here – when Tregubov’s suggestion of 13 Qb2 may give White a slight advantage in a
rather complex position.
9 Ba3 Bxa3 10 Nxa3
NOTE: The knight may appear a little awkwardly placed on a3, but it remains on track for the ideal
e5-square: White would like, time permitting, to play Na3-c2-e1-d3.
It’s also possible, though, to keep the knight on b1 and 10 Rxa3!? g6?! 11 a5 Nbd7 12 cxd5 exd5 13 a6!
wrecked Black’s queenside with some effect in B.Gulko-L.Milov, Bern 1994, but Black’s 10th was rather
misguided; 10...a5 being called for.
10...Ne4 11 Nc2 (Diagram 30)

Diagram 30 (B)
Nc2-e1-d3 is the plan

11...Nd7 12 Qe1
White might also consider Yusupov and Dvoretsky’s suggestion of an immediate 12 Nd2!?, preparing to
chase Black out of e4 with f3.
12...a5! 13 Nd2 b6 14 f3 Nxd2 15 Qxd2 Nf6
Black had to avoid 15...Bb7?! 16 e4! dxe4 17 fxe4 fxe4 18 Bxe4 with some advantage since 18...Nf6 is
met by 19 Qg5!. This is exactly the type of scenario Black must avoid; a well-timed e4-break can easily
highlight his positional weaknesses, including his backward e-pawn.
16 Rfd1 Ba6 17 Ne1! (Diagram 31)

326
 Diagram 31 (B)
Continuing to improve the knight

Thematic: Bf1 and Ne1-d3-e5 is the plan.

17...Rfc8 18 Rac1 Qb7 19 Qf2?! c5! 20 Qg3
White would prefer to play 20 cxd5 Nxd5 21 e4, but here Black has 21...Nf4 with active counterplay. This
is why White’s queen needed to stay on d2 and so 19 Bf1 was more accurate.
20...cxd4 21 exd4 dxc4 22 bxc4 Qc7!
Forcing the queens off after which Black’s pressure against the hanging pawns should balance out the
weakness of e6.
23 Bf1 Qxg3 24 hxg3 (Diagram 32)

Diagram 32 (B)
Targeting c4 is not so easy
327
 24...Ra7?
Tempting and logical, but this allows Korchnoi to make use of a little tactic to regroup his pieces with
effect. As pointed out in Yusupov and Dvoretsky’s Informant notes, Black should have preferred 24...Ne8!.
25 Nc2! Rac7
The point is that 25...Bxc4 26 Na3 Ba6 27 Bxa6 Rxc1 28 Rxc1 Rxa6 29 Rc8+ Kf7 30 Nb5 (Yusupov and
Dvoretsky) is pretty grim for Black; with his rook imprisoned on a6, his extra pawn is pretty meaningless,
although it is not that easy for White to make further useful progress.
26 Na3 Nd5 27 Re1 Kf7 28 Re5 Nb4 29 Rce1
This is exactly the sort of scenario Black needs to avoid: e6 is rather weak and White can easily increase
the pressure.
29...Rc6 30 Nb5! Bxb5
Not ideal, but after 30...Rd8 31 c5! bxc5 32 dxc5 (Yusupov and Dvoretsky), White’s knight lands on d6.
31 axb5 Rd6 32 c5! Rxd4 33 cxb6 Rd6 34 b7 Rb8 35 Bc4 (Diagram 33)

Diagram 33 (B)
e6 collapses

35...Kf8
Trying to avoid losing his pawns with check and refusing to fall for 35...Nd5?? 36 Rxe6!.
36 Bxe6 Rxb7 37 Rxf5+ Ke8 38 Bd7+?
In the heat of the scramble, White sets a cheapo (38...Kxd7? 39 Rf7+ Kc8 40 Re8+ wins the exchange and
the game), but lands up losing valuable time with his bishop. Much better was 38 Bg8+.
38...Kd8! 39 Rf8+ Kc7 40 Bc8 Rb8
The time control has been reached and it becomes clear that Korchnoi has mangled his co-ordination,
while greatly improving Yusupov’s. The passed a-pawn is now a very useful asset for Black.
41 Rf7+ Kb6 42 Bf5 Rf6! 43 Rxf6+ gxf6 (Diagram 34)

328
 Diagram 34 (W)
Black has good counterplay

44 Bxh7 a4 45 Ra1 Ra8 46 Be4 Ra7 47 Ra3 Kxb5

If anyone has to be careful now it’s White, although Korchnoi manages to make a draw without any real
difficulty.
48 Kf2 Na6 49 g4 Nc5 50 Bd5 Rd7
After 50...Kb4 51 Re3 a3 52 Ba2 Rd7 53 Ke2 the a-pawn is firmly blockaded and there doesn’t appear to
be a way for Black to penetrate with either his knight or rook.
51 Bg8 Kb4 52 Re3 Rd8 ½-½
Black can win the bishop with 52...Rd8 53 Bf7 Rd7 54 Bg8 Rd2+ (54...Rd8 forces a repetition) 55 Kg3 a3 56
Kf4 a2 57 Bxa2 Rxa2, but after 58 Kf5 White has plenty of counterplay on the kingside. Even if Black
managed to somehow force rook and knight against rook, Korchnoi was never going to lose.

