JriBh|Cyrus Lakdawala

/Nepomniachtchi
move by move
 Cyrus Lakdawala

Nepomniachtchi
move by move

EVERYMAN CHESS
www everymanchess com
First published in 2021 by Gloucester Publishers Limited, London.

Copyright © 2021 Cyrus Lakdawala

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

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

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

British Library Cataloguing-in-Publication Data

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

ISBN: 978 1 78194 631 2

Distributed in North America by National Book Network,

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

Distributed in Europe by Central Books Ltd.,

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

All other sales enquiries should be directed to Everyman Chess.

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

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

Everyman Chess Series

Commissioning editor and advisor: Byron Jacobs

Typeset and edited by First Rank Publishing, Brighton.

Cover design by Horatio Monteverde.
Printed by TJ International Limited, Padstow, Cornwall.
About the Author
Cyrus Lakdawala is an International Master, a former National Open and American Open Champion,
and a six-time State Champion. He has been teaching chess for over 40 years, and coaches some of the
top junior players in the U.S.

Also by the Author:

1... b6: Move by Move
1... d6: Move by Move
A Ferocious Opening Repertoire
Anti-Sicilians: Move by Move
Bird’s Opening: Move by Move
Botvinnik: Move by Move
Capablanca: Move by Move
Carlsen: Move by Move
Caruana: Move by Move
First Steps: the Modern
Fischer: Move by Move
Korchnoi: Move by Move
Kramnik: Move by Move
Larsen: Move by Move
Opening Repertoire:... c6
Opening Repertoire: Modern Defence
Opening Repertoire: The Sveshnikov
PetroffDefence: Move by Move
Play the London System
The Alekhine Defence: Move by Move
The Caro-Kann: Move by Move
The Classical French: Move by Move
The Colle: Move by Move
The Four Knights: Move by Move
The Modern Defence: Move by Move
The Nimzo-Larsen Attack: Move by Move
The Scandinavian: Move by Move
The Slav: Move by Move
The Trompowsky Attack: Move by Move
 Contents
About the Author
Bibliography
Introduction

1 Prodigy: 2003-2009
2 Gathering Power: 2010-2018
3 The Road to Magnus: 2019-2021
4 What Would Nepo Do?
5 Solutions

Index of Complete Games

 Bibliography
ChessBase Magazine
Chess Life #6 2021
Chess Today
Chess.com
Chess24.com
Chessgames.com
Chesspro.ru
Chesspublishing.com
New in Chess #7 2016
Suomen Shakki #10 2008

Annotations by Goran Arsovic, Alex Baburin, Vladimir Barsky, John Bartholomew, Nigel Davies,
Andrei Deviatkin, Glenn Flear, Mikhail Golubev, Gawain Jones, Victor Mikhalevski, Danilo
Milanovic, Maxim Notkin, Richard Palliser, Tom Rendle, and Branko Tadic.
 Introduction
Before writing this book, my knowledge of Ian Nepomniachtchi was scant, as he was somewhat under
my radar. Since then, I have learned the following:

1) His name is unbelievably difficult to memorize for spelling purposes, so much so that GM Anish
Giri wrote the following in New in Chess'. “People usually stop writing Ian Nepomniachtchi’s name
after four letters because of the chaos that ensues around letter number 8!” Maybe the proper title for
this book is Nepomniachtchi: Letter by Letter.
2) Every online commentator refers to him affectionately by his nickname “Nepo”.
3) He is a gifted video game player, at a semi-professional level. Nepo was introduced to the video
game Dota in 2006 and became a member of the team that won the ASUS Cup in the Winter 2011 Dota
tournament. He also served as a commentator at the ESL One Hamburg 2018 Dota 2 tournament, using
the handle FrostNova. He also plays Hearthstone and introduced fellow Russian grandmaster Peter
Svidler to the game, after which the two GMs provided feedback to Hearthstone developers.
4) Perhaps influenced by the Buddha, Nepo stylishly wears his hair in a man-bun.
5) Ian majored in journalism and graduated from Russian State Social University.
6) Nepo excels in attack and open and irrational positions, so it isn’t such a big surprise that he cites Tai
as his all-time favourite player and the one who most influenced his chess.
7) He is one of those stratospheric players who have actually accumulated a plus score over Magnus
Carlsen in classical time control games (four wins to one, with six draws) - although, for full
disclosure, a number of them were played when both were children. In the coming world championship
match, we are about to see a clash of two great players in their primes. We can be certain that Magnus
is salivating to settle a long, unavenged vendetta, since world champions insist on a plus score against
everyone! On the other hand, there is no greater animosity than that of number 2 for number 1, so Nepo
will be equally determined to unseat Magnus and take his crown.

Ian Alexandrovich Nepomniachtchi was bom on July 14th 1990 in Bryansk to a Jewish/Russian family,
still during the Soviet Union era, over a year before its collapse. He was bom into a literary family, and
grandfather Boris Iosifovich was a well-known lyricist. Nepo learned chess at the relatively early age of
four and it soon became apparent that the child displayed extraordinary talent. He was that kid for
whom every answer in class came easily. So much so that his natural abilities made “ordinary”
prodigies feel dumb by comparison.
His first teacher was his uncle, Igor Nepomniashchy. GM Valentin
Evdokimenko then took over, coaching him from age five to thirteen. Right
from the start, Ian looked like he was on a clean trajectory to potential world
champion. He won European Junior Championships three times: the under-10
category in 2000, and the under-12 in both 2001 and 2002. In 2002, Ian also
won the World Junior Championship U12 category, edging out another
prodigy, Magnus Carlsen, on tiebreaks. In their individual game Magnus
played the Black side of Alekhine’s Defence and lost a level game to Nepo.
At that time Ian’s Elo rating was almost 100 points higher than Magnus’, and
Nepo was considered the more gifted prodigy of the two. So it must rankle
with Nepo to play second fiddle to Magnus, since there is no horror greater
than being passed by a player who was once lower rated than you.
Ian became an International Master at age 13. In 2007, he finished second in the C group at Wijk
aan Zee, earning his first Grandmaster norm. Nepo picked up his second and third norms later the same
year: at the European Individual Championship in Dresden, followed by the World’s Youth Stars
tournament in Kirishi, which Nepomniachtchi won on tiebreak, having scored 2!4/3 against his three
rivals. So by age 16, Ian became Grandmaster Ian. (For context, Abhimanyu Mishra holds the record
for youngest ever GM title at the age of 12 years and 4 months, while Magnus became a GM at 13
years and 4 months!) He kept steadily improving, raising his rating and winning numerous tournaments
(I’m not going to fill the next page with a list of all his victories from age 16 to 28!). In December
2020, he won the Russian Championship with 734/11, edging out former Carlsen challenger, Karjakin,
by half a point. In 2021, Nepomniachtchi won the 2020/21 Candidates tournament with 834/14 (+5-
2=7), this time half a point above Máxime Vachier-Lagrave - which is the impetus for this book, since
that tournament victory gave Nepo the right to challenge Magnus Carlsen for the World Championship.
His FIDE ratings at the time of writing are: 2784 at the classical time control, 2778 rapid, and 2785
blitz.

The Match Against Magnus in Dubai

This book is about Ian Nepomniachtchi, yet the giant shadow of Magnus Carlsen looms over it. For us,
chess is a game. When it is played for the world championship, it is no longer just that. Magnus’
persona is that of a rock star who fdls every seat in the stadium. In other words, he is already a player
whose name stands with the all-time greats. Nepo has yet to prove himself at the very highest level of a
world championship match. Nepo’s obstacles are formidable, yet if he wins the match his name will
become enshrined in the elite group, alongside Lasker, Capablanca, Alekhine and so on; if he loses then
he gets downgraded to challengers who tried for the supreme title and failed, like Tarrasch, Bogoljubow
and Korchnoi. No pressure!
How does one qualify for the process of chess canonization? To be a chess
saint, one must become world champion. To get there, a player must pull off
multiple miracles:

1) They must first reach an elite level in the chess world.

2) They must win the Candidates tournament, which picks the challenger, and
in so doing must push aside a group of ridiculously talented grandmasters.
3) As challenger, they must dethrone the reigning world champion.

Of these, number three is the most difficult to achieve - and how much more
so when your opponent is considered to be a strong candidate for all-time
greatest chess player?
This is a book about a player whose most important games haven’t even
been played yet (unless of course you buy it after Nepo’s match with
Carlsen).
The first rule of a match is to be observant of the opponent’s habits,
stylistic likes and dislikes, strengths and weaknesses. Those observations can
then be weaponized. This is the part of the movie where the manager
massages his underdog fighter’s shoulders and gives the inspiring pep talk in
the ring, just moments before the fighter faces the heavyweight champion.
Is Nepo going to be the next world champion, or will he be another name
on the list of challengers dispatched by Magnus Carlsen? I don’t really know,
but it is at least possible to consider whether they are competing on an even
playing field. Let’s analyse:

1) Openings/preparation: I give the edge to Magnus for two reasons.

Firstly, Magnus’ repertoire is more varied and therefore more difficult to
prepare against, meaning that Nepo must spread his pre-match preparation
more thinly. Secondly, Nepo specializes in high-risk openings such as the
Najdorf Poisoned Pawn (here is the Poisoned Pawn for Dummies game plan:
get in, grab the b2-pawn, get out as quietly as possible, don’t lose your queen,
don’t get mated!) and Griinfeld, either of which may not be that well suited
for match play, where the goal is generally play for a win with White and
hold a draw with Black. When it comes to these two lines, Nepo is a true
believer and protector of the faith. Magnus’ repertoire, on the other hand,
contains both risky and safe options. But maybe Nepo doesn’t want to play it
safe, even with Black. Who knows, he may even score wins with the black
pieces.
2) Endgame technique: Carlsen, of course, since every other player’s
endgame technique is peasant-like in comparison. I rank Magnus in the top
three endgame players of all time. Nepo is a brilliant endgame player but still
a level below Magnus.
3) Attack: Nepo! In my opinion Nepo is the most skilled attacker on the
planet and his games are studies in pre-meditated acts of violence.
4) Irrational positions: Nepo thrives on anarchy and chaos, and if I had to
pick the top player in the world in irrational positions, it would be him. On
multiple occasions he has burned Magnus when the world champion
attempted to generate head-spinning complications.
5) Strategic understanding: Magnus is the strongest strategist in the world,
perhaps tied with Fabiano Caruana.
6) Open positions: Nepo excels in open games, and I predict that Magnus
will try his best to evade them in the match.
7) Initiative: Nepo. Nobody in the world is more lethal with the initiative.
8) Intuition: What we all want, don’t have, and can never get, is near-perfect
chess intuition. Only a handful of players in chess history have had it and
only one player alive today. Magnus is the new Capablanca and nobody in
the world possesses his level of intuition. Magnus moves a piece to a certain
square and then, 15 moves later, by magic, it just happens to be perfectly
placed.
9) Defence: You can’t kill your enemy twice, unless he is Magnus Carlsen, in
which case you must, since he has a strong claim of being the greatest
defensive player in chess history. Even though Nepo is an excellent defender,
Magnus dominates this category. Anyone who is capable of surviving 125
classical time control games without a single loss, against a rating average
over 2700 Elo, is one of the greatest - if not the greatest - defenders of all
time. So in this match we come upon the paradox/rivalry between the
irresistible force and the immovable object.
10) Assessment abilities: Kasparov once noted that the main reason Magnus
is number one is his ability to assess even the most baffling positions.
11) Nerves/Coolness under pressure: The tension of a world championship
match is almost beyond comprehension. Make one powerful move in a key
game and your name lives in chess history; botch the move and your name
becomes a footnote. This category is a tie. Neither player is the nervous type
and both are phenomenal under pressure. Of course, the rules tend to alter in
world championship matches. The expectations for Carlsen are crushing;
anything below dominance is interpreted by the critics as failure. Magnus
was criticized (unfairly!) for his super-safe play in his matches against
Karjakin and Caruana, both of which he failed to win in the classical portion
of the match. What the critics failed to factor in is that Carlsen was unafraid
of a tie at that point, since he was rightfully confident he would beat both
players in the rapid and blitz games. And he did just that. For Nepo, it’s now
or never, since you generally don’t get a second shot at a world championship
title (unless your name is Korchnoi). So the pressure on him is also
unimaginable.
12) Clutch player: Magnus is the Michael Jordan of chess, in that he always
seems to dig deeply and pull out a win when he needs it. In his match against
Karjakin, Magnus was a point down with four games to play and managed to
level the score and then easily win the rapid/blitz games. The last time I saw
him play online, at the Crypto Cup, he faced elimination five times. Magnus
won all five of those games when he most needed to and ended up winning
the tournament. If that isn’t a description of a clutch player, then what is?
13) Creativity/risk-taking: Neither player is afraid of the dark and both tend
to allow free reign to their imaginations. Both are willing to give up a lot to
seize initiative, attack, or even complicate. These factors are not necessarily
an advantage; in a match setting they may even be a minus! Sometimes a
super-aggressive risk-taker may be their own worst enemy, even more so
than the actual opponent. I rate Magnus as the more creative player, while
Nepo and Magnus are tied in the appalling-risk taking category. Restraint is
neither player’s strong suit! And there are other types of courage besides
sacrifice. It also takes courage to grab material and allow your opponent the
initiative or an attack for it, and both indulge in this - Nepo, right from the
opening, with his Poisoned Pawn Najdorf. The new 14-game format means
there may be more risk-taking, since there is more time to recover from a
loss; whereas a loss with fewer games to play is a potential disaster. I have no
idea whether either player will go high risk this match, but if one or both do,
the result is impossible to predict.
14) Natural ability: Are great chess players born or do they arise from great
toil? The answer is both, but the former seems more important, since there
have been, by their own admission, some comparatively lazy world
champions (Capa and Spassky come to mind). Although both players are
gifted with staggering levels of natural talent, my guess is that Magnus
possesses more of it.
15) Self-confidence: A world champion’s super-power - especially a
dominant champion like Carlsen - is that he tends to instil fear in his rivals.
But what if one rival isn’t intimidated? Magnus’ “invincibility” factor took a
hit when he sunk into a slump at the end of 2020 and well into 2021. I call
this one a tie. It is the psychological nature of a world champion to have
supreme belief in his own ability. Yet Nepo must enter the match with
confidence as well, since he has a plus lifetime score over Magnus and knows
he can be beaten. On the other hand, Fischer had never beaten Spassky before
their world championship match and yet won with relative ease.
16) Motivation: This category is a tie. Nepo has trained for this moment his
entire life, while Magnus is playing a difficult opponent and wants to prove
that he can beat anyone.
17) Psychological advantage: Nepo has the edge here. It wasn’t until 2019
when Magnus finally won his first classical time control game against Nepo,
whose lifetime score remains at 4-1 in his favour. Magnus confessed: “And
yes, my score against him is bad. Even in games from 2011 onwards, he leads
2-1. So I do think about turning this around to a positive score during the
match.” Furthermore, Nepo has in the past been Magnus’ second, which
means he probably got a closer look into the inner workings of the world
champion’s mind than the other way around.
18) Match experience: This is a no-brainer. Magnus wins this category by a
mile, since he has already faced Anand (twice), Caruana and Karjakin. So he
understands the pressures involved in a world championship match, whereas
this will be new to Nepo and we have no idea how he will handle it.
19) Rapid time control skills: Advantage Magnus, whose stats are higher.
The ratio: Magnus leads 20-10, with 32 draws.

One infallible law of life is that if you obtain something precious, there is
always someone who seeks to take it from you. That someone this time is
Nepo, who wants Magnus’ title, which no previous challenger has been able
to take. Does Nepo have the power to bring about Magnus’
Gotterdammerung? Although there have been only a handful of upsets in
world championship history, no player - no matter how great - is exempt
from an upset loss in a title match against a top-ranked challenger. The
following come to mind:

Alekhine taking down an over-confident and under-prepared Capa.

Euwe defeating an alcohol-weakened Alekhine.
Kramnik defeating Kasparov, who refused to give up trying to break down
Kramnik’s Berlin Wall.

Will this match be another upset win for the challenger? I saw an article
about an Indian clairvoyant (who doesn’t even play chess!) who claims that
Nepo will win the match. Every time a world championship match
approaches, Everyman expects me to predict the outcome, as if I were both
clairvoyant and a prophet. Obviously I’m neither, but let’s give it a shot
anyway. If I have to choose, I’d make Magnus a 60-40 favourite, to win the
match by two points with, say, a score of three wins and one loss. I also
checked the Chess.com survey and a whopping 84% predicted Carlsen will
win the match and retain his title, while only a fringe 16% predicted a new
champion. Having said that, nobody actually knows the scope of Nepo’s true
power until the match is over. Can Nepo prove the 84% majority wrong?

The games in this book are arranged in chronological order, so

Nepomniachtchi essentially becomes stronger and stronger as you progress
through it - until you reach 2021, where he becomes Magnus Carlsen’s
official challenger for the World Chess Championship.
May your own understanding of chess rise from the study of Nepo’s games.

Many thanks to Ville Matias (Father Vasilios) for his help in researching
Nepo’s life; and thanks as usual to Nancy for proof reading.

Cyrus Lakdawala, July 2021

 Chapter One
Prodigy: 2003-2009
In this chapter we look at games from Nepomniachtchi’s childhood to the end of his teens. From early
on we see his spark of genius with the initiative, when attacking, and in irrational positions. We begin
with an early clash between two kids who go on to become great rivals. I would never guess this was
one of their games since, uncharacteristically, it is virtually devoid of tactics. I just wanted to add one
very early game, to get a glimpse of how they played as children.

Game 1
I.Nepomniachtchi-Ma.Carlsen
World Junior Championships, Halkidiki 2003
Sicilian Defence

1 e4 c5 2 c3
Today, Nepo meets all Sicilians with the Open versions.
2 ... d5
2 ... Nf6 is Black’s main alternative.
3 exd5 Qxd5 4 d4 g6!?
This is Black’s fifth choice here. 4 ... Nf6 is the main line. Perhaps young
Magnus intended to throw Nepo out of normal preparation.
5 Nf3
I’m just wondering: if Nepo had played 5 dxc5, would Magnus have
played it safe with 5 ... Qxc5 or would he gambit with the riskier 5 ...
Qxdl+!? - ?
5 ... Bg7 6 Na3
Now Black must watch out for both Bc4 and Nb5 tricks.
6 ... cxd4 7 Bc4 Qe4+ 8 Be3
We are still well within theory’s sphere.
8 ... Nh6
Magnus, probably even as a newborn infant, wouldn’t fall for 8 ... dxe3??
which hangs the
queen to 9 Bxf7+ Kxf7 (or 9 ... Kf8 10 Qd8+ etc) 10 Ng5+.
9 Nb5
The line 9 cxd4 0-0 10 0-0 Bg4 11 Bxh6 Bxh6 12 Rel is also played.
. 0-0
9..
Black finally threatens to take the e3-bishop. 9 ... dxe3?? is again a major
goof-up: 10 Nc7+ Kf8 11 Qd8 mate.
10 cxd4
 As always, Nepo is happy to take on potential structural weakness in
exchange for piece activity. Now Nc7 becomes an actual threat. 10 Nbxd4?!
is weak, allowing 10 ... Ng4 and Black already stands better.
10 ... Na6
Instead:
a) 10 ... Nc6 11 Nc3 Qf5 12 h3 Qh5 13 d5 Ne5 14 Nxe5 Qxe5 15 0-0 is a
version of a reversed QGD Tarrasch, D.Pavasovic-S.Bromberger, Austrian
League 2010.
b) 10 ... Nf5!? is playable: 11 Nc7 Nc6 12 Nxa8 Ncxd4! 13 Nxd4 Nxd4
(threatening a fork on c2) 14 0-0 and now Black can draw with 14 ... Bh3! 15
gxh3 Nf3+ 16 Khl Nd2+. All four games in the databases ended quickly in
perpetual, since 17 f3!? seems exceedingly risky. After 17 ... Qxe3 18 Nc7
Nxc4 19 Nd5 Qc5 Black has loads of compensation for the exchange.
11 Nc3
The immediate 11 Qb3! scores well for White.
 .11.. Qc6 12 Qb3
Targeting f7.
12 ... Nf5
The engines say dead even, yet Black has scored just 20% from this
position.
13 0-0 Qb6?’

Mystery writer Raymond Chandler wrote: “The swans of our childhood

were probably just pigeons.” We wince in the present when we go over the
games we played as children.

Question: Why did Magnus deliberately allow damage to his

queenside structure?

Answer:! agree with you. White gets the better ending since Black’s queenside pawns are weaker
than the d4-isolani. The move is essentially a shady pawn sacrifice, for which Black gets insufficient
compensation. 13 ... Nc7 looks better.
 14 Qxb6 axb6 15 Na4!
Immediately targeting b6.
15 ... Nd6 16 Bb3 Nb4!
Magnus correctly gives away the forward b-pawn for piece activity. 16 ...
b5 17 Nb6 Rb8 looks depressing for Black.
17 Nxb6 Ra6 18 Nd5
It’s understandable that kid-Nepomniachtchi wants to remove Black’s
annoying b4-knight. I would have bagged the bishop pair with 18 Nxc8 Rxc8
19 Rfdl, when Black’s activity fails to make up for White’s extra pawn.
18 ... Nxd5 19 Bxd5 Bf5
After 19 ... Nf5! 20 Rfdl Rd8 21 Rael Rad6 22 Bb3 Be6 23 Bxe6 Rxe6 24
Rc3 it wouldn’t be so easy for White to convert.
20 Rfcl?!

This is an odd choice. It looks more natural to place the al-rook on cl and
the fl-rook on dl.
 20 ... Ra5 21 Bb3 Be4
Magnus seizes control over d5.
22 Nd2 Bc6 23 Bc2’?
I don’t understand the motivation behind this move. Surely 23 Nc4 is more
natural.
23 ... Rd8 24 Bd3 Rd5’?
From my experience, 12- and 13-year-old kids aren’t inspirational
strategists, even ones who go on to become immensely strong players.
25 Nf3 Rh5
I’m not really certain that White should be afraid of ... Bxf3. Black may
benefit from a swap of light-squared bishops, so maybe he should offer it
with 25 ... Bb5.
26 Be2
26 a4! is more principled, pushing on his majority wing. Then 26 ...
Bxf3?! 27 gxf3 Rh4 28 Rc7 is good for White, since 28 ... Bxd4? is met by
29 Bg5!.
26 ... Rd5
Now ... Nf5 is threatened.
27 g4!?
 Question: Is this act of self-inflicted structural vandalism necessary or wise?

Answer: Well, probably neither, but we begin to see a lifelong Nepomniachtchi trait: he isn’t afraid
to weaken his pawns if it restricts the opponent’s piece activity. The alternative is to return with pawn
for the bishop pair with 27 Ne5.
27 ... Nb5 28 Bc4 R5d6 29 Ne5 e6?’
This is a clear strategic error. Magnus was better off backing down with 29
... Be8.
30 Bxb5?’
It was unwise to hand over the bishop pair. After 30 Nxc6! bxc6 31 a4!
Nxd4 32 Ra2 Black’s position doesn’t look so great since he has to cope with
White’s outside passed a-pawn.
30 ... Bxb5 31 a4
Here 31 g5 would discourage ... f7-f6. If 31 ... Bxe5 32 dxe5 Rc6 33 b3,1
like White’s chances despite the opposite-coloured bishops, since he remains
a pawn up and Black’s dark squares look weaker than White’s light squares.
.
31.. Be8 32 b4

Exercise (planning): What is Black’s best defensive plan?

Answer: Sacrifice the exchange on d4, after which Black’s picks up two pawns.
32 ... Rxd4! 33 Bxd4 Rxd4 34 Rc8 Kf8
White is a full exchange up, but not for long, since numerous pawns hang
simultaneously.
35 Rel Rxb4
Magnus proceeds with a sense of mission. Objectively best is 35 ... Bxe5!
36 Rxe5 Rxg4+ 37 Kfl Ke7! (evading the strategic trap 37 ... Rxb4?? 38
Rb5!, when Black is completely busted) 38 Rxe8+ (if 38 Ra5 Rxb4 39 Ra7
Bxa4 40 Rc7+ Kf6 41 Raxb7 Rxb7 42 Rxb7, there is a zero chance Black
will lose) 38 ... Kxe8 39 Rb5 Kd7! 40 Rxb7+ Kc6! 41 Rxf7 Rxb4 42 Rxh7
Rxa4 and the game will end in a draw.
 36 h3 Rxa4 37 Nc4
Threatening Nd6, which is easily dealt with.
37 ... Ke7 38 Rdl

38 ... Rxc4?’
Allow curiosity to simmer long enough and there is no way to decline
temptation. Let’s put it all on the line with an unprovable assumption, shall
we? Ah yes, the limitless optimism of youth. Magnus refuses to cool off,
goes too far, and misevaluates the resulting ending, which is in White’s
favour. From this point, Black gets one of two results: a draw or a loss. In
chess we often (especially in youth) feel the urgency to do something. Then
what happens if your best course is to do nothing?
Rather than this radical measure, Black should play 38 ... b5! 39 Nd6 (or
39 Rc7+ Kf8 40 Nd6 Be5 41 Rb7 Bxd6 42 Rxd6 Ra8 and a draw is a virtual
certainty) 39 ... Bd7 40 Rc7 Rd4! 41 Rxd4 Bxd4 42 Nxb5 Be5 43 Rb7 g5!,
when the passive d7-bishop has a double: its evil and powerful twin on e5. I
don’t see how White makes the slightest progress.
39 Rxc4 Bc6 40 f4 f5’?
Black wants to remove the mobility of White’s kingside pawns. He may
have been better off with 40 ... h6 and then just shuffling.
41 Kf2 fxg4’?
Here 41 ... h6 is met by 42 g5 hxg5 43 fxg5 e5 44 Rh4 Bf8 45 Rh7+ Ke6
46 h4 f4 47 Rcl e4 48 h5 gxh5 49 g6 and White has chances to convert; e.g.
49 ... Be7 50 g7 Bh4+ 51 Kgl Kf7 52 Rai Bf6 53 Ra5! wins a piece and the
game.
42 hxg4 h5?
The plan to eliminate all of White’s pawns fails. Even now I’m not sure
how White makes progress after 42 ... h6!.
43 gxh5 gxh5 44 Rc5! h4 45 Rh5 Bf6
His best chance was 45 ... Kf6 46 Rxh4 Kf5 47 Ke3 Bf8.
46 Rh7+
Black’s king is boxed in further, driven to the back rank.
46 ... Kf8
Or 46 ... Ke8 47 Rd6 Bd5 48 f5!, when 48 ... Be7 fails to 49 Rh8+ Kf7 50
fxe6+ Bxe6 51 Rh7+ Kf6 52 Rh6+, picking off the e6-bishop.
47 Rd6 Be4 48 Rc7 Bd5 49 Ke3
Stronger was 49 Rdd7! Ke8 50 Rh7 Kd8 51 f5! and Black won’t hold the
game.
49 ... Ke8 50 Rh7 Be7?’
Magnus had to try 50 ... h3 51 Rxh3 Ke7.
 Exercise (planning): The defence is overwhelmed by sheer weight in numbers.
What is White’s conversion plan?

Answer: Simplification. Returning one exchange results in an easy win.

51 Rxd5!
Now Black’s multiple fears fuse into a single one: he won’t be able to stop
White’s newly passed f-pawn from moving forward. Certainly not 51
Rxe7+?? Kxe7 52 Rb6 h3 53 Rb2 Kf6 54 Re2 (or 54 Rh2 Bg2) 54 ... b5 and
White can resign.
51 ... exd5 52 Kd4 Bd6 53 f5 1-0

We now jump forward to 2006, where young Nepo stunned the defending
Russian Champion with a brilliant upset win.

Game 2
I.
Nepomniachtchi-S.Rublevsky
Russian Championship Superfinal, Moscow 2006
Sicilian Defence

1 e4 c5 2 M3 e6 3 d4 cxd4 4 Nxd4 a6 5 Nc3 Qc7 6 Qf3

Question: Aren’t we supposed to play f2-f4 before the queen is brought to f3?

Answer: Yes, that would be normal. In Open Sicilians, Nepo has a quirky tendency to post his
queen on f3 in front of his f-pawn, maybe to keep future options of both f2-f4 and Í2-Í3.
.6.. Nf6
An early classic game in this line went 6 ... Nc6 7 Nb3 Nf6 8 Bf4 d6 9 Be2
Be7 10 Qg3 0-0 11 Rdl Rd8 12 0-0 b5 13 Bh6 Ne8 14 Bd3 b4, when Black
already stood slightly better and eventually won, V.Korchnoi-M.Taimanov,
USSR Championship, Riga 1958.
7 Bg5
 A novelty at the time of the game, this is now White’s main line. 7 Qg3 is
an alternative, but after 7 ... d6! (swapping queens gives White an edge in the
ending) 8 Bd3 b5 9 a3 Bb7 10 0-0 Nbd7 11 Bd2 g6! Black stood no worse,
K.Yayloyan-A.Mikaelyan, Poti 2015.
7 ... Qe5!?

Question: Isn’t this a violation of the principle: Don’t move the same piece
twice in the opening, unless necessary!

Answer: I agree that Rublevsky’s optimistic queen move is in the unnecessary column. It targets the
white bishop and knight, and the e4-pawn will also be under fire after ... Bb4; but on the downside,
Black loses time and may lose even more in the future with attacks on his queen. 7 ... Be7 is a safer
alternative.
8 Be3
Black stands well if White hands over the bishop pair and dark square
control with 8 Bxf6 gxf6.
8 ... Bb4
Not 8 ... Be7? 9 0-0-0 with Bf4 to follow, when White has a dangerous
lead in development.
9 0-0-0!

Question: Isn’t this an invitation for Black to wreck White’s queenside

structure and then win the e4-pawn?

Answer: Welcome to Nepo’s universe of no rules. Such a move, which immediately drops the
pretence of friendliness, is a stylistic identity statement, offering his opponent both a pawn and
structural advantage. In return White gets the initiative, the bishop pair, control over the dark squares
and attacking chances - not such a bad deal. Whereas after 9 Bd3?! d5 10 exd5 Nxd5 11 0-0 Nxe3! 12
fxe3 0-0 13 Qh3 g6, Black’s bishop pair and superior structure clearly outweigh White’s vague
attacking hopes.
9 ... Bxc3
There is no backing down now. 9 ... Nc6? 10 Nxc6 bxc6 11 Bd4 Qc7 12
e5 Nd5 13 Ne4 0-0 14 Bd3 looks quite wretched for Black.
10 bxc3
 .10.. 0-0!
Rublevsky wisely develops, rather than go pawn grabbing:
a) 10 ... Nxe4?? hangs a piece to the simple 11 Bf4 and if 11 ... Qd5 12
Nb5! axb5 13 Rxd5 exd5 14 Qg4!, Black can resign.
b) 10 ... Qxe4?! 11 Qg3 0-0 12 Bd3 Qg4 13 Qd6 Nc6, D.Kerigan-
G.Quparadze, Turkish League 2019, and 14 h4 gives White a nasty-looking
attack.
c) 10 ... d5!? is also risky since Black opens the game when behind in
development. After 11 exd5 Nxd5 12 Bd2 0-0 13 Rel White has plenty of
compensation for the missing pawn and bad structure, in the form of
development, the bishop pair and attacking potential.
11 Bd3 d6
11 ... d5 has been played more often, when 12 exd5 Nxd5 13 Bd2 again
gives White plenty of compensation, M.Alinasab-Mo.Asadi, Iranian
Women’s Championship, Tehran 2017.
 12 Nb3!
Not 12 Bf4? Qc5 13 Nb3 Qc7 14 g4 e5 15 Bg5 Ne8! 16 h4 Be6 17 h5 a5!
and I prefer Black’s chances.
12 ... d5
Black’s development lag becomes serious. The same applies after 12 ...
Bd7 13 Kb2 Bc6 14 Bd4 Qh5 15 Bxf6! Qxf3 16 gxf3 gxf6 17 Rhgl+ Kh8 18
Na5 as even here White’s development is more meaningful than Black’s
extra pawn.
Note that 12 ... Qxc3? isn’t such a great idea: 13 Bd4 Qc6 14 Bxf6 gxf6 15
Qxf6 Nd7 16 Qh6 e5 17 Be2! (threatening Rd3!) 17 ... d5 (forced) 18 Qxc6
bxc6 19 exd5 cxd5 20 Rxd5 and now White has an extra pawn, still with the
initiative.
13 exd5!
Nepomniachtchi said he considered 13 Bf4!? dxe4 14 Bxe5 exf3 15 gxf3
Nbd7 16 Bd6 Re8 17 c4. After 17 ... b6 White has compensation for the
missing pawn and rotten structure, but no more than that. The text move is
better.
13 ... exd5
On Chesspro.ru GM Andrei Deviatkin said he preferred White after 13 ...
Nxd5. One possibility is 14 Bc5 Nxc3 15 Rhel Nxa2+ 16 Kd2 Qc3+ 17 Ke2
and Black cannot save the exchange in view of 17 ... Re8? 18 Rai or 17 ...
Rd8?? 18 Bxh7+ etc.
14 h3
 Preventing ... Bg4, while preparing to attack Black’s king.
.
14.. Nc6!
This is probably the best defensive choice. After other continuations Black
has serious problems. For example:
a) 14 ... Be6? 15 Bf4 Qh5 16 g4 Qh4 17 Nd4 with a winning attack for
White.
b) 14 ... Qe7? 15 Bc5 Qd8 16 Bxf8 Qxf8 17 Rhel and Black is busted
c) 14 ... Nbd7 15 Rhel Qd6 16 Kbl Re8 17 Bf4 Qf8 18 Qg3! and Black is
in deep trouble, unable to play 18 ... Nh5? due to 19 Qh4 g6 20 Rxe8 Qxe8
21 Bd2 Nf8 22 Rel with a winning attack.
d) 14 ... Qxc3 (sure, let’s go pawn grabbing when desperately behind in
development) 15 Bc5! (with the dual threats of Bxh7+ and Bxf8; this is
stronger than 15 Bd4 Qc6) 15 ... Qe5 16 Bxf8 Kxf8 17 Rhel Qc7 18 g4 and
Black won’t survive.
15 Bf4 Qe7
 The only move. Not 15 ... Qh5? 16 g4, and if 16 ... Ne5 17 Bxh7+! Kxh7
18 Qe3 Qg6 19 Bxe5, Black is again in deep trouble. And two question
marks are not nearly enough for 15 ... Qxc3??, when the simple discovered
attack 16 Bxh7+ picks off Black’s queen.
16 Bg5
More accurate than 16 Rhel Qa3+ 17 Kbl Be6 18 Bg5 Ne4! 19 Bxe4
dxe4 20 Rxe4 Rfc8, when Black is starting to get some play, although White
stands better even here.
.
16.. Be6
Instead, 16 ... Qa3+ 17 Kbl Ne4 18 Bxe4 dxe4 19 Qxe4 Be6 20 Be3!
(threatening Bc5) 20 ... Rfe8! 21 Bc5 Qxa2+! 22 Kxa2 Bxb3+ 23 cxb3 Rxe4
24 Rd7 b5 25 Rhdl leaves Black fighting for the draw. Deviatkin pointed out
the line 16 ... Ne5 17 Qf4 Nxd3+ 18 cxd3 Qa3+ 19 Kbl, when Black remains
in trouble after 19 ... Nd7 20 Qc7!.

17Qg3
 Question: Why not go for a nice ending by chopping on f6?

Answer: The universal law is that 99.999% of all teenagers prefer to play for mate rather than a
superior technical ending, which Nepo could have reached with 17 Bxf6 Qxf6 (not 17 ... gxf6? 18 Qh5)
18 Qxf6 gxf6 19 Rhe 1 Rfe8 20 Nd4.
.17.. Rfe8
After 17 ... a5 18 Rhel a4 19 Nc5! Rfe8 (or 19 ... Qxc5 20 Bxf6 g6 21 Re3
Qa3+ 22 Kd2) 20 Re3 Nh5 21 Bxe7 Nxg3 22 Bh4 Ne4 23 Nxe6 Rxe6 24
Bb5 g5 25 Bg3 White stands better in the ending. Not now 17 ... Qa3+ 18
Kbl Ne4? 19 Bxe4 dxe4 because of 20 Bh6 g6 21 Qf4! f5 22 Bxf8 Rxf8 23
13 and White is winning.
18 Kbl
Not 18 Qh4? Qa3+ 19 Kbl Ne4! 20 Bxe4 dxe4 and suddenly Black stands
better.
18 ... a5! 19 Rhel a4
It isn’t easy to foresee at this juncture whether White’s initiative lives or
dies. In a few moves the answer becomes completely clear.
20 Nc5!
This knight goes on to do great things.
20 ... Ra5?
Rublevsky should have gone for 20 ... Qxc5! 21 Bxf6 g6 22 Qh4 (not 22
Re3? Qa3) 22 ... Qa3 (threatening ... d5-d4) 23 Qh6 Qf8 24 Qxf8+ Kxf8,
when Black has excellent chances to draw.
 Exercise (combination alert): Nepo found a brilliant shot here. What did he play?

Answer: Overloaded defender/pinned pieces.

21 Nd7!!
Threatening an explosion on the pinned f6-knight. Unbelievably, White’s
knight is perfectly safe on d7, a square on which three black pieces are
trained! In a position where most players would reel from the disorienting
geometry, Nepo continues to find the optimal moves. This shot is embedded
in amber in my mind and I still remember it from going over this game 15
years ago.
21 ... Nh5!
The only move. 21 ... Qxd7?? is instantly punished by 22 Bxf6 g6 23 Qf4
and Black must hand over the queen to evade mate on g7.
22 Bxe7!
Not 22 Bxh7+? Kh8! and mass liquidation follows: 23 Bxe7 Nxg3 24 Bb4
Nxb4 25 cxb4 Rb5 26 Bd3 Rxb4+ 27 Kcl Bxd7 28 fxg3 with a likely draw.
22 ... Nxg3 23 Bb4!
This clever zwischenzug is the deep point of Nepo’s combination.
23 ... Ra7
The rook must back off.
24 Bc5
24 Nc5! is stronger, as at move 26 below. 24 Rxe6?! is imprecise due to 24
... Rxe6 25 fxg3 b6! (now the knight is trapped) 26 Nf8 Nxb4 27 Nxe6 Nxd3
28 Nf4 Nxf4 29 gxf4 f6 30 Rxd5 Re7 and Black has chances to hold the
game.
24 ... Ra5
Technically 24 ... b6 is better, though after 25 Nxb6 Ne4 26 Bxe4 dxe4 27
Rd6 White’s advantage is overwhelming.
25 Bb4
Going back for another go. Even better here is 25 Rxe6! fxe6 26 fxg3 Rd8
27 Rfl! (White’s knight is immune due to the back rank threat on f8; weirdly,
neither ... h7-h6, nor ... g7-g6 help Black. 27 ... a3 28 Kcl b6 29 Bd6 Ra7 30
Rf8+! Rxf8 31 Nxf8 Kf7 32 Nxh7, when the knight has miraculously escaped
and Black is dead lost.
25 ... Ra7 26 Nc5!
 26 ... Nh5
If 26 ... Nxb4 27 cxb4 Nh5, then 28 Bb5! Re7 29 Rxd5! wins.
27 Bb5!
Black must watch out for Rxd5 and the undermining Nxb7!.
27 ... Nf4 28 g3!
More accurate than 28 Nxb7 Nxb4 29 Bxe8 Rxb7 30 cxb4 Rxb4+ 31 Kai
Nxg2, when Black can at least play on.
28 ... Nxh3
If 28 ... Ng6 then 29 Rxd5 wins.
29 f4!
Stronger than the immediate 29 Rxd5. Black is now faced with threats of
f4-f5 and Nxb7, as well as Rxd5.
29 ... Raa8 30 Nxb7 Rec8 31 Nd6 Nf2
31 ... Rc7 32 f5 Bd7 33 Rxd5 is hopeless for Black.
32 Nxc8 Rxc8
 After 32 ... Nxdl 33 Nd6 Nxb4 34 Rxdl! Na6 35 f5! Black loses apiece.
33 Rd2 Ne4

Exercise (combination alert): How did Nepo finish his opponent off?

Answer: Clearance/overloaded defender/weak back rank.

34 Rxe4! dxe4 35 Bxc6
The c6-bishop is immune since Black’s rook must cover d8, which in turn
means he loses a piece.
35 ... e3 36 Rd4! e2 37 c4 Rb8 38 Kcl! 1-0
White remains a piece up. This time the b4-bishop is safe from capture due
to Black’s weak back rank.

Game 3
I.
Nepomniachtchi-M.Bosboom
Wijk aan Zee 2007
Sicilian Defence

1 e4 c5 2 Nc3
Today Nepomniachtchi almost always goes for 2 NI3.
2 ... d6 3 f4 e6
3 ... g6 4 Nf3 Bg7 5 Bc4 gives White a slightly favourable version of a
Grand Prix. The point is that if Black plays ... e7-e6 and ... d6-d5, White will
be a tempo up over normal lines, since Black took two moves to engineer ...
d7-d5. Nonetheless, it can’t be that bad if opening specialists like Boris
Gelfand are willing to play Black’s side.
4 Nf3 a6 5 a4
Suppressing... b7-b5.
5 ... Nf6 6 g3

Wait. It isn’t a Grand Prix Attack after all. Instead, the game transposes to
a sort-of Closed Sicilian. Alternatively, 6 d4 cxd4 7 Nxd4 switches into a
Scheveningen, which would suit Nepo’s style better, since with him open
equals stronger.
6 ... Nc6 7 Bg2 Rb8
The immediate 7 ... Be7 is more common.
8 0-0 Be7 9 h3
Nepo keeps open the possibility of d2-d4.
9 ... b5
Black has to find his chances for counterplay on the other side of the
board.
10 axb5 axb5 11 d3 b4
11 ... d5!? looks premature: 12 Ne5! Nd4 13 exd5 Nxd5? (13 ... exd5 is
correct) 14 Nxd5 exd5 15 c3 (chasing away the defender of the c6-fork
square) 15 ... Nf5 16 Nc6 won material and eventually the game, J.Kettner-
P.Zuse, German League 2017.
12 Ne2 h5’?
 Question: Is it so scary for Black just to castle?

Answer: I agree and think 12 ... 0-0 would have been a better bet, despite White’s coming pawn
storm after 13 g4. Black’s last move is a highly committal decision since it commits the king to life in
the centre.
13 b3
Spassky defeated Geller over and over in their Candidates’ match with the
Closed Sicilian plan of b2-b3, Be3 and Rcl. This way only c2 has to be
defended on the queenside, White’s weak wing. Spassky then played Qel-h4,
f4-f5 and Bh6 with a strong attack. But this plan assumes Black has
fianchettoed on the kingside. In this version Nepo has the option of Bb2,
alongside Be3.
13 ... Bb7 14 Bb2
14 Be3 is the alternative.
.
14.. Qb6 15Rf2’?
This is a strange choice. Most players would have opted to move the king.
15 ... d5
He doesn’t want to wait around for Nepo to build his attack, so Bosboom
strikes in the centre. 15 ... 0-0!? is still playable, when White will play for g3-
g4.
16 e5
16 Ne5 can be met by 16 ... 0-0 17 Nxc6 Bxc6 18 e5 Nd7 19 f5 d4 20 f6
gxf6 21 Bxc6 Qxc6 22 exf6 Nxf6 23 Nf4 and the engine says chances are
balanced.
16 ... Nd7 17 f5!
 It’s funny how in a chess game the thin veneer of a single move is all that
separates civilization from barbarism. Black’s goal is to preserve, White’s is
to destroy. Sometimes the most efficient way to deal with our opponent’s
subtlety is to push back in heavy-handed fashion. This move is vintage Nepo,
who offers a pawn to pry open the centre and get at Black’s king. In this case
the offer is completely sound. A quieter move, such as 17 Kh2, can be met by
17 ... g6 and Black looks okay.
.
17.. exf5?
Hostilities commence via mutual agreement, yet Bosboom underestimates
his opponent’s coming initiative. Black holds the balance by declining: 17 ...
c4! 18 fxe6 (18 d4 exf5 is a better version than Black got in the game) 18 ...
fxe6 19 d4 0-0 20 Nf4 h4 21 Nxe6 hxg3 22 Rfl Rf5 23 Nfg5 Rxfl+ 24 Qxfl
Bxg5 25 Nxg5 Ncxe5 (exploiting the pin on the d4-pawn) 26 Qf5 Qg6! and
Black doesn’t stand worse.
18 Nf4!
 Stronger than the more huffy 18 e6!? fxe6 19 Nf4 Kf7 20 Qe2 Nd4 21
Nxd4 cxd4 22 Bxd4! Bc5 23 Bxc5 Nxc5 24 Nxh5 Rbe8 25 g4 g6, when
Black looks okay.
18 ... Nd8
18 ... d4!? blunts the b2-bishop, at a cost: 19 Nd5 Qd8 20 Qe2 Nf8 21 Nd2
g6 22 Nc4 Ne6 23 g4! and an attack is brewing for White.
19 Nd2!
Uncovering a double attack on d5 and h5.
19 ... h4
Vladimir Barsky assesses 19 ... d4 20 Nc4 Qc7 as okay for Black, whereas
I’m not a fan of Black’s position after 21 Bxb7 Qxb7 22 Qe2, threatening e5-
e6 or 22 ... Ne6 23 Nd6+! Bxd6 24 exd6 0-0 25 Nxe6 Rbe8 26 Nxc5! Nxc5
27 Qxh5 Re3 28 Bxd4 Rxg3+ 29 Kh2 with advantage to White.

Exercise (critical decision): White’s choice is between bypassing with 20 g4

or grabbing the d5-pawn with 20 Nxd5. One of these moves falls for
 Black’s trap, so choose carefully.

Answer: 20 g4!
The correct decision. If you picked 20 Nxd5?? then it is my painful duty to
tell you that you fell for the trap: 20 ... Bxd5 21 Bxd5 hxg3 22 Rxf5 c4+ 23
d4 c3 - oops! - and White loses a piece for not nearly enough compensation.
.
20.. c4!
Here 20 ... d4? 21 Nc4 Qh6 22 gxf5 looks horrible for Black. Bosboom
decides he has compromised enough. Principle: Counter in the centre when
your king is under fire. The problem is that Black’s king becomes more and
more exposed as the centre opens.
21 bxc4 Bc5 22 d4 Bxd4 23 Bxd4 Qxd4

I flunked remedial math in high school, so don’t ask me what’s going on.
The position whirls in possibility.

Exercise (critical decision): In this disorienting environment, do you want to

 play 24 cxd5 or go for the other pawn with 24 gxf5 - ? One line gives
White
a clear advantage; the other allows Black back into the game.

24 cxd5?
The fact that you proceed without a trace of self-doubt doesn’t necessarily
mean you are on the right track. Nepo goes astray with a move which fails to
satisfy the position’s specifics.
Answer: 24 gxf5! is the correct path; e.g. 24 gxf5! dxc4 (or 24 ... Nxe5 25 cxd5 0-0 26 Rbl and
White will soon generate a strong attack) 25 Ra5! c3 26 Bxb7 cxd2 (not 26 ... Rxb7?? 27 Rd5! Qa7 28
Ne4 and wins) 27 Bg2 and White is clearly better. Note that the e5-pawn is immune in view of 27 ...
Nxe5?? 28 Rxe5+! Qxe5 29 Re2 etc.
24 ... fxg4!
Threatening g4-g3, winning a rook. The engine assesses this mess at dead
even. Not 24 ... Nxe5? 25 gxf5 0-0 26 f6! with a strong attack.
25 Ne2
Alternatives are also okay for Black. For example:
a) 25 c3 bxc3 26 Ra4 Qc5 27 Ne4 c2 28 Qcl Qb6 29 Nd6+ Kf8 30 Qxc2
g3 31 Ng6+ Kg8 and White must take perpetual check with 32 Ne7+ Kf8 33
Ng6+ etc.
b) 25 e6 g3 26 exd7+ Kf8 27 Ng6+ Kg8 28 Ne7+ Kf8 29 Ng6+ is another
perpetual.
c) 25 hxg4 can be answered by 25 ... h3! 26 Bxh3 Bxd5 27 Nb3 Qxdl+ 28
Rxdl Bxb3 29 cxb3 Nxe5 and Black stands a shade better.
25 ... Qxe5 26 Nc4 Qe7?
This is not a position conducive to a trial-and-error method. The queen is
not well placed on the e-file. Black should be fine if he goes for 26 ... Qc7!,
when 27 Qd4 g3 28 Nxg3 hxg3 29 Re2+ Kf8 30 Rael (threatening mate on
e8) 30 ... Nf6 31 Re7 Ne6! (interference: Black survives by returning the
extra piece; not 31 ... Qc8?? 32 Nd6 Qc3 33 Re8+ and mates) 32 Rlxe6 Qd8
33 Re5, is assessed at even by the engine, although I still prefer White by a
hair’s breadth.
27 Nd4!
Threatening Re2. Black’s king is in huge trouble.
 27 ... Kf8
If 27 ... 0-0 then 28 Qxg4, threatening Nf5 with a decisive attack for
White.
28 Qxg4 Nf6 29 Qf4
Gaining a tempo on the b8-rook.
29 ... Ra8 30 Rxa8 Bxa8 31 Qb8!
The immediate 31 Nf5! is White’s optimal continuation, chasing away an
important black defender. After 31 ... Qa7 32 Qd6+ or 31 ... Qc5 32 Qb8, the
d8-knight is lost.
31 ... Bxd5 32 Nf5! Qel+
Here 32 ... Qe8 33 Qxb4+ is much the same, as is 32 ... Qd7 33 Ne5 Qe8
34 Qxb4+ Kg8 35 Ne7+ and wins.
33 Rfl Qe8 34 Qxb4+ Kg8
 Exercise (combination alert): One side’s argument cancels the other’s.
How does White force the win?

Answer: Step 1: Give check on e7.

35 Ne7+ Kh7
Or 35 ... Kf8 36 Nxd5+ Kg8 37 Rxf6!.
36 Ne3! 1-0
Formerly disarrayed pieces begin to cohere into a more recognizable
formation.
Step 2: Clearance. Black is threatened with mate on h4 and must give up
heavy material to evade it.

Compare the next game with Game 2. There we witnessed young Nepo’s
uncanny genius with initiative and tactics. This time we take a look at
potential weakness: closed strategic positions, where tactics are downplayed.
Game 4
Z.Andriasian-I.Nepomniachtchi
World Youth Stars, Kirishi 2007
French Defence

1 e4 e6
This may be a blasphemous opinion, but I’ve always wondered why pure
tacticians like Nepomniachtchi, Morozevich and Nakamura play the French
Defence, whose nature is generally a closed position. All three of the players
I mentioned are clearly superior in open games. Yet since childhood, Nepo
has played the French with the quasi-religious fervour of a true believer.
Having said that, I checked Nepo’s stats on the Black side of the French and
they are decent, so maybe in the end results take precedence over stylistic
issues. Our love for an opening is conditional, in that we must do well (or at
least okay) with it. We basically say: “I love you, but only if you love me
back.”
2 d4 d5 3 Nd2 Be7
This strange-looking waiting move has gained popularity over the past few
decades.

Question: What is Black’s idea?

Answer: One is that White’s f3-knight is diverted from the possibly superior e2-square, as in the
line 3 ... Nf6 4 e5 Nfd7 5 Bd3 c5 6 c3 Nc6 7 Ne2. Another is that if White plays e4-e5, Black’s g8-
knight can go to f5, via ... h7-h5 and ... Nh6.
4 Ngf3
Instead, 4 Bd3 stalls the development of the gl-knight, hoping to move it
to e2 later, as after 4 ... Nf6 5 e5 Nfd7 6 Ne2!. So Black responds with 4 ...
c5, when 5 dxc5 Nf6 6 Qe2 0-0 7 Ngf3 a5 is the most usual continuation.
4 ... Nf6
There’s no need to defer this move once White has committed the knight
to 13.
5 e5 Nfd7 6 Bd3 c5 7 c3 b6
 I also confess to a soft spot for this line. Black’s idea is to give White
space, but at a price: a future exchange of light-squared bishops, which
reduces White’s attacking chances and leaves him with a technically bad
bishop, with a pawn chain on the same colour. Black is ensured a
strategically pleasant position - unless he gets mated first! The main
alternative is 7 ... Nc6 8 0-0, followed by 8 ... a5, 8 ... Qb6, or 8 ... g5!?.
8 0-0
Note that 8 Qe2 doesn’t prevent the exchange of bishops since Black can
toss in 8 ... a5 first.
8 ... Ba6 9 Bxa6
By exchanging on a6 White denies the b8-knight easy access to c6.
9 ... Nxa6 10 Rel b5
Nepo is quick to generate counterplay on the queenside.
II Qe2
Nepo twice faced 11 a3 earlier the same year. 11 Nfl is also normal.
 11 ... c4
It’s clear this will be a kingside versus queenside battle.
12 a3
A later game saw 12 Nfl Nb6 13 Ng3 (intending Nh5) 13 ... h5! 14 Qe3
(preventing ... g7-g5 ideas), A.Berelowitsch-R.Srinath, Riga 2017. Here
Black might borrow Nepo’s idea below and play 14 ... Kd7!. I would be
happy taking on this position.
12 ... Nc7 13 Nfl Nb6 14 g3 h6 15 h4 Kd7!

Question: What!?

Answer: This idea was sometimes played by Botvinnik in the French Winawer. Why would Black
make a seemingly random king move?
1. Black’s territorial advantage is on the queenside, so his king is actually
safer there, rather than on the kingside.
2. Black will activate his queen with the odd manoeuvre ... Qg8! and ...
Qh7!, increasing light square power.
16 N3h2 Qg8! 17 Nd2 Qh7!
Thank you Mikhail B for this idea!
18 b3!?

Bravery and foolishness have a way of merging. If our bold idea is

successful, we are heroes; if it fails, we look like a fool. I wouldn’t have
played this move, which is in violation of the principle: Don’t create
confrontation on your opponent’s strong wing.

Question: Isn’t the move thematic, since it opens lines against Black’s king.

Answer: Yes, it does indeed open lines against Black’s king. The cost is heavy damage to White’s
structure:
1. A vulnerable a-pawn.
2. A vulnerable c-pawn, backward, on an open file.
3. Black’s knights get access to newly created holes on a4 and c4.
 18.. . cxb3
There was actually no rush to play this; 18 ... Rhb8!? was also possible,
since 19 bxc4 Nxc4 leaves Black in control on the queenside.
19 Nxb3 Rhc8
Targeting c3.
20Rdl?!
I don’t understand this move, which looks slow. Maybe it is time for the
radical 20 Nc5+! Bxc5 (ignoring the provocation with 20 ... Ke8!? is possibly
superior) 21 dxc5 Nc4 22 a4 bxa4 23 Rxa4 Nb5 24 Rxc4 (pretty much
forced) 24 ... dxc4 25 Qxc4 a6 26 Nf3 and the engine claims close to full
compensation for the exchange.
20.. . Ke8?’

Question: Why is Nepomniachtchi moving his king back to the kingside?

Answer: Good question, since it looks like the wrong plan. For the next few moves both sides
shuffle pieces randomly with feline stealth but without clear purpose. Black should keep his focus on
the queenside and his king on d7, such as with 20 ... Na4 21 Rd3 Rab8 with a clear advantage.
21 Rd3 Rab8 22 Nfl
Still fishing for an active plan.

Question: Shouldn’t White act with something more forceful like 22 f4 - ?

Answer: That would further weaken his light squares. Black can respond 22 ... h5 23 Rf3 Qe4! 24
Qdl Qf5 25 Nd2 Nc4 26 Nhfl a5 with advantage.
22 ... Nca8!
 Question: What is Black up to now?

Answer: Nimzowitsch’s contorted heart would glow with pride at Nepomniachtchi’s move. The
idea is to reroute the knights to occupy both a4 and c4.
23 Ne3 Na4
The pressure on c3 mounts, which in turn slows White’s hoped-for
kingside counterplay.
24 Ndl N8b6 25 f3
Perhaps he feared a future ... Qe4.
25 ... Rc7 26 Kg2 Rbc8 27 Bd2 a6 28 Ra2 Qg8?’
Hey, make up your mind! The queen was well placed on h7. 28 ... Nc4 29
Bel Kd7 makes more sense.
29 Nal’?
Nimzowitsch-approved. The knight can now reinforce the a3-pawn and
b4-square from c2.
 29 ... Nc4 30 Bel Qh7
Weren’t you just there a moment ago?
31 Nb3 Kf8’?
This looks like the wrong direction for Black’s king, who may be safest on
d7. But Nepo has conceived a plan to reverse his queen’s tracks to support
actions on the queenside, which requires her consort to clear out of the way
first.
32 Kh2
White is just waiting to see how Black makes progress.
32 ... Kg8 33 Nal Na5 34 Bd2 Kh8 35 Nc2 Nc4 36 Bel Qg8’?

If this game had a name it would be the Immortal Non-commitment Game.

Question: Do you get the sense that both sides believe they have all eternity
to decide?

Answer: Have you ever seen an early King’s Indian game from the 1950s, where Black takes
twenty moves to prepare the thematic ... f7-f5, break, when he could have played it in a single turn?
This game reminds me of that. I might backtrack and return the king to d7!
37 Nb4 a5 38 Nc2 Qe8 39 Kh3
All that is necessary for the triumph of evil is that good men and women
do nothing. It’s high time White did something. How about 39 f4 to try and
get an attack going on the kingside where the black king now resides?
.
39.. Ncb6
Uncovering against c3.
40 Bd2 Qc6 41 Nce3?’
The knight was fulfilling a useful defensive task on c2. Here too 41 f4
looks better, intending 41 ... Nc4 42 Bel once more.
41 ... Nc4 42 f4!?

Andriasian is willing to give up hope on the queenside, if in return he is

able to generate an attack against Black’s king. Still, this might have been
more effective if played earlier.
At this point 42 Bel no longer holds things together; e.g. 42 ... Nxe3 43
Bxe3 Qe8 44 Bd2 Nb6 45 f4 Nc4 46 Bel Qf8! 47 a4 b4 and Black’s
queenside attack is faster than anything White can generate on the other side.
42 ... Nxa3
Of the two prospective attacks - queenside vs. kingside - it is clearly not a
case of parallel evolution. Black’s is both faster and more potent. The all­
knowing engine says it was even better to capture the other way with 42 ...
Bxa3!. For example: 43 g4 Nxd2 44 Qxd2 b4 45 f5 Qb5! 46 f6 Bel! 47 Qe2
(or 47 Qxcl Qxd3 48 Rxa4 Rxc3! 49 Qal Qd2 50 Nxc3 Qxe3+ 51 Kh2 Rxc3
and so on) 47 ... b3 48 Rxa4 Qxa4 49 g5 g6 50 h5 Qa2 51 Qg4 Qbl 52 hxg6
Qxd3 53 gxf7 Bxe3 54 g6 Bg5+ 55 Kg2 Qxg6 and White can resign.
43 g4
I suppose the legal term for White’s iffy attack is “reasonable doubt”.
White only has vague chances to get at the black king. Meanwhile, Black’s
progress on the queenside is very real. If 43 f5 exf5 44 Nxf5 Nc4 45 Q13 Bf8
46 g4 Nxd2 47 Raxd2 b4 48 g5 Qe6, threatening ... g7-g6, I don’t see
White’s attack going anywhere.
43 ... b4
Inserting 43 ... Qa6! first looks even stronger; e.g. 44 g5 b4 45 gxh6 g6 46
h5 b3 47 Rxa3 Bxa3 48 hxg6 fxg6, when ... b3-b2 is coming and White’s
kingside attack has gone dead.
44 cxb4 axb4 45 f5?
45 Rb3 was necessary.
45 ... Qb5!
Tangling up White’s d3-rook and queen.
46 f6 b3!
 Notice how Nepo comes to life the moment the game turns tactical and
more open. Despite the high rate of absenteeism for the black king’s
defenders on the kingside, the king himself remains relatively safe.
47 Rxa3
Desperation. 47 Rai loses to the simple 47 ... b2.
47 ... Bxa3 48 fxg7+ Kxg7 49 Nf5+’?
When desperate, are we the bravest of the brave or the craziest of the
crazy? Everything else loses too; e.g. 49 Qf3 b2 50 Qf6+ Kg8 51 Nf5 Qxd3+
52 Nde3 exf5.
49 ... Kf8
The human solution. The greedy engines show that Black can grab the
knight with safety, despite acceptance appearing to court death: 49 ... exf5!
50 gxf5 Kh7 51 Qe3 Bf8 52 Qf3 b2 53 Nxb2 Qxb2 54 f6 Rc2 55 Qf5+ Kh8
and once again White’s attack has gone dead.
50 Qe3 exf5 51 Qxh6+ Ke8 52 Rf3 Qe2!
 Elementary chess tactics: the fork.
53 Rxf5 Qxdl 0-1
Since 54 Qh8+ Bf8 55 Bh6 is too slow: after 55 ... Rc3+ White is mated.

Game 5
Le Quang Liem-I.Nepomniachtchi
Aeroflot Open, Moscow 2008
Semi-Slav Defence

1 d4 d5 2 c4 c6 3 NB e6 4 Nc3 Nf6 5 Bg5 h6 6 Bh4

With this move White offers a gambit. 6 Bxf6 is a tamer, more strategic
variant.
6 ... dxc4 7 e4 g5
My former student IM Keaton Kiewra and I advocated for the Moscow
Gambit in our book Opening Repertoire:... c6. For both sides, it’s a line only
for those who don’t fear a sudden and violent death. 7 ... b5 8 e5 g5 9 Nxg5
hxg5 10 Bxg5 is a cousin variation.
8 Bg3 b5

Many of us, including me, switch openings the way we change the TV
channel when a movie we are watching bores us. Nepo normally plays the
Griinfeld against Queen’s Pawn openings. This game he experimented with
the heavily disputed Anti-Moscow Gambit, which leads to some of the most
irrational positions imaginable. We all do this: read a book and study a few
games from the database and then believe ourselves an expert in the line,
which is the same as getting a mail-order dot.com doctorate in one month. It
just isn’t the same as earning the doctorate from Cornell, Yale or Harvard.
Only years of (often painful!) experience in a line earn us the true doctorate.
9 Be2
 The most frequent continuation. 9 Ne5 and 9 h4 are also played.
9 ... Bb7 10 0-0
10 h4 is White’s main alternative.
10.. . Nbd7 11 Ne5 h5
This is Black’s number two choice. 11 ... Bg7 is the main line.
12 Nxd7 Qxd7 13 Be5
13 Qcl Rg8 14 Rdl Bb4 15 Qe3 is also seen occasionally.
13 ... Rh6
So far, so book.
14 Qcl Rg6
14 ... Ng4 15 Bxg4 hxg4 16 Qxg5 Rg6 is another theoretical line.
15 B
But this move, reinforcing the centre, is comparatively rare. 15 Rdl is
White’s main line.
15.. . Qe7
After 15 ... h4 16 f4 gxf4?! (16 ... 0-0-0 is correct) 17 Qxf4 Bh6? (Black
should try 17 ... Bg7 18 Qxh4 Qe7) 18 Qxh4 Bg5 19 Qf2 Black found
himself in serious trouble, E.L’Ami-I.Cheparinov, Wijk aan Zee 2008.
16 a4
Whether he knew it or not, this was a small novelty. In games before and
since White has played 16 f4 at once; e.g. 16 ... gxf4 17 Bxf4 Ng4 18 h3 e5!
19 hxg4 exf4 20 gxh5 Rxg2+ ZAA in view of perpetual check, M.Ragger-
A.Jedinger, correspondence 2009.
 Question: Does White get full compensation for the sacrificed pawn?

Answer: In this line it’s not so difficult to predict that somebody will kill somebody. But who is the
killer and who the victim is still in question. The idea is to loosen Black’s queenside pawns to make
castling queenside riskier. So what does White have for the pawn?
1. A powerful centre.
2. The safer king.
3. The more aggressively posted pieces.
So the conclusion is: yes, White does receive full compensation for the
pawn.
16 ... a6
There is no reason to commit himself with 16 ... b4!?.
17 Qc2?’
The scavenger hunt for a plan begins. Presumably Le Quang Liem wants
to play f3-f4, without fear of... b5-b4 and ... Nxe4. Yet it may be a slight
waste of time since ... b5-b4 isn’t a real threat. He shouldn’t worry about his
e-pawn and should just go for it with 17 f4! gxf4 (not 17 ... b4? 18 f5! exf5
19 exf5 Rh6 20 Ndl Black’s position is loose) 18 Qxf4 Bh6 19 Qf2 Bg7 20
Bf3 0-0-0, when chances look approximately balanced.
17 ... h4 18 f4?
This is mistimed, as the opening of Black’s g-file takes precedence over
the opening of White’s f-file. At this point 18 h3 looks better.
18 ... gxf4 19 Bxf4
19 Bf3 is met by 19 ... h3!, exposing White’s king.
.
19.. c5!
Nepomniachtchi’s position comes to life as he seizes the initiative.
20 d5 b4!

21 e5
After 21 d6 Qd7 22 Be5 Bg7 White is busted.
21 ... Nxd5 22 Nxd5
 Not 22 Ne4? Nxf4 23 Rxf4 Qc7 24 Nf6+ Rxf6! 25 Rxf6 Qxe5 26 Rafi
Bd6 and White is again busted.
22 ... Bxd5 23 Rf2 b3
Black could also play 23 ... c3! at once, intending 24 bxc3 b3 25 Qd2 c4
with a deeply entrenched, protected passed pawn on b3, as occurs in the
game.
24 Qcl Qb7
Continuing to target g2.
25 Qfl c3! 26 bxc3 c4
Threatening ... Bc5.
27 Be3

Question: Does White have anything for the missing pawn now?

Answer: Let’s examine:

1. White is a pawn down.
2. Black exerts massive pressure on g2, which can be increased still further
if his other rook makes it to g8.
3. White’s king is in far greater danger than Black’s.
4. Even if Black’s attack goes nowhere, he is still winning due to his deep,
protected passed b-pawn.
Conclusion: White’s game is a strategic wreck and he has zippo
compensation for the missing pawn.
27 ... Rc8 28 Bh5 Rg7 29 Qe2
29 Bd4 was probably necessary, though even here White’s game looks
awful after 29 ... Bc5.
 Exercise (combination alert): What is Black’s strongest continuation?

29 ... Kd7’?
What is it with Nepo and the move ... Kd7 - ? This move is dramatic but
not actually necessary.
Answer: 29 ... Bxg2!! (annihilation of defensive barrier) 30 Rxg2 b2! 31 Rbl Qe4! was very strong.
After 32 Rxb2 Rxg2+ 33 Qxg2 Qxe3+ 34 RÍ2 (or 34 Khl Qcl+ 35 Qgl Qxb2) 34 ... Bc5 35 Qg8+ Kd7
36 Qxf7+ Kc6 37 Qxe6+ Kb7 38 Qf7+ Kb6 39 Qf6+ Ka5! White must resign, in view of 40 Qf3 Qel+
41 Kg2 Bxf2 42 Qxf2 (or 42 Qd5+ Bc5) 42 ... Rg8+ 43 Kf3 Rf8+ etc.
30 Rdl Kc7 31 Bh6
 Exercise (combination alert): How should Black continue now?

Answer: Annihilation of defensive barrier again. The exchange sacrifice starting with 31 ... Rxg2+
grabs the initiative and a winning position.
31 ... Rxg2+! 32 Rxg2 Bxh6 33 Rf2 Qb6
Continuing on with the king was stronger: 33 ... Kb6! 34 Rxf7 Qc6 35 Bf3
Bxf3 36 Rxf3 Rg8+ 37 Kfl Ka5! 38 Kf2 Rg4!, threatening ... Qc5+, and
White is busted.
34 Kfl Kb8 35 Bf3! Be3 36 Rg2
The engine comes up with 36 a5! Qa7 37 Bxd5 exd5 38 Rf6, though after
38 ... Rg8 39 Qh5 Qb7 40 Rxf7 Qc8 (threatening to slip into h3) 41 Qf3 Bc5
Black still retains a big advantage.
.
36.. b2?
Too hasty. When the complexity level is high and the time on the clock
short, not even the best human players in the world are able to maintain their
standards. Black should play 36 ... Bc5! (now ... h4-h3, followed by ... b3-b2,
is in the air) 37 Bxd5 exd5, intending 38 Rxd5? h3 and wins, since Black will
play ... b3-b2 if the rook moves off the second rank.
37 Rbl Bel

Exercise (planning): Find a defensive idea which saves White.

Answer: Sham queen sacrifice on b2.

38 Rf2?
How much money are you willing to part with if it would save your life?
Think carefully about your answer since, to some people, death is a more
merciful state than poverty.
White could survive in a pawn-down rook ending, by giving up the
exchange for Black’s deadly b-pawn: 38 Qxb2! Bxb2 39 Rgxb2 Qxb2 40
Rxb2+ Kc7 41 Bxd5 exd5 42 Rf2 Re8 43 Rxf7+ Kb6 44 Rf5 Ka5 45 Kf2
Kxa4 46 Ke3 Kb3 47 Kd4 and White holds the draw, despite Black’s extra
pawn; e.g. 47 ... a5 48 Rf7 Ra8 49 e6 a4 50 Rb7+ Kc2 51 e7 a3 52 Ra7! Re8
53 Kxd5 Kxc3 54 Rxa3+ Kb2 55 Ra7 c3 56 Rb7+ Kc2 57 Ke6 h3 58 Kf7
Rc8 59 e8Q Rxe8 60 Kxe8 Kd2 61 Ke7 c2 62 Rd7+ Ke2 63 Rc7 Kd2 64 Kf6
clQ 65 Rxcl Kxcl 66 Kg5 Kd2 67 Kg4 Ke3 68 Kxh3 Kf4 and draws.
Note that 38 Rxb2? is a false simulacrum of the real thing, as 38 ... Bxb2
39 Qxb2 Kc7 is still in Black’s favour.
.
38.. Rg8?’
Here 38 ... Bxf3! 39 Qxf3 Rd8 leaves Black in command.
39 Qdl?!
Allowing Black a second try. 39 Bxd5! exd5 40 Q13 offers better chances
to save the game.
39 ... Bxf3 40 Qxf3 Rg5
As above, 40 ... Rd8! is stronger.
41 Qe4 Qb7! 42 Qd4 Rg8
Covering against a queen check on d8.
43 Qxh4 Qd5
Missing the more deadly 43 ... f5! 44 exf6 Qhl+ 45 Ke2 Rd8!, which the
engine spots in a millisecond.
44 Qd4 Qxd4!
Nonetheless, the ending should be a win for Black.
45 cxd4 c3 46 Rc2
 Exercise (combination alert): On the surface it seems as if White may yet survive.
How should Black continue?

Answer: Move the bishop to e3, with dual threats of... Rgl+ and ... Bxd4.
46 ... Be3! 47 Ke2
Forced. If 47 Rxc3? then 47 ... Rgl+ 48 Ke2 Rxbl 49 Rxe3 Rhl! wins.
47 ... Bxd4 48 Kd3 Bxe5
Black has way too many pawns for the exchange, so the position is a
technical win for Nepo.
49 Re2 Rg5 50 Kc2 f6 51 h4 Rh5 52 Re4 Kc7 53 Rdl Rf5?
A slip. Either 53 ... Kb6 or 53 ... a5 was superior.
 Exercise (critical decision): Should White give a rook check on c4 or toss in
54 a5 first? One of the lines holds the draw.

Answer: The correct order is to check on c4.

54 a5?
After 54 Rc4+! Kb6 55 a5+! Kxa5 (or 55 ... Kb5 56 Kb3) 56 Kb3 White
draws, since 56 ... Rf2 57 Ra4+ Kb5 58 Rb4+ Kc5 59 Rc4+ Kb6 60 Rb4+
Ka7 61 Rd7+ Ka8 62 Rd8+ is perpetual check.
54 ... Rf2+
Now Black is winning again.
55 Kb3
Threatening perpetual check with Rc4+ as above. Nepo’s next move
disables the idea.
55 ... Rd2! 56 Rc4+
Certainly not 56 Rxd2? blQ+.
 56 ... Kd6
Now Black’s king is safe from repeated checks.
57 Rhl f5 58 h5
We hear the bang of the starter pistol and the contest ends before it begins.
White has no chance in the promotion race since Black has just too many
extra pawns.
58 ... Rh2! 59 Rdl+
The rook is taboo due to 59 Rxh2? blQ+ again.
59 ... Ke7 60 Rc6 f4 61 h6 f3 62 h7 blQ+!

63 Rxbl Rb2+! 64 Rxb2 cxb2

Black’s pawns can’t all be halted.
65 Kc2 f2 66 Rc7+ Bxc7 67 h8Q blQ+ 68 Kxbl flQ+ 0-1

The next three games demonstrate Nepo’s slippery defensive abilities,

especially in complex positions and under attack. In all three he is on the
Black side of the Najdorf Poisoned Pawn.

Game 6
E.Najer-I.Nepomniachtchi
Mainz (rapid) 2008
Sicilian Defence

1 e4 c5 2 M3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Bg5 e6 7 f4 Qb6 8 Qd2

For players who allow the Poisoned Pawn and then chicken out with 8
Nb3, maybe the London System is a better fit for you!
. Qxb2
8..

When we were six years old and our mother told us “No! You can’t have
that”, it was in our nature to find a way to get it on the sly.

Question: Is the Poisoned Pawn Najdorf still viable today?

 Answer: Black scores 51.9% - well above average - from this position, so we can confidently
conclude that the variation is alive, well and resting comfortably, from a theoretical standpoint. It’s a
line which defies conventional classification. Should it be filed under “adventurous”, or is “suicidal”
more accurate? The answer depends upon the following factors:
1. Black’s defensive skill level.
2. Black’s ability to avoid disorientation in the coming complications.
3. Black’s ability to counter-attack at the appropriate moment.
Nepo has delighted in grabbing the b2-pawn since childhood, so this is
familiar territory for him. I tried (and failed miserably!) to play the line too as
a kid, in imitation of my hero Bobby Fischer. But it must be played with a
proofreader’s eye. Unfortunately, I’m a lousy proofreader.
9 Rbl Qa3
Checking the database, to my amusement I found three games where
Black played the controversial moves:
a) 9 ... Be7??, hanging the queen to 10 Rxb2.
b) 9 ... Qxc3??, hanging the queen to 10 Qxc3.
c) 9 ... Qb3??, which hangs the queen four different ways. I am tempted to
quote Nancy Reagan: “Just say no to drugs!”
10 e5
This pawn push has morphed into White’s main line. Others are 10 f5 (see
Game 36), 10 Bxf6, and 10 Be2. The last of those was used by Caruana to
defeat Vachier-Lagrave in the Candidates’, thus derailing the latter’s chances
to be Magnus’ 2021 challenger.
.
10.. h6
Nepo favours this over 10 ... dxe5, which can transpose after 11 fxe5 Nfd7
12 Ne4 h6 13 Bh4. Presumably he regards possible deviations to be less
troublesome in the 10 ... h6 line.
11 Bh4 dxe5 12 fxe5 Nfd7
Also played are 12 ... Nd5 and 12 ... g5, which Nepo tried against Anand
(see Game 8).
13 Ne4
Please, be my guest and take a second helping.
13 ... Qxa2 14 Rdl Qd5 15 Qe3 Qxe5
 Good lord, even a Ferengi Grand Nagus would be impressed by Black’s
raw greed for pawns.

Question: Which factor is more important: White’s monster lead in

development and accompanying attacking chances, or Black’s three (!)
extra pawns and relatively weakness-free position?

Answer: Thankfully this is a book on Nepomniachtchi, not the Najdorf Poisoned Pawn, so I can
dodge the question with an “I don’t know”. When we ask the engines, they assess at a totally unhelpful
“0.00” - dead even!
Instead, 15 ... Bc5 16 Nxe6 Bb4+ 17 c3 Qxe6 18 cxb4 0-0 19 Be2 Nc6 20
0-0 Qxe5, as in Em.Berg-E.Najer, FIDE World Cup, Khanty-Mansiysk 2007,
is met by 21 Bg3! Qe6 22 Rd6 Qe7 23 Bf4! with a dangerous attack, which
led to a win for White in a few correspondence games.
16 Be2 Bc5 17 Bg3
Please be patient. We are still well within book, having reached a trendy
tabiya position.
17 ... Bxd4
Every piece removed from the board benefits Black, yet this one comes
with a cost: White’s dark square power is enhanced. 17 ... Qd5 is the main
alternative; e.g. 18 c4 Bxd4 19 Rxd4 Qa5+ 20 Rd2 0-0 21 Bd6 f5 22 Bxf8
Nxf8 23 Nd6 Nbd7 24 g4 and the engine very slightly prefers White, though
most games from this point end in draws.
18 Rxd4 Qa5+ 19 Rd2 0-0 20 Bd6 Nc6
“Black’s most sensible approach, offering the exchange in a bid to
untangle,” writes editor and Everyman cousin IM Richard Palliser.
Alternatives include 20 ... Rd8 and 20 ... f5.
21 0-0

As usual in this line, White gets his way with a monster development lead
and potential for attack, but only at ruinous expense to his bank account. If
White’s initiative fizzles later on, he is just down too much material to
survive, so there is tremendous pressure on both sides.
21 ... f5!?
Nepo engages his favourite Poisoned Pawn defensive move, ... f7-f5.
Black has also tried 21 ... Qd8, 21 ... Re8, and 21 ... Nce5.
22 Bxf8 Nxf8 23 Nd6 Ne5?’
A novelty, but it isn’t a good one. What I resent most about our
engine/database dominated era is that we the ordinary, who aren’t gifted with
photographic memories, are forced to memorize opening lines like long,
rehearsed speeches. Today 23 ... Qe5!? is considered the best move.
24 Khl?!
Unnecessary prophylaxis. Simply 24 Rfdl! leaves Black tangled up since
24 ... Bd7 is met by 25 Nxf5! exf5 26 Rd5! Qb4 27 Rxe5, which the engine
assesses as heavily in White’s favour.
24 ... Nfg6
24 ... Kh7 25 Rfdl Bd7 26 Nxb7 Qc7 27 Nd6 Neg6 28 c4 also looks better
for White.
25 Nc4
Why retreat? Again 25 Rfdl is better.
25 ... Nxc4 26 Bxc4 Kh7 27 Bb3
Obviously not 27 Bxe6?? Bxe6, since White’s rook on d2-hangs if the
bishop is recaptured.
27 ... Ne5 28 h3
Preventing ... Ng4.

Question: How is Black going to develop his queenside?

Answer: It’s not going to be easy. The engine says Black is okay here, but only if he pushes his b-
pawn next move.
28 ... Bd7?
This is the wrong way to develop as it hangs the e-pawn. 28 ... b5! was
correct; e.g. 29 Rd6 Nc4 30 Bxc4 bxc4 31 Qd4 Qc7 32 Rbl a5 and chances
are balanced.
29 Rd6!
Threatening to win a piece with 30 Rel.
.
29.. Rc8
 Exercise (critical decision): Does 30 Rel still win a piece? If not, then what
is White’s best move?

30 Rel?
Falling for Nepo’s trap. You can’t find something if you are unaware it is
lost. Najer goes from winning to clearly worse, in just a single move.
Answer: White is winning if he clips the e-pawn: 30 Bxe6! Rc3 31 Qd4 Nc6 32 Qd5 Qxd5 33 Bxd5
Be8 34 Rxf5, threatening Rf8! followed by Bg8+. Black is busted.
.30.. Nc4!
This move proves the Earth is flat and that all space flights - including
missions to the Moon and Mars - are a hoax perpetrated on the public! From
what looked like a resignable position, Nepo pulls off a clever swindle. White
isn’t necessarily lost at this point, but the sudden shock from thinking he is
winning, to fighting for his life, was too much for Najer to deal with in this
game.
 31 Bxc4 Rxc4 32 Rxd7
The best of a bad situation. 32 Qd2? steps out of the trap, yet makes
matters worse after 32 ... Qxd2 33 Rxd2 Bc8. Three pawns is too many for an
exchange and White is lost.
32 ... Re4!

This is the sting in the calculation which Najer must have missed earlier.

Exercise (critical decision): Should White play 33 Qxe4, picking up two rooks
for the queen; or 33 Rxg7+, followed by exchanging rooks, to try and
defend
a queen ending?

33 Qxe4?’
The engines don’t dislike this move, but I think Najer had better chances
in the queen endgame, since it is difficult for a human player to coordinate
the rooks with required precision.
 Answer: Hence 33 Rxg7+! Kxg7 34 Qg3+ Kf7 35 Rxe4 fxe4 36 Qf4+ and then something like 36
... Qf5 37 Qc7+ Kf6 38 Qxb7 e3 39 Qxa6 Qxc2 40 Qal+ Kf5 41 Qa5+ e5 42 Qd5, when it won’t be so
easy for Black’s king to evade eternal checks, at least from a practical standpoint.
33 ... fxe4 34 Rxe4 Qb5 35 Rc7?
Najer was undoubtedly flustered by the change of fortunes and this cost
him the game. The c2-pawn is irrelevant. He should have played 35 Rf7!
Qbl+ (or 35 ... Kg8 36 Re7 Qbl+ 37 Kh2 Qxc2 38 R4xe6) 36 Kh2 Qxc2 37
Rxe6 Qb2 38 Rf5!, when it is hard to see how Black can advance the
queenside pawns successfully.
35 ... Qb6 36 Rec4?
The final and fatal error. Here he had to find 36 Re7 Qbl+ 37 Kh2 Qxc2
38 R4xe6 Qb2 39 Re2 Qb4 40 R2e4 Qd6+ 41 Khl b5 42 R4e6 Qdl+ 43 Kh2
a5 44 Rb6 b4 45 Reb7 and White survives with a miracle, since 45 ... h5 (or
45 ... a4 46 Rxb4 a3 47 Rg4!) 46 Rb5 a4 47 Rxb4 a3 48 Rbl Qd6+ 49 Khl
Qd4 50 R7b5 a2 51 Rxh5+ Kg6 52 Rhb5 axblQ+ (or 52 ... alQ etc) 53 Rxbl
is a fortress draw once the rook gets to f3, which the tablebases confirm.
36 ... Qf2 37 Re4
Unfortunately, 37 Rxb7 hangs a rook to 37 ... Qfl+.
37 ... b5 38 Rc6 a5
The pawns will now race home.
39 Rexe6 a4 40 Re7 a3 41 Rcc7 Qf6 42 Ra7
 Exercise (combination alert): Work out Nepo’s forced win.

Answer: Step 1: Push the a-pawn. The a7-rook is unable to capture, since it must simultaneously
keep watch over its e7 counterpart.
42 ... a2!
The coronation ceremony was just delayed, not cancelled.
43 c3 Qxe7! 0-1
Step 2: Overloaded defender (again!). Chop the e7-rook.

Game 7
P.Smirnov-I.Nepomniachtchi
Novokuznetsk 2008
Sicilian Defence

1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Bg5 e6 7 f4 Qb6 8 Qd2

Qxb2
Poisoned Pawn, round 2! My nine-year-old Akita, Emma, still digs up
bones she buried as a puppy. In the same way, Nepo has stuck with his
beloved Poisoned Pawn variation since childhood.
9 Rbl Qa3 10 e5
Same line as last time.
10 ... h6 11 Bh4 dxe5 12 fxe5 Nfd7 13 Ne4 Qxa2 14 Rdl Qd5 15 Qe3
Qxe5
Black is willing to suffer the discomfort of development deprivation to
achieve his goal of collecting lots of white pawns!
16 Be2 Bc5 17 Bg3 Bxd4 18 Rxd4 Qa5+ 19 Rd2 0-0 20 Bd6 Nc6 21 g4’?

The Poisoned Pawn Najdorf is a circus where you never know what
strange act will appear. This was a theoretical novelty at the time. White
intends g4-g5, going after Black’s king. The usual moves are 21 Bxf8 and 21
0-0 (as in Game 6).
 .21.. Qb6!
As Charles Darwin did not in fact write: “It is not the strongest of the
species that survives. It is the one that is most adaptable to change.” Nepo,
without the aid of pre-game analysis, refuses to be flustered by Smirnov’s
novelty and finds Black’s strongest response.
22 Qh3
When you play a line like this, mate is foremost on your mind, so 22
Qxb6?! goes against the position’s grain. After 22 ... Nxb6 23 Bxf8 Kxf8 24
0-0 Nd7 25 c4 Ke7 26 c5 f6 I prefer Black’s chances.
22 ... f5!
Principle: Counter in the centre when attacked on the wing.
23 g5! ?
Refusing to dial it back, even a little, Smirnov is in the grips of a frenzy of
attacking largesse. The question arises: is this sacrifice a stairway to heaven
or a highway to hell?
The engine opts for 23 gxf5 Rxf5 24 Qg3 Nd4 25 Rgl g5!? and only very
slightly prefers Black.
23 ... fxe4 24 Qxe6+
24 gxh6? is too slow: 24 ... Q12+ 25 Kdl Qf6! (threatening ... Qal mate)
26 Bg4 Qxh6 and White doesn’t have enough for the material deficit.
24 ... Kh8
 Question: Is Black in danger of getting mated?

Answer: If you believe that Black’s king is some terrified rodent who burrows underground, then
think again. The engine begins to prefer Black’s position, despite the practical dangers to his king.
25 gxh6?
This natural capture surrenders the initiative. The correct way to proceed
was 25 Bxf8! Nxf8 26 Qe8 Qc5 27 Rfl Bh3! 28 Qxa8 Bxfl 29 Bxfl e3 30
Rd3 Ne5 31 Rd8 Kh7! 32 Bd3+ (not 32 Rxf8?? Qc3+ 33 Ke2 Qd2 mate) 32
... Nxd3+ 33 Rxd3 Qb4+ 34 Ke2 Qg4+ 35 Kxe3 Qxg5+ and the game would
probably be drawn.
25 ... Qf2+
Here 25 ... Rf6! 26 hxg7+ Kxg7 27 Qxe4 Nf8 may be even stronger, when
Black has excellent chances to consolidate.
26 Kdl Qg2! 27 hxg7+ Kxg7 28 Rel Rf6 29 Qb3?’
This allows Black to unravel further. He should have played 29 Qd5 to
retain control over e5.
29 ... Nde5

Nepo is a piece up and his king looks safe.

30 Bxe5
If 30 Qb6, threatening Rgl, then 30 ... Qf2 31 Bc5 Qf5 32 Rgl+ Kh7 and
White doesn’t have a good way to continue attacking.
30 ... Nxe5 31 Qe3
Threatening Rgl again.
31.. . Ng6 32 h4 Qf2 33 Qc3 Qb6
Both 33 ... Qxh4!? and 33 ... Bf5 are strong too.
34 Kcl Bf5 35 h5 Rc8 36 Bc4
36 Qal? loses instantly to 36 ... e3, removing the defender of c2.
36.. . Qc6!
 Chaos is slowly being transformed into order. White is totally tied up in a
pin and basically runs out of pieces to even attempt to get at Black’s king.
37 Rd4 b5 38 hxg6 bxc4 39 Rfl
Threatening Rxf5.
39 ... Kxg6 40 Qg3+ Kf7
White has nothing except a few final, desperate checks.
41 Qg5 Rg8 42 Qh5+ Bg6 43 Rxf6+ Qxf6 44 Qd5+ Kg7 45 Qd7+ Qf7
46 Qh3 Qf4+ 47 Kbl Rb8+ 48 Ka2 c3!
 Clearance. This move allows the bishop to enter the attack.
49 Qxc3 Bf7+ 50 Ka3 Qe5! 0-1
White has no good way to untangle.

Game 8
V.Anand-I.Nepomniachtchi
Mainz (rapid) 2009
Sicilian Defence

1 e4 c5 2 M3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Bg5 e6 7 f4 Qb6 8 Qd2

Those in power who espouse a doctrine tend to hunt down heretics.
Former World Champion Vishy Anand was the most deadly opening
theoretician of his reign (2007-13), so he was undoubtedly happy to face
Nepo in such a topical line.
. Qxb2
8..
 No, this isn’t an opening book on the Najdorf Poisoned Pawn. Are you
ready for that gripping daytime drama, the Poisoned Pawn season 3? It takes
some serious chutzpah to risk playing the line against a reigning world
champion, as Nepo did in this game, especially against one as well prepared
as Vishy Anand. We all remember what happened to Bobby Fischer when he
had the nerve to play it in his world championship match against Boris
Spassky, who trapped Bobby’s queen and humiliated him in a virtual
miniature.
9 Rbl Qa3 10 e5 h6 11 Bh4 dxe5 12 fxe5 g5!

An excellent practical decision. Nepo wisely deviates from his normal 12

... Nfd7, which Anand undoubtedly expected.
13 Bf2’?
This is White’s most radical option. He gives Black a free tempo for ...
Ng4 and then gives away the e5-pawn for open lines and increased
development. 13 exf6 and 13 Bg3 are also played.
 13 ... Ng4 14 Bg3 Nd7
White’s e5-pawn can’t be protected, but of course Anand knew that.
15 Be2 Ngxe5 16 0-0
The immediate 16 Nxe6!? is another tricky line which was unplayed at the
time of the game.
16 ... Bg7 17 Nxe6!

Question: Is Anand’s sacrifice sound?

Answer: An unsound attack/sacrifice is merely a subset of a delusion of grandeur. This one

however, is both sound and necessary, otherwise White will just be two pawns down. At the time this
shot had only been played once, in an obscure email game, which Anand may have seen, but I’m
guessing Nepo hadn’t. Despite that possible handicap. Nepo managed to find the right defensive
moves, stay afloat, and even outplay the 15th world champion.
17 ... fxe6 18 Ne4!
Threatening Nd6+, among other things.
.
18.. Rf8
 The unplayed (even today) 18 ... Nf7 is a serious alternative, intending
simply to castle. White should therefore prevent that: 19 Bd6! (not 19 Bh5??
0-0! and White doesn’t have anywhere near enough compensation) 19 ...
Nxd6 20 Nxd6+ Kd8 21 Nf7+ Ke7 22 Rb4! (interference; White has seized
control over d6) 22 ... Nf6 23 Qd6+! Kxf7 24 Rxb7+ (discovered attack) 24
... Bxb7 25 Bh5+ Kg8, when White should decline the “free” queen and be
satisfied with perpetual check with 26 Qxe6+! Kh7 27 Qf5+ Kg8 28 Qe6+.
19 Bh5+?

Question: Why would you give this natural move a question mark?

Answer: The move is a spam call, in that it annoys, yet fails to inflict damage. The check may be
natural, but it actually grants Black a favour since the king wants to go to e7 anyway.
White should prefer 19 c4 (or 19 Khl b6 20 c4 Rxfl+ 21 Rxfl,
transposing) 19 ... Rxfl+ 20 Rxfl b6 21 Khl Kd8! (Black’s slippery king is a
tough target to pin down to one spot) 22 Rdl Qe7 23 Nd6 Rb8 24 c5 Bb7 25
Nxb7+ Rxb7 26 c6 Rc7 27 Qb2!? Qc5 28 Bf3 b5, which the engines call
dead even and led to a draw in four correspondence games.
19 ... Ke7 20 Rb3
 Exercise (criticaldecision).-Now what? White’s queen is about to enter d6.
How did Nepo avoid all danger and even emerge with the advantage?

Answer: 20 ... Rxfl+ 21 Kxfl Nc4H

Suddenly, White’s queen is the fool who does business with a drug dealer
and then is shocked when double crossed. White is denied the queen check
on d6 with this clever retort, which returns the piece for a pawn-up ending
and all the winning chances.
Anand may have expected 21 ... Qa4?? 22 Qd6+ Kd8 23 Nc5 Qc4+ 24
Rd3! Nxd3 25 Qc7+ Ke7 26 Bd6+ Kf6 27 Nxd7+ Bxd7 28 Qxc4 and Black
can resign.
22 Rxa3
With deep regret Anand is forced to abandon his attack and enter the
inferior ending, since 22 Qel?? goes nowhere. For example, 22 ... Qa5 23
Qe2 Nde5 24 Bel Qd5 25 Bb4+ Kd8 26 Bf3 a5 27 Nf6 Qd4 28 Bc3 Qc5 29
Ne4 Qc7 and Black has consolidated the extra piece.
22 ... Nxd2+
What a relief. With queens off the board, Black’s king is granted absolute
immunity from arrest.
23 Nxd2 Be5?’
Don’t you hate it when the most natural, logical-looking move is actually
incorrect? This move screams to be played since it follows the principle: The
pawn-up side in an ending benefits from swaps. The problem is that it allows
White’s rook to infiltrate f7, complicating Black’s task of conversion.
Instead, 23 ... b5! 24 c4 Bb7 25 Bf3 Bxf3 26 gxf3 bxc4 27 Nxc4 Rc8 leaves
White fighting for the draw.
24 RB! Bxg3 25 Rf7+! Kd6 26 hxg3 b5

27 Rh7?’
Don’t you hate it when the most natural, logical-looking move is actually
incorrect? Oh, wait, I already said that. The temptation to chop Black’s weak
h-pawn should be avoided since it surrenders the initiative. Instead, 27 Bf3!
Ra7 28 Ne4+ Ke5 29 Nf6! leaves Black tangled up and White with excellent
chances to hold.
27 ... Bb7?’
Both sides miss the promising pawn sacrifice 27 ... Nf6! 28 Rxh6 Nxh5 29
Rxh5 Bb7! 30 c4 (not 30 Rxg5? Rc8 and White is in deep trouble; e.g. 31
Rg7 Bxg2+! 32 Kxg2 Rxc2 33 K13 Rxd2 34 a4 bxa4 35 Ra7 Ra2 36 Rxa6+
Ke7! with a winning endgame) 30 ... Rc8 31 cxb5 axb5 32 Rxg5 Rc2 33 Ke2
Rxa2 and Black’s tactically protected b-pawn may yet cost White the game.
28 Bf3!
It helps White to eliminate Black’s powerful bishop.
28 ... Rf8 29 Ke2 Bd5 30 Bxd5 exd5 31 Rxh6+ Rf6 32 Rh5?’
This is a waste of time. 32 Rh8! is preferable, when it isn’t easy for Black
to make progress; e.g. 32 ... Re6+ 33 Kdl Re3 34 Rh6+ Re6 35 Rh8 and so
on.
32 ... Re6+ 33 Kdl g4!
 Multi-purpose:
1. Black’s g-pawn is secure now since it can be protected via ... Nf6 or ...
Ne5.
2. If Black’s g-pawn is safe, then White is essentially a pawn down since
the one on g2 hardly counts.
3. White’s knight is denied use of the important f3-square.
34 Nb3
White is better off playing 34 Rh7, with ideas like 34 ... Nf6 (or 34 ... Re3
35 Rh6+) 35 Ra7 Ke5 36 Nb3 Rc6 (or 36 ... Ne4 37 Kcl Nxg3 38 Nc5) 37
Nel! Kd4 38 Nd3 Ne4 39 Kcl Nc3 40 Rg7 and so on.
34 ... Nf6 35 Rf5
At this point it’s hard to see an exchange of rooks saving White; e.g. 35
Rh6 Ne4 36 Rxe6+ Kxe6 37 Nd4+ Ke5 and White is busted, since 38 Ne2?
d4 leaves his position hopelessly passive.
35 ... Ne4 36 Kcl Nc3 37 a3?
 Anand had to try 37 Kd2! Ne4+ (37 ... Nxa2 38 Rg5 should draw) 38 Kcl
Rf6 39 Rf4 Rg6 40 Rf8 Nxg3 41 Ra8 Ke5 42 Nc5 and White still has
chances to hold the game.

Exercise (planning): White’s position has passed the survivability threshold.

Prove why his last move was an error.

Answer: 37... Re3!

Black now wins both white g-pawns.
38 Rf6+ Ke5 39 Rxa6 Rxg3 40 Nc5 Rxg2
Black’s passed g-pawn, supported by rook, king and knight, is decisive.
41 Nd3+ Kd4 42 Nel Re2 43 Nd3 Rh2 44 Nel Ne4 45 Kb2 Rhl 46 Nd3
Nc3!
The threat of... Rbl mate forces further concessions from Anand.
47 a4
Or 47 Nel Na4+ 48 Kbl g3 49 Rg6 Rgl 50 Rg8 Ke3 (threatening ... Kd2)
51 Rd8 g2 and wins.
47 ... Nxa4+ 48 Ka3 Ral+ 49 Kb4 Rbl+ 50 Ka3 Nc5!

Simplification. The rook ending removes White’s cheapo potential and is

an easy win for Black.
51 Nxc5 Kxc5 52 Rg6 Rgl 53 Kb3 g3 54 Rg4 g2 55 Kb2 b4 56 Rg5
Kd6 0-1
Black’s king walks over to the kingside and helps his g-pawn promote.
 Chapter Two
Gathering Power: 2010-2018
This chapter covers the period of Nepomniachtchi’s young adulthood, from 2010 to 2018, to the cusp
of when he became a world-class player. The finish of the first game below just blew my mind, as
Nepo hit his opponent with a seemingly endless series of combinations.

Game 9
I.Nepomniachtchi-E.Inarkiev
European Championship, Rijeka 2010
Ruy Lopez

1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 d3

Question: Another Anti-Marshall. Doesn’t anyone allow Black a Marshall

Gambit in the Lopez anymore?

Answer: Apparently FIDE must have banned the line, given its relatively rare occurrence compared
with its popular Anti-Marshall cousins. But 5 d3 isn’t just an Anti-Marshall; it’s an Anti-Everything.
5 ... d6
In later games Inarkiev switched to a Moller set-up with 5 ... b5 6 Bb3
Bc5.
6 c3
This way White’s light-squared bishop is preserved from ... b7-b5 and ...
Na5 ideas.
. g6
6..
 Question: Why fianchetto, rather than develop the bishop to e7?

Answer: The fianchetto is logical since White has yet to apply pressure on e5. So, instead of the
traditional, long-winded ... Be7, ... 0-0, ... Re8, ... Bf8, ... g7-g6 and ... Bg7 manoeuvres, as seen in
many Closed Lopez lines, Black cuts out the middleman and saves time by putting the bishop on g7
straight away.
7 Nbd2 Bg7 8 0-0 0-0 9 Rel
Honestly, I don’t know why so many top GMs are happy to embrace this
position for White, which, to my eyes, is already equal. But maybe achieving
a “+=“ edge against 1 e4 e5 is an impossibility, since AlphaZero pretty much
proved that, with perfect play, Black equalizes in all lines of the double king
pawn openings.
9 ... b5
9 ... Nd7, reinforcing e5 and preparing for a future ... f7-f5, is also popular.
10 Bc2
The bishop drops right because 10 Bb3 helps Black to expand on the
queenside with 10 ... Na5 11 Bc2 c5.
.
10.. Re8
Black more often plays 10 ... Bb7 first. The text is directed against a quick
d3-d4.
11 a4
Now 11 d4 might be met by 11 ... exd4 12 cxd4 Nb4 13 Bbl c5 with play
against the white centre.
11 ... Bb7
The alternative is 11 ... b4 12 a5 Rb8.
12 b4
White gains a touch of territory on the queenside and rules out... b5-b4.
12 ... Qe7
A novelty at the time, but it hasn’t caught on. 12 ... Nd7 and 12 ... Qd7 are
more usual, the latter also seen in I.Nepomniachtchi-G.Kamsky, European
Cup, Plovdiv 2010 (which Black won).
13 Nb3!
 Question: The knight looks useless on b3. What is its purpose?

Answer: It’s a prophylactic move, for the following reasons:

1. Black’s future ... d6-d5 ideas are discouraged, since White’s knight will
jump into the c5-hole.
2. The b3-knight also eyes a5, which means that Black’s c6-knight is
discouraged from leaving its square.
13 ... Rad8?’
This feels wrong to me. I don’t understand what the rook does on d8, since
... d6-d5 would likely come with unfortunate strategic consequences for
Black, due to the aforementioned hole on c5. 13 ... Nd7 looks preferable, as
in G.Monnisha-S.Ghukasyan, World Junior Championships, Durban 2014.
14 Qe2 Qd7

Question: What is the point of Black’s last move, given that he could have
 played the queen to d7 in the first place?

Answer: We reached one of those strange situations where there is no useful plan, except to await
White’s intent, since taking action with ... d6-d5?! would be clearly detrimental.
15 Bd2!
Preparing a4xb5 and c3-c4.
15 ... Rb8
An admission that his earlier 13 ... Rad8 was a waste of a time. Indeed,
Black could have played ... Qd7 and ... Rab8 in two moves rather than four.
16 axb5 axb5 17 c4 Ne7?’
White achieves a clear strategic advantage after this second inaccuracy. 17
... bxc4 18 dxc4 Qc8 keeps White’s edge to a minimum.
18 cxb5! Qxb5 19 Na5!

Threatening Ba4. White has seized a queenside initiative.

19 ... Qd7 20 Ba4! c6 21 Rael
Unpleasant pressure mounts on c6.
 .21.. Rec8 22 d4! exd4 23 Nxd4
Here comes another guy, going after c6.
23 ... Ne8?’
A passive stance only increases White’s grip. Black needs counterplay, so
it would be better to strike back in the centre with 23 ... d5!, despite its
drawbacks: 24 exd5 Nfxd5 25 Ndb3 (eyeing the newly created hole on c5,
while threatening a double attack on black queen and b7-bishop) 25 ... Ba8
26 Nc5 Qd8 and Black’s position is a bit less wretched than the one he got in
the game.
24 Qd3!
Preventing... d6-d5 tactically.

Question: Why doesn’t Nepomniachtchi grab the bishop pair with Nxb7 - ?

Answer: Even though the position is open - which would normally make a bishop the more
valuable minor piece - in this instance White’s towering a5-knight, which applies unrelenting pressure
on c6, is clearly superior to Black’s cringing light-squared bishop.
24 ... Ba8
Not really necessary since Nepo had no intention of swapping on b7, but
it’s hard to find any good moves for Black now. 24 ... d5?? fails to 25 exd5
Qxd5 (or 25 ... Nxd5 26 Ndxc6! Nec7 27 Ne7+!, winning the queen) 26 Rxe7
Bxd4 (threatening ... Bxf2+ and ... Qxd3) 27 Khl! Ba8 28 Bb3! and wins.
25 h3
This cuts out... Qg4 ideas.
25 ... Rb6 26 Nf3!
 The d6-pawn is a new target.
26 ... Nc7
Maybe Black should consider 26 ... d5 at this point, though after 27 Nb3
Qa7 28 Nc5 dxe4 29 Qxe4 Nd5 30 Ne5 his position remains under heavy
strain.
27 Nc4 Ra6 28 Bb3
Targeting both d5- and f7-squares.
28 ... Rd8 29 Bf4 Ne6! 30 Bg3!
Now Nxd6 is a deadly threat. Nepomniachtchi avoids the strategic trap 30
Bxd6?, which allows Black to mix it up a bit with 30 ... c5! 31 Redi cxb4 and
his position has improved.
30 ... d5
There is nothing better.
31 Nce5 Qb7 32 Qe2?’
Nepo misses the combination he finds next move. I can’t tell you what it
is, or I give away the answer to the coming exercise.
32 ... Rb6?
Black worries about his future, when he should be worrying about his
present. 32 ... Qb5! was forced.

From this point on until the end of the game, we get a glimpse of
Nepomniachtchi’s power when on the attack, where he sometimes mimics a
supernatural being who has the ability to walk through walls.

Exercise (combination alert): White has a deep and crushing combination here.
Try to see further than just the first move and work it out to its conclusion.

Answer: Step 1: Undermining/annihilation of defensive barrier. Sacrifice a piece on f7.

33 Nxf7!!
Now nothing can stanch the flow of Nepo’s initiative, as white attackers
begin to populate the kingside with demographic inevitability. Far weaker is
33 exd5?! Nxd5 34 Nxf7 Qxf7 35 Qxe6 Qxe6 36 Rxe6 Rxb4, when Black
may yet hold the game through the power of prayer.
33 ... Kxf7 34 exd5
Step 2: Open the e-file. This move isn’t exactly a tough call; our eyes just
tell us that Black will at some point collapse.
.
34.. cxd5
The b6-rook covers the knight. Now comes the second wave of Nepo’s
sacrificial idea.
35 Rc7!
Step 3: Interference. It was crucial for you to see this move before starting
the combination. Not 35 Ng5?? Nxg5 36 Rc7 which fails to 36 ... Re6!.
35 ... Rd7
This move keeps Black alive for the moment, at the cost of pushing his
position halfway to bankruptcy. Black must hand over the queen since 35 ...
Qa6 36 Ng5+ or 35 ... Nxc7 36 Qxe7+ Kg8 37 Qxd8+ Bf8 38 Bxc7 wins
even more.
36 Rxb7 Bxb7
 Exercise (combination alert): Nepo’s attack is not yet placated. Not even close.
Find the fourth wave of the combination.

Answer: Removal of the guard.

37 Ba4!
Black’s d7-rook is attacked, yet at the same time is unable to move as it
covers the e7-knight.
37 ... Ba6
Black’s position is an old man, weary of life’s pain. 37 ... Rd8? loses
instantly to 38 Ng5+!, while 37 ... Bc6? of course fails to 38 Qxe6+.
38 Ne5+
38 Qd2 also wins; e.g. 38 ... Rdb7 39 Rxe6! Rxe6 40 Ng5+ and if 40 ...
Kf6 then 41 Qd4+! Kxg5 42 Qf4+ Kh5 43 Qh4 mate.
.
38.. Bxe5
Or 38 ... Kg8 39 Qf3! (threatening to infiltrate f7) 39 ... Ng5 40 Qg4 and
wins.
39 Qxe5 Ra7 40 Qh8!
The queen enters the soft underbelly of Black’s position.
40 ... Nf8 41 Bh4!
Another attacker enters. White threatens Bg5!, Bh6, and Qg7 mate. 41
Be5! Nf5 42 g4! was even stronger, if there is such a thing as stronger than
completely winning
41 ... g5
Complete desperation. 41 ... Nf5? isn’t such a great defensive idea: 42 Be8
mate.
42 Bxg5 Rg6

Exercise (combination alert): Nepo found yet another combination to put

his opponent away.

Answer: Step 1: Transfer the queen to d4, attacking the a7-rook.

 43 Qd4! Rb7 44 Qf4+! 1-0
Step 2: Check on f4, winning more material; while after 44 ... Kg8 45
Bxe7 Black is soon mated.

Game 10
LNepomniachtchi-Ba.Jobava
European Championship, Rijeka 2010
Caro-Kann Defence

1 e4 c6 2 d4 d5 3 B
The Fantasy Variation of the Caro-Kann. It used to be played as a surprise
weapon and has now gone mainstream.
3 ... Qb6
I have faced the Fantasy eight times, all from players rated 2300 or higher,
and noticed that every one where I played 3 ... Qb6 was drawn, while I won
all the games when I responded with 3 ... e6, turning it into a French. Of
course, eight games are hardly an empirical sample. Looking at the stats in
the database, I see the opposite result, with White scoring 55% against 3 ... e6
and a dismal 44.2% against 3 ... Qb6. So maybe this is Black’s optimal line,
after all.
4a4’?
 These kids, with their new-fangled opening ideas! Nepo decides to veer
away from civilization, into a relatively uncharted land. It’s not quite a
novelty, as Nepo thought, because it had featured in two games in the same
tournament the day before. 4 Nc3 is White’s main move.

Question: Does 2+2 now equal 5? Why would Nepomniachtchi choose 4 a4

over a developing move?

Answer: Clearly this is a loose interpretation of the line’s original intent. White’s problem in the ...
Qb6 line is that Black may play ... e7-e5 at some point, and if White replies d4xe5, then Black has ...
Bc5, dominating the important gl-a7 diagonal. If White tosses in a2-a4-a5, this disrupts Black’s
intention by displacing the queen from b6.
.4.. e5’?
Jobava plays it anyway, hoping to foment an early crisis. The two earlier
games both saw 4 ... dxe4 5 a5 Qc7 6 fxe4 e5, whereas 4 ... e6 is now more
often played, as three months later in I.Nepomniachtchi-V.Ivanchuk, Havana
2010 (which Black won). The engine actually prefers Jobava’s choice, but
not as played.
5 dxe5 dxe4?
This isn’t Black’s best continuation. Instead:
a) 5 ... Bc5? isn’t effective either due to 6 a5, when 6 ... Bf2+ 7 Ke2 Qd4 8
Qxd4 Bxd4 9 f4 dxe4 10 Nd2 is in White’s favour; for instance, 10 ... f5 (10
... Bf5 11 Ra4! wins a pawn) 11 g4! g6 12 gxf5 gxf5 13 Nb3 Bxgl 14 Rxgl
is awful for Black.
b) 5 ... a5! 6 Bd3 Bc5 is how Stockfish wants to proceed, which looks
decent for Black and led to a win in Ch.Bauer-M.Ragger, French League
2013.
6 a5 Qc7
6 ... Qc5 7 f4 Ne7 8 Nc3 looks good for White too.
7 f4!

Question: Black is one tempo up in development. Does this mean he is at

least even or slightly better?
 Answer: Let’s assess:
1. Black’s e4-pawn is vulnerable to capture.
2. If Black picks off the a5-pawn in return, this is not an equitable
arrangement, since Black loses time while swapping a valuable centre pawn
for a lesser wing pawn.
3. Black’s slight lead in development is not enough to make up for points
one and two.
Conclusion: I slightly prefer White; and if Black’s position were a house,
it would be a fixer-upper.
7 ... Nh6 8 Nc3 Bb4 9 Bd2 e3
In his own notes Nepo gives this move a question mark. I’m not sure
Black has any good moves at this stage:
a) 9 ... Bf5 is met by 10 Nxe4 Bxd2+ 11 Nxd2 Nd7 12 Nc4! and Black
doesn’t get full compensation for the missing pawn.
b) 9 ... Bxc3 10 Bxc3 0-0 11 Qd4 Bf5 12 h3!, intending g2-g4, is also in
White’s favour.
c) 9 ... Bxa5 10 Nxe4 Bxd2+ 11 Qxd2 0-0 12 NI3 Rd8 13 Nd6 looks
difficult for Black.
10 Bxe3 0-0
If 10 ... Bxa5 11 Nf3 0-0 12 Bd3, White may soon whip up a dangerous
attack on Black’s king.
11 Nf3 Rd8 12 Bd3
 12 ... Na6
Intending ... Nc5. After 12 ... Bf5 13 Qe2 Bxd3 14 cxd3 Bxa5 15 d4 Nf5
16 0-0 White’s imposing centre gives him a clear advantage.
13 Qe2 Bxa5 14 0-0
So point two in the notes to 7 f4 has come to pass. This means White gets
decent chances to launch an attack on Black’s slightly under-defended king.
Nepo is not the type to reduce his initiative for structural benefit with a line
like 14 Bxa6 bxa6 15 0-0 Rb8.
14 ... Nb4
If 14 ... Bf5, White might well change his mind, since 15 Bxa6 bxa6 16
h3! Rab8 17 Na4 is strategically uncomfortable for White.
15 Be4 Bf5 16 Khl Bb6 17 Bxb6!
 Exercise (critical decision): Should Black recapture on b6 with his a-pawn
or queen?

Answer: Recapturing with the queen is the only move.

17.. . Qxb6!
The b4-knight requires protection. Jobava alertly avoids the trap 17 ...
axb6?? (we are not afforded the luxury of creative latitude in forcing
positions) 18 Rxa8 Rxa8 19 Bxf5 Nxf5 20 Qe4 (double attack) 20 ... Nxc2 21
Ndl! and White wins a piece.
18 Ng5!
Now Black must factor in tricks like Bxf5, followed by either Qe4 or Qh5.
18.. . c5
Jobava rightfully worries about his king. His last move opens lines of
communication between the black royalty.
19 Rael
 Nepo ominously continues to mass pieces on the kingside.
.
19.. Qg6
The engine suggests the desperate 19 ... Bxe4 20 Ncxe4 Qa6 21 Qf2 Nf5
22 Qxc5, when White has won a pawn, while remaining with the superior
position.
20 Bxb7 Nxc2?’
This makes matters worse; but 20 ... Rab8 21 Be4 leaves White a pawn up
for nothing.
21 Nd5!
At least as strong as taking the a8-rook. The text threatens Ne7+ and
Nxg6. Even worse is that Black’s queen has no safe square to run to.
21 ... Rxd5
Forced. Not 21 ... Re8? 22 Bxa8 and Black is unable to recapture due to
the knight fork on e7.
22 Bxd5 Bd3
22 ... Rb8 allows 23 Nxf7! Nxf7 24 e6 Nd6 25 Qe5! Nxel 26 Qxd6 Re8
27 Rxel.
23 Qf3
The engine likes 23 Qd2 even more.
23 ... Re8
 Exercise (combination alert): White has two winning lines. Find one of them.

Answer #1: Push the e-pawn, which creates havoc on Black’s camp.
24 e6!
Answer #2: 24 Nxf7! Nxf7 25 f5 also wins; e.g. 25 ... Qg5 26 Bxf7+ Kxf7 27 Qd5+ Kf8 28 f6!
Nxel 29 Qxc5+ Kf7 30 e6+! Kg6 31 Qxg5+ Kxg5 32 f7 Rf8 33 e7 etc.
24 ... Nxel 25 Rxel fxe6
25 ... Kh8 loses to the simple 26 Bc6! Rxe6 27 Nxe6 fxe6 28 Be4 with an
easy conversion for White.
26 Rxe6! Kh8
 Exercise (planning): What is White’s most efficient winning plan?

Answer: Step 1: Simply make luft for White’s king, eliminating Black’s back rank tricks and
threatening Rxg6.
27 h3!
27 Rxg6?? would be a colossally unwise decision in view of 27 ... Rel+
28 Qfl Rxfl mate.
27 ... Rxe6 28 Bxe6
Step 2: Clearance. White’s bishop moves from d5, enabling Qa8+.
28 ... Bb5
Intending to block on e8 with his queen, if required.
29 f5!
Step 3: Gain a tempo on Black’s queen.
29 ... Qe8 30 f6!
Step 4: Push the f-pawn again.
 .
30.. Qf8

31 f7 1-0
Step 5: Push to f7, threatening Qe4; e.g. 31 ... a6 32 Qe4! g6 33 Qe5+ Qg7
34 f8R+ Ng8 35 Nf7 mate. For some people a practical joke is the goal of
life. Another option was to play 31 Qe4! at once, intending 31 ... g6 32 Qb7!
and Black is defenceless.

Young Ian was a 100-point underdog when the following game was
played, so his win was a bit of an upset.

Game 11
N epomniachtchi-A.Grischuk
I.
Russian Team Championship, Dagomys 2010
Sicilian Defence
1 e4 c5 2 M3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 h3

This has become a trendy line.

Question: Isn’t h2-h3 a passive move which wastes time?

Answer: I agree that it’s not a developing move, yet it is hardly passive, since White plans to thrust
forward with g2-g4. A secondary function is that it enables Be3 without allowing the annoying ... Ng4.
6.. . Nc6
Black has a million viable responses: 6 ... e5 in Najdorf style; 6 ... e6 in
Scheveningen; 6 ... g6 is a version of the Dragodorf; 6 ... b5 prepares to
fianchetto.
7 Be3
The immediate 7 g4 is White’s main move.
7.. . e5
The combination of... e7-e5 and ... Nc6 turns the position into a hybrid of
the Boleslavsky Sicilian. 7 ... e6 is Black’s most popular choice.
 8 Nf3

Question: Why block the f-pawn?

Answer: The f-pawn may not need to move, since with this structure a future f2-f4 can be met by ...
e5xf4, after which Black occupies e5 with the c6-knight.
8.. . Be7 9g4!?
The committal thrust was a novelty in this exact position. The push
weakens dark squares, especially f4. But as we know by now, Nepo is never
one to hold back on a structural concession, if he gets the promise of the
initiative in return. 9 Bc4 is White’s main move here.
9.. . Be6
The idea is to prevent Bc4. Of course it invites Ng5. The only other game
to take this path deviated with 9 ... h6 10 Nd5 (10 Bc4 is met by 10 ... Na5)
10 ... Nxd5 11 exd5 Qa5+ 12 c3 Nb8 13 Bg2 Qc7 14 Qe2 f5 15 gxf5 Bxf5 16
Nd2 Nd7 and Black was okay, G.Poli-B.Fajs, correspondence 2010.
10 Ng5
White can also ignore the provocation and play 10 g5!? Nd7 11 Nd5,
though even here Black looks fine.
 .
10.. 116!

Question: Why is Black giving away the bishop pair?

Answer: He also slightly weakens his light squares. In return Black gets enhanced central control,
especially over d5, and receives an open f-file for his rooks, which isn’t such a bad deal.
In any case it’s too late to back down now with 10 ... Bd7? because of 11
Nxf7! Kxf7 12 g5, when Black must return the piece with an inferior
position, since the greedy 12 ... Ne8?? loses to 13 Bc4+ Kg6? (or 13 ... Be6
14 QI3+) 14 Rgl, when h3-h4-h5 is coming and Black’s king doesn’t have a
prayer of survival; e.g. 14 ... Na5 15 h4! Nxc4 16 h5+ Kf7 17 g6+ Kf8 18
Qd5 and Black is mated.
11 Nxe6 fxe6 12 Bc4 Qd7

Question: Where does Black’s king belong?

Answer: That is the position’s big question. At this stage, intuitively I would have said on the
queenside. In the game Grischuk decided his king was safest in the centre.
13 h4!?

Question: Isn’t this a bit reckless?

Answer: You are technically correct, but we must also factor in style. Nobody would have bought
Tai: Move by Move if they intended to play ultra-safely in every game, with Colles and Londons as
White, and Caro-Kanns and Petroffs with Black!
A more strategically reserved player would have gone for something like
13 Qd3 Na5 14 Bb3 Nxb3 15 axb3 Qc6 16 0-0-0 b5 17 f3 Nd7, when Black’s
position is at least even.
13 ... b5
Natural but not best. Black can improve with 13 ... Bd8!, intending to re­
route the passive bishop to the a5-el diagonal; e.g. 14 f3 Ba5 15 Qd3 d5 16
exd5 exd5 17 Qg6+ Qf7 18 Qxf7+ Kxf7 19 Bb3 Nd4! 20 Kf2 Nxb3 21 axb3
Bb4 22 g5 hxg5 23 hxg5 Rxhl 24 Rxhl Bxc3 25 bxc3 Nd7 and I prefer
Black’s chances in the ending since White’s structure is a touch loose.
14 Bd3
After 14 Bb3 Na5 Black stands no worse.
14 ... Nd4
14 ... Nb4 is a decent alternative.
15 Rgl g5’?
 The engine doesn’t like this, but Grischuk’s move has its points;:
1. It removes the mobility from White’s kingside pawns.
2. It fixes g4 and 13 as potential future targets.
3. It stabilizes Black’s powerful d4-knight by cutting out I2-f4 ideas.
At least it does those things if Black maintains the pawn on g5.
16 Rhl
Here 16 hxg5! hxg5 17 Nbl! d5 18 Nd2 may be best, when I slightly
prefer White’s position since Black’s powerful d4-knight is about to get
booted by c2-c3. Note 17 Bxg5? fails to 17 ... Rh3! (threatening ... N13+) 18
Rg3 Rhl+! 19 Bfl b4 20 Bxf6 Bxf6 21 Ne2 Bh4! 22 Nxd4 (forced, since 22
Rd3? allows 22 ... Qf7) 22 ... Bxg3 and White is the exchange down and in
deep trouble.
.
16.. Rf8?’
The engine doubles down on its opinion by releasing the blockade at once:
16 ... gxh4! 17 Rxh4 Rg8 18 Bxh6 0-0-0 with active counterplay; e.g. 19 Be3
Nxe4 20 Nxe4 Bxh4 21 c3 Nf5! (21 ... Nc6? 22 a4! gives White a wicked
attack for the exchange) 22 gxf5 exf5 Black stands no worse.
17 hxg5 hxg5 18 a3
Dare I accuse Nepo of excess caution? 18 a4! gives White the advantage
after 18 ... b4 19 Nbl! Qc6 20 Nd2. Not 18 Bxg5?! because of 18 ... Nxg4!
19 Qxg4 N13+ 20 Ke2 Bxg5 and Black stands at least equal.
18 ... Qb7 19 Rh3
Reinforcing 13.
19 ... Kd7’?

Question: Does every player in this book strive to move their king to d7?

Answer: It certainly seems that way! Castling was probably a better option: 19 ... 0-0-0! 20 a4 b4
21 Bxd4 exd4 22 Qe2 (threatening Bxa6, winning the queen) 22 ... Kb8 23 Bxa6 Qc6 24 Na2 d5 25
exd5 Nxd5 with full compensation for the sacrificed pawn.
20 Kd2!
 Me too! Nepo pulls the same trick. This time 20 Bxg5?! is met by 20 ...
Rh8! 21 Rxh8 Rxh8 22 Kd2 Nxg4! 23 Be3 (forced) 23 ... Nxf2! 24 Bxf2
Bg5+ 25 Be3 Bxe3+ 26 Kxe3 Rh2! 27 Qgl Rh3+ 28 Kf2 Rf3+ 29 Kel Rxd3!
(clearance, threatening ... Nf3+) 30 Qg7+ Kc6 31 Qxb7+ Kxb7 32 cxd3 Nc2+
33 Kd2 Nxal and after all that Black emerges with an extra pawn and serious
winning chances, in view of the endgame principle: A knight ending tends to
be the worst one possible for the pawn-down side.
20 ... Rf7
20 ... Rh8 21 Qhl Rxh3 22 Qxh3 Rf8 looks okay for Black. Grabbing the
g-pawn never seems to work for White. Here 23 Bxg5? fails to 23 ... Nxe4+
24 Nxe4 Nf3+ 25 Ke2 Nxg5 and Black’s central majority gives him the
advantage.
21 13
Reinforcing both e4 and g4. Still not 21 Bxg5?! Nxg4 22 Qxg4 Rg7 23
Rahl Rxg5 and Black regains the piece with the upper hand.
21 ... Raf8 22 Ne2
The irksome d4-knight is finally sent packing. The engine suggests the
inhuman 22 Kcl!? unafraid of 22 ... Ne8 due to 23 a4! b4 24 Bxd4 exd4 25
Nbl!, aiming to unravel with Nd2, Rh5, Bc4, Qe2, Kbl-a2 and so forth,
while 25 ... b3?! 26 Ra3 only helps White.
22 ... Nc6 23 c3
23 Bxg5? Nxe4+! 24 Bxe4 Bxg5+ leaves White in deep trouble on the
dark squares. We should just stop analysing Bxg5 and assume it doesn’t work
in any version.
23 ... Ne8’?
The engine doesn’t like this either, preferring the more aggressive 23 ...
Na5.
 24 Qhl?!
This removes her Majesty from the action on her own flank. 24 Nel! was a
better way to defend the f3-pawn, when 24 ... Na5 25 Kc2 Nc4 26 Bgl Qc6
27 a4 looks better for White, since the black rooks are misplaced on the f-file.
24 ... Na5 25 Rdl?!
Why give up the bishop pair and allow Black to open the b-file? 25 Kc2!
Nc4 26 Bgl Qc6 27 b3 Nb6 28 Kb2 still looks a shade better for White.
25 ... Nc4+ 26 Bxc4 bxc4 27 Kcl Nc7
Both kings appear relatively safe for the moment.
28 Rh7 Qb8!
Grischuk gives his queen potential access to the kingside, while waiting to
see what White does next. Not 28 ... Rxf3? because of 29 Bxg5 (hooray! - a
variation involving Bxg5 finally works for White) 29 ... R8f7 30 Bxe7 Kxe7
31 Qxf3! Rxh7 32 g5, when the g-pawn is dangerous and Black struggles.
29 Rxf7 Rxf7 30 Ngl’?
 With three ideas:
1.13 is secured.
2. White may later target Black’s g5-pawn with Nh3.
3. White’s second rank is cleared, which can be of help both offensively
with Rd2 and Rh2, or defensively with Qh2, covering b2.
The drawback is seen after Black’s reply.
.
30.. Qb3!
Seeking to worry his opponent with ideas of... Nb5 and a sacrifice either
on a3 or c3.
31 Rd2 Qa2’?
I’m not sure a lone queen can do that much. Grischuk plays for the win.
Instead, 31 ... Nb5 would induce 32 Ne2, when Black stands better.
32Kdl!

White’s king is safer on the other side. Have you ever seen a game where
one player castles by hand - twice!?
 32 ... Qb3+?’
Having decided to come in, Black should continue with 32 ... Qbl+ 33
Ke2 and then 33 ... Ne8, intending ... Nf6 and ... Rh7 to gain the h-file, after
which the knight can return to the queenside again and Black is no worse.
33 Ke2
Nepo is not interested in proposing a draw by moving the king back to cl.
In that case, 33 Kel looks more accurate, leaving the second rank clear for
the rook with ideas of Rh2-h7.
33 ... Qb8 34 Kf2’?
Intending Kg3 and Rh2.
34 ... Kc6 35 Kg3 d5’?
Both sides are playing for the win.
36 Qh6
With the strategic threat of 37 exd5+, which Grischuk immediately
removes by pushing on.
.
36.. d4!
Another option was 36 ... Qg8!?, intending 37 exd5+ Kb7! 38 d6 Rh7, as
pointed out by GM Neil McDonald. White should still hold after 39 Qxh7
Qxh7 40 dxe7 Qxe7 41 Nh3 with rook, bishop and (soon) a pawn for the
queen; or he could just play 37 Rh2 with equality.
37 cxd4 exd4 38 Bxd4!
At first this looks like a terrible error. It isn’t. Which is just as well, since
38 Bxg5? Bxg5 39 Qxg5 Nd5+ 40 Kf2 Nf4 41 Qa5 (not 41 Rxd4?? Qxb2+)
41 ... d3 42 Qxa6+ Qb6+! 43 Qxb6 Kxb6 44 Ke3 Kc5 is very much in
Black’s favour, despite the two pawn deficit; e.g. 45 Rh2 (or 45 e5 Kd5 46 b3
cxb3 47 Ne2 Ng2+ 48 Kf2 Kc4 and so on) 45 ... e5 46 Rh8 Rd7 47 Kd2 Rb7
48 Rc8+ Kd6 49 Kc3 Rh7, followed by ... Rhl, winning the knight.
38 ... Nb5+ 39 Kg2 Nxd4
 Exercise (combination alert): White is a piece down, and 40 Rxd4 is
unplayable due to 40 ... Qxb2+ and 41 ... Qxd4. Do you see Nepo’s
hidden defensive resource?

Answer: Removal of the guard.

40 Ne2!
White saves the game with this trick. Not 40 Rxd4?? Qxb2+ 41 Kfl Qxd4,
since 42 Qxe6+ Kb5 43 Qxf7 fails to 43 ... Bc5 44 Qb7+ Ka5 45 Qc7+ Bb6
46 Qg3 Qxgl+! (simplification) 47 Qxgl Bxgl 48 Kxgl c3 etc.
40 ... Bf6??
This looks like a time pressure blunder. Hanging on to his extra piece is
disastrous for Black. He should have conceded the draw: 40 ... Nxe2! 41
Qxe6+ Kb5 42 Qd7+ Kb6 (not 42 ... Ka5?? 43 Rd5+ Kb6 44 Qe6+ Kb7 45
Rd7+ and wins) 43 Qe6+ Kb5 44 Qd7+ with perpetual check.
41 Nxd4+!
 The correct recapture, whereas 41 Rxd4? Qxb2! 42 Rxc4+ Kd7 43 Kf2
Qb6+ would have enabled Black to scramble a draw thanks to his active
queen.
41 ... Kb6
Not 41 ... Bxd4? 42 Qxe6+ Kb5 43 Qxf7 and wins.
42 Nxe6
Suddenly, White is two pawns up, with the far safer king.
42 ... Qe5 43 Qg6!
Another clever shot.
43 ... Re7
43 ... Qxe6 44 e5! (pinned piece/overloaded defender) 44 ... Kc7 45 Rc2!
would have seen White forcing helpful simplification, after which Black has
no chance of saving the game.
44 Nc5!

Another stunner. The knight can’t be taken by either piece: 44 ... Kxc5? 45
Rd5+ or 44 ... Qxc5? 45 Qxf6+ Kb7 46 Rd5 Qc7 47 Qxg5 wins easily.
44 ... Qf4 45 Rd5
Covering both hanging pieces.
45 ... Ka7 46 e5!
The shots keep coming. I like this better than the mundane 46 Qf5!, which
also wins.
.
46.. Qe3
After 46 ... Rxe5 47 Rd7+ Kb6 48 Ne4 Rxe4 49 Qxe4 Qxe4 50 fxe4 Bxb2
51 Rd8, White has a winning endgame.
47 Qc2 Rxe5 48 Rd7+ Kb6 49 Ne4
Black must give up the exchange or risk a near certain mate; e.g. 49 ... Be7
50 Qxc4 and Black’s king is fried.
49 ... Rxe4 50 fxe4 Be5
Or similarly 50 ... Bd4 51 Rxd4! Qxd4 52 Qf2! Kc5 53 Qxd4+ Kxd4 54
Kf3 with a winning pawn endgame, since 54 ... a5 55 a4 is zugzwang.
51 Qf2!
 Simplification.
51 ... Qxf2+ 52 Kxf2 Bxb2 53 Ke3 1-0
Since 53 ... Bxa3 54 Kd4 ends it.

Game 12
A.Giri-I.Nepomniachtchi
Wijk aan Zee 2011
Grünfeld Defence

If you listen to some of the social media know-it-alls, they try and convince
you that Dutch GM Anish Giri is this super-cautious, modern-day Karl
Schlechter, whose only goal is to avoid loss. If you actually bother to
examine Giri’s games and stats, it’s easy to ascertain this type-casting is a
nonsensical fiction. To my mind Giri is every bit as dynamic as any other top
ten player. His games are as exciting and as hard fought as any of his
colleagues’, and his draw ratio is no higher either.
1 d4 Nf6 2 c4 g6 3 Nc3 d5
In case you didn’t already know, Nepo is a Griinfeld devotee. The
Griinfeld suits his style perfectly, since it often leads to complicated and open
positions, in both of which he excels. The big question on my mind is: will he
dare to play it against Magnus in their upcoming match? I certainly hope so!
4 cxd5 Nxd5 5 e4 Nxc3 6 bxc3 Bg7 7 Nf3 c5 8 Be3 Bg4’?

This line is a touch offbeat from the usual 8 ... Qa5 or 8 ... 0-0.
9 Rcl

Question: White has an open b-file. Why would he post his rook on cl-rather
than bl?

Answer: bl is also possible, but keep in mind that c3 is White’s most vulnerable point, since ... Qa5
is almost certain to follow. Furthermore, 9 Rbl is met by 9 ... Nc6! and if 10 Rxb7 0-0, theory says that
Black receives full compensation for the sacrificed pawn.
 9 ... BxB
Nepo is willing to create an imbalance by handing over the bishop pair in
exchange for damaging White’s structure on the kingside.
10 gxf3 cxd4 11 cxd4 0-0 12 f4
A good move which suppresses Black’s ... e7-e5 ideas. Now White will
probably set up with Bg2 and e4-e5. 12 d5 has been played more often here.
12 ... e6
Restraining d4-d5 ideas.
13 Bg2 Nc6 14 e5
14 d5!? is risky: 14 ... exd5 15 exd5 Qa5+ 16 Qd2 Qxd2+ 17 Kxd2 Rfd8
18 Rhdl Nb4 19 Kel Nxa2 20 Rc7 a5! and it becomes a battle between
White’s passed d-pawn and Black’s passed a-pawn.
14 ... Qa5+!
Clearing d8 for a black rook.
15 Qd2
15 Kfl is unplayed, probably because after 15 ... Rfd8 16 Bxc6 bxc6 Black
looks just fine.
15 ... Qa6!
 It benefits Black to keep queens on the board.
16 Qe2 Qa5+ 17 Qd2 Qa6
This is effectively a draw offer.
18Bfl!?
Giri gives his opponent the silent treatment and declines.
.
18.. Qa4 19 Rc4 Qb5’?
Black can taunt with this move since White has no useful discovery with
his c4-rook.
20 Rc5
After 20 Be2 Ne7 21 Rc7 Qbl+ 22 Qcl Qb4+ White would probably have
to take the draw, since he may stand a shade worse if he risks 23 Kfl!? Rfc8.
20 ... Qbl+ 21 Ke2
This way Nepo’s irritating checks run out.
21 ... Qe4! 22 13
22 Rgl is a bit slow after 22 ... Rfd8 23 Rc4 Qd5 (not 23 ... Bxe5?? 24
Bg2) 24 Qd3 b5, when White has to liquidate with 25 Rxc6 Qxc6 26 Bg2
Qd7 27 Bxa8 Rxa8 28 Rcl in order not to be worse.
22 ... Qf5’?
Black might also insert 22 ... Qbl 23 Kf2 (threatening Rb5 and Rxb7; here
23 Bg2 Qb6 24 d5 Rad8!? 25 d6 g5! gets messy) 23 ... Rfd8!? (ignoring it) 24
Rb5 Qf5 and now if 25 h4 (threatening Bh3, whereas 25 Rxb7 Rac8 gives
Black a dangerous initiative for the pawn), then 25 ... Rxd4! 26 Bxd4 Rd8 27
Rxb7 Bh6 28 Bd3 Rxd4 29 Bxf5 Rxd2+ 30 Ke3 Rxa2 and I don’t believe
Black can lose.
23 h4!

Threatening Bh3-g4 and f4-f5, hunting down Black’s queen.

23 ... Qh5’?

Question: Isn’t Nepomniachtchi in grave danger of either losing his queen

or getting it buried?
 Answer: One must be an incredibly confident calculator to play such a move, since Black’s queen is
indeed close to being trapped on h5. But when it comes to queen traps, “close” doesn’t cut it. All the
queen needs is one safe square and Nepo sees that she has it.
After 23 ... Qbl 24 Bg2 Qb6 25 d5 Qa6+ 26 Kf2 exd5 27 Bfl Qa3 28
Rxd5 Rad8 29 Bc4,1 slightly prefer White.
24 Bf2
White should prefer 24 d5! (eliminating a potential target pawn) 24 ...
exd5 25 Rxd5 Rad8 26 Kf2 and looks a touch better.
24 ... Rfd8
Here 24 ... Ne7! 25 Bh3 Rac8 is more accurate.
25 Bh3!
Now the Bg4, h4-h5 idea is in play.
25 ... Bf8!

Various elements slide into place with this triple-purpose move:

1. It leaves a loophole for Black’s queen to escape via h6 and g7.
2. Black gains a tempo on White’s rook.
 3. In some instances Black may be able to engineer ... Bc5, adding
unbearable pressure to White’s d-pawn.

Question: What is the more potent factor: Black’s queen in danger or

White’s centre in danger?

Answer: A bit of both. The engine assesses this as currently even.

26 Bg4 Qh6 27 Rc4 Rd5

Question: Why isn’t Nepomniachtchi occupying the hole by transferring

his knight here via e7?

Answer: He prefers to keep his knight on c6 to pressure d4.

28 h5 Rad8
There is no need to move the queen as long as White’s hl-rook is
unprotected.
29 Rh3?
29 Qc3! was necessary, with even chances.
 Exercise (combination alert): White’s position contains a hidden design flaw.
How should Black continue?

Answer: Moving the bishop to c5 adds intense pressure on White’s centre.

29 ... Bc5!
The changeling turns from defender to attacker in a single move.
30 hxg6 Qxg6 31 f5!?
Unsound, yet completely understandable. Giri is desperate to try and
confuse the issue, since with normal play his d4-pawn falls and with it his
position. 31 Qc2 is objectively better, but still losing for White: 31 ... Qxc2+
32 Rxc2 Bxd4 33 Bh4 R8d7 with excellent chances to convert.
31 ... exf5 32 Bh5 Nxd4+!
Zwischenzug.
33 Kfl
Not 33 Bxd4? Qg2+ or 33 Rxd4? Qa6+! and White heavy material.
 Exercise (combination alert): What is Black’s only winning move?

Answer: Discovered attack.

33 ... Nxf3!
Nepo is in his element, with a tactical position requiring engine-level
calculation skills. He doesn’t fall for the trap 33 ... Qa6?, which is tempting
but fails to 34 Kg2! Nxf3 (not 34 ... Qxc4?? 35 Bxf7+ Kxf7 36 Rxh7+ Ke8
37 Qh6 and wins) 35 Qf4! Nxe5 36 Rxc5, when White doesn’t stand worse.
34 Qxd5 Rxd5 35 Bxf3
35 Bxg6 Nd2+! 36 Kg2 Nxc4 37 Bxh7+ Kg7 38 Bxc5 Rxc5 39 Bxf5 Nxe5
leaves Black two clean pawns up.
35 ... Bxf2! 36 Bxd5 Qgl+ 37 Ke2 Qel+ 38 Kf3
38 Kd3 Qfl+ picks off the h3-rook.
 Exercise (combination alert): Black has two ways to win. Find one of them.

Answer #1:38 ... Qfl! 0-1

In view of 39 Rc8+ Kg7 40 Bc4 (or 40 Rh5 Bg 1+! 41 Kg3 Qf2+ 42 Kh3
Qh2 mate) 40 ... Qxh3+ 41 Kxf2 Qh2+ 42 Kfl Qxe5 and White is completely
lost.
Answer #2:38 ... Qe3+! 39 Kg2 Qd2! also wins; e.g. 40 Rc8+ Kg7 41 Bxb7 Bb6+ 42 Khl Qdl+ 43
Kg2 Qgl+ 44 Kf3 Qf2 mate.

Magnus and Nepo were barely out of their teens during the next game, yet
Magnus was already over 2800, the youngest player ever to reach such
heights. Despite being the underdog, Nepo proved that he wasn’t intimidated
by the player everyone knew would soon be world champion.

Game 13
Ma.Carlsen-I.Nepomniachtchi
Wijk aan Zee 2011
Sicilian Defence

1 e4 c5 2 M3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Be2 e5

This true Najdorf move is played more by a 3:1 ratio over 6 ... e6.
7 Nb3
7 N13 is a lesser seen sideline.
7 ... Be7 8 0-0 0-0 9 Khl
The immediate 9 Be3 is White’s main move.
. Nc6
9..
9 ... b6 and 9 ... Be6 are also played here.
10f4b5 11Be3

There’s no need to defend the e4-pawn just yet, since 11 ... b4?! 12 Nd5
Nxe4? is an unwise pawn grab: 13 B13 f5 14 Nb6! wins the exchange, due to
the dual threats of Nxa8 and Qd5+.
11 ... Bb7 12 a4
All still theory.
12 ... exf4 13 Rxf4 Ne5 14 Qd4

Question: Why didn’t Magnus win a pawn by simply capturing on b5?

Answer: 14 axb5 axb5 15 Bxb5 only wins the pawn temporarily, since 15 ... Ng6! regains it at once
with at least equality.
14 ... Nc6 15 Qd2 Ne5 16 Qd4 Nc6 17 Qd2 Ne5
Now White has a draw if he wants one, but of course Carlsen doesn’t, so
he has to think of something else to do.
18 axb5
No draw!
18 ... axb5

19 Rel
 The first new move. Whether the players knew it or not, 19 Rxa8 Qxa8 20
Qd4 Nc4 21 Bxc4 bxc4 22 Qxc4 d5 23 Qc7 Bd8 24 Qd6 Re8 25 Rfl dxe4 led
to a balanced game in Mberggren-Juggernaut, playchess.com 2007.
19 ... Ng6! 20 Rffl?!
The rook should have gone forwards: 20 Rf5!, when 20 ... Nxe4 (here 20
... b4 21 Nd5 Nxe4 can be met by 22 Qxb4! Bxd5 23 Rxd5 Nf6 24 Rb5 and I
prefer White’s position) 21 Nxe4 Bxe4 22 Rxb5 Ne5 is about equal.
.
20.. b4!
Stronger than 20 ... Nxe4 21 Nxe4 Bxe4 22 Bxb5 with a level game.
21 Nd5 Nxe4 22 Nxe7+
It’s tempting to grab the bishop pair, especially in an open position. In this
instance it allows Black to seize the initiative. 22 Qd4 was probably better,
though after 22 ... Bh4! 23 Rdl (not 23 Qxe4? Bxel 24 Rxel Re8 and Black
wins a piece, leaving White the exchange down) 23 ... Re8 24 BI3 Nf6, even
here the engines prefer Black.
22 ... Qxe7 23 Qxb4 Nh4!
The target is g2.
24 Bf3
An unfortunate necessity. Not 24 Rgl?? Nf5 25 Bb6 Qh4 (threatening
mate on g3) 26 Rgfl Neg3+ (discovered attack) 27 Kgl Qxb4 and wins.
24 ... Nxf3 25 gxf3?’
Here 25 Rxf3 Qd7 26 Nd4 Rfe8 keeps Black’s edge to a minimum.
 Question: Why did Carlsen go for self-inflicted damage to his kingside pawns?

Answer: This would be a brilliant idea, if it actually worked! Magnus is really asking for it in this
game and his move is based on a tactical miscalculation.

Exercise (combination alert): What is Black’s strongest move?

Answer: Leave the knight hanging on e4 and move the queen to d7.
25 ... Qd7!
Magnus admitted he had overlooked this move. And otherwise Black is in
trouble:
a) 25 ... Nf6?? 26 Bg5 Qd7 27 Bxf6 gxf6 28 Qf4 Qd8 29 Rgl+ Kh8 30
Nd4 and White has generated a winning attack.
b) 25 ... Nc5?? 26 Bxc5 dxc5 27 Rxe7 Bxf3+ 28 Rxf3 cxb4 29 Nd4 with
an extra piece.
26 Bf4
 The only move. Nepo’s knight is untouchable: 26 fxe4?? runs into 26 ...
Ra4! and White must hand over his queen, since 27 Qd2 Bxe4+ 28 Kgl
allows 28 ... Qg4+ 29 Kf2 Qg2 mate.
26 ... Ra4 27 Qb6 Nf6 28 Qxd6

Magnus has won a pawn. That is where his good news ends, since Black
has a dangerous attack.
.
28.. Qg4?’
A smart-looking move, but inaccurate! The correct continuation was 28 ...
Qh3! 29 Nd4 Rc8 30 c3 h6 31 Bg3 Nh5 32 Kgl Ra6! and White is busted.
29 Nd4 Rxd4 30 Qxd4 Bxf3+ 31 Rxf3 Qxf3+ 32 Kgl Qg4+ 33 Khl
White should be okay here, despite his exposed king, since Black just
doesn’t have enough fire-power to make a difference.
33 ... Qc8’?
A psychologically telling moment. Nepo could take perpetual check and
score a draw with the black pieces, yet now opts to play for the full point.
The conclusion: he is one of the few players in the world who isn’t
intimidated by Magnus!
34 Qf2?’
This allows Black to reclaim attacking chances. After 34 Bg5! Qc6+ 35
Kgl Nh5 36 Be7 Rc8 37 c4! Qxc4 38 Qxc4 Rxc4 39 b4, White has full
compensation for the pawn with the superior minor piece and dangerous b-
pawn.
34 ... Qb7+ 35 Kgl Ne4!
35 ... Qxb2 36 Be5 Qb4 37 Bxf6 gxf6 is equal and most likely drawn.
36 Qd4 Re8 37 Re2

Exercise (planning): Come up with an attacking plan for Black.

Answer: Step 1: Create luft for his king.

At first it may feel as if White’s position is invulnerable to criticism. When
we look deeper, we see Black’s coming ... Re6 and ... f7-f5. Eventually,
White’s king has a good chance of coming under attack from either ... Rg6
or, if circumstances allow,... g7-g5.
37.. . h6!
Now Black’s rook can be lifted without fear of a back rank mate.
38 h3 Re6!
Step 2.
39 Kh2
White needs counterplay, so he should really play 39 c4!.
39.. . f5
Step 3. Though 39 ... Kh7 first was more precise.
40 b4?!

Question: Aren’t White’s queenside pawns enough to generate sufficient

counterplay?

Answer: The queenside pawns are synthetic happiness, for two reasons:
1. White simply doesn’t have the time to push them now.
2. They are far more vulnerable to capture than they look.
Instead, he might have disrupted Black’s plan with 40 Qc4! Qf7 41 Qd5
g5 42 Bg3, when it would be much harder for him to make progress.
40 ... Kh7!
Now that chance has gone.
41 Re3
After 41 Rg2 g5 White gets pushed about; e.g. 42 Be5 Qe7 43 Bb8 Qe8 44
Bc7 Re7 45 Ba5 Rd7 46 Qal Qb8+ 47 Khl Qf4 48 Bb6 Qf3 49 Kh2 g4 50
hxg4 Ng5! with a winning attack.
41 ... Rg6!
 Step 4. Nepo’s slow motion attack begins to pick up steam.
42 Re2 Qb5 43 Rel Rc6
43 ... Qa4 may be even stronger; e.g. 44 Qb2 Rc6 45 Re2 Rc3!.
44 Rxe4!
This is White’s best practical chance as it forces Black into a technical
ending, but the position is still lost, despite Magnus’ spirited protests to the
contrary.
After 44 c3 Rg6 (not 44 ... Nxc3? 45 Re7 Rg6 46 Qxc3 Qfl 47 Bg3 f4 48
Rxg7+! Rxg7 49 Qc2+ with perpetual check) 45 Qe3 Qa6 White is quickly
running out of moves: 46 Qf3 Qa4 47 Re2 Qdl 48 Qe3 Qfl and White is
busted. If 49 b5? then 49 ... Rgl! 50 Qxgl Qxf4+ 51 Kg2 Qg3+ 52 Kfl Qf3+
wins.
44 ... fxe4 45 Qxe4+ Rg6 46 Bg3 Qd7 47 h4
Threatening h4-h5.
47 ... h5 48 c4 Qd2+ 49 Kh3
 Exercise (critical decision): Is 49 ... Qxb4 Black’s best move? If not, then what
should he play instead?

Answer: Zwischenzug.
.49.. Qc3!
White’s b-pawn is lost, without hanging the h5-pawn. Whereas after 49 ...
Qxb4 50 Qd5! Black must hand over the h5-pawn, or face perpetual check.
50 Qf4 Qxb4 51 Qf5 Qxc4 52 Qxh5+ Rh6 53 QB
If 53 Qf5+ Kh8, there is no perpetual check since Black can block with his
queen on g8.
53 ... Qe6+ 54 Kh2 Rf6 55 Qd3+ Rf5 56 Qc2 Qd5 57 Bf2 Kh6 58 Be3+
Kg6
Black must make threats against the white king, while simultaneously
guarding his own king against checks.
59 Bf2 Kf6 60 Bg3 Rfl
 Threatening ... Qhl mate.
61Bf2
Swapping queens with 61 Qg2 Qxg2+ 62 Kxg2 leads to a technical win
for Black, whose king will eventually infiltrate to g4.
.
61.. Rdl

We see the gradual replacement of squares which previously belonged to

White.
62 Qc3+
This allows queens to be removed from the board, which is the equivalent
to resignation. But a queen swap couldn’t be avoided forever; for example,
after 62 Bel!? (the bishop is tactically immune due to the queen check on 12)
62 ... Rd3 63 Q12+ Kg6 64 Qe2 Q13 65 Qb2 Kh6 66 Qb6+ Kh7 and White’s
checks have run out.
62 ... Qe5+ 63 Qxe5+
The exchange of queens squeezes the position dry of any hope for White.
 63 ... Kxe5 64 h5
Or 64 Kg3 Kf5 65 Bb6 Rd3+ 66 Kg2 Kg4 67 Bf2 Rd2 68 Kgl Rxf2 and
Black wins the pawn ending.
64 ... Kf6 65 Bh4+ Kf5 66 Be7 Rd7 0-1
Since 67 Bc5 Kg5 picks off the h-pawn.

The next game is an example of two factors:

1. Nepo’s tendency towards recklessness.
2. How difficult it is to beat him, even when his position stinks!

Game 14
N epomniachtchi-E .N aj er
I.
Russian Rapid Championship, Olginka 2011
Scotch Game

1 e4 e5 2 Nf3 Nc6 3 d4 exd4 4 Nxd4 Bb4+

Question: Doesn’t this lose a tempo?

Answer: Yes and no. Black wants to provoke c2-c3, so that the bl-knight is denied that square. The
most commonly played lines are 4 ... Bc5 and 4 ... Nf6.
5 c3 Bc5 6 Be3 Bb6
Avoiding 6 ... Nf6?? 7 Nxc6 bxc6 8 Bxc5.
7Qg4!?
 Question: Isn’t this a gross violation of principle, to bring the queen out early
in the opening to go pawn hunting?

Answer: It’s a violation, but I wouldn’t label it as “gross”, since it’s an exception to the principle
and fully playable. Kasparov first uncorked this sharp idea against GM Wolfgang Unzicker. White’s
most popular choices are 7 Bc4 and 7 Nf5.
7 ... Nf6
The most aggressive reply. Black sacrifices a pawn for a few tempi.
Instead:
a) 7 ... Qf6 8 Qg3 Nxd4?! 9 cxd4 Bxd4?! 10 Bxd4 Qxd4 11 Nc3 was
G.Kasparov-W.Unzicker, Zürich 2001, where Black was unable to withstand
White’s lead in development.
b) 7 ... g6 doesn’t seem to equalize either. For example: 8 Nd2 Nf6 9 Qh4
Nh5 10 Qxd8+ Nxd8 11 Nc4 d6 12 Nb5 Ne6 13 Nxb6 axb6 14 Bc4 and
White is better with the bishop pair and more space, I. Salgado Lopez-
L.Fressinet, European Championship, Aix les Bains 2011.
8 Qxg7 Rg8 9 Qh6 d5!
An interesting and strong near novelty, having only been played in a 2009
email game.
10 Nxc6
10 Bb5 is the main move today. D.Cawdery-L.Aronian, FIDE World Cup,
Tbilisi 2017, continued 10 ... Bd7 11 Bxc6 bxc6 12 e5 Rg6 13 Qh4 Ng4 14
Qxd8+ Rxd8 15 0-0? (15 Bf4 was necessary) 15 ... Nxe5 (15 ... Nxe3 looks
good too) 16 Ne2 Bxe3 17 fxe3 Nc4 and White landed in a difficult position.
10 ... bxc6 11 e5?!

Nepo recklessly allows his development deficit to increase with this

dubious move. 11 Nd2 keeps White even.
11 ... Rg6! 12 Qh4
Even worse is 12 Qf4? Nh5 13 Qa4 (13 Qf3?? Bg4 traps the queen) 13 ...
Bxe3 14 fxe3 Qg5 and White is in trouble.
 12 ... Bxe3 13 fxe3
Nepo is forced to accept a wretched pawn structure, as after 13 exf6??
Rxf6 14 fxe3? Rxfl+ White loses his queen again.
13 ... Ng4 14 Qxd8+
If Black could have kept queens on the board, White would be dead lost,
too far behind in development. As it is Black has to be content with winning
a pawn back in a slightly better position.
14 ... Kxd8 15 Kd2?’
I read about a Kentucky judge from the 1950s who would routinely arrive
in his court room drunk and then, just to show how much he respected the
law, promptly fine himself for public drunkenness. There is innocent
violation of the law and there is the wilful sort, which we see here. White’s
position becomes increasingly unstable and Nepo’s opening experiment has
clearly ended in disaster. His only “developed” piece is his king!
White’s best chance for survival was something like 15 Bd3 Rg5 16 Nd2
Nxe5 17 Bfl Ng4 18 e4 dxe4 19 h3 Ne3 20 Kf2 Nxfl 21 Rhxfl f5 22 g4 Rb8
23 b3 fxg4 24 hxg4 Rbb5 25 Nxe4 Rxg4 26 Ke3, when he still has chances to
hold the draw.
15 ... Rb8 16 b3 Nxe5 17c4’’
 “Those who don’t believe in magic, will never find it,” wrote Roald Dahl.
From this wretched position, Nepo finds a shocking/brilliant defensive idea.
This pawn sacrifice allows White to develop his queenside, while ruining
Black’s structure.
.
17.. dxc4!
Acceptance is Black’s optimal path. After 17 ... Bf5 18 cxd5 Be4 19 dxc6
Bxg2 20 Bxg2 Rxg2+ 21 Kc3 Rb6 22 Nd2 Rxc6+ 23 Nc4 White has
reasonable chances to hold the draw.
18 Kc3!

Question: What is Nepomniachtchi doing, giving away pawns in an endgame,

with zero development?

Answer: He isn’t investing in a lost cause. This sacrificial idea is undoubtedly extracted from his
vast Grünfeld experience, where Black plays a similar sacrifice. But that one is in the opening while
this one is in an ending! Here are the points behind Nepo’s enterprising defensive sacrifice:
 1. If Black declines, White will be given time for Nd2 and Nxc4.
2. If Black accepts, then after a2xb3 White’s al-rook is freed, attacking
Black’s a7-pawn. On top of which, Black’s structure is suddenly one with
five isolanis.
18 ... cxb3 19 axb3
Black’s pawn structure now resembles a tattered rag doll.
.
19.. Ng4!
A reminder to White that he is in a state of zero development and his king
may not be so safe.
20 Bd3
A white piece is finally developed!
20 ... Rg5
More accurate is 20 ... Rh6 21 Nd2 Nxe3 22 Rxa7 Be6 23 Be4 Nd5+,
when White is still fighting for his life.
21 Rel!
Nepomniachtchi correctly protects his e-pawn, while offering his h2-pawn.
21 ... Nxh2
After 21 ... Nf2 22 Bxh7 Rxg2 23 Nd2 Rxh2 24 Bc2 Rb5 25 Rxa7, White
should hold the draw.
22 Nd2!
 It is hard to find another game where one side sacrifices so many pawns in
an ending, just for piece activity. A few moves earlier, Nepo had zero pieces
developed, compared to his opponent’s entire army. Just a short while later, it
is White who is technically ahead in development.
22 ... Rxg2 23 Be4 Rg3!
Adding heat to e3.
24 Nc4

Question: Nepomniachtchi could capture either black a or h-pawn. Why does

he refrain?

Answer: There is no rush. Even though White is three pawns down, it is certain that a few of
Black’s pawns will soon fall. The idea of his last move is actually to launch an attack (!) against
Black’s king with Ne5, hitting both c6- and f7-pawns.
24 ... Ng4
Bringing the knight back towards the centre, increasing the pressure on e3,
and preventing Ne5, so Nepo decides to take a pawn after all.
25 Rxa7
Stronger than 25 Bxh7? Nf6 26 Bc2 a6 27 Ne5 Nd5+ 28 Kb2 Nb4 and
Black has serious winning chances. White doesn’t have time for 29 Nxf7+??
due to 29 ... Ke8 30 Ne5 Rg2 31 Rael Rb5! 32 Nf3 Nd3+, winning the
exchange.
25 ... Be6! 26 Na5?’

This tempting move is not White’s best but is at least tricky; whereas the
simple 26 Bxc6 Bxc4 27 Rdl+ Ke7 28 Rxc7+ Kf6 29 bxc4 Nxe3 30 Rd3 h5
leaves Black with straightforward chances.
26 ... Kc8?
This mistake, likely due to time trouble, throws away almost all of Black’s
advantage. He should have played 26 ... Kd7! 27 Bxc6+ Kd6 28 Be4 Nxe3
with a probably winning position.
27 Nxc6 Nf6?’
 Another inaccuracy which allows White full equality. The idea is correct,
yet Black’s move order is off. He should move the knight the up the board,
rather than down, and play 27 ... Nf2! which covers against both Rdl and
Bhl ideas.
28 Bhl! Rxb3+ 29 Kc2 Nd7?
Now Black’s king is in grave danger. After 29 ... Kd7! 30 Rdl+ Bd5! 31
Rxd5+ Nxd5 32 Kxb3 Nxe3 the game is headed for a draw.
30 Ra8+! Nb8??
We all live in regret for the remainder of our lives when we recall our past
hideous blunders. Here Najer had to give up material with 30 ... Rb8! 31
Nxb8 Nb6! 32 Na6+ Nxa8 33 Bxa8, when Black still has a prayer to try and
hold the draw.

Exercise (combination alert): What did Black overlook on his last move?

Answer: Mating net.

 31 Rdl! 1-0
Threatening two separate mates in one with Rd8 or Ne7. All of Black’s
defences fail; e.g. 31 ... Kb7 32 Rxb8+ Ka6 33 Ral+ and mates, while 31 ...
Bd7 32 Kxb3 leaves White a rook up. Nepo pulled off a defensive/counter
attacking miracle.

Just because Nepo is an incredibly gifted tactician and attacker doesn’t

disqualify him from strategic power. In the following game he dismantles one
of the great strategists of his era, with almost no tactics or fireworks of any
kind.

Game 15
N epomniachtchi-G.Kamsky
I.
FIDE World Cup, Khanty-Mansiysk 2011
English Opening

1 c4 c5 2 Nf3 Nf6 3 Nc3 d5 4 cxd5 Nxd5 5 g3 g6 6 Bg2 Bg7 7 h4’?

 The trendy push of the h-pawn is that poorly lit subdivision of the military
industrial complex. Sometimes I wonder why we waste time with inferior
first moves like 1 e4 or 1 d4, when we can open with 1 h4!!.

Question: Is this a natural continuation or an aberration? Why would

Nepomniachtchi push the h-pawn before completing development?

Answer: The advance of the h-pawn actually makes sense here since Black’s knight is no longer on
f6, which means that he must react to White’s potential h4-h5.1 checked the database and was surprised
to find that 7 h4 is the highest scoring line, with an impressive 73% for White. The main move 7 0-0
lags way behind in the low fifties.
7 ... Nc6
Kamsky ignores the h4-h5 possibility. I would probably have gone for 7 ...
h6, intending to bypass with ... g6-g5 if White pushes the h-pawn further.
8 h5 Nxc3
The engine says White only has an edge, yet White’s score jumps to
91.7% after this move. 8 ... Bf5 is preferable, as in I.Nepomniachtchi-P.Leko,
Beijing (blitz) 2013, though Nepo won anyway.
9 bxc3
Principle: Capture towards the centre. Oddly enough the engine prefers
the dove-like 9 dxc3!?.
9 ... Bf5
A novelty, and also Fat Fritz's top choice. Black has tried several moves
here, of which 9 ... Qc7 has been seen most often.
10 Qb3
Going after b7.
.
10.. Qb6
Kamsky probably didn’t like 10 ... Qd7 11 Ng5, which induces Black to
castle kingside. It actually looks okay to me, despite the coming opening of
the h-file.
11 Ng5 Qxb3?’
A serious concession since it benefits White’s structure. Black should
probably risk 11 ... 0-0 even now.
12 axb3
 Advantage Nepo, for the following reasons:
1. White has a single pawn island to Black’s two.
2. White exerts pressure down the hl-a8 diagonal, and Black must watch
out for an opportune Bxc6+.
3. White’s rook is conveniently placed on the open a-file.
12 ... 0-0 13 hxg6
13 h6 Bh8 14 Bxc6 bxc6 15 e4 Bd7 16 Ba3 Rfb8 17 Bxc5 Rxb3 18 f4 is
also a pleasantly favourable ending for White.
13 ... hxg6 14 Ne4 Bxe4
This gives up both the bishop pair and light square control. 14 ... Rac8
might be preferred, though it still looks highly favourable for White after 15
Nxc5 Nb4 16 Ra4 Nc2+ 17 Kdl Rxc5 18 g4, winning the piece back with
advantage.
15 Bxe4 Rfc8 16 e3
So that his king gets access to a nook on e2.
 16 ... Rc7 17 Ke2 a6 18 g4
Nepo begins to expand.
18 ... e6 19 Ba3 Na7! 20 f4 Nb5 21 Bb2 Nd6 22 Bf3 Rd8 23 Ra2 c4’?

Kamsky hopes to generate some activity in mid-board, but I don’t like it.

Question: Why not?

Answer: The strategic cost is to allow the swap of Black’s c-pawn for White’s less valuable b-
pawn. The move also opens the b-file for White’s rooks. 23 ... f5 looks a more resilient choice.
24 bxc4 Nxc4 25 Bal
White’s centre is about to roll forward.
25 ... Nd6 26 Rbl Bf6
Black’s position seems to be getting worse and worse. Maybe it is time for
radical measures with 26 ... g5!? 27 fxg5 Kh7, intending ... Kg6.
27 g5 Bg7 28 Rb6!
The pressure continues to increase.
 28 ... Rdd7 29 Rb3
Not really necessary. He could also go for the immediate 29 d4 and if 29 ...
Nc4 then 30 Rbl.
29 ... Re7 30 d4!

Question: Doesn’t this create a hole on c4?

Answer:Nepo correctly gauges that the hole is outweighed by White’s monster centre. In fact, it’s
really a fake hole since White can seize control over the c4-square with Kd3 and Rb4.
30 ... Nc4 31 Kd3 Nd6 32 e4 Red7 33 e5!
Good judgment. Nepo realizes that control of f5 is of little benefit to
Black.
33 ... Nc8
Depressing, but this is the only way to hang on to his queenside pawns. 33
... Nf5 34 Rab2 b5 35 Ra2! Ra7 36 Rba3 wins the a-pawn.
34 c4
 Here they come.
34 ... Ne7 35 Bc3 b5’?

Desperation. Kamsky hopes that control over d5 will afford him some
measure of compensation.
36 cxb5 axb5 37 Ra8+ Rc8 38 Rxc8+ Nxc8 39 Rxb5 Bf8
A mistake in an already hopeless position. 39 ... Kh7 stretches the game
out, though the result is not in doubt. Here is an example of White’s
technique: 40 Rb8 Ne7 41 Bb4 Nd5 42 Bxd5 Rxd5 43 Bc5 Rd7 44 Kc4 Rd5
45 Kb5! Rd7 46 Kc6 Rd5 47 Bb6! Bh8 48 Rf8 Kg7 49 Rd8! Rxd8 50 Bxd8
Kf8 51 Bf6 Bg7 52 d5 exd5 53 Kxd5 Kg8 54 Bxg7 Kxg7 55 e6 Kf8 56 Kd6
fxe6 (or 56 ... Ke8 57 e7 with zugzwang) 57 Kxe6 Kg7 58 Ke7 Kg8 59 Kf6
Kh7 60 Kf7 and wins.
40 Rb8 Rc7
Or 40 ... Ne7 41 Bb4 Rc7 42 Bb7! etc, as in the game.
41 Bb7! 1-0
 Kamsky’s forlorn position is given up for dead, as the win doesn’t require
laborious effort: 41 ... Ne7 (41 ... Nb6 loses to 42 Ba5) 42 Bb4 Rd7 43 Bc5
Rd7 (or 43 ... Kg7 44 Kc4 followed by 45 Kb5 etc) 43 Bc5 Rc7 44 Kc4 Rd7
45 Kb5 Re7 46 Kb6 Rd7 47 Bc6! and wins.

Game 16
N epomniachtchi-R.Ponomariov
I.
ACP Cup, Riga (rapid) 2013
Queen’s Gambit Declined

1 c4 e6 2 Nc3 d5 3 d4 Be7

Question: This move feels a bit cart-before-the-horseish. Why bring the

bishop out to e7 when 3 ... Nf6 is available?

Answer: In this version if White plays the popular Exchange Variation, he is denied the normal
Bg5. If White temporizes with 4 Nf3, he rules out Exchange lines where the knight is developed to e2
in conjunction with a later 12-13.
4 cxd5
The major alternative is 4 Nf3 Nf6 5 Bg5 or 5 Bf4.
4 ... exd5 5 Bf4
As mentioned above, 5 Nf3 allows Black equality since the Exchange
loses a lot of its dynamism when White develops the knight to f3.
. c6
5..
If you don’t mind an isolated queen’s pawn, then 5 ... Nf6 6 e3 0-0 7 Bd3
c5 is another way to play.
6e3
6 Qc2 prevents ... Bf5 for right now, but Black can respond with 6 ... g6 7
e3 Bf5 8 Qd2 Nf6 9 f3 h5! and I doubt that White has much, if anything.
6 ... Bf5
Black logically occupies the bl-h7 diagonal before White plays Bd3.
7g4!?
 It was Jane Austen who pointed out that manners are the glue which holds
society together. If you aren’t familiar with this line, then get prepared for a
Queen’s Gambit Exchange version like you have never seen it before.
7 ... Be6
7 ... Bg6?! is considered inaccurate due to 8 h4! and Black must push the
h-pawn, one or two squares, since 8 ... Bxh4?? loses to the trap 9 Qb3! b6 10
Rxh4! Qxh4 11 Nxd5!.
8 h4
Simply 8 h3 is a major alternative.
8 ... Nd7
8 ... Bxh4 is a line governed by both unpredictability and violence: 9 Qb3
g5 10 Be5 f6 11 Bh2 Bxg4 12 Qxb7 Qe7 13 Qxa8 Qxe3+ 14 Be2 Bxf2+ 15
Kfl Bh4 is equal, Ding Loren-W.So, Wijk aan Zee 2015. Yes, this is an
actual theoretical position.
9g5
Nepo might be trying to mess with his opponent’s head, since Ponomariov
had won with this move twice earlier the same year. 9 h5 is the alternative.
9.. . h6
Otherwise Black is unable to develop the g8-knight or the h8-rook.
10 g6!
This pawn offer is White’s main idea. Black would love 10 gxh6?! Nxh6.
10.. . Ngf6
Instead, 10 ... fxg6 11 Bd3 Ngf6?! (11 ... Bf7 was necessary) 12 Bxg6+
Bf7 13 Bxf7+ Kxf7 14 Nf3 gave White a clear advantage in R.Ponomariov-
G.Sargissian, World Blitz Championship, Khanty-Mansiysk 2013.
11 gxf7+ Bxf7 12 Bd3 Ne4’?
12 ... Nh5 is normal here.
 13 Bxe4’?
A theoretical novelty and almost certainly home preparation from Nepo,
who deviates from 13 Nxe4 dxe4 14 Bxe4 Nf6 15 Bf3 0-0, as in
R.Ponomariov-A.Riazantsev, World Rapid Championship, Khanty-Mansiysk
2013, where Black got full compensation for the sacrificed pawn and the
engine even slightly prefers his side.

Question: Doesn’t swapping off the light-squared bishop hand Black the
light squares?

Answer: Somewhat, yes. But Nepomniachtchi’s idea is to not capture the e4-pawn, leaving it in
place to lessen the influence of Black’s light-squared bishop. He envisions the set-up Nge2, Qc2, 0-0-0
and Rgl, in some order or another.
13 ... dxe4 14 Nge2 0-0

Question: Is this a case of castling into it?

 Answer: Black is sure to come under attack on the g-file. But he just doesn’t have time to castle
queenside. Nor will it be safe for his king if he does.
If instead 14 ... Bxh4 15 Nxe4! (now he will take the e-pawn, threatening
Nd6+) 15 ... Be7 16 Qc2 Nf6 17 N4c3, White has dangerous attacking
chances.
15 Qc2 N16?!
This natural-looking move is inaccurate and allows White an attack. After
15 ... Qe8 16 Rgl Kh8 17 Qxe4 White is perhaps only a shade better.
16 Rgl
Threatening Bxh6.
16 ... Kh8 17 Be5! Qa5?’
The engine gives the line 17 ... Bd6 18 0-0-0 Bxe5 19 dxe5 Nd5 20 Qxe4
Bh5 21 Nf4! Rxf4 22 exf4 Bxdl 23 Rxdl Qxh4 24 Nxd5 cxd5 25 Rxd5 Qxf2
26 Rdl Qc5+ 27 Kbl Qc6. I don’t much like Black’s position, but it’s better
than the one Ponomariov got in the game.
18NÍ4
Black is in deep trouble after 18 0-0-0! Bxa2 19 Nf4 Bf7 20 Qbl!,
intending to double rooks on the open g-file.
18 ... Bb4!
Ponomariov fights back and begins to generate some counterplay.
19 Bxf6
Here 19 0-0-0!? Bxc3 20 Qxc3 Qxc3+ 21 bxc3 is just a slightly inferior
ending for Black.
19 ... gxf6 20 Qxe4!
 .20.. Rg8!
Not 20 ... Bxc3+? 21 bxc3 Qxc3+ 22 Ke2 Qc4+ 23 Kf3! with a winning
position.
21 Rxg8+ Rxg8 22 Kd2! f5?’
There was no reason to weaken. 22 ... Bd6 is better.
23 Qf3!?
A Nepomniachtchi trait: attacks take precedence over a favourable ending,
which he could have reached with 23 Qe5+ Qxe5 24 dxe5 Re8 25 Nd3.
23 ... Bc4?’
White’s queen should not be allowed free access to h5. After 23 ... Bd6 24
Nd3 White’s advantage is less than in the game.
24 Qh5
Of course Nepo is happy to slide his queen closer to Black’s king, while
threatening mate on the move.
24 ... Kh7
 Exercise (combination alert): What is White’s strongest attacking continuation?

Answer: Interference.
25 d5!
Black’s queen is cut off from the f5-pawn.
25 ... Bxd5
Or 25 ... Bxc3+ 26 bxc3 Qa3 27 Qxf5+ Kh8 28 Qe5+ Kh7 29 Rcl! (now
Qe4+ and Qxc4 is a real threat) 29 ... Bxd5 30 Nh5 Rg2 31 Rc2! and Black is
busted. If 31 ... Rxf2+? then 32 Kel Rf7 33 Nf6+ wins.
26 Qxf5+ Kh8 27 Ng6+!
From this point on the knight’s path is as sinuous as the inner chambers of
an Egyptian Pyramid.
27 ... Kg7 28 Ne7! Bxc3+
Nothing else is any better:
a) 28 ... Rd8 29 Kc2 and with Rgl+ coming Black is losing.
 b) 28 ... Bxe7 29 Rgl+ and Black can resign, since 29 ... Kh8 is met by 30
Qe5+ Kh7 31 Qxe7+ Kh8 32 Qf6+ Kh7 33 Qf5+ Kh8 34 Rxg8+ Kxg8 35
Qg6+, when the h-pawn falls with check.
29 bxc3 Rd8 30 Qe5+! Kf7

Exercise (combination alert): Black threatens terrible discovered checks with

his bishop and it seems like White is in deep trouble. Look closely and you
may find Nepomniachtchi’s stunning response.

Answer: Move the knight to f5 and allow Black the discovered check!
31 Nf5!!
White’s knight is in the right place at the right time. The amazing idea is to
meet any of Black’s discovered checks with Nd6 counter-check, with a
discovered attack of his own on Black’s then loose queen on a5.
Not 31 Nc8? Qa3! 32 Kc2 Qa4+ with a draw, or 31 Kel?? Re8! 32 Nxc6
bxc6 33 Qf5+ Ke7 34 Qe5+ Kd8 35 Qd6+ Kc8, when White has run out of
checks and is losing.
31 ... Rd7
Disasters occur rarely and nobody believes “rarely” will happen to them.
Unbelievably, nothing works for Black. If 31 ... Be6+ (the discovered check
with White’s queen hanging is the extent of Black’s good news; now comes
the bad) 32 Nd6+ (it is Black, not White, who loses material via a discovered
check) 32 ... Rxd6+ (otherwise Black loses his queen) 33 Qxd6, White will
convert, up the exchange.
32 Nd6+ Kf8 33 Qe8+ 1-0
Black’s rook hangs with check next move. Even worse, his king is soon
mated.

How is it possible to defeat Vassily Ivanchuk in 20 moves? Just keep

watching.

Game 17
N epomniachtchi-V.Ivanchuk
I.
SportAccord Rapid, Beijing 2013
Scotch Game

1 e4 e5 2 Nf3 Nc6 3 d4 exd4 4 Nxd4 Qf6’?

 Question: Why bring out the queen early when Black can attack the knight
by moving the bishop to c5?

Answer: Part of the line’s allure may be its theoretical ambiguity. Black’s early queen move avoids
4 ... Bc5 5 Nb3, gaining a tempo on the bishop. Secondly, Black’s queen eyes f2, which may come
under fire after ... Bc5.
5 Nb3
This move doesn’t gain a tempo in this version. On the other hand, Black’s
bishop is kept off the dangerous gl-a7 diagonal. 5 Nxc6 and 5 Be3 are
White’s main lines.
Note that 5 Nb5 is ineffective: 5 ... Bc5 6 Qe2 Bb6 7 Nlc3 (intending
Nd5) 7 ... Nge7 8 Be3 Ba5! 9 0-0-0 a6 10 Nd5 Nxd5 11 exd5 axb5 12 Bd4+
(or 12 Bg5+ Qe5) 12 ... Qe7 13 dxc6 dxc6 14 Bxg7 Qxe2 15 Bxe2 Rg8 and
Black looks just fine, M.Perunovic-N.Grandelius, European Cup, Budva
2019.
 5 ... Qg6 6 13
6 Nc3 is answered conveniently by 6 ... Bb4.
. Bb4+’?
6..
Black wants to induce c2-c3, then play ... Bd6 and develop the queenside
with... b7-b6.
In a later game Nepo faced an immediate 6 ... Bd6 7 g4!? (as you may
have noticed by now, reasonable compromise is not really Nepo’s strong
suit!) 7 ... h5 8 g5 f6 9 Nc3!? fxg5 10 Rgl g4 11 f4 Nge7 12 Be3 0-0?! (I
prefer Black after 12 ... Bb4) 13 Qd2 a5?! 14e5?(14 Bc4+ Kh7 15 0-0-0 is
much better) 14 ... Nxe5? (and this is just unsound; 14 ... Bb4 should still be
played) 15 fxe5 Bxe5 16 0-0-0 a4 17 Nd4 a3 18 Bc4+ Kh8 19 Rdfl! axb2+
20 Kbl Bf6 21 h3 d5 22 Bd3 Qe8 23 hxg4 Bxg4 24 Rxg4 (so glaringly
obvious that it doesn’t even require an exclam) 24 ... hxg4 (acceptance allows
mate) 25 Qh2+ Kg8 26 Qh7+ Kf7 27 Qh5+ 1-0, as it’s mate in three more
moves, I.Nepomniachtchi-I.Cheparinov, World Blitz Championship, St.
Petersburg 2018.
7 Bd2! a5?’
He should swap bishops, even though 7 ... Bxd2+ 8 Qxd2 d6 9 Nc3 Be6
10 Nd4 is a solid edge for White.
8 Nc3! Nge7
8 ... a4 is met by 9 Nb5! Bd6 (not 9 ... axb3?? 10 Nxc7+ Kd8 11 Nxa8 and
Black is busted because he isn’t going to trap the knight) 10 Nxd6+ Qxd6 11
Bc3! Qxdl+ 12 Rxdl f6 13 Nel! d6 14 Nd3 with advantage to White, who
owns the bishop pair.
9 Nb5! Bxd2+?
The problem with this move is that it dooms Black’s king to the centre,
which means eternal vulnerability to an attack. He had to try 9 ... Bd6 10
Nxd6+ Qxd6 11 Bc3 Qxdl+ 12 Rxdl f6 13 g4!, when White’s bishop pair
and space give him a favourable ending.
10 Qxd2
There’s now only one way to protect c7 and it’s not a pleasant one!
10 ... Kd8
 Question: Is Ivanchuk determined to commit suicide?

Answer: It isn’t suicide if every option loses. Equally hopeless is handing over a pawn and
wrecking his game strategically with 10 ... 0-0 11 Nxc7 Rb8 12 0-0-0, when White is a pawn up with a
crushing position.
11 0-0-0 d6
After 11 ... a4 12 Nc5 d6 13 Nd3 Qf6 14 f4 Bg4 15 Rel Be6 16 g4! White
has a decisive attack. Note that 16 ... Bxg4 is met by 17 e5.
12 Kbl

Question: Why waste a tempo when attacking?

Answer: There is no rush. Sliding the king to bl ensures that Black doesn’t get tricks to force White
to exchange queens.
12 ... Rf8
Ivanchuk wants to distract his opponent with central counterplay and so
prepares ... f7-f5. Instead:
a) 12 ... Qh6 is met by the simple 13 f4!, preventing the exchange of
queens and leaving White with a big positional edge.
b) 12 ... f5 is answered by 13 Nc5! b6 14 Na4!, when Black must watch
out for sacrifices on b6, d6 and even c7.
13 g4
Nepo logically stifles the break, while gaining kingside space.
13 ... Qf6?

This move appears to be a case of Stockholm Syndrome, where the victim

relates to his kidnapper. It’s made with the false assumption that... Qxf3 is an
actual threat. In reality, the move is of vast benefit to White. Instead, 13 ... a4
is met by 14 Nc5!, but this had to be tried.
14 g5!
Nepo, who makes it look so easy, is never shy about giving away pawns
for initiative or attack. Ivanchuk probably hoped for the passive 14 Be2?
Ng6, when Black’s defensive chances rise due to his kingside dark square
blockade.
14 ... Qxf3
The theory “a little something is better than nothing” doesn’t apply here.
White’s open lines infinitely outweigh Black’s extra pawn. Still, it’s too late
to back down, since if he declines, Black just gave away a tempo; e.g. 14 ...
Qe6 15 f4 Bd7 16 e5 d5 17 Nc5 Qg4 18 h3 Qg3 19 Bg2 and Black doesn’t
have a hope of surviving.
15 Bg2 Qg4
White pounces on the fact that the black queen’s coverage of the e5-square
is dislodged.
16 e5!
Nepo ruthlessly rams through in the middle. Black’s king no longer has a
prayer.
.
16.. Nxe5
 White’s pawns are the cursed mummy’s treasure, which bring misfortune
and death to those unlucky enough to possess it.

Exercise (combination alert): What is White’s optimal attacking continuation?

Answer: Sacrifice on d6, after which Black’s position collapses.

17 Nxd6!
Not 17 Nxc7??, which at first appears crushing, because it actually walks
into Black’s trap: 17 ... Nc4! (zwischenzug) 18 Qc3 Kxc7 19 Bf3 Qf4 20 Rd4
Qxg5 21 Rxc4+ Kb8 and while White has attacking compensation for the two
sacrificed pawns, he certainly doesn’t stand better.
17 ... cxd6 18 Qxd6+ Nd7
If 18 ... Bd7? 19 Qxe5, Black must resign since 19 ... Qxg2 is met by 20
Nc5.
19 Rhel Qxg5
As good as anything.
a) 19 ... Nc6 allows 20 Qxf8+.
b) 19 ... Qxg2 20 Qxe7+ Kc7 21 Qd6+ Kd8 22 Qb6 is mate.
c) 19 ... Re8 is a wonderful move, but only if you are a composer of help­
mate problems: 20 Qb6 mates again.
20 Nc5 1-0
 Black must hand over his queen to evade mate, since 20 ... Ra6 is met by
the simple 21 Nxb7+ Ke8 22 Qxa6 and if 22 ... Qxg2 then 23 Rxd7! Kxd7 (or
23 ... Bxd7 24 Nd6+ Kd8 25 Qb6 mate) 24 Qd6+ Ke8 25 Qd8 mate.

Game 18
LNepomniachtchi-V.Potkin
Yaroslavl (rapid) 2014
Sicilian Defence

1 e4 c5 2 Nc3
Today Nepo pretty much exclusively plays Open Sicilians.
2 ... e6 3 Nge2 Nc6 4 g3
I guessed wrong and expected the switch back with 4 d4 cxd4 5 Nxd4.
4 ... Nf6
Here 4 ... d5 5 exd5 exd5 6 Bg2 d4 7 Nd5 Nf6 8 Nef4 Nxd5 9 Nxd5 Bd6
10 0-0 0-0 11 d3 Be6 12 Nf4 Bf5 is considered as at least even for Black.
R.J.Fischer-B.Spassky, 17th matchgame, Belgrade 1992, continued 13 h3
Rb8 14 Bd2 Re8 15 Rel Rxel+ 16 Qxel Qd7 17 g4 Re8 18 Qdl Bxf4 19
Bxf4 Be6 20 QI3 Nb4 21 Qxb7 Qxb7 22 Bxb7 Nxc2 23 Rcl Nb4 24 Be4
Bxa2?! (24 ... c4! 25 dxc4 Bxg4! is still even) 25 Bd2 Bd5 26 Bxd5 Nxd5 27
Rxc5 Nb6 28 Kfl and Bobby went on to milk his slight edge to a win.
5 Bg2 Be7 6 0-0 0-0 7 f4
It’s not too late for an Open Sicilian with 7 d4.
7 ... d5 8 e5 Nd7 9 d3

Question: Is this a Closed Sicilian or a French Defence?

Answer: It’s a bit of a hybrid between the two, where White sets up in Closed Sicilian fashion
against Black’s French formation. Or the other way round, depending on your point of view.
9 ... Rb8
Intending to expand on the queenside, his strong wing, with ... b7-b5.
 10 a4 a6 11 g4
I have learned from long experience with students that patience can’t be
taught. The push of the g-pawn is more forceful and also more risky than 11
Khl. White’s idea is to begin massing pieces for a kingside attack with Ng3
and Nce2.
11 ... b5 12 axb5 axb5 13 Ng3 b4 14 Nce2

14 ... f6
Principle: Respond to the opponent’s wing attack with a counter in the
centre.
Fat Fritz suggests the counter-intuitive 14 ... h6, which actually may have some merit since White’s
pushes on g5 and f5 are both suppressed, at least for now.
Instead, 14 ... Bb7 looks a bit slow: 15 g5! g6 16 h4 Qb6 17 h5 Ra8 18
Rbl Nd4?! 19 Nxd4 cxd4 20 b3 Qc7 21 Bb2 Rfc8 22 Rf2 Ra2 23 Bh3 and
Black’s queenside attack stalled, while White can continue to build on the
kingside, J.Carrasco Martinez-J.Fernandes, Lisbon (rapid) 2019.
15 exf6 Nxf6 16 g5!?
 From this point on there is no pretext of civilization, and barbarism is the
norm.

Question: Didn’t this move create a hole on f5?

Answer: It did. Nepo did so with the plan to sacrifice a pawn there to plug the hole and open up f4
for his pieces. Instead, 16 h3 Bd6 17 f5 Ne7 looks fine for Black.
16 ... Ne8 17 h4 Nd6?’
This most natural of moves is inaccurate because Black’s price for
fortifying f5 is a weakening of d5. 17 ... Nc7 looks equal.
18 f5!

Clearance.
18 ... Nxf5
18 ... Nb5 is met by 19 Nf4.
19 Nxf5 exf5
There is no benefit in a rook swap with 19 ... Rxf5 20 Rxf5 exf5 21 Nf4
Ne5 22 Bxd5+ Kh8 23 Qh5, as Black’s position is even worse than the one
he gets in the game.
20 Nf4
Threatening Bxd5+, followed by Bxc6.
20 ... Ne5 21 Qh5’?
I would grab the d5-pawn first: 21 Bxd5+! Kh8 22 Qh5 Qb6 23 Rel Bd6
24 Kfl Bd7 and now 25 Rxe5! Bxe5 26 g6 h6 27 Ne6! (with the deadly
threat of Bxh6) 27 ... f4 28 Qxe5 Bxe6 29 Qxe6 leaves Black busted.
.
21.. g6
A move like this cannot be played without serious misgivings, since it
weakens the dark squares around the black king, which play a huge role later
on in Nepo’s attack. Nonetheless, Potkin isn’t panicked enough to give up an
exchange to lessen White’s attack with 21 ... Bb7 22 Ne6 Qb6 23 Nxf8 c4+
24 Khl Rxf8 25 Qe2, which is lost for Black anyway.
22 Qe2?’
Again I would play 22 Bxd5+ first, intending 22 ... Nf7 23 Nxg6! or 22 ...
Kh8 23 Qe2 as in the game. Black might have to resort to 22 ... Rf7 just to
keep breathing.
22 ... Bd6 23 Bxd5+ Kh8?
Here Black had the option of 23 ... Nf7!, which keeps immediate
damnation at bay, though he remains clearly worse.
 Exercise (planning): From this point Nepo found a forced win. Demonstrate how,
step by step.

Answer: Step 1: Opening the long al-h8 diagonal for the dark-squared bishop.
24 b3! Re8 25 Bb2 Qc7 26 Rfel Bd7 27 Ra6 Rb6
It’s a mistake to lock your doors when the killer is already inside your
home. Black’s position is on the threshold of disaster and a combination for
White screams to be played.
28 Ra7!
Step 2: Decoy. Where did that come from!? There now follows a brutal
cascade of sacrifices from Nepo, which flare into incandescence.
28 ... Qxa7
If 28 ... Qb8 then 29 Rxd7 wins; while 28 ... Qd8 29 Qxe5+! Rxe5 30
Rxe5! leaves Black helpless to prevent mate.
 29 Qxe5+!
Step 3: Queen sacrifice.
29 ... Bxe5
Or 29 ... Rxe5 30 Rxe5! Bf8 31 Re8+ Rf6 32 Bxf6 mate.
30 Rxe5! 1-0
At this point I strongly encourage mass hysteria for the black pieces, since
collapse along the al-h8 diagonal is clear to see. Notkin in fact resigned in
view of 30 ... Rf8 (or 30 ... Rf6 31 Rxe8+ Bxe8 32 Bxf6+ Qg7 33 Kf2,
winning easily) 31 Re8+ (Step 4: Discovered check) 31 ... Rbf6 32 Bxf6
mate, since ... Rxf6 is illegal.

Game 19
I.
Nepomniachtchi-S.Sjugirov
Russian Team Championship, Sochi 2016
PetroffDefence
1 e4 e5 2 M3 Nf6 3 Nxe5 d6 4 Nf3 Nxe4 5 c4’?

Question: Why this move, rather than the normal 5 d4 or 5 Nc3 - ?

Answer: The idea behind the early push of the c-pawn is to discourage ... d7-d5, thereby grabbing
space in the centre.
5 ... Be7 6 d4 0-0 7 Bd3 Ng5
Principle: Swaps benefit the side with less space. 7 ... Nf6 is Black’s other
option.
8 Nc3 Bg4?’
An inferior product sells for only a short period of time. When its
inferiority is outed, people stop buying it. It’s exactly the same for a suspect
opening which is playable only until some spoilsport GM refutes it, and then
everyone abandons it. This particular line is now considered unplayable due
to Nepomniachtchi’s devastating novelty coming up in a couple of moves.
 Black should prefer 8 ... Nxf3+ 9 Qxf3 Nc6 10 Be3, which offers White an
edge due to central space, but no more than that.
9 Bxg5!
After 9 Be2 Nxf3+ 10 Bxf3 Bxf3 11 Qxf3 Nc6 12 Be3 White has allowed
too many piece trades for his space to mean much, V.Tseshkovsky-
S.Makarichev, USSR Championship, Minsk 1979.
9 ... Bxg5

Exercise (combination alert): In this position Nepo unleashed a devastating

novelty, which sent Black’s line to the theoretical graveyard. What would
you play here?

Answer: Step 1: Greek Gift sacrifice.

10Bxh7+’’
At first this looks totally unplayable for two reasons:
1. Black’s queen still covers g5.
 2. White’s f3-knight is pinned.
My friend GM Max Illingworth told me that the concept of a Greek Gift
sacrifice with the f3-knight pinned just blew him away when he first saw this
game.
10 ... Kxh7 11 h4!
Step 2: Zwischenzug.
11 ... Bd2+?’
This leads to a pawn down position with zero compensation, but nothing
equalizes:
a) 11 ... Bxf3? 12 hxg5+ Kg8 13 Qxf3 Qxg5 14 Rh5 Qg6 15 0-0-0 is a
winning attack for White.
b) 11 ... Bxh4?! 12 Qd3+ Kg8 13 Rxh4 (threatening mate on h7) 13 ... f5
14 Rh2 Bxf3 15 Qxf3 c6 16 0-0-0 is still a nasty attack.
c) 11 ... Bh6! (the first of Black’s three best chances for survival) 12 Ng5+
Qxg5 13 hxg5 Bxdl 14 Rxdl Nd7 (or 14 ... Nc6 15 gxh6 gxh6 16 Rh4) 15
gxh6 Rae8+ 16 Kd2 gxh6 17 Rh4 Re6 18 Nd5 c6 19 Nf4 Rf6 20 Rel and
White has a clear advantage in the ending.
d) 11 ... Re8+! 12 Kfl Bh6 13 Ng5+ Qxg5! 14 hxg5 Bxdl 15 Rxdl is
similar to line ‘c’, where Black has a bad ending, yet is not busted.
e) 11 ... Qe7+! 12 Kfl Bh6 13 Ng5+ Qxg5! 14 hxg5 Bxdl 15 Rxdl is -
yes, you guessed it - a similar ending to lines ‘c’ and ‘d’.
12 Qxd2 Re8+ 13 Kfl Bxf3 14 Qd3+!
Zwischenzug.
14 ... Kg8
14 ... Be4? is dumb since it gives White’s knight a free move to enter the
attack with 15 Nxe4.
15 Qxf3
 When we examine the aftermath, we see that White has won a clean pawn.

Question: Is one extra pawn enough to win when the defending side
otherwise has no weaknesses?

Answer: While the number one doesn’t constitute an overwhelming majority, with all other things
being equal it is enough to claim a winning position for White. So as sweeping generalizations go, this
one appears to be true.
15 ... Nd7
This allows the knight to go to f8 in case White plays Ne4 and Ng5.
16 Rdl
Not 16 Qxb7?! Rb8 17 Qxa7 Rxb2 and Black’s monster rook on the
seventh rank pretty much compensates for the two pawns.
.
16.. Qf6’?

Question: Why is Black offering a queen swap when he is a pawn behind?

 Answer: He is worried about his king. In any case, after 16 ... c6 17 g3 Nf6 18 d5 c5 19 Kg2 a6 20
a4 Qd7 21 Rhel Ng4 22 Qf4 Re5 23 Rxe5 Nxe5 24 b3 Black lacks counterplay and remains a pawn
down.
17 Qxf6
This may actually be more accurate than 17 Rh3 Qxf3 18 Rxf3 Nf6, as
White’s rook is not necessarily well placed on the third rank.
17 ... Nxf6 18 f3!
Now White can play Kf2 and then challenge for the e-file.
.
18.. d5!?
Black is desperate for counterplay.
19 c5 b6 20 cxb6 axb6 21 Kf2 b5’?

Otherwise, there is no plan for Black in this featureless expanse. The

option advocating neutrality and non-interference is to play 21 ... c6 and then
just wait.
22 a3 b4
This was Black’s idea, trying to generate play on the queenside.
 23 axb4 Rab8 24 b5 c6 25 Rhel cxb5 26 Rxe8+ Rxe8 27 Rcl
There is no reason to give Black hope with 27 Nxb5? Rb8 28 Nc3 Rxb2+
29 Ne2, when White’s win is complicated by fact that there aren’t many
pawns left on the board.
27 ... Ra8
27 ... Rc8 28 Ke3 Kf8 29 g4 Ne8 30 Ne2 is lost for Black, whether he
swaps rooks or not.
28 Nxb5 Ra4
If 28 ... Rb8 then 29 Rc5 keeps the extra pawn, since 29 ... Nd7? is
unplayable: 30 Rxd5 Nf6 31 Rf5 g6 32 Rg5 Nh7 33 Re5 and White hangs on
to both extra pawns.
29 Rc8+ Kh7 30 g4!
Once again we see Nepo’s uncanny eye for mating geometry, even with
minimum material left on the board. The threat is Nd6, Nxf7, Rh8+ and Ne5
mate.
30 ... Rb4 31 Nd6
 31 ... Rxd4
Other ways to lose include:
a) 31 ... Rxb2+ 32 Kg3 Rd2 33 Nxf7 and if 33 ... Rxd4? then 34 Rh8+ Kg6
35 Ne5 mate.
b) 31 ... Kg6 32 h5+ Nxh5 33 gxh5+ Kxh5 34 Nxf7 with an extra piece.
32 Kg3 1-0
Since 32 ... Rb4 33 Nxf7 (threatening mate in two) 33 ... g6 34 Ng5+ Kg7
35 Ne6+ Kf7 36 Nd8+ Kg7 37 Rc7+ Kh8 38 Rb7! consolidates the two
pawns with an easy win.

Embedded within the next game is possibly the deepest combination in the
entire book.

Game 20
L.Dominguez Perez-I.Nepomniachtchi
Russian Team Championship, Sochi 2016
Sicilian Defence

1 e4 c5 2 M3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 f3 e5

This more solid version is played almost twice as much as the riskier 6 ...
e6.
7 Nb3 Be6 8 Be3 Be7 9 Qd2 Nbd7
Less committal than 9 ... 0-0, although Black nearly always castles
kingside in this line.
10 g4 b5
Both 10 ... 0-0 and 10 ... h6 are major alternatives.
11 g5
White seizes control over d5, at the cost of weakening f4.
11 ... Nh5
I like this line more than 11 ... b4 12 Ne2 Nh5 13 Ng3 Nf4 14 h4 h6 15
gxh6! Rxh6 16 0-0-0 Bxh4, as in Wang Hao-Bu Xiangzhi, Chinese
Championship, Wuxi 2006. when White’s chances seem better after 17 Nf5
Bxf5 18 exf5.
12 0-0-0 Nb6 13 Nd5
 13 ... Nxd5

Question: Doesn’t this lose a tempo?

Answer: The loss of a tempo is less important than loss of the light squares after 13 ... Bxd5? 14
exd5 Rc8 15 Na5! Nxd5 16 Qxd5 Qxa5 17 Bh3 Rc4 18 Qa8+ Qd8 19 Qxa6. Black scores an unhappy
10% from this position.
14 exd5 Bd7 15 Na5!
With two purposes:
1. Black must now watch out for Nc6, even if it loses a pawn, since White
would gain dominance over the crucial light squares.
2. Black’s ... a6-a5 attacking push has been stifled, at least temporarily.
15 ... Qc7 16 Kbl 0-0
At long bloody last! I’m wondering what the record is for latest castling. I
do recall a game between IMs Lodewijk Prins and Lawrence Day, where the
game ended with 31 ... 0-0-0 mate!
 17 Rgl
17 Bd3 is a decent alternative.
17.. . g6
Denying any g5-g6 sacrificial ideas, while giving Black’s knight a possible
escape route via g7 and f5.
18 c4!

Question: What is White doing, creating confrontation on Black’s strong wing?

Answer: It’s actually quite debatable that the queenside is Black’s preserve. Instead, Black may
look for counterplay on White’s weakened kingside, while opening the c-file reinforces the idea of
White’s knight jumping into c6.
18.. . f6!?

This is awfully dangerous. Nepo gambles that White’s weaknesses on the

f-file, and 13 in particular, outweigh the fact that he is prying open the g-file
against his own king!
 18 ... Rfc8 looks safer, but not necessarily better. After 19 Bd3 Nf4 20
Bxf4 exf4 21 Nc6! Bf8 22 Re 1,1 still prefer White.
19 Rcl
Threatening c4xb5, followed by b5-b6.
19 ... Bf5+ 20 Kai Bd8!
That a5-knight was driving Nepo crazy, so he expends a tempo to dislodge
it.
21 Nc6 fxg5 22 cxb5
A novelty on move 22! Previously White had always played 22 Nxd8
Qxd8 24 Bxg5.
22 ... axb5 23 Bxb5!

Question: Is it my imagination or are both sides deliberately prying open

their own kings?

Answer: Both sides proceed with a kind of Schrodinger’s opposite wing attack, which
simultaneously is and isn’t real. The truth is that both kings are safer than they look, and both sides
attack on the same side as their own king.
23 ... Nf4
 24 Nxd8!
Clearing c6 for another piece. 24 Rc3?! was no improvement: 24 ... Bf6 25
Rb3 Qf7 26 Bxf4?! (26 Rdl or 26 Bc4 was necessary) 26 ... gxf4 27 Rel
Rfe8 was clearly better for Black, I.Zhukov-A.Tsvetkov, St. Petersburg 2018.
24 ... Qxd8 25 Rc6!
Dominguez invites a future ... Bd7 and ... Bxc6, which would give away
all Black’s light squares for the exchange. I like White’s position better here
and feel his queenside chances are superior to Black’s on the kingside. GM
Leinier Dominguez Perez is Cuba’s highest rated player and is worshipped by
his fellow countrymen. My evidence for that assertion is several Cuban
Facebook friends having asked me if I would ever write Dominguez: Move by
Move.
.
25.. Qe7
The engine doesn’t like this move but I don’t blame Nepo, since White has
a big advantage in the ending after 25 ... Qa5 26 Qxa5 Rxa5 27 Bc4 Bd3 28
b4! Ra4 29 Bb3 Rxb4 30 Rxd6.
26 Bc4! Kh8
To avoid White’s Bc5 tricks.
27 Qb4!
Targeting d6.
27 ... Rfd8 28 Bb6

White has a decisive strategic advantage and I experience difficulty

pinpointing just where Black went wrong. It feels as if Nepo played all the
logical moves and is now busted!
28 ... Rd7 29 a4!
White’s king is perfectly safe. 29 b3 h5 30 a4! is similar.
29 ... Qf7 30 Qb5 Rb7 31 a5!
A carried gun need not be discharged to make it threatening. It’s still a
threat, even in silence. As mentioned earlier, White isn’t afraid to hand over
the exchange after ... Bd7 and ... Bxc6.
 31 ... Bd7 32 Rcl Rbb8 33 Ba2 Qf5
When will Black be delivered from his present torment? The answer is: on
the very next move.
34 a6?
Although it’s incredibly difficult to see why this natural move is an error,
in a single move Dominguez goes from winning to losing. Instead, 34 Qc4!
Rc8 35 Qe4! leaves White in a winning position, as the passed b-pawn soon
marches up the board. I learned from watching forensic analysis murder
shows on TV that the dead do indeed speak. We must examine the corpse of
our losses with the same kind of detail. But I’m wondering just what
Dominguez will learn from this loss, since it feels as if he did everything
right and then lost anyway!

Exercise (combination alert): With his last move White planted the seeds of
his future destruction. How can Black gain a decisive advantage? (Be
warned:
 this one is brutally difficult. Out of around 15 students who tried it, none
got
the entire combination right.)

Answer: Step 1: Move the knight to d3, allowing White to pin it with Bbl.
34.. . Nd3’’
A miracle occurs. The success of such a move hinges upon a staggering
level of calculation. When you envisage a line as long and as convoluted as
this one, you had better correctly predict what lies on the other side. And
Nepo does just that.
Instead, 34 ... Qd3? 35 Bc4 Qd2 36 Rc3 g4 37 fxg4 Nxd5 38 Bxd5 Qdl+
39 Ka2 Rxa6+ 40 Qxa6 Qxd5+ 41 b3 Bxc6 42 Kb2 would offer White decent
chances of survival.
35 Bbl
On the surface it appears as if Nepo’s last move was a double question
mark blunder. His next move reveals it the opposite. Dominguez had no
choice anyway, since 35 Rlc3?? is refuted by 35 ... Nc5 36 a7 Rxa7 37
R3xc5 dxc5 38 Qxc5 Raa8 39 Bc7 Re8 and White is completely busted.
35.. . Qf4!
Step 2: Unpinning, while threatening ... Qxcl!.
36 Qxd3
White puts up more resistance in the line 36 Bxd3 Qxcl+! 37 Rxcl Bxb5
38 Bxb5 Rxb6 39 Bc4 Raxa6+! 40 Bxa6 Rxa6+ 41 Kbl, though Black should
still win; e.g. 41 ... Ra5 42 Rdl Kg7 43 Kc2 Rc5+! (forcing White’s king to
block his own passed b-pawn) 44 Kb3 h5 45 h3 Kf6 46 Ka4 Rc4+ 47 b4 g4
48 fxg4 hxg4 49 hxg4 e4! 50 Ka5 Ke5 51 b5 Rc2 52 Rbl Kxd5 53 b6 Kc6!
and wins, in view of 54 b7 Ra2+ 55 Kb4 Kxb7.
36 ... Bxc6 37 Rxc6 Qa4+!
The a6-passer falls.
38 Qa3
 .38.. Qxa6!
Threatening ... Qb7. Nepo alertly evades the trap 38 ... Rxa6?? 39 Bc5!
and White is suddenly winning again.
39 Bc7
It’s a bad sign when you keep giving away things against your will. White
has no choice but to surrender his queen.
a) 39 Qxa6? Rxa6+ 40 Ba2 Raxb6 41 Rc2 Ra8 is an easy endgame win for
Black.
b) 39 Rxd6 Qe2 also wins, since 40 Ba5 is refuted by 40 ... Rxa5 41 Rd8+
Rxd8 42 Qxa5 Rd6! (threatening ... Ra6) 43 Qa8+ Kg7 44 Qb7+ Kf6 45 Ka2
Ra6+ 46 Kb3 Qe3+ 47 Kc2 Rb6! 48 Qxh7 Qe2+ and White is mated.
39 ... Qb7 40 Bxb8
Again no choice. 40 Ba5? Rxa5 41 Rc8+ Qxc8 42 Qxa5 Ra8 wins the
queen and the game.
40 ... Rxa3+ 41 bxa3 Qxb8
 The final phase begins, where Black must create a passed pawn before
White’s a-pawn moves up the board.
42 Ka2 Kg7 43 Be4 Kh6! 44 Rc2 Qb5 45 Rc6 g4! 46 fxg4 Qe2+ 47 Bc2
e4 0-1
Black’s passed e-pawn will win the game.

Game 21
I.
Nepomniachtchi-Bu Xiangzhi
Danzhou 2016
Ruy Lopez

1 e4 e5 2 M3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 0-0 b5 6 Bb3 Bb7

6 ... Be7 and 6 ... Bc5 are two other major lines.
 7d3
Nepo’s default setting is to toss in an early d2-d3 in most Lopez lines. 7
Rel and 7 c3 are also played.
 Question: Isn’t White winning with the Fried Liver Attack type 7 Ng5 - ?

Answer: This is actually an inferior version for White due to 7 ... d5 8 exd5 Nd4!. Not 8 ... Nxd5?
because 9 Qh5! g6 10 Qf3! Qxg5 11 Bxd5 wins due to a double attack on c6 and f7, which is why
White provoked ... g7-g6; without that Black would have ... Qf6 as a defence.
7 ... Be7 8 Nc3
This is logical since White can now toss in a2-a3, whereas Black doesn’t
get to challenge White’s powerful light-squared bishop with ... Be6, since his
bishop is already committed to b7.
. d6
8..
Now ... Na5 is on since the e5-pawn is protected.
9 a3 0-0 10 Rel
Five years later Nepo tried 10 Ne2 on Magnus, continuing 10 ... d5! 11
exd5 Nxd5 12 Rel Kh8! 13 Ng3, and now 13 ... Nf4! 14 Bxf4 exf4 looks fine
for Black, as in Z.Almasi-V.Erdos, Hungarian Championship (rapid playoff),
Ajka2018.
10 ... h6 11 Ne2 Re8 12 Ng3 Bf8 13 c3 Qd7 14 d4
 14 ... Na5
This standard Lopez move was actually a novelty here. In such Closed
Lopez positions, we go through the opening ritual of the first 15 to 30 moves
with the enthusiasm of a third-grader, forced to recite a poem in front of the
class. It’s almost as if the game begins from move 15 on, rather than move
one. Instead, 14 ... Rad8 was P.Svidler-A.Mikhalchishin, Warsaw 1999,
where White looks comfortably better after 15 h3.
15 Bc2 Nc4
The idea of transferring the a5-knight to c5 after 15 ... c5 16 d5 c4 is a bit
slow: 17 h3 Bc8 18 Nh2 Nb7 19 f4! Nc5 20 f5 and White’s extra territory on
the kingside ensures an advantage.
16 a4 Rad8 17 b3 Nb6 18 a5! Nc8 19 d5 c5?’
One cannot look down upon servitude if you want to play the Black side of
a Closed Ruy. When things begin to go wrong for Black in this line, you get
the feeling you’re in one of those dead-end jobs, with twice the work and half
the pay. My intuition says that this move is the main reason for Bu’s strategic
downslide. Black has no counterplay with a sealed queenside and centre. He
can only wait while White builds for a kingside attack. I would have kept the
centre as fluid as possible with 19 ... c6 20 c4 Ne7.
20 h3 g6
The engines don’t like this natural move either, preferring 20 ... Ne7 21 c4
b4 22 Nh4 Bc8.
21 c4 b4 22 Ra2

Nepo foresees a transfer of this rook to the f-file to feed his kingside
attack.
22 ... Bg7 23 Bbl Ne7 24 Nh2
Now f2-f4 is coming, unless Black wants to commit to the radical ... g6-
g5-
24 ... g5!?
Bu obviously feels that a half-measure won’t do here. He was probably
thinking “now White will never achieve f2-f4” when making this move. This
assumption is a best-case scenario which, as we all realize, rarely happens in
real life.

Question: Which do you think is the more important factor: preventing f2-f4
or the holes created by pushing the g-pawn to the fifth rank?

Answer: The question is: should we be loyal to our style or to the empirical evidence? The latter is
your better choice. I’m inclined to think that Bu’s strategically risky move is closer to a dubious mark
than an exclam. Yet I can’t bring myself to give it one since the alternative was passive defence and
allowing White to keep building on the kingside.
25 Nh5
There is no need for a key when the opponent opens the door for you. The
idea is to eliminate the defender of the h5-hole. I would be more inclined to
add light square fire-power in the direction of the kingside with 25 Bd3,
intending Be2 and possibly Bg4.
25 ... Bc8?’
With the benefit of hindsight Black is better off playing 25 ... Ng6!, since
it would prevent Nepomniachtchi’s coming combination.
26 Nxf6+
Now White’s queen has an appointment with the h5-square.
26 ... Bxf6 27 Qh5 Kg7 28 Ng4 Rh8
 Exercise (combination alert): What is White’s optimal way forward?

Answer: Annihilation of defensive barrier. White can actually blast his way through with f2-f4!!,
despite Black’s strenuous attempts to prevent it.
29 f4!!
Sometimes it feels as if Nepo’s will simply overrides the position’s reality.
The fact that White is winning after such a move is a contradiction to what
our eyes tell us. The immediate f2-f4 is far stronger than the slow build-up
with 29 g3 Ng6 30 f4 Rdg8 31 f5 Nf8, when Black’s position is better than in
the game.
29 ... gxf4
There is nothing better:
a) 29 ... exf4?? walks into 30 Nxf6 and if 30 ... Kxf6 then 31 Bb2 mate!.
b) 29 ... Ng8 fails to 30 fxe5 Bxe5 31 Nxe5 dxe5 32 Bxg5! hxg5 33
Qxg5+ Kf8 34 Qxe5 and Black is strategically busted.
 30 g3!
You cold-blooded monster! This second line-opening shot is one of the
brilliant corollaries to 29 f4!!.
.
30.. Ng6
An admission that things are falling apart.
a) 30 ... fxg3? allows 31 Bxh6+ Rxh6 32 Qxh6+ Kg8 33 Nxf6 mate.
b) 30 ... Rdg8 31 gxf4 Kf8 32 Rg2 Rxg4 33 Rxg4 Ng6 34 f5 is ultimately
hopeless; e.g. 34 ... Nf4 35 Bxf4 exf4 36 e5! Bxe5 37 Rxe5! dxe5 38 f6 Qd6
39 Bf5 Bb7 40 Be6! fxe6 41 Rg7 forces mate in four.
31 gxf4 Qe7
If 31 ... exf4 then 32 Nxf6 Kxf6 33 e5+! dxe5 34 Rg2 Rhg8 35 Bxg6 fxg6
36 Qxe5+ Kf7 37 Qxf4+ Qf5 38 Qc7+ Rd7 39 Qxc5 wins.
32 Nxf6
Black’s dark squares soon collapse with the removal of this bishop.
32 ... Qxf6 33 fxe5 dxe5
The worsening of our position tends to be in exact ratio to our growing
depression. Now White is given a protected and passed d-pawn. But 33 ...
Qxe5? hangs the queen to 34 Bb2, while 33 ... Nxe5? allows 34 Rg2+ Kf8 35
Rfl Qe7 36 Bg5! Qd7 37 Rf5! Qe8 38 Bf6 Bxf5 39 Qxf5 and Black must
give up a full rook, since 39 ... Rg8 40 Qh7! forces mate in three.
34 Rg2 Qd6 35 Rfl Rde8 36 Khl Re7 37 Be3 Rc7
 Exercise (planning): Come up with a winning plan for White.

Answer: Double rooks on the f-file and then plant one on f6.
38 Rgf2! Rh7
38 ... Qd7 doesn’t change anything: 39 Kh2 Qe7 40 Rf6 wins.
39 Rf6! Qxf6
Avoiding this swap loses too; e.g. 39 ... Qd8 40 Bc2 Qe8 41 Bdl Qd8 42
Bg4 Bxg4 43 Qxg4 Qxf6 (or 43 ... Rh8 44 Rxa6) 44 Rxf6 Kxf6 45 d6 Rc6
(or 45 ... Ra7 46 Bxc5) 46 Qd7 traps the rook.
40 Rxf6 Kxf6 41 Bxh6
Threatening Qg5 mate. Now we see painful evidence of the absence of
Black’s dark-squared bishop.
41 ... Rxh6
No choice since 41 ... Ke7? hangs as full rook to 42 Bg5+.
42 Qxh6 Re7
 Black doesn’t have a hope in hell of establishing a fortress.
43 h4 Re8 44 Qg5+ Kg7 45 h5 Rh8

Intending ... Rh6, ... f7-f6, and ... Rxh5+. Nepo alertly derails this idea on
his next move.
46 Bc2!
Note that 46 Kgl?! doesn’t win the knight because Black has 46 ... Rh6 47
Kf2 f6 48 Qgl Rxh5.
46 ... Rh6 47 Bdl!
The once bad light-squared bishop relishes its newfound respectability.
47 ... f6 48 Qgl! 1-0
Black’s position falls apart when the c5-pawn is taken.

Game 22
N epomniachtchi-E.Tomashevsky
I.
Tai Memorial, Moscow 2016
Scotch Game

1 e4 e5 2 M3 Nc6 3 d4
This rustic-looking move, a by-product of the Great Romantic era, is for
those of us who feel aversion for the long siege warfare of the Closed Lopez.
Have you ever seen a child carrying a doll or Teddy bear and thought they
were too old for that? This is how Ruy Lopez players tend to view adults who
play the kid-friendly Scotch Opening. But why not, if you keep winning with
it, as Nepo does?
3 ... exd4 4 Nxd4
Having studied Nepo’s devastating handling of his signature Scotch, I’m
tempted to recommend it to students who display Nepoish stylistic
tendencies.
4 ... Nf6 5 Nxc6 bxc6 6 e5

. Qe7
6..
 Question: Everyone seems to toss in 6 ... Qe7. Is it obligatory? Why not play
6 ... Nd5 instead?

Answer: 6 ... Qe7 is played at a higher than 10:1 ratio and is indeed superior. After 6 ... Nd5 White
gets a pleasant edge with 7 c4 Nb6 8 Bd3 Bb4+ 9 Nd2 d5 10 0-0, as in I.Nataf-M.De Wit, European
Cup, Albufeira Algarve 1999.
7 Qe2 Nd5 8 c4 Ba6
All opening lines have their unique customs and rituals.

Question: Is it better to play ... Ba6 or just move the knight to b6?

Answer: Everyman pays me to answer questions, but for some of them I come up blank. I have
never been able to decide whether it’s better to pin White’s c-pawn as here, or play 8 ... Nb6 at once,
reserving the light-squared bishop for b7, to be activated with ... c6-c5. If we check the stats then 8 ...
Ba6 wins, scoring higher than its 8 ... Nb6 counterpart.
9 b3 g6
This is now established as Black’s main line.
10 f4 Bg7 11 Qf2
Unpinning the c-pawn and hence attacking the d5-knight.
11 ... Nf6
Black plays 11 ... Nb6 more often, but 11 ... Nf6 is popular too.

Question: Where is this knight going to go when White unpins the e-pawn?

Answer: The knight can advance to e4, followed by ... f7-f5. In this game Black was given time to
castle and the knight retreated to e8.
12 Ba3
The main line runs 12 Be2 Ne4 13 Qe3 f5 14 Ba3 d6 15 Nd2 Nxd2 16
Qxd2 0-0-0, and here the engine suggests a novelty in 17 Rdl.
 12 ... d6?
This seemingly natural move gets Black into deep trouble, shockingly
quickly. Black’s position requires swift course-correction with 12 ... Ng4! 13
Qe2 (not 13 Qf3?? Nxe5! and wins) 13 ... Qe6 14 Nd2 d6 15 h3 Nh6 16 0-0-0
0-0-0 17 exd6 Qf6 18 Qe5 Qxe5 19 fxe5 Bxe5 20 c5! Bb5 and White only
has an edge, P.Gyugyi-F.McDermott, correspondence 2016.
13 Nc3 0-0
Black can’t castle queenside since a7 would hang.
14 0-0-0 Ne8 15 g3 Bb7
Black can’t play 15 ... c5 because of 16 Bg2 Rb8 17 Rhel and White is
winning.
16 Bg2
Black can barely move.
16 ... f6 17 exd6 Nxd6?
This blunder is predicated upon the misassumption that Black’s queen is
safe. It isn’t. As strategically awful as it looks, Black had to try 17 ... cxd6 18
Rhel Qc7 19 f5.
18 c5! Nf5
Other squares are no better:
a) 18 ... Nf7 19 Rhel and Black must hand over his knight.
b) 18 ... Nb5 19 Nxb5 and Black loses a piece whether he recaptures or
not.

Exercise (planning): Black’s fortress is in reality a prison. Find the correct idea
and White wins heavy material.

Answer: Step 1: Activate the hl-rook with tempo, chasing Black’s queen to the fatal f7-square.
19 Rhel Qf7 20 Bfl!
Suddenly, disaster yawns before Tomashevsky. Black’s expectations are
dashed by this meek-looking, undeveloping move, which is much more
difficult to see than you might imagine, since our mind tends to dismiss
passive-looking retreats.
Step 2: Threaten Bc4, trapping Black’s queen. Black must hand over at
least an exchange to deal with the threat.
20 ... Rfd8
Tomashevsky has no choice but to brace for the coming hit.
21Rxd8+
21 Bc4 Rd5 22 g4 Ne7 23 Nxd5 Nxd5 24 Bxd5 cxd5 25 c6! Bxc6 26 Re7!
is another way to win.
21 ... Rxd8 22 Bc4
The bishop arrives with ill-tidings and Black’s rook is forced to impale
itself to save its queen.
22 ... Rd5
We only appreciate something when we are about to lose it.
23 Qe2! 1-0

Step 3: Simplification. Black receives a second rebuke. Not only will

Black lose the exchange, queens are coming off the board as well, leaving
Black no reason to play on.

Here is an example of a Nepo win, purely by opening preparation in the

home laboratory.

Game 23
I.
Nepomniachtchi-Li Chao
FIDE Grand Prix, Sharjah 2017
PetroffDefence

1 e4 e5 2 M3 Nf6 3 Nxe5 d6 4 Nf3 Nxe4 5 d4 d5 6 Bd3 Bd6

This is the aggressive posting of the bishop. 6 ... Nc6 and 6 ... Be7 are
slightly safer.
7 0-0 0-0 8 c4 c6 9 Nc3 Nxc3
Black can also swap off a pair of rooks via 9 ... Re8 10 Rel Nxc3 11 bxc3
Rxel+ 12 Qxel h6.
10 bxc3 dxc4
This leads to a typical hanging pawns structure.
11 Bxc4
 11 ... Bf5
I know this is played a lot, but I am of the opinion that it is inferior to 11 ...
Bg4.

Question: What could possibly be wrong with posting the bishop on such
a natural square?

Answer: The bishop on f5 is vulnerable to Nh4! tricks, as Nepo demonstrates in the game.
12 Bg5!
12 Rel Nd7 13 Bg5 is similar.
12 ... Qc7?’
Black is already in trouble after this move. Instead, 12 ... Qa5! 13 d5 Qc7
14 Rel h6 gave Black a playable position in I.Nepomniachtchi-A.Esipenko,
Russian Championship, Moscow 2020. Or if 13 Nh4 then 13 ... Be6! 14 Bxe6
Qxg5 15 Nf3 Qa5 16 Bb3 Nd7! (not 16 ... Qxc3?? 17 Ng5! with threats on f7
and h7, as well as Ne4; e.g. 17 ... Nd7 18 Ne4 Qb4 19 a3! wins a piece) 17
Nd2 with only an edge for White.
13 Rel h6
Or 13 ... Nd7 14 Nh4! Bg6 15 Nxg6 hxg6 16 h4! with a difficult position
for Black, Ad.Horvath-N.Vardan, Budapest 2018.
14Nh4’’
This was a novelty, and a deadly one, after which Black is already fighting
for his life. Previously White played 14 Bh4, giving Black time for 14 ... Nd7
with equality.
14 ... Bh7

Exercise (planning): What is White’s strongest continuation?

Answer: Annihilation of defensive barrier.

15 Bxh6!
On the surface it seems as if Black can defend, since he controls the dark
squares. Nepo’s analysis has seen deeper into the position. Incidentally,
reversing these moves with 14 Bxh6! gxh6 15 Nh4! was also very strong.
15 ... Bxh2+
Black cannot hope for anything by declining the piece. He had to try 15 ...
gxh6 16 Qg4+ Kh8 17 Nf5 Bxf5 (not 17 ... Rg8?? 18 Re8!, which wins on
the spot) 18 Qxf5 (threatening Bd3, as well as a deadly queen check on f6) 18
... Bxh2+ (and not 18 ... f6? 19 Re6! Nd7 20 Rael Nb6 21 Bd3, threatening
Re7, or if 21 ... Qd7 then 22 Qh5 wins) 19 Khl Qf4, when White can choose
between a middlegame after 20 Qh3 Nd7 21 Re3 Nf6 22 R13 Qg5 23 Qxh2 or
an endgame after 20 Qh5 b5 21 Bb3 Nd7 22 Re7 Nf6 23 Qxh2 Qxh2+ 24
Bxh2, either way with a definite advantage.
16 Khl Bf4?
This was the last chance for 16 ... gxh6 17 Qg4+ Kh8 18 Nf5 Bxf5 19
Qxf5, transposing to the previous note.

Exercise (combination alert): In this case, Black’s safer move isn’t safe at all.
How did White exploit it?
 Answer: Annihilation of defensive barrier.
17 Bxg7’’ Kxg7 18 Qg4+ Kh8
After 18 ... Bg6 Black’s defensive wall is paper thin; e.g. 19 Re4! (19 Re6!
is good too) 19 ... Bd2 20 Nf5+ Kf6 (forced; all other king moves lead to
mate) 21 f4! (threatening Qg5 mate) 21 ... Rh8+ 22 Kgl Rh5 23 Ng3 Qa5 24
Rbl b6 25 Re8 Bxc3 26 Rb5!! (interference) 26 ... Bxd4+ 27 Kfl Rxb5 (the
hopeless 27 ... Qxb5 only delays the mate) 28 Qh4+ Kg7 29 Qh8 mate.
19 Nf5! Bxf5
The knight must be eliminated. The problem is that the light-squared
bishop was the black king’s key defender. 19 ... Rg8? 20 Re8! wins as before.
20 Qxf5 Qd6 21 g3!
Intending Kg2 and Rhl+.
21 ... Bh6 22 Kg2 b5
Or 22 ... Nd7 23 Rhl Kg7 24 Rh4! Nf6 25 Rxh6! Kxh6 26 Rhl+ Kg7 27
Qg5 mate.
23 Bb3
Sometimes threatening mate isn’t such a great solution: if 23 Bd3?? then
23 ... Qd5+ removes the queen from the board and Black suddenly stands
better.
23 ... Qg6
 Exercise (planning): What is White’s strongest plan in the position?

Answer: Swap queens, even though White is a piece down!

24 Qxg6’’

Question: Isn’t this in violation of principle?

Answer: It’s a valid exception. Nepo understands that his attack is far more lethal with queens off
the board, since he also removes Black’s most potent defender. The second exclam is for the fact that
such a move is psychologically difficult even to consider; when we have sacrificed a piece for an
attack, the last thing on our mind is a queen swap.
This was in fact the only good move, since 24 Qh3?! Kg7 25 Re2 Bg5 26
Bc2 f5 27 Rael Rh8 28 Bxf5 Rxh3 29 Bxg6 Rh6 30 Bf5 Na6 gives Black
good chances to survive.
24 ... fxg6 25 Re7!
Intending Rhl. White’s cupboards are not bare; he has enough fire-power
remaining on the board for a successful mating attack, the reason being that
Black’s queenside rook and knight lie in a comatose state.
25 ... g5
Planning to answer Rhl with ... Rf6, though even that doesn’t save Black.
26 Re6!
The simple 26 Rhl Rf6 27 f4! is still deadly; e.g. 27 ... gxf4 28 g4 Rg6 29
Bc2 Rd6 (or 29 ... Rxg4+ 30 Kf2) 30 g5 c5 31 g6!? Nc6 32 Rxh6+ Kg8 33
Reh7 f3+ 34 Kxf3 Rf8+ 35 Kg4 c4 (stopping Bb3+ isn’t enough) 36 Rh8+
Kg7 37 R6h7+, winning the f8-rook and forcing mate in four.
26 ... Kg7 27 Rhl Rh8 28 Re7+ Kg6
Or 28 ... Kf6 29 Rhel! g4 30 Rle6+ Kg5 31 Bc2! Rg8 (covering g6) 32
Rf7! (threatening Rf5 mate or Re5 mate) 32 ... Kh5 33 f4! gxf3+ 34 Kxf3
Na6 (34 ... Rf8 35 g4+ forces mate in two) 35 Rf5+ Bg5 36 g4+ Kh4 37
Rh6+! (decoy/clearance) 37 ... Bxh6 38 Rh5 mate.
29 Bc2+ 1-0
Li Chao didn’t need to see anymore. In fact it’s mate in six, beginning 29
... Kf6 30 Rhel g4.
 Exercise (calculation): Without moving the pieces, work out in your mind’s
eye White’s forced mate.

Answer: 31 Rle6+ Kg5 32 Rg6+ Kh5 33 Re5+ Bg5 34 Rexg5 mate.

In the final position, Black’s queenside rook and knight are still sitting on
their original squares, never having moved. So in reality it was White, not
Black, who was material up throughout the game.

Game 24
Nepomniachtchi-Hou Yifan
I.
FIDE Grand Prix, Geneva 2017
Ruy Lopez

1 e4 e5 2 M3 Nc6 3 Bb5 Nf6 4 d3 Bc5 5 Bxc6

This line is all the rage right now, even though its advocates are members
of that dismal brotherhood/sisterhood, the Exchange Lopez.

Question: Why give away the bishop pair unprovoked?

Answer: The idea is that the structure suddenly goes rigid, which may favour White’s knights. This
line is second in popularity to 5 c3.
5 ... dxc6 6 Nbd2
White leaves open the option of castling long. Not 6 Nxe5?? Qd4 (double
attack) 7 Be3 Qxe5 8 d4 Qxe4 9 dxc5 Qxg2 10 Rfl Bh3 and it’s time for
White to resign.
6 ... 0-0 7 Qe2
Now Nxe5 is a real threat.
. Re8
7..
7 ... Nd7 8 Nc4 Re8 transposes.
8 Nc4 Nd7 9 Bd2 Nb6’?

This pawn offer is a known line. Hou Yifan probably was following the
Howell-Adams game from the previous year. 9 ... b5 is Black’s main move,
when 10 Ne3 Nf8 11 h4 a5 12 a4 Bb6 offers balanced chances.
10 Na5
A novelty, given as a one-move possibility by GM Adams in his own
annotations.

Question: A knight on the rim is dim?

Answer: Not all that dim here since a black piece must babysit the b7-pawn. This means that if
White castles queenside, it will take Black time to unravel and push the queenside pawns.
Instead:
a) 10 Ne3 a5 11 h4 Bd4 12 0-0-0 Be6 13 Kbl f6 14 g4!? (14 Nf5 may be
an improvement) 14 ... Bxe3 15 Qxe3 Bxg4 16 Rdgl was D.Howell-
Mi.Adams, British Championship, Bournemouth 2016, where White received
full attacking compensation for the sacrificed pawn, although Black can force
the position with 16 ... Nc4 17 dxc4 Bxf3 18 Qxf3 Qxd2 19 Qxf6 Qel+!
(decoy) 20 Rxel gxf6 to reach an equal ending (as also noted by Adams).
b) 10 Nfxe5 Bd4 11 f4 f6 12 Nxb6 axb6 13 c3 Bxe5 14 fxe5 Rxe5 is equal
too.
c) 10 Ncxe5? is a mistake due to 10 ... f6 11 Nc4 Nxc4 12 dxc4 Bf5, when
Black regains the pawn with a clear advantage due to the powerful bishop
pair and the white queen’s awkward placement on the e-file.
.
10.. Qe7 11a3 Bd6 12 h4’?
Nepo is clearly not interested in playing it safe by castling short.
12 ... Na4
Attacking b2.
13 h5!?
 Which Nepomniachtchi ignores!

Question: Isn’t Nepo just giving up his b-pawn for nothing?

Answer: One must have faith to play such moves, the way some people rely on their horoscopes to
guide their lives. Admittedly, Nepo’s pawn offer - for the veiled promise of a future attack on the black
king - is a bit of a bluff, but it ends up working. Whereas 13 b3 Nc5 14 h5 Ne6 15 Nc4 just looks
balanced.
13 ... Qf8’?

Question: Why didn’t Black accept the b-pawn?

Answer: Bluffs often work. The question facing Black was whether to enter the subterranean
enclosure on b2 with its dual promise of treasure or death? Hou must have seen ghosts. Her best course
was indeed to accept the offer with 13 ... Nxb2! 14 h6 g6 15 a4 b6 16 Nxc6 Qd7 17 Ncxe5 Bxe5 18
Nxe5 Qd4! (threatening ... Nxd3+) 19 0-0 Qxe5 20 Qel (threatening Bc3) 20 ... Nc4! (inflicting
damage to White’s structure) 21 dxc4 Qxe4 22 Qxe4 Rxe4 23 Rfel Bf5 24 Rxe4 Bxe4 25 Bf4 and the
game seems headed for a draw.
14 Nc4
 14 h6 looks more accurate.
.
14.. b5!?
And here 14 ... h6! 15 g4 f6! may be a better defensive set-up.
15 Nxd6 cxd6 16 b3 Nc5 17 Nh4 Ne6 18 Nf5
Covering against... Nd4.
18 ... c5 19 0-0

Question: Why push the h-pawn and then chicken out and castle kingside?

Answer: It’s not a matter of being “chicken”. It’s too risky to castle long since Black’s queenside
pawns are already in motion.
19 ... Nf4 20 QB g6!
20 ... Bxf5?! 21 exf5 wins a pawn since the knight has no place to run: 21
... Qe7 22 Bxf4 exf4 23 Qxf4 with winning chances for White.
21 g3!

Now matters get more complex.

 21 ... Bxf5
a) 21 ... Nxh5?? is a major goof-up due to 22 Bh6 and Black’s queen is
trapped.
b) 21 ... gxf5! was the best way to capture; e.g. 22 gxf4 Kh8 23 fxe5 dxe5
24 exf5 Qe7! 25 f6 (not 25 Qxa8? Rg8+ 26 Qg2 Bb7 and White’s king is in
grave danger) 25 ... Rg8+ 26 Kh2 Qd7 27 Rgl Bb7 28 Qh3 Qc6 and Black
stands no worse.
22 gxf4 Bd7 23 Kh2
Nepo’s once submerged intent reveals itself:
1. White will use the open g-file to attack with rooks and the potential
prying mechanism f4-f5.
2. White counts on the principle: Opposite-coloured bishops favour the
attacker.
Instead, 23 hxg6 fxg6 24 Qg3 exf4 25 Bxf4 Re6 is only slightly in White’s
favour.
23 ... f6?
Black should apply the principle: Meet the opponent’s wing attack with a
central counter, i.e. 23 ... f5! 24 fxe5 fxe4 25 Qxf8+ Rxf8 26 dxe4 Rae8! 27
hxg6 hxg6 28 Rael Rxe5, reaching an even ending.
24 Rgl Kf7!
This is still Black’s optimal defensive plan. Hou correctly realizes that the
kingside will soon be too dangerous for her king. However, her king isn’t
completely safe in the centre or on the queenside either, since White can open
with b3-b4.
25 Rg3 Qe7 26 Ragl Rg8 27 Rlg2 Rh8 28 Kgl
Now the g6-pawn is really attacked.
28 ... Rag8
 Exercise (planning): What is the best way to get at Black’s king?

Answer: Open up a second front on the queenside.

29 b4! c4
If 29 ... cxb4 30 axb4 Qe6 31 fxe5 dxe5 32 Be3 a6 then 33 d4! leaves
Black’s king in serious danger.
30 Be3
It’s logical to keep the centre fluid. The engine suggests the inhuman 30
f5!?, allowing Black to close the kingside, and yet is still heavily in White’s
favour; e.g. 30 ... g5 31 h6 Ke8 32 a4! and Black’s king remains unsafe.
.
30.. Bc6!
Worrying White about a future ... f6-f5.
31 Rh2
More accurate is 31 Bd2 cxd3 32 cxd3 Qe6 33 Qe3 with advantage.
31 ... a6 32 Kfl Qe8?
 Hou makes a couple of nothing moves, confident in her defences. This was
a mistake. After 32 ... exf4! 33 Bxf4 Qe6 Black looks okay.
33 Rh4!
The idea is to reinforce the e4-pawn which frees the d-pawn to advance.
33 ... Qe7?’
Here 33 ... cxd3 34 cxd3 Qe6 was better.
34 fxe5! dxe5 35 d4!
White’s attack begins to pick up traction.
35 ... exd4 36 Bg5!
This unexpected non-capture threatens Bxf6! and the tension in Black’s
position becomes palpable. Hou may have expected 36 Bxd4? g5! 37 Rhg4
Qe6, when Black stands no worse.
36 ... Rf8 37 Rf4!
Even stronger than winning the queen with 37 Bxf6 Qxf6 38 Rf4.
37 ... Ke6 38 Qg4+! Kd6
 If White’s position is a Porsche, then Black’s is the 1978 Ford Pinto. You
know, the one that catches on fire and then explodes.

Exercise (combination alert): How can White go after Black’s king?

Answer: Push the e-pawn forward. Black has three ways to take, all of which lose!
39 e5+! Kc7
Instead:
a) 39 ... Qxe5? hangs the queen to 40 Rxd4+ Ke7 41 Re3 or 40 ... Kc7 41
Bf4.
b) 39 ... Kxe5? 40 Qe2+ Kd6 41 Rxd4+ wins the queen again.
40 Bxf6 Rxf6
40 ... Qe8 41 Bxh8 is equally hopeless for Black.
41 exf6 Qf7 42 Rxd4
White is the exchange up, with Black’s king remaining in danger.
42 ... Qxf6
 43 Qf4+!
Swapping queens is an excellent practical decision since Black has no
chance of holding the ending. Refusal to simplify with 43 Re3!? is a Tal-ish
alternative.
43 ... Qxf4 44 Rxf4 gxh5 45 Rg7+ Kb6
Or 45 ... Kc8 46 Rf6 Be4 47 c3 Rd8 48 f3 Bb7 49 Rxh7 Rd7 50 Rxd7
Kxd7 51 Kf2 Bd5 52 Rf5 and wins.
46 Rf6 h4 47 Rh6 Rd8 48 Rf7
The rook transfers to f6, paralysing Black.
48 ... h3
This desperate diversion is not enough.
49 Rff6 1-0
The only way Hou’s position gets out of this mess is on a stretcher, inside
a body-bag. After 49 ... Rc8 50 Rxh7 Rc7 51 Rxh3 there is no reason for
Black to play on.

Game 25
I.
Nepomniachtchi-E.Inarkiev
FIDE Grand Prix, Geneva 2017
Ruy Lopez

1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 0-0 Be7 6 d3 d6 7 c3 0-0 8 h3

I doubt that ... Bg4 is an actual threat, since White can meet it with the
manoeuvre Nbd2, Rel, Nfl, h2-h3 and Ng3, either picking up the bishop pair
or gaining time on Black’s light-squared bishop. Nonetheless, h2-h3 is
usually useful at some point. 8 Rel is the main line.
8 ... Re8 9 Rel Bf8 10 Nbd2 b5 11 Bc2 h6 12 a4 b4
12 ... Bb7 is also possible.
13 a5
 Question: Doesn’t this move artificially isolate White’s a-pawn and therefore
render it vulnerable?

Answer: Yes, a bit. On the other hand, Black’s now fixed a6-pawn is a target as well, and by
pushing to a5 White enables future ideas like Ba4 or Ra4.
13 ... Rb8 14 Nc4 g6
A small novelty and a logical one. Black’s bishop is well placed on g7
since White will almost certainly attempt to push d3-d4 later on. After 14 ...
Bd7 15 Bd2 bxc3 16 bxc3 Qc8, as in A.Naiditsch-L.Lenic, European Team
Championship, Reykjavik 2015,1 slightly prefer White.
15 Bd2 bxc3 16 bxc3 Bg7
Black should avoid 16 ... Be6 17 Ne3 d5?! 18 exd5 Nxd5 19 Ng4, which
favours White since Black must hand over the bishop pair and light square
control with 19 ... Bxg4 20 hxg4.
17 d4’?
 Nepo’s instincts are to confront and open the game. Given that Black’s
entire position is set up to meet d3-d4, I’m not so sure this is the right choice.
White may have been better off building with 17 Ne3 Bd7 18 Bb3.
.
17.. exd4!

Question: Why did Black voluntarily cede the centre?

Answer: This is a good decision for the following reasons:

1. White’s chances to attack Black’s king are reduced with an open centre.
2. White must watch over his now slightly vulnerable e-pawn.
3. White must also watch for Black’s ... d6-d5 tricks, breaking up the
centre altogether.
4. After c3xd4, the b4-square opens for Black’s knight, which in turn
enables both ... Nb4 and ... c7-c5 possibilities.
18 cxd4 Nb4?’
Going after the defender of e4, yet inaccurate. I wouldn’t call this a hostile
gesture as much as a quasi-military operation designed to push the opponent
around. Black should seize the opportunity for 18 ... d5!, which looks like an
equalizer after 19 Nce5 Nxe5 20 Nxe5 Bb7.
19 Bbl d5!?
This isn’t as effective as it was on the previous move, since it runs into
tactical issues.
20 Nce5!
Threatening Bxb4, followed by a knight fork on c6.
.
20.. Re6?’
Inarkiev is worried about the c6-square, yet his move is in violation of the
principle: Don 7 lift your rook in a crowded middle game, since it may be
vulnerable sooner or later.
His best course was 20 ... dxe4! 21 Bxb4 Rxb4 (not 21 ... exf3?? 22 Nc6)
22 Ba2! (here 22 Nc6 Qd6 23 Nxb4 exf3 is fine for Black) 22 ... Be6 23 Nc6
Qd6 24 Nxb4 Qxb4, when 25 Ne5 c5! offers Black counterplay for the
exchange.
21 exd5! Nfxd5 22 Qcl! Qf8
Covering h6, while evading a potential fork on c6.
23 Be4! Rb5
23 ... Nf6? 24 Qxc7 Rb5 25 Rael is heavily in White’s favour.
24 Rbl c6 25 Qc4! Qd6 26 Rb3!
Nepo with the initiative is a terrifying sight. He threatens to build with
Rebl, or even swing the rook over to the kingside if the f3-knight moves.
26 ... Bf8?
The correct move order was 26 ... Rf6 27 Rebl Bf8.
 Exercise (combination alert): Black’s timing is off and he has overlooked
something with his last move. How should White continue?

27 h4?
Answer: Nepo uncharacteristically misses (or at least misassesses) a tactical opportunity with 27
Nxf7!! (annihilation of defensive barrier) 27 ... Kxf7 28 Ne5+ Rxe5 (or 28 ... Kg7 29 Rg3! g5 30 h4
with a decisive attack) 29 dxe5 Qc5 30 e6+! Ke8 31 e7!! (threatening Bxg6+ and e8Q+, which means
White’s queen isn’t really hanging) 31 ... Bxe7 32 Qxc5 Rxc5 33 Bxg6+ Kd7 34 Rxb4! and Black is
completely busted.
27 ... c5?
That which can be excused in a strategic situation cannot in a position
which demands perfect calculation, such as this one. Black’s status shifts
from miserable to unbearable after this error. 27 ... Nf6 was necessary.
28 h5! g5
 Exercise (combination alert): White has two different combinations here.
Find one of them.

Answer #1: Move the bishop to f5, which wins the exchange.
29 Bf5!
Answer #2: 29 Bxg5! (annihilation of defensive barrier), since acceptance with 29 ... hxg5? 30
Nxg5 Rb7 31 Bh7+ Kg7 32 Rg3 leads to a crushing attack for White.
29.. . Ne7
29 ... Re8 fails to 30 Bxc8 Rxc8 31 Bxb4! and, amazingly, Black has no
good way to recapture the bishop.
30 dxc5!
Undermining the b4-knight. White wins material.
30.. . Rxc5
Or 30 ... Qc7 31 Bg6! fxg6 32 Bxb4 with a completely winning position
for White.
31 Bxb4!
 Even stronger was 31 Bxe6! Rxc4 32 Bxf7+ Kg7 33 Nxc4 Qc7 34 Bxb4
Kxf7 35 Bd6! (White’s c4-knight is tactically protected by a knight fork on
e5) 35 ... Qd8 36 Bxe7 Bxe7 37 Nce5+ and wins since Nc6 picks up another
piece.
31 ... Nxf5
31 ... Rxc4 32 Bxd6 Nxf5 33 Bxf8 also wins a piece, Black doesn’t have
time to recapture due to the hanging rook on c4.
32 Bxc5 Qxc5 33 Qxc5 Bxc5

Black’s position has been found guilty. Now comes the sentencing phase.
34 Rc3! 1-0
Chess players need to eat, just like everyone else. Black loses one of his
bishops. 34 Rcl! is equally good.

Game 26
I.
Nepomniachtchi-L.Aronian
FIDE Grand Prix, Geneva 2017
English Opening

1 c4 Nf6 2 Nc3 e5 3 Nf3 Nc6 4 g3 d5

The Reversed Dragon with Aronian’s coming set-up scores decently for
Black.
5 cxd5 Nxd5 6 Bg2 Nb6
6 ... Be7?? is a common trap at the club level: 7 Nxe5! (discovered attack)
7 ... Nxc3 8 Bxc6+ bxc6 9 dxc3 and Black doesn’t get enough compensation
for the missing pawn.
7 0-0 Be7 8 a3 a5
Suppressing b2-b4. 8 ... 0-0 9 b4 is also completely playable.
9 d3 0-0 10 Be3
Black must watch out for Bxb6 or Ne4-c5 ideas in the future.
.
10.. BeóllRcl a4’?
Otherwise White might put a piece on a4, as with 11 ... Qd7 12 Na4.
 Behind one door is safety, behind the other, adventure. Every speculative
venture begins with a leap of faith. Aronian offers a pawn in the hopes of
complicating. He soon gets his wish.
12 Nd2’?
Nyet, spasibo. Nepomniachtchi said he couldn’t remember anything and
hence declined the offer, but Aronian is insistent.

Question: Doesn’t 11 ... a4 lose a pawn at once to the simple Bxb6 and Nxa4 - ?

Answer: Black is okay with this sacrifice since full compensation is generated. The engine calls it
even at the end of the line 12 Bxb6 cxb6 13 Nxa4 e4! 14 Nel Bg5 (rather than 14 ... Nd4? 15 Nc3 Bg5
16 e3 Nb3 17 Bxe4 Nxcl 18 Qxcl and White is clearly better, I.Nepomniachtchi-D.Bocharov, Apatity
rapid 2011) 15 Rai exd3 16 Nxd3 Nd4 (threatening ... Bb3) 17 Nc3 Nb3 18 h4 Bh6 19 Bxb7 (not 19
Rbl?! Nd2) 19 ... Rb8 20 Bg2 Nxal 21 Qxal and Black is not worse.
12 ... f5!?
Aronian offers the pawn again, under only slightly different circumstances.
12 ... Nd5 is safer.
 13 Bxb6
Nepo accepts this time as there was nothing better.
13 ... cxb6 14 Nxa4 Bg5!
Threatening ... Bxd2, overloading White’s queen.
15 Nc3 e4

16Rbl!
This line has been played a few times since, but White has not managed to
improve on Nepo’s original move, simply breaking the pin on the d2-knight.
For instance, 16 Khl Qd7 17 Rbl Rad8 18 Nc4?! Qf7! 19 b3 (19 Nxb6? Bb3
20 Qel exd3 21 e3 Qe6 is awful for White) 19 ... exd3 20 exd3 f4! gave
Black more than enough compensation in A.Giri-Ma.Carlsen, Shamkir 2018,
where Giri was unable to withstand Carlsen’s initiative.
.
16.. Ne5?’
Pressure mounts on d3, yet this move loses the initiative. Instead, after 16
... Rf7! 17 Nc4 Rd7 18 b3 Bf6 19 Nb5 exd3 20 exd3 Nd4! (20 ... Rxd3?! 21
Qc2 favours White) 21 Nxd4 Bxd4 22 a4 the position looks dynamically
balanced, S.B.Sri-R.Mohite, Mumbai 2020.
17 Nb3! Ng4’?
Now ... Nxf2 is in the air. If 17 ... exd3 18 exd3! Nxd3 (18 ... f4 19 d4
looks good for White too) 19 Qc2 Rf7 20 Rfdl Rd7 21 Bxb7! Rxb7 22 Rxd3
Qc7 23 Rbdl, Black doesn’t get enough for the sacrificed pawn even with the
two bishops.
18 Qc2?’
Not the best response. Nepo should call his bluff and play 18 h3!,
intending 18 ... Nxf2 19 Kxf2! Qd6 20 Nb5 and if 20 ... e3+ 21 Kgl Qxg3 22
N3d4 Bd5 23 Nf3,1 don’t see Black’s compensation for the piece.
.
18.. Be3’’

You again? Where did that come from?! Black threatens ... Nxf2, followed
by ... f5-f4, with a powerful attack. Aronian is one of the most
creative/combative players in the world’s elite and I hope one day to get the
chance to write Aronian: Move by Move.
Nepo is still okay after this shot, yet from a practical standpoint White’s
game is difficult to play without error. This is a position so difficult to
analyse that you either conquer it or it conquers you. There is no middle
ground.
19 dxe4?!
The problem with our digital age is that the engines constantly mock us for
our inferiority. Here we see a rare example of Nepomniachtchi becoming
confused in heavy complications.
a) 19 fxe3? Nxe3 20 Qcl Nxfl 21 Nal, supposedly trapping Black’s
knight, runs into 21 ... Nxh2! 22 Kxh2 f4! and Black’s attack is more
significant than White’s material advantage.
b) 19 Bh3! is the only way to defuse the bomb: 19 ... Nxf2 20 Rxf2 (Black
is unable to play ... f5-f4 here, since his bishop on e6 hangs with check) 20 ...
Qc8 21 Rbfl f4 22 Bxe6+ Qxe6 23 g4! f3 24 exf3 Rxf3 25 Nxe4 Qxg4+ 26
Khl Rxf2 27 Rxf2! Bxf2 28 Nxf2 Rc8 29 Qd2 with a sharp and dynamically
balanced position.
.
19.. Qg5?
Aronian begins to oversell his product, going wrong in turn and allowing
White a winning position. In order to negotiate, leverage is required. Correct
was 19 ... Nxf2! 20 Rxf2 fxe4 21 Rbfl Qc7 22 Bxe4 Rxf2 23 Rxf2 Qf7!
(double attack) 24 Bf3 Bxf2+ 25 Kxf2 Bxb3 26 Qe4 Rf8, when White
doesn’t have enough for the exchange.
20 fxe3 Qxe3+ 21 Khl Qh6
 22 h3?
The complexity level is too much for either player. This move in turn
throws away a winning advantage, which Nepo could have exploited with 22
h4! Ne3 (or 22 ... f4 23 Qd2) 23 Qd2 Bxb3 24 Rf3 f4 25 gxf4 Nc4 26 Qcl
and Black’s attack has gone dead.
22 ... Ne3 23 Qd2 f4! 24 gxf4 Nxfl?
Grabbing material completely surrenders the initiative. 24 ... Nxg2? is also
bad after 25 Kxg2 Qxh3+ 26 Kf2 Qh4+ 27 Ke3 Bxb3 28 Qd4 Rad8 29 Qe5
and White’s king is weirdly safe mid-board, so Black doesn’t have full
compensation for the missing pawn.
The best option was 24 ... Rxf4! 25 Rxf4 Qxf4 26 Nd4 Bxh3! 27 Bxh3
Qg3 28 Qxe3! Qxe3 29 Nf3 Rd8, which the engine calls balanced, though I
would prefer White’s three minor pieces to Black’s queen.
25 Rxfl Bxb3 26 e5 Rae8
Threatening ... Rxe5. 26 ... Rad8 is a touch better.
 27 Ne4

One glance makes it clear that White’s two pawns and knight are worth
more than Black’s rook.
27 ... Kh8
It might have been a good idea to eliminate the monster knight with 27 ...
Ba4 28 e3 Bc6 29 Kh2 Bxe4 30 Bxe4. This is still heavily in White’s favour,
yet less so than in the game.
28 Kh2 Bg8 29 e3 Re6
White’s king has too many able-bodied defenders for Black’s would-be
attack to gain strength.
30 Nd6 Qh4 31 Qd4
Not 31 Bxb7?? Rxe5! and all of a sudden Black no longer stands worse.
.
31.. Rg6
Black is unable to attack with such meagre resources. This mating threat is
easily defused.
 32 RI3 Qel
No better is 32 ... Rh6 33 f5 Qxd4 (or 33 ... Qg5 34 Qf4 Qe7 35 f6!) 34
exd4 Rxd6 35 exd6 Rd8 36 Rd3 Rxd6 37 Bxb7 and White is winning.
33 f5
Nepo’s pawns and pieces have completely taken over in the centre.
33 ... Rg5
33 ... Rxg2+ 34 Kxg2 Qe2+ 35 Rf2 is not even close to perpetual check.
34 h4! Rh5 35 Rg3!

White covers against... Rxh4+ and White threatens either e5-e6, or f5-f6,
targeting g7.
.
35.. Be6
We are all allowed a little joke before resigning, which Black would soon
have to do anyway:
a) 35 ... Qcl 36 f6 g5 37 f7! Rxh4+ 38 Qxh4 gxh4 39 fxg8R+ Rxg8 40
Nf7 mate.
 b) 35 ... Rhxf5 36 Nxf5 Rxf5 37 e6 (threatening mate on g7) 37 ... Rf6 38
e7 wins.
36 fxe6 1-0

The sure-fire way to generate days-long heated arguments and name­

calling on my Facebook page is to ask the dreaded question: “Who is the
greatest chess player of all time?”
The candidates who always pop up, in no particular order, are Paul
Morphy, Bobby Fischer, Garry Kasparov, and Magnus Carlsen. And then you
also get the “eccentrics” who claim Samisch or Bogoljubow or Flohr was the
GOAT. In this book, Nepomniachtchi had the good fortune to play and defeat
two out of the four candidates. (It would have been tough to arrange a game
with either Morphy or Fischer.)
In the 2017 St. Louis tournament, despite a twelve-year layoff, Kasparov
surprisingly held his own against the world’s elite - in fact, he outplayed
many of them and only later threw away wins and draws in time pressure -
which gives us an indication of his massive residual strength. This is a
champion who from 1984 until his retirement in 2005 was the highest rated
player in the world. It wasn’t until 2013 when Carlsen finally passed
Kasparov’s peak rating of 2851 (absolutely staggering for 1999), and Magnus
himself regarded this at the time as the result of rating inflation. If Kasparov
could still hang with the best in the world, more than 20 years after his prime,
just imagine the Kasparov of 1999, the year of his peak rating, seeded into
this field. I would put all the money in my bank account ($37.50!) on the
Kasparov/Magnus duo to dominate.
In the St. Louis tournament I was rooting for this great past champion in
every game (we old guys need to stick together) and when he messed up a
few excellent positions from a combination of rust and time pressure, I felt I
was the one who lost, rather than Garry. In the following game it all went
wrong for him after move thirty.

Game 27
G.Kasparov-I.Nepomniachtchi
St. Louis (rapid) 2017
Grünfeld Defence

1 d4 Nf6 2 c4 g6 3 Nc3 d5
 Kasparov himself plays the Griinfeld with Black.
4 Nf3 Bg7 5 h4! ?

Oh no. Not you too, Garry? I suppose this should be dubbed the AlphaZero
Griinfeld.
. c6
5..
This is Black’s main response now.

Question: Isn’t the move passive?

Answer: Black switches to Schlechter Slav style, with two ideas:

1. To meet c4xd5 with ... c6xd5, thereby denying White the centre, which
is normally the birthright on the White side of the Griinfeld.
2. If White allows it, Black may consider grabbing the c-pawn and
propping it up with ... b7-b5.
6Bg5’?
Great players are not accustomed to being held accountable. Kasparov
goes with the high-risk version. Instead, 6 cxd5 cxd5 7 Bf4 Nc6 8 e3 0-0
looks even since White’s h2-h4 doesn’t make as much sense with a rigid
pawn structure. Nepo has defended this position successfully three times in
both rapid and long play.
6.. . dxc4
Nepo is up for the challenge. After 6 ... Ne4 7 e3 Nxg5 8 hxg5 e6 9 e4
dxe4 10 Nxe4 Qa5+, S.Mamedyarov-S.Karjakin, Beijing (rapid) 2013, White
should perhaps risk 11 Nfd2!? and if 11 ... Bxd4 12 Qf3! Bxb2 13 Nd6+ Ke7
14 c5! Qxc5 15 Qxf7+ Kd8 16 Rdl, I don’t believe in Black’s long-term
survival chances.
7e4
Now the game resembles a weird kind of Geller Gambit, with a Schlechter
Slav set-up from Black.
7.. . Be6’?
Novelty. I stand corrected. It isn’t like a Geller Gambit after all. Hitherto
Black generally played 7 ... b5 8 e5 Nd5, when 9 h5 h6 10 hxg6! hxg5? (the
sacrifice should be declined with 10 ... fxg6) 11 Rxh8+ Bxh8 12 gxf7+ Kxf7
13 Nxg5+ Kg6 14 QfB! gave White a winning attack in I.Nepomniachtchi-
A.Giri, Beijing (blitz) 2013.
8e5
 Question: Why deliberately create a hole on d5?

Answer: In order to get something of value, we must often give up something of value. For the hole
conceded, White has gained:
1. Even more space and enhanced attacking chances, with the e-pawn
pushed to the fifth rank.
2. Black’s fianchettoed bishop hits a wall on e5.
3. By chasing Black’s knight away from f6, White enables h4-h5.
4. The e4-square opens for White’s knight.
8 ... Nd5 9 h5
The consistent continuation. 9 Bxc4?? would be a colossal oversight, since
9 ... Nxc3 wins a clean piece.
9 ... Nd7 10 h6!
Black’s bishop is driven back.
.
10.. Bf8 11Ne4f6
 Black must fight back in the centre or else face slow asphyxiation.
12 Bd2 b5 13 a4 Bf5 14 Ng3 e6 15 Nxf5
Kasparov picks up the bishop pair and enhanced control over the light
squares, at the cost of improving Black’s structure.
15 ... exf5 16 axb5 cxb5 17 Be2 Be7
If Black wants to push his a-pawn two squares, then 17 ... a5! ? is probably
the moment to do so.
18 0-0 a5?’
Here 18 ... a6 was preferable.

Exercise (combination alert): Nepomniachtchi overlooked something on his

last move. How should White continue?

Answer: Step 1: Annihilation of defensive barrier.

19 b3!
White temporarily sacrifices a piece to destroy Black’s queenside
structure.
.
19.. c3
Not 19 ... cxb3?? 20 Qxb3 N7b6 21 Bxb5+ Kf8 22 Bxa5 and Black can
resign.
20 Bxb5!
Step 2: Zwischenzug.
20 ... cxd2 21 e6
Step 3: Pinned piece. White regains the knight with a huge advantage,
while Black’s position is freighted with defensive responsibility.
21 ... 0-0

22 exd7?’
This is the wrong way to capture, since it essentially loses the pawn. After
22 Bxd7! Nb6 23 Bc6 Rc8 24 Bb5 Qd5 25 Qe2 Qxb3 26 Rxa5 White has a
considerable advantage.
22 ... Nc3!
 Double attack/removal of the guard.
23 Bc4+ Kh8 24 Qxd2 Ne4 25 Qe3
25 Qd3 Qxd7 26 Ra2 is still slightly in White’s favour, whereas now
chances are balanced.
25 ... Qxd7 26 Nd2 Nxd2 27 Qxd2 Bb4 28 Qd3
The game looks headed for a likely draw, due to the opposite-coloured
bishops.
28 ... Qd6 29 Ra2 Rae8
Threatening ... Qf4 and ... Qxh6.
30 Re2
Here 30 Rdl! Qf4 31 d5! Rd8 32 Qh3 is even.
.
30.. Qf4
If 30 ... Rxe2 31 Qxe2 Qxd4 32 Qe6 Qd6 33 Qxd6 Bxd6 34 g3 f4 35 Rdl
Rd8 36 g4 Be7 37 Rxd8+ Bxd8 38 Kg2 Be7 39 Kf3 Bf8 40 Kxf4 Bxh6+ 41
Ke4, Black is unable to convert the extra pawn.
 31 g3?!
White can still remain approximately level with 31 Rxe8! Rxe8 32 Qh3
Qxd4 33 Qf3 Qh4 34 Qc6 Rf8 35 Qc7 Qxh6 36 Rdl Bc3 37 Qc5 Be5 38 Qe7
Qg7 39 Qe6, when Black is too tied down to make progress, despite his two
extra pawns.
31 ... Qxh6 32 Kg2 f4?!
Black has better chances with 32 ... Rxe2! 33 Qxe2 Rd8 34 Qe6 Qf8 35
Rdl Rd6 36 Qe3 Qd8 37 Rd3, though it still won’t be easy to make progress.
33 Rhl?
The first real mistake. Instead, 33 Rxe8! Rxe8 34 Bf7 Rd8 35 Rhl
(threatening Bxg6 if Black’s queen moves) 35 ... f3+! (the only move; if 35 ...
Qg7? 36 Bxg6 h6 37 g4, Black is going to have a tough time saving the
game) 36 Qxf3 Qg7 37 Be6 f5 38 d5 is equal.
33 ... Qg5 34 Rxe8?’
The start of an incorrect combination. 34 Re6! is White’s best chance to
hold the game.
.
34.. Rxe8
If your opponent is intent on suicide, then be helpful and assist him.
35 Bf7??
 It looks as if Kasparov has outsmarted Nepo as White threatens both Bxe8
and Bxg6. In reality it’s a time-pressure blunder and a sign of rust from not
having played for so long.

Exercise (combination alert): How should Black respond?

Answer: Interference.
.35.. Re3!
In a single crushing move, White’s position is disgorged of dark squares.
36 fxe3
You go on vacation to a beautiful destination where you take some
amazing photos, only to realize on returning home that you forgot to take the
lens cap off. This is how Kasparov must have felt, since on the previous
move he probably believed himself to be winning.
Black wins on both acceptance and decline of the rook sacrifice. 36 Qa6
allows 36 ... 13+ 37 Kh2 Qh5+ 38 Kgl Rel+ and mates in two.
 36 ... Qxg3+ 37 Kfl f3! 0-1
White is unable to cover the dual mating threats on g2 and el, while 38
Rxh7+ (38 Qc2 covers g2, but not 38 ... Qel mate) 38 ... Kxh7 39 Qxg6+
Qxg6 40 Bxg6+ Kxg6 is just winning for Black; e.g. 41 Kf2 f5 42 Kxf3 (it is
White’s rotten luck that Black is left with the correct bishop for the rook’s
pawn) 42 ... Kf6 43 e4 fxe4+ 44 Kxe4 Ke6 45 d5+ Kd6 46 Kd4 Bel 47 Kc4
Ke5 48 Kc5 Bb4+ 49 Kc6 Kd4! 50 d6 Bxd6 51 Kxd6 Kc3 52 Kc5 Kxb3 etc.

Game 28
Ma.Carlsen-I.Nepomniachtchi
London Classic 2017
Slav Defence (by transposition)

1 Nf3 c5 2 c3’?
 Question: Why would someone as supremely tactically gifted as
Magnus Carlsen play in such a ridiculous, dull manner?

Answer: First of all I insist on an apology, since this opening is named after your writer. My friend
GM Jesse Kraai christened it (probably sarcastically!) “The Lakdawala Accelerated”, after he saw me
bore a GM to tears with 2 c3 at the Far Western States Open, Reno 2006. Magnus may have wanted to
keep the game strategic and controlled, rather than mess with Nepo in tactical, open or irrational
positions. It’s not as dumb as it looks anyway, since d2-d4 will inevitably follow, which means Black
must decide in advance what sort of Queen’s Pawn Opening he wants to defend. I have been playing 2
c3 for two decades and have yet to lose, so it can’t be that bad, can it?
2 ... d5
Instead, 2 ... Nf6 3 d4 e6 4 Bg5 turns the game into a Torre Attack; e.g. 4
... d5 5 e3 h6 6 Bh4 Nc6 7 Nbd2 a6 8 Bd3 Be7 9 0-0 Nd7 10 Bxe7 Nxe7 11
Ne5 cxd4 12 exd4 Nxe5 13 dxe5 Bd7 14 Rel Rc8 15 Nf3 with an edge for
White, who owns extra space and controls the d4-square, Ma.Carlsen-
J.Xiong, Isle of Man 2017.
3 d4 cxd4

Question: Why is Nepomniachtchi willing to enter the dull Exchange Variation

of the Slav?

Answer: It is both a sign of respect and of confidence. Nepo indicates, firstly, that he doesn’t mind
a draw with the world champion, especially with the black pieces; and secondly, that he is confident of
making one in even a dry, technical position. Perhaps overconfident as things turn out.
4 cxd4 Nc6 5 Nc3 Nf6 6 Bf4 Nh5’?
 The standard moves are 6 ... Bf5 and 6 ... a6.
7 Bd2 Nf6
The unspoken words are: “I offer a draw, if you want to keep repeating the
position by moving the bishop back to f4.” Magnus is not about to do that
with the white pieces, or the black pieces for that matter.
8e3

Question: Isn’t this a big concession, hemming in White’s dark-squared bishop?

Answer: I don’t believe so. For it Magnus was gifted an extra tempo. Why? Because White’s
bishop moved twice to reach d2, while Black’s knight moved three times to reach f6.
. e6
8..

Question: Why did Nepomniachtchi not play his bishop outside the pawn chain?

Answer: 8 ... Bf5 9 Ne5! is difficult for Black, since 9 ... e6? runs into 10 g4! Be4 11 f3 Bg6 12 h4!
h5 13 Nxg6 fxg6 with a strategically won game for White due to the bishop pair, light square control
and a juicy target on g6, which makes Black’s king move after 14 Bd3 Kf7.
9 Bd3!
Seizing control over e4.
. Bd6
9..
White is a move up in an otherwise dull position, which reminds us of one
of those plotless novels which puts us to sleep by page 50. Magnus’ next
move changes everything.
10 e4!

Principle: The side ahead in development should create confrontation and

open the position. Siegbert Tarrasch was mean and hurt my feelings when he
declared that a person who is afraid to take on an isolani (i.e. me) should give
up chess. 10 0-0 would be normal.
.
10.. dxe4
10 ... Nb4? hangs a pawn to 11 Bb5+ Nd7 12 a3 Nc6 13 exd5 exd5 14
Qe2+ Qe7 15 Nxd5.
11 Nxe4 Be7’?
 Nepo gives up another tempo! 11 ... Nxe4 12 Bxe4 helps White fight for
control over d5 with a slight advantage, H.Alf-J.Perez Gutierrez,
correspondence 2008.
12 0-0 0-0
After 12 ... Nxd4?! 13 Nxd4 Qxd4 14 Bc3 Qd5 15 Q13 0-0 16 Rfdl
White’s lead in development is more important than Black’s extra pawn.
13 Qc2’?

Question: Isn’t this an unusual decision, placing the queen on the open c-file,
rather than on e2?

Answer: Agreed. Carlsen’s ideas behind his unorthodox move:

1. His queen and bishop battery may provoke ... h7-h6, which then
provides White with a sacrificial target on h6.
2. It may provoke ... g7-g6, which weakens the dark squares around
Black’s king.
 13 ... h6!?
Nepomniachtchi provokes back, offering up h6 as the target. Instead:
a) 13 ... g6 14 Radi Qb6 15 a3 Bd7 16 Bh6 Rfd8 would be a more
standard isolani position, although I still prefer White, a fact which would
bring joy to Siegbert Tarrasch’s heart.
b) 13 ... Nxe4?! 14 Bxe4, with a double attack on h7 and c6, wins a pawn
if White wants it.
14 Radi! Bd7
Here too 14 ... Nxd4?! 15 Nxd4 Qxd4 16 Bc3 Qb6 17 Nxf6+ Bxf6 18
Bxf6 gxf6 19 Qd2 offers White dangerous attacking compensation.
15 a3
Preventing ... Nb4.
15 ... Rc8
Now ... Nxd4 is a real threat.
16 Nc3!
Principle: The side with the isolani should retain pieces on the board. In
an earlier game, 16 Qbl Qb6 looked okay for Black, E.Safarli-Z.Mammadov,
Baku 2012.
16 ... a6
Intending to generate queenside play with ... b7-b5-b4. I prefer the plan of
16 ... Bd6!, followed by ... Ne7 and ... Bc6, seizing control over d5 while
transferring a helpful defender to the kingside.
17 Qcl
 Now Nepo must watch out for Bxh6 sacrifices.
.
17.. Re8!
Clearing the way for ... Bf8.
18 Rfel

Question: Did Carlsen miss an opportunity for a promising sacrifice on h6?

Answer: Hardly. 18 Bxh6? is premature: 18 ... gxh6 19 Qxh6 Bf8 20 Qg5+ Bg7 and White’s
pseudo-attacking chances are not worth the piece he sacrificed.
18 ... Bf8 19 Bf4 b5!
Nepomniachtchi has achieved dynamic equality and now worries White
with ideas of... b5-b4 and ... Na5-c4.
20 Qd2 b4!
Here 20 ... Na5 21 Ne5 Nc4 22 Bxc4! bxc4 23 Re3! looks dangerous for
Black, since White’s attack is coming quickly.
21 axb4 Nxb4 22 Ne5!!
 Question: Isn’t Carlsen’s last move a gift for Black, since it gives Nepo both
the bishop pair and the light squares?

Answer: Just look at the broken principles:

1. Don’t offer the opponent the bishop pair in an open position.
2. Don’t allow swaps if you are the one with the isolani, since your
attacking chances reduce with each exchange.
Magnus correctly assesses this position to be an exception for these
reasons:
1. By allowing ... Nxd3, Black is deprived of a powerful piece, his b4-
knight.
2. This is Carlsen’s incredibly deep point: White’s attack is dark square­
based, so his light-squared bishop isn’t as critical a component.
3. White plans to lift major pieces to the third rank and ... Nxd3 hands
White a tempo/free attacking move.
 Pretty much everyone else in the universe would have played the
automatic 22 Bbl?! Nbd5 23 Be5 Bb4, when Black stands slightly better.
22 ... Nxd3 23 Qxd3 a5
Slightly inaccurate. The comp suggests giving the a-pawn away, to reduce
White’s attacking force and weaken d5, with 23 ... Bb5! 24 Nxb5 (otherwise
the bishop will go on to c4 and d5) 24 ... axb5 25 Qxb5 Nd5, when Black has
full compensation for the pawn due to:
1. Control over the d5-hole, which gapes wide as a toothless mouth.
2. A towering knight on that square.
3. Isolated pawn targets on b2 and d4.
4. White’s attack has gone dead.
24 Qf3!
It’s invisible moves like this which win games. Advantage White:
1. He may later probe with Bg3 and Bh4.
2. Black’s f6-knight is frozen in place due to the weakness of 17.
3. White followed the principle: Fight for control over the square in front
ofyour isolani.
4. Black’s ... Bc6 is prevented for now.
24 ... Bb4?
Correct was 24 ... Qe7! to cover f7 and free the black knight.
 Exercise (combination alert): Prove why Black’s last move was an error.

Answer: It allowed White a promising piece sacrifice.

25 Re3’?
While this move is dangerous, far stronger was 25 Bxh6H (annihilation of
defensive barrier) 25 ... Bxc3 26 bxc3 gxh6 27 Qg3+ Kh7 (forced; if 27 ...
Kf8? 28 Rd3! Bc6 29 Qf4 Ke7 30 c4!, there is no cure to the coming d4-d5)
28 Nxf7 Qe7 29 Qd3+ Kg7 30 Ne5 (threatening Qg6+) 30 ... Rf8 31 Qg6+
Kh8 32 Qxh6+ Qh7 33 Ng6+ Kg8 34 Qg5! Qh5 35 Qxh5 Nxh5 36 Nxf8
Kxf8 and now 37 Re5 (the engine prefers 37 d5!?) 37 ... Nf4 38 Rxa5 Rxc3 is
simplest, leaving White with rook and three pawns for two minor pieces,
which should be decisive.
25 ... Bxc3’?
Nepo is understandably worried about the momentum of Magnus’ attack,
so he reduces material. He concedes a lot for it:
 1. He greatly improved White’s structure.
2. He gave up his bishop pair.
3. Now he must constantly watch for the ideas c3-c4, followed by either
c4-c5 or d4-d5.
Then again, 25 ... Bf8 is also difficult after 26 Bg3! Qe7 27 Bh4
(threatening Ng4) 27 ... g5 28 Bg3 Bg7 29 h4 and Black is in trouble since
his kingside cover is shaky and he must still keep watch over d4-d5 tricks.
26 bxc3
Magnus has played brilliant strategic chess but now loses concentration,
and his game plunges quickly. He said afterwards: “I failed to predict a single
one of his moves.” From this point on Nepo pulls out his Santa Claus-sized
bag of defensive tricks.
26 ... Ba4!
Intending ... Bc2! and ... Bh7, adding a key defender to his king.
27 Rai?!
This doesn’t make sense, as it just chases the black bishop where it wants
to go. 27 Rcl! is better for White.
27 ... Bc2! 28 h3!?
White’s king is given luft, at the steep price of denying his rook or queen
the attacking h3-square. The alternative is 28 Rcl Bh7 29 c4! Qxd4 30 Rdl
Qc5 31 Bxh6 and the engine calls it even.
28 ... Bf5?!

Question: What? Didn’t Nepo gift his opponent g2-g4, a useful attacking move,
for free?

Answer: He did just that and I don’t understand the motivation, unless Nepo thought that g2-g4
favours Black - which it doesn’t. Black should chance 28 ... Nd5! 29 Bxh6 Qe7 30 Qg3 f6 31 Re2
Rxc3 32 Nf3 Qc7 33 Qxc7 Rxc7 34 Rxa5 (threatening Rxd5) 34 ... Bdl! 35 Rel Bxf3 36 gxf3 KÍ7!,
after which he stands no worse.
29 g4!
 Thanks for the tempo!
29 ... Bh7?
This intended move is now a mistake. 29 ... Bg6 was necessary, to cover f7
and enable ... Nd5, when 30 c4 (30 Nxg6 fxg6 is nothing to worry about) 30
... Qxd4 31 Rdl Qc5 32 Bxh6 Nh7! 33 Nxg6 fxg6 34 Rd7 Re7 35 Re5 Rxd7
36 Rxc5 Rxc5 37 Qe3 Rc6 should still be defensible.
30 c4’?
Magnus gives up a pawn to prevent... Nd5, which isn’t necessary. He has
a strong attack with the simple and stronger 30 g5! Nd5 (if 30 ... hxg5 31
Bxg5 then Ng4 is a dangerous threat) 31 Nxf7! Qb6 32 Nd6 Nxe3 33 Qxe3
hxg5 34 Be5 and White is clearly better.
30 ... Nd7
Nepo senses that a bomb is about to detonate and is desperate to put
distance between himself and the shrapnel blast. 30 ... Qxd4!? is not as
effective here, as after 31 Rdl Qa7 32 Bxh6 Bg6 33 h4 White has a
dangerous attack brewing.
31 Nc6?’
Tempting, yet inaccurate. Going for the a5-pawn costs White his initiative
and bind. Magnus incorrectly rejects 31 c5! Nxe5 32 Bxe5 f6 33 Bd6,
keeping the bind and all the winning chances.
31 ... Qf6 32 Nxa5 Nb6!
Suddenly both c4 and d4 hang simultaneously and Black looks fine.
Magnus may have expected 32 ... Qxd4?! 33 Rdl Qa7 34 Nb7! Rxc4 35 Nd6
Rxf4 36 Qxf4 Rf8 37 Qd4, when the extra exchange would allow him to keep
pressing for a long time.
33 c5?
This is now incorrect. He should have played 33 Rea3 Nxc4 34 Nxc4
Rxc4 35 Ra8 Rcc8 36 Qb7, when a draw is a virtual certainty.

Exercise (combination alert): Prove why White’s last move was a blunder.
 Answer: Black can just take the pawn since the d4-pawn is pinned.
33 ... Rxc5!
Even a reigning world champion is still human and may on rare occasions
overlook a basic tactic due to any of ... fatigue, time pressure, loss of
concentration, inaccurate assessment, or paranoia.
34 dxc5?!
The shock of the first error produces a second. After 34 Be5 Qxf3 35 Rxf3
White should still save the game.
34 ... Qxal+ 35 Kh2 Qxa5 36 Qc6?
Here he had to play 36 cxb6 Qxb6 and then suffer for the next 40 moves,
praying for a draw.

Exercise (combination alert): Magnus hallucinated for the third time.

Prove why his double attack fails miserably.

Answer: X-ray defence.

 36.. . Qa4!
In a single magical move, Black’s pieces reform into perfect alignment.
The formally loose e8-rook is covered, while at the same time the white
bishop hangs on f4. White loses a piece, no matter how he plays it.
37 Qxa4
Or 37 Qxb6 Qxf4+ and White resigns.
37 ... Nxa4 38 c6 Nb6 39 c7 f6!
Threatening to block out the f4-bishop with ... e6-e5.
40 Rb3 Nc8 0-1
White’s passed c-pawn isn’t going anywhere.

Game 29
Nepomniachtchi-D .Howell
I.
World Rapid Championship, Riyadh 2017
Nimzo-Indian Defence

I have had the pleasure of listening to GM David Howell’s commentary

many times on Magnus’ site, Chess24.com.
1 d4 Nf6 2 c4 e6 3 Nc3 Bb4 4 Qc2
The Capablanca Variation is perhaps White’s most solid choice against the
Nimzo-Indian.
4 ... d5
Alternatives are 4 ... 0-0 and 4 ... c5.
5 cxd5 exd5 6 Bg5
The position has shifted to a sort of Ragozin.
6.. . c5
Black can play also 6 ... h6 first.
7 Nf3
Here 7 dxc5 h6 8 Bh4 g5 9 Bg3 Ne4 10 e3 Qa5 11 Nge2 Bf5 12 Be5 0-0
13 Nd4 is more popular, and then it’s a battle between your engine and the
opponent’s.
7 ... 0-0
After 7 ... h6 8 Bxf6 Qxf6 9 a3 Bf5 10 Qa4+ Bd7 11 Qb3 Bxc3+ 12 Qxc3
cxd4 13 Qxd4 Qxd4 14 Nxd4 Nc6 15 Rdl, Black’s lead in development
compensates for White’s superior structure, as in Bu Xiangzhi-A.D.Nguyen,
Macau 2007, and later S.Mamedyarov-Bu Xiangzhi, World Rapid
Championship, Riyadh 2017.
 8 e3 h6 9 Bh4
This indicates that Nepo is willing to allow the game to go crazy. 9 Bxf6 is
the tamer version.
9 ... g5
The main alternatives are 9 ... c4 and 9 ... Nc6. As for the latter, maybe it’s
stylistic bias but I don’t trust Black’s position after 10 dxc5 d4 11 0-0-0.
10 Bg3

.10.. Ne4’?
Nobody can accuse GM Howell of lacking a can-do attitude. This move is
completely thematic and structurally unsound. Howell reasons that inaction
in such a position is tantamount to suicide. But is it? If you put his position
though the engines, they all say the same thing: White’s structure is of greater
value than Black’s development.
Instead, 10 ... c4! 11 Nd2 Nc6 12 Be2 Re8 13 a3 Bxc3 14 bxc3 Ne7 15 e4
Nxe4 16 Nxe4 dxe4 17 Bxc4 Nd5 18 0-0 Be6 looks balanced,
D.Batkovskyte-P.Guichard, European Women’s Team Championship, Porto
Carras 2011.
11 Bd3 Bf5
I like White after 11 ... Nc6 12 dxc5 Bxc3+ 13 bxc3 Nxc5 14 h4!,
L.Javakhishvili-A.Vakhania, Georgian Women’s Championship, Tbilisi
2015.
12 dxc5 Nd7
The idea is to go after the d3-bishop. White must react vigorously.
13 Nd4

Gaining a tempo on the f5-bishop.

13 ... Bxc3+?’
In business it’s a bad idea to hire the very first applicant for a job. This
novelty is not a good one, since Black’s position sinks fast without coverage
of the dark squares. Instead:
a) 13 ... Nxc3? is even worse: 14 Nxf5! Ne4+ 15 Kfl Ndxc5 16 13 and
Black is in deep trouble.
b) 13 ... Bg6 14 f3 Nxc3 15 bxc3 Bxd3 16 Qxd3 Nxc5 17 Qc2 Ba5 was
V.Shishkin-R.Doros, Arad 2013, where I like White’s chances after 18 h4!.
c) 13 ... Nxg3! 14 hxg3 Bxd3 15 Qxd3 Nxc5 16 Qdl Qf6 should probably
be played, when White only stands a shade better after 17 Qf3 Qxf3 18 gxf3
Na4 19 a3 Bxc3+ 20 bxc3 Kg7 21 Kd2.
14 bxc3 Bg6 15 f3!
Eliminating Black’s best piece, while forcing the opening of the h-file.
15 ... Nxg3 16 hxg3 Bxd3
16 ... Qe8 (covering g6 and menacing e3) is answered by 17 Ke2 h5 18
Nb5 (threatening a fork on c7) 18 ... Rd8 19 Nc7 Qe5 20 Bxg6 Qxc7 21
Rxh5! Qxg3 22 Rahl! Qxg2+ 23 Kel Qxc2 24 Bxc2 (threatening Rxg5 mate)
and if 24 ... f6 then 25 Bf5! wins since 25 ... Nxc5 26 Rh8+ Kf7 27 Rlh7+
Ke8 28 Bg6 is mate.
17 Qxd3 Nxc5 18 Qc2!
 White leaves f5 vacant for his knight. The momentum has quickly drained
away from Black’s temporary initiative. The opening did not go well for
Howell and in fact he must be lost here. Black is about to lose a pawn since
he has no good way to protect h6, while White’s king will be safe on f2.
18 ... Qb6 19 Nf5 Rfe8 20 Nxh6+ Kf8 21 Nf5
Threatening mate on the move.
21 ... QÍ6 22 Kf2 Ne4+’?
When we are hopelessly trapped, remember that there is still a method of
escape: Suicide. Howell’s move is unsound but can we blame him? The dead
weight of certain defeat sits upon his chest, no matter what he plays. 22 ...
Re5 23 g4 Rc8 24 Rh6 Rxf5 25 Rxf6 Rxf6 26 Rdl is also lost for Black.
23 fxe4 Rxe4 24 g4!
Now Black’s attack is sent back to the nothingness from whence it arose.
24 ... Rxg4
After 24 ... Re5 25 Rafi Black can resign since there is zero compensation
for the missing piece.
 Exercise (planning): Black’s attack is close to a catatonic state, such that simply
25 Kgl wins. Do you see an even stronger path?

Answer: Step 1: Gain a tempo on Black’s queen, while breaking the pin on the knight.
25 Rh6! Qe5 26 Rahl 1-0
Step 2: Activate the al-rook, threatening back rank mate. Howell doesn’t
need to see 26 ... Kg8 27 Nd6! (threatening Qh7+ and Qxf7 mate, among
other things) 27 ... Qg7 28 Qf5 (threatening Rh8+) 28 ... Rh4 29 Rlxh4 gxh4
30 Qh5 (threatening Rh8+ again) 30 ... K18 31 Nf4 and there’s no longer
anything to be done. Nepo makes it all look so obvious and easy.

Game 30
Nepomniachtchi-E.Sutovsky
I.
Poikovsky Karpov 2018
French Defence
1 e4 e6 2 d4 d5 3 Nc3 Nf6 4 e5 Nfd7 5 f4 c5 6 Nf3 Nc6 7 Be3 cxd4
Black has three popular moves in this basic position from the Steinitz
French: 7 ... a6, 7 ... Be7 and the text.
8 Nxd4 Qb6’?

This is less usual; 8 ... Bc5 is the main line. Playing a move like 8 ... Qb6
against the normally blood-thirsty Nepo requires the élan of a matador who
flourishes the red cape to further provoke the already enraged bull.

Question: Isn’t it insanity to deliberately place your queen on the same

diagonal as your opponent’s bishop?

Answer: Black goes pawn grabbing early on. Just like its cousin, the Najdorf Poisoned Pawn, this
line is well suited for those with suicidal tendencies and an endless capacity to endure defensive pain,
which is probably why I play it from Black’s side. The engines say it’s completely playable and,
strangely enough, is often deployed as a drawing weapon at the highest levels. The problem arises with
practical chances: Black’s position is a lot more difficult to play than White’s.
 9 Qd2
This is the principled choice, offering the b2-pawn to gain a lead in
development. It’s not made out of a benevolent mind state. Of the alternatives
- 9 a3, 9 Na4 and 9 Ncb5 - the first two are wimpy and the third has been
worked out to equality for Black.
9 ... Qxb2 10 Rbl Qa3 11 Bb5
Theory has coalesced around this simple developing move, which gives
Black more problems than either 11 Ndb5 or 11 Ncb5.
11 ... Nxd4
Exchanges benefit Black. The Nimzowitschian 11 ... Ndb8!? is a specialty
of GM Sergey Volkov, but optically isn’t all that tempting.
12 Bxd4 a6!

Question: Isn’t Black giving the pawn back with this move?

Answer: Yes, but he gets something important for it in White’s powerful light-squared bishop.
Today it’s considered wise to return the pawn to deflate White’s attack. In the olden days we used to
play 12 ... Bb4.
13 Bxd7+
The passive retreat 13 Be2 doesn’t make much sense; after 13 ... Bc5
Black already stands clearly better.
13 ... Bxd7
13 ... Kxd7?? takes love of material to new heights: 14 Rb3 Qe7 (forced)
15 Ne4! gives White a crushing attack.
14 Rb3!
This finesse chases Black’s queen to an unfavourable square. 14 Rxb7?!
Bb4, as in D.Aldama-C.Lakdawala, San Diego (rapid) 2016, is a far superior
version for Black, whose queen is a million times better off on a3 than on e7.
14 ... Qe7

Question: Why e7 when a5 was available?

Answer: It is my unpleasant duty to inform you that you fell for a trap if you wanted to play 14 ...
Qa5??, as after 15 Bb6 the queen then has no place to go.
15 Rxb7 Rc8
 We reach a tabiya position in the variation.
16 f5
A slightly unusual move order. More often White plays 16 0-0 first.
16 ... exf5 17 0-0!
More accurate than 17 Nxd5 Qh4+ 18 Bf2 Qe4+ 19 Be3 Be6 20 Rb8!
Qxe5 21 Rxc8+ Bxc8 22 0-0 f6! 23 Rel Kf7, when I like Black’s chances.
.
17.. Qd8
Transposing back to a standard position, at least for one move. Here Black
might also play 17 ... Qe6, when 18 Rfbl Qc6 19 Qe3 f4! 20 Qxf4 Bc5 21
Ne2 0-0 is equal, O.Lizarzaburu-N.Bernal Varela, correspondence 2019.
18Qf2
A novelty. 18 Qd3 is normally played, when 18 ... Bc5 19 Bxc5 Rxc5 20
Nxd5 Bb5 21 Qxf5 0-0 has led to a lot of draws. Nepo decides to try
something else.
18 ... Rc4
 Perhaps not the nest response. After 18 ... Be6 19 Rfbl Qa5 20 Ra7 Be7
21 Nb5 0-0! 22 Nd6 Rb8! 23 Rfl Bd8 Black looks okay.
19 Rfbl Bc8 20 Ra7 Be7?
Here 20 ... h5! was necessary, still with a playable position.

Exercise (combination alert): What did Black overlook on his last move?

Answer: Step 1: Queen trap.

21 Bb6!
I told you this line is difficult for Black to navigate.
21 ... Rxc3
Sadly forced. Sutovsky may have hallucinated, thinking he could play 21
... Qxb6??, followed by ... Bc5+, regaining the queen. Unfortunately, White
has the devastating zwischenzug 22 Rxe7+! which completely short circuits
Black, who must resign.
22 Khl!
 Step 2: Another zwischenzug. After 22 Bxd8?! Bc5 Black again regains
the lost queen, although 23 Qxc5 Rxc5 24 Be7 is no picnic either.
22 ... Qd7
22 ... Bc5 23 Bxc5 saves the queen, at the cost of a dead lost position for
Black.
23 Rxd7 Bxd7
Black has rook, bishop and pawn for the queen, which is just about enough
materially. The problem isn’t material as much as an insecure king.
24 Bd4 Rc4 25 Rb7
Preventing castling.
25 ... Bc8 26 Ra7 h5!
This manoeuvre is actually typical of the variation. Black activates his
sleeping h8-rook.
27 c3 Rh6 28 Ra8 Rhc6 29 h3

It’s wise to make luft for White’s king.

 29 ... Bd8 30 Qf3!
Double attack. White wins a pawn.
.
30.. Be6
There are mass desertions from the battlefield. Black begins to bleed
pawns over the next fifteen moves.
31 Qxh5
There goes pawn number one. The threat is Qh8+.
31 ... Kd7
The black king bolts with the urgency of a jackrabbit evading a hungry
wolf. 31 ... Rc8 hangs another pawn to 32 Rxa6.
32 Qh8!
Double attack, winning pawn number two.
32 ... Rc8 33 Ra7+!
More accurate than 33 Rxc8 Rxc8 34 Qxg7 which is also winning for
White, who will eventually push his newly passed h-pawn down the board.
33 ... R4c7 34 Rxa6
Yet another black pawn falls. Black doesn’t have the time to cover g7.
34 ... Be7
Not 34 ... g6? 35 Rd6+ Ke7 36 Be3!, threatening a deadly check on g5, to
which there is no good defence.
35 Qxg7
 There goes pawn number 3. How depressing for Black, who since move
30 lost his a-, h- and g-pawns.
35 ... f4 36 Qh7
Bringing the queen back round, while covering bl against a black rook
invasion.
36 ... Rb7 37 Qc2 Ke8 38 Qf2 Bg5 39 Qf3 Kd7 40 Kh2 Rc4 41 Rd6+ 1-
0
The conversation is now past tense and White’s win is just a matter of
time.

The next game is an example of Nepomniachtchi’s almost supernatural

intuition in open positions.

Game 31
N epomniachtchi-V.Kr amnik
I.
Dortmund 2018
Ruy Lopez

1 e4 e5 2 M3 Nc6 3 Bb5
There is always debate on my Facebook page when someone, looking to
pick a fight, announces that the Scandinavian is unsound or that 1 ... b6 is
unplayable. The Ruy Lopez, on the other hand, is as invulnerable to criticism
as our religious or political beliefs, which almost never change.
3 ... a6 4 Ba4 Nf6 5 0-0 Bc5

The ... Bc5 versions of the Lopez are anything but rusty with disuse.
Carlsen, Caruana, Aronian and Kramnik all have faith in this Lopez subset,
which is now officially a mainstream line. I count nearly 5,000 games with
this position in the database.

Question: Isn’t the bishop misplaced on c5, allowing White to gain a

future tempo with c2-c3 and d2-d4 - ?
 Answer: Black gives up something to get something. Yes, the bishop on c5 is exposed to d2-d4. On
the other hand, it is far more aggressively posted than on the traditional e7-square.
6 c3
6 Nxe5 looks a lot stronger than it actually is: 6 ... Nxe5 7 d4 b5 8 Bb3
Bxd4 9 Qxd4 d6 (setting up a Noah’s Ark trap with ... c7-c5-c4) 10 f4 c5 11
Qc3 Neg4 12 e5 c4 13 exf6 0-0! 14 Bxc4 bxc4 15 fxg7 Re8 16 h3 Bb7! 17
hxg4 Re2 18 Qg3 (18 Rf2?? loses to 18 ... Qb6) 18 ... Rxg2+ 19 Qxg2 Bxg2
20 Kxg2 Qh4 and Black stood no worse, A.Feher-M.Godeña, Sarajevo 2010.
6 ... b5
After 6 ... 0-0 7 d4 Ba7 8 Bg5 h6 9 Bh4 exd4 Black looks fine; e.g. 10 e5
g5 11 Bxc6 dxc6 12 Nxg5 hxg5 13 Bxg5 dxc3 14 Qxd8 Rxd8 15 Nxc3 Bf5
16 exf6 Bd4 with full compensation for the sacrificed pawn, I.Duzhakov-
G.Bazeev, St. Petersburg 2018.
7 Bb3 d6 8 a4
The immediate 8 d4 is White’s main line, when 8 ... Bb6 9 a4 is a tabiya
position.
8.. . b4’?
8 ... Rb8 is a safer version.
9 d4 Ba7 10 Bg5
Threatening Bd5. This was actually a novelty. Instead:
a) 10 dxe5 Nxe5 11 Nxe5 dxe5 12 Qxd8+ Kxd8 doesn’t give White much,
if anything.
b) 10 h3 0-0 11 Be3 is another option. Note that White’s e-pawn is most
certainly not hanging, since 11 ... Nxe4?? 12 Bd5 wins a knight. I really wish
ChessBase would add a three-question mark option in their annotation pallet.
10.. . Rb8!
 Kramnik, as usual, demonstrates a perfect understanding of the position’s
strategic elements and finds Black’s best response over the board. The rook
covers b4 while pre-empting a pin with Bd5. The tempting 10 ... h6?! is
incorrect: 11 Bxf6! Qxf6 12 Bd5 Bb7 13 cxb4! is in White’s favour since 13
... exd4?? loses instantly to 14 Qc 1!, winning a piece in view of 14 ... Nd8 15
Bxb7 Nxb7 16 Qc6+.
11 Bd5!?
Nepo plays it anyway. 11 a5 0-0 12 Ba4 Ne7! is fine for Black, as 13 dxe5
is answered by 13 ... Nxe4 14 Bh4 d5.
.
11.. Ne7!
Better than 11 ... Bd7, when 12 Bxc6 Bxc6 13 dxe5 dxe5 14 Qe2
(threatening Qxa6) 14 ... Qc8 15 Nxe5 Ba8 16 Bxf6 gxf6 17 Nd3 is in
White’s favour.
12 dxe5
12 Bxf6 gxf6 doesn’t offer White any advantage.
 12 ... Nfxd5 13 exd5 0-0!
Kramnik correctly offers a pawn in return for development and the bishop
pair in an open position. He had little choice in any case, since 13 ... dxe5? 14
Nxe5 (threatening a fork on c6) 14 ... Bb7 15 Bxe7 Kxe7 16 Rel Kf8 17 c4 is
awful for Black.
14 exd6
14 Rel dxe5 15 Bxe7 Qxe7 16 Nxe5 Qh4 offers Black full compensation
for the pawn.
14 ... Qxd6 15 c4 Nf5 16 Nbd2 f6 17 Bh4 Qf4?!
A correct and good idea, but the wrong implementation. Kramnik either
missed, or underestimated, the power of Nepo’s 20th move. After 17 ... g5!
(or 17 ... c6) 18 Bg3 Nxg3 19 hxg3 Qxg3 20 c5 Qf4 Black stands at least
even.
18 Bg3 Nxg3 19 hxg3 Qxg3

At this stage things look exceedingly promising for Black. After all, he
owns the bishop pair against a knight pair in an open position, controls the
dark squares, and his queen is menacingly close to the white king. This is all
an illusion.

Exercise (planning): Find one powerful idea and we realize that it is Black,
not White, who is in danger.

Answer: Push the c-pawn, after which Black’s initiative is seen to be non-existent.
20 c5!
This smothering move is Nepo’s “Open Sesame” moment where his once
dormant pieces are given entry to powerful posts.
.
20.. Qg6
The c-pawn is immune due to 20 ... Bxc5?? 21 Ne4, forking and winning a
piece.
21 Rcl

Question: Why is White’s position so wonderful when Black still owns the
bishop pair?

Answer: The position is deceptive, in that it isn’t easy to assess correctly. Unusually, White’s
knights at least hold their own here. Black’s dark-squared bishop is in quarantine as long as the c5-
pawn remains in place, and White’s central pawns are exceedingly dangerous. Black must constantly
calculate lines involving d5-d6 or c5-c6.
21 ... Qf7
Instead:
a) 21 ... Rd8? is both natural and incorrect: 22 d6! cxd6 23 c6! (threatening
c6-c7) 23 ... Ra8 24 Nc4 d5 25 a5! Bh3 26 Nh4 Qg4 27 Qxg4 Bxg4 28 Nb6
wins and since Black can’t afford to take the knight.
b) 21 ... Bh3 22 Nh4! Qg4 23 Rc4! Qxdl 24 Rxdl Bc8 (or 24 ... Bd7 25 c6
Bc8 26 Ne4 and White’s pieces dominate the centre) 25 Nb3 and the
exchange of queens has not helped Black.
22 Ne4
The centre bustles with activity. My inclination would be a more stable
square for the knight with 22 Nb3 Bb7 23 c6 Bc8 24 Qd3, cutting off... Bf5,
with advantage for White.
22 ... Re8
22 ... Bg4 can be met by 23 Qb3. White doesn’t fear 23 ... Bxf3? 24 Qxf3
since the d5-pawn is tactically protected.
23 Rel Bf5 24 Ng3! Rxel+ 25 Qxel! Bg4
 Again the d5-pawn isn’t really hanging since 25 ... Qxd5? runs into 26
Nxf5 Qxf5 27 Qe7, when c7 is under attack and White also threatens Nd4,
followed by Nc6. If instead, 25 ... Bg6 26 Qd2 Rd8 27 d6! cxd6 28 c6 Qc7 29
Qxb4, the c6-passer ties Black up.
26 Qe4 h5 27 Nf5!

Nepo’s forward knight mounts a robust response to the claim that Black’s
bishops are the superior minor pieces. The threat is Ne7+, followed by Nc6.
27 ... Re8 28 Qd3 Bb8
Long odds creep closer and closer to the level of impossible. Black’s
caged-in dark-squared bishop is desperate to emerge and there is no way to
achieve this. 28 ... Qd7 29 N3h4! Re5 30 Ne3 is winning for White, while 28
... Qg6?? hangs the queen to the simple 29 Ne7+.
29 N3h4!
Now f2-f3 is a strategic threat.
29 ... Be2
 29 ... g5?? isn’t so well conceived since it hangs the queen to 30 Nh6+.
30 Qd2 Bg4
If 30 ... a5, White wins with 31 d6! cxd6 32 cxd6 (threatening d6-d7,
followed by Qxe2) 32 ... Bc4 33 Ne7+ Kh7 34 d7 Rd8 35 Nhg6 Be6 36 Rc8!
Rxc8 (or 36 ... Rxd7? 37 Rh8 mate) 37 dxc8Q Bxc8 38 Qc2! Be6 and now 39
Nf8+ Kh6 40 Qh7+ Kg5 41 Nxe6+ Qxe6 42 Qg6+ Kf4 43 Qg3+ Ke4 44 Qe3
mate.
31 Ne3?
A slip. White should play 31 Qxb4! at once. Black is unable to respond
with 31 ... Qxd5?? due to the trick 32 Qxb8!, followed by Ne7+ again if the
queen is taken.
31 ... Bd7?
Kramnik’s position falls apart quickly after this. 31 ... f5! could be met by
32 g3! a5 33 Rc4, but Black has more hope than in the game.
32 Qxb4 a5
Desperation.
33 Qxa5 c6 34 d6
The dark-squared bishop is a dead piece now.
34 ... Qb3 35 Qc3 Qxa4 36 Nhf5 Qe4 37 Ne7+ Kh8 38 Qb3
Eyeing the f7-square.
38 ... Rf8 39 Qc2!
 Simplification. An accurate summation of Black’s position in only two
words is: it stinks. All forms of resistance end if queens are removed from the
board.
.
39.. Qxc2
If 39 ... Qb4 then 40 Ng6+ wins.
40 Nxc2!
The knight meanders to its goal of d4.
40 ... Kh7 41 Nd4 1-0
Now c6 falls and the game is over.

Game 32
N epomniachtchi-Geo.Meier
I.
Dortmund 2018
French Defence
1 e4 e6 2 d4 d5 3 Nc3 dxe4
No one can force a person to be an extrovert. GM Georg Meier may well
be the world’s leading exponent of the Rubinstein French.
4 Nxe4 Nd7 5 Nf3 Ngf6 6 Nxf6+
Theory considered this to be White’s best chance to extract an edge.
. Nxf6
6..

7Ne5’?
This is White’s sixth choice in the position, which likely means that Nepo
didn’t want to challenge Meier to a theoretical dual in the more popular lines
7 Bd3 c5, or 7 c3 c5 8 Ne5 a6, or 7 Bg5 h6 8 Bh4 c5.
7 ... Bd6’?
Meier responds with a rare choice which is actually Fat Fritz's top pick in
the position. The threat is to swap on e5, trade queens, and then play ... Ng4
with a double attack on f2 and e5-pawns. 7 ... Be7 is Black’s main move.
8Bg5
Fat Fritz says the game is equal, whereas in the database, out of nine games, White wins 100%. So
who do you believe, the engine or the stats?
8 ... h6 9 Bh4 0-0 10 Bd3
10 Ng4?! is just a waste of time after 10 ... Be7.
.
10.. c5

One of the major points of the Rubinstein is that Black gets a Caro-
Kannish position, with one difference: In the Caro-Kann it always takes two
moves to play the ... c5 freeing break, whereas in the Rubinstein it can be
done in one go, as we see here.
11 Qe2

Question: Doesn’t this just give away the d4-pawn.

Answer: The d-pawn is sacrificed in the name of the attack. In any case, 11 c3? is unsound: 11 ...
cxd4 12 cxd4 Qa5+ 13 Kfl Bxe5 14 dxe5 Qxe5 and White doesn’t have anything for the pawn.
11 ... Qa5+!
Meier wants to make it more difficult for Nepo to castle queenside. After
11 ... cxd4 12 0-0-0 White has decent attacking chances and will continue
with 12-f4 and g2-g4.
12 c3
If 12 Kfl!? cxd4 13 Nc4 Qd5 14 Nxd6 Qxd6 15 Bxf6 gxf6 16 Qg4+ Kh8
17 Qh4 Kg7, White has nothing better than to take the draw.
12 ... cxd4 13 Nc4 Qc5?’
This is the wrong square for the queen. 13 ... Qd5 was correct, when White
has two tries:

a) 14 Nxd6 Qxd6 15 Bxf6 gxf6 16 Qg4+ Kh8 17 0-0-0!? (17 Qh4 Qe5+!
18 Kdl Kg7 19 Rel Qg5 20 Qxd4 e5 is not good for White) 17 ... dxc3 18
Qh4 Kg7 19 Bc2 Qc5 (threatening ... Qg5+) 20 Kbl e5! (stronger than 20 ...
Qg5 21 Qb4) 21 Rd3 Bf5 22 Rg3+ Bg6 23 Qh5 f5 24 Qxf5 Qb6 25 Qxe5+
Kg8 26 Rxc3 Rac8 and the engine calls it dead even.
b) 14 Bxf6 gxf6 15 Be4 Qc5 16 Nxd6 (or 16 Qg4+ Kh8 17 Qh4 Kg7 18
Nxd6 Qxd6 19 0-0-0 Qe5 20 Rxd4 Qg5+) 16 ... Qxd6 17 Rdl e5 18 Qh5 f5
19 Bxf5 Bxf5 20 Qxf5 Qg6 21 Qxg6+ fxg6 22 cxd4 exd4 23 13 (23 Rxd4?!
allows 23 ... Rae8+ 24 Kdl Rxf2) 23 ... Rad8 and White has only a nominal
edge in the endgame.
14 Bxf6 gxf6 15 Qg4+ Kh8 16 cxd4
This is why Black’s queen should have gone to d5, rather than c5. White
gains a precious tempo for his attack.
.
16.. Qb4+ 17Kfl?’
Not best. The rooks should be connected with 17 Ke2! Bc7 18 Qh4 Kg7
19 Rael, when White has a dangerous attack brewing. There’s no need to
worry about the b2-pawn even if Black manages to take it with check. For
example: 19 ... Rd8 20 Rhdl b5? (or 20 ... Bb6 21 Bbl) 21 a3! Qe7 22 Ne5!
Bxe5 23 dxe5 Rd5 24 Qg3+ Kh8 25 exf6 Qxf6 26 Be4 Qxb2+ 27 Rc2 Qe5
28 Qxe5+ Rxe5 29 K13 Rb8 30 Rd8+ and White wins.
17 ... Bc7 18 Qe4 f5 19 Qh4?’
This is not the best either. 19 Qe3 Kg7 20 g4! f4 21 Qe4 Bd7 22 Ne5 looks
more uncomfortable for Black.
 Black must now defend the h6-pawn with the king. But how? The problem
with ... Kh7 is that it allows White g2-g4 ideas where the f5-pawn is pinned;
while ... Kg7 places the king on the g-file which later may be opened.

Exercise (critical decision): One move leads to equality, while the other gives
White a strong attack. So pick: ... Kh7 or ... Kg7 - ?

19 ... Kg7?
Moving the king here ignores the position’s dire warnings. The king is
actually far more vulnerable on the g-file.
Answer: 19 ... Kh7! saves the game, albeit only with very precise play: 20 g4 b5! 21 g5 (or 21 gxf5
bxc4 22 Rgl Rg8!) 21 ... Bf4! 22 Qxh6+ (or 22 Qxf4 bxc4 23 Qh4 Qd2! 24 Qxh6+ Kg8 25 Be2 Rb8!
26 Rgl c3! and if 27 bxc3 then 27 ... Rb2 28 Rel Ba6! 29 Qh5 Bxe2+ 30 Qxe2 Qxc3) 22 ... Kg8 23
Rgl (or 23 Qh5 Kg7! 24 Qh6+ Kg8 etc) 23 ... bxc4 24 Rg4 Qb7! (eyeing hl and Black’s second rank)
25 Bc2 (not 25 Rh4?? Qhl+ 26 Ke2 cxd3+ 27 Kxd3 Qe4+ and White gets mated) 25 ... c3! 26 Kgl
(forced) 26 ... f6 27 gxf6+ fxg4 28 Qg6+ Kh8 29 Qh5+ Kg8 with a draw, as White is unable to clear the
back rank for his remaining rook.
 20 g4!
All those who like Black’s position, please raise your hand. That’s what I
thought. No hands raised. From this point on, Nepo’s play is the chess
version of a Shock and Awe campaign.
20 ... f4
Black has no time for 20 ... b5?? here because 21 exf5 bxc4 22 f6+ mates
very quickly. Or 20 ... Bd8 21 Qh5 Qe7 22 h4 (preventing ... Qg5) 22 ... Qf6
23 gxf5 exf5 24 Rgl+ Kh7 25 Ne5 Rg8 26 Rg5!, when Ng4 is in the air and
Black’s position is under tremendous strain.
21 Rgl!
Now g4-g5 is coming.
21 ... Bd7
If 21 ... Rg8 then 22 Rcl Bd7 23 g5 h5 24 g6 f5 25 Ne5 Bd8 26 Qxf4 and
Black won’t survive.

Exercise (planning): We reach the exact calculation stage when we no longer

 deal in hypotheticals. Come up with a plan to give White a decisive attack.

Answer: First move the queen to h5 and only then play g4-g5.
22 Qh5!
This ensures that the g-file will be opened. I wonder whether Meier had
only envisaged 22 g5? h5!, after which the engines prefer Black’s position.
22 ... Rh8
Or 22 ... Kf6 23 g5+ Ke7 24 gxh6 Bc6 25 Rg7 with a winning position.
23 g5 hxg5 24 Qxg5+ Kf8

Exercise (planning): Construct White’s remaining attacking plan.

Answer: Step 1: Move the queen to f6 with tempo.

25 Qf6! Rxh2
If 25 ... Rg8 then 26 Rxg8+ Kxg8 27 Qh6! (threatening mate in four, with
Bh7+ next) 27 ... f5 28 Ke2! (clearing the way for White’s rook to enter the
attack) 28 ... Qf8 29 Rgl+ Kf7 30 Qh7+ Ke8 31 Rg8 wins the queen.
26 Rg7!
Step 2: Play Rg7, threatening mate on f7.
26 ... Be8 27 Bh7! 1-0
Step 3: Transfer the bishop to h7, threatening mate on g8. On a misery
index from 1 to 10, this position is at least an 11 for Black, who is completely
helpless; e.g. 27 ... Rhl+ (or 27 ... Qxc4+ 28 Kgl Rhl+ 29 Kxhl Qd5+ 30
Kgl) 28 Kg2 Bc6+ 29 13 Bxf3+ 30 Kxf3 Rh3+ 31 K12 Rh2+ 32 Kgl and the
checks have run out so Black is mated.

It isn’t easy to be a heavy 350-point underdog, as Lopez was below, and

not get intimidated when playing a potential future world champion. What
the underdog should always remember is that there is an equal amount of
pressure on the favourite to win, since any other result is interpreted as a
disaster.

Game 33
A.Astaneh Lopez-I.Nepomniachtchi
Batumi Olympiad 2018
Nimzo-Larsen Opening

1 b3

Question: Why would someone open with the b-pawn, when 1 e4 and 1 d4
main lines offer more chance of an advantage?

Answer: The main reason people go for 1 b3 is because they want a one-on-one fight, rather than
memorizing and reproducing 25 moves of theory. After I wrote Nimzo-Larsen Attack: Move by Move, I
was amused by one negative reviewer who declared that the book was informative and well written, yet
still gave it a poor review for the sole reason that 1 b3 is a lame opening which only grovellers like.
Tell that to GM Baadur Jobava, the world’s leading exponent of the Nimzo-Larsen! Strangely enough,
when I did stats on my own games, I was shocked to discover that 1 b3 was my highest scoring opening
with White, with an over 90% ratio.
.1.. e5 2 Bb2 Nc6 3 e3 Nf6 4 Nf3
White’s main move is 4 Bb5, to which Black’s main reply, strangely
enough, is 4 ... Bd6. One example: 5 Na3 a6 6 Bxc6 (I always play this,
rather than 6 Be2 b5 7 c4 b4 8 Nc2 which looks as if it was played by two
beginners) 6 ... dxc6 7 Nc4 Qe7 8 Nf3 (I have also tried 8 Ne2 0-0 9 0-0,
aiming for f2-f4) 8 ... Bg4?! (8 ... e4 is better) 9 Bxe5! (my own novelty,
inflicting damage on Black’s pawn structure) 9 ... Bxf3 10 Bxd6 Bxdl 11
Bxe7 Bxc2 12 Bxf6 gxf6 13 Rcl Bf5 14 fB Rg8 15 Kf2 and Black was
unable to overcome his structural deficit, C.Lakdawala-K.Griffith, San Diego
(rapid) 2013.
. e4
4..
More aggressive than 4 ... d6 and probably superior.
5 Nd4

The position resembles a reversed Alekhine’s Defence.

5 ... Bc5
5 ... Nxd4 is more often played.
6Nf5’?
Principle: Don 7 provoke and pick a fight with an opponent twice your size.
This looks like a waste of time. I would have taken the safer route 6 Nxc6
dxc6 7 d4 Bd6 8 Be2 0-0 9 Nd2, intending 9 ... c5?! 10 Nc4, when White is a
shade better, Ba. Jobava-D.Brodsky, PNWCC (blitz) 2020.
6.. . g6

Question: Isn’t this a huge weakening of Black’s dark squares?

Answer: It gains time in return, which is a fair deal. The trouble with 6 ... 0-0 is that it allows 7
Nxg7!? Kxg7 8 Qg4+ Kh8 9 Qf4 Be7 10 g4! d5 11 g5, which the engine calls even.
7 Ng3 d5 8 d3
White must chip away at Black’s centre at some point. 8 Nc3!? would
probably be the Jobavian choice, aiming to castle queenside, since Jobava
absolutely loves to play Nc3 with his c-pawn still on c2.
8.. . Qe7!

Question: Shouldn’t this be a question mark, rather than an exclamation mark?

Didn’t Nepomniachtchi’s last move hang a pawn?
 Answer: A player of Nepo’s level doesn’t fall for a simple tactic. It was a deliberate sacrifice, for
which Black receives more than enough compensation. I prefer Nepo’s choice over 8 ... exd3 9 Bxd3
Qe7 10 0-0 Bd7 11 c4 dxc4, when the players agreed to a premature draw, A.Lenderman-R.Sadhwani,
PNWCC (blitz) 2020.
9 dxe4 dxe4 10 Nd2
Winning a pawn but not without cost. Note that 10 Nxe4?? is a terrible
miscalculation, since 10 ... Nxe4! 11 Bxh8 Nxf2! leaves White busted.
10 ... Ba3!
This move seriously weakens White’s queenside dark squares.
11 Bxa3
If 11 Bxf6 Qxf6 12 Ngxe4 Qe7, Black gets a lead in development, the
bishop pair and power on the dark squares for the pawn.
11 ... Qxa3 12 Ndxe4!?
White bravely accepts the sacrifice. It might have been more prudent to
chicken out with 12 Qcl Qe7 13 Qb2 0-0 14 Bb5 Bd7 15 Bxc6 Bxc6 16 0-0,
when I only slightly prefer Black.
12 ... Nxe4 13 Nxe4 Qb4+!
Gaining a tempo and sending White’s knight to an uncomfortable post on
d2.
14 Nd2 Bf5
Clearing the way for ... 0-0-0.
15 Bd3?
This doesn’t clog the d-file, as he hopes. White should insert 15 Qcl Qc3!
before 16 Bd3. At least in this version White gets to castle, although the
engine still likes Black after 16 ... 0-0-0! 17 0-0 Ne5!.
.
15.. 0-0-0!
 16 a3
Not yet 16 0-0?? since 16 ... Bxd3 17 cxd3 Rxd3 wins a piece; while after
16 Qe2 Nd4! 17 Qdl