Points to Remember
1. After recapturing on a3 with his knight, White should gradually aim for the manoeuvre Na3-c2-e1-d3.
2. This type of Stonewall position can be deceptively tricky to play; always watch out for little tactics which
can turn the game as we saw here.
3. Try to avoid time trouble!

329
Index of Complete Games
Alekhine.A-Euwe.M, World Championship (Game 17), Eindhoven 1935
Benderac.A-Le Thanh Tu, Turin Women’s Olympiad 2006
Berg.K-De Firmian.N, Danish League 1999
Bogdanovich.G-Ortmann.D, Ditzingen 2002
Colle.E-Aguilera.J, Barcelona 1929
Colle.E-Delvaux, Terneuzen 1929
Colle.E-Kashdan.I, Bled 1931
Colle.E-Landau.S, Rotterdam 1931
Colle.E-O’Hanlon.J, Nice 1930
Cristia.J.M-Lynch.J, Buenos Aires 1931
Danner.G-Novikov.I, Graz 1996
Divinsky.N-Holowach.W, Canadian Championship, Saskatoon 1945
Gausel.E-Lie.K, Norwegian Championship, Asker 2000
Georgiev.K-Sokolov.I, Sarajevo 2002
Haag.M-Doennebrink.E, German League 1997
Harrison.P-Pritchard.D, Cheltenham 2004
Kashdan.I-Steiner.H, Pasadena 1932
Koltanowski.G-Defosse.M, Belgium 1936
Korchnoi.V-Yusupov.A, Montpellier Candidates 1985
Landau.S-Book.E, Kemeri 1937
Mamedyarov.S-Volokitin.A, Calvia Olympiad 2004
Olafsson.H-Knezevic.M, Neskaupsstadur 1984
Pelikian.J-Coelho.L, Santos 2000
Piantoni.R-Popa.F, Verona 2002
Root.D-Rachels.S, Los Angeles 1991
Root.D-Silman.J, 1st matchgame, California 1990
Rygaard.M-Lugovoi.A, European Club Cup, Chalkidiki 2002
Smyslov.V-Romanishin.O, Tbilisi 1976
Straeter.T-Hladik.J, Pardubice (Game 3) 2006
Topalov.V-Kramnik.V, World Championship (Game 9), Elista 2006
Yusupov.A-Zarnicki.P, Minneapolis 2005

330
Table of Contents
Bibliography 5
Introduction 6
1 Black Plays ...c5 and ...Nc6 20
2 The Main Line with 9...Qc7 88
3 Black Plays ...c5 and ...Nbd7 166
4 Other Black Set-ups after 3...e6 229
5 Black’s Third Move Alternatives 251
6 Black Delays ...Nf6 305
Index of Complete Games 330
Alekhine.A-Euwe.M, World Championship (Game 17), Eindhoven 1935 282
Benderac.A-Le Thanh Tu, Turin Women’s Olympiad 2006 65
Berg.K-De Firmian.N, Danish League 1999 221
Bogdanovich.G-Ortmann.D, Ditzingen 2002 266
Colle.E-Aguilera.J, Barcelona 1929 246
Colle.E-Delvaux, Terneuzen 1929 33
Colle.E-Kashdan.I, Bled 1931 92
Colle.E-Landau.S, Rotterdam 1931 195
Colle.E-O’Hanlon.J, Nice 1930 181
Cristia.J.M-Lynch.J, Buenos Aires 1931 124
Danner.G-Novikov.I, Graz 1996 152
Divinsky.N-Holowach.W, Canadian Championship, Saskatoon 1945 291
Gausel.E-Lie.K, Norwegian Championship, Asker 2000 38
Georgiev.K-Sokolov.I, Sarajevo 2002 259
Haag.M-Doennebrink.E, German League 1997 84
Harrison.P-Pritchard.D, Cheltenham 2004 116
Kashdan.I-Steiner.H, Pasadena 1932 208
Koltanowski.G-Defosse.M, Belgium 1936 133
Korchnoi.V-Yusupov.A, Montpellier Candidates 1985 325
Landau.S-Book.E, Kemeri 1937 198
Mamedyarov.S-Volokitin.A, Calvia Olympiad 2004 160
Olafsson.H-Knezevic.M, Neskaupsstadur 1984 310
Pelikian.J-Coelho.L, Santos 2000 80
Piantoni.R-Popa.F, Verona 2002 185
Root.D-Rachels.S, Los Angeles 1991 136

331
Root.D-Silman.J, 1st matchgame, California 1990 223
Rygaard.M-Lugovoi.A, European Club Cup, Chalkidiki 2002 99
Smyslov.V-Romanishin.O, Tbilisi 1976 176
Straeter.T-Hladik.J, Pardubice (Game 3) 2006 49
Topalov.V-Kramnik.V, World Championship (Game 9), Elista 2006 277
Yusupov.A-Zarnicki.P, Minneapolis 2005 234

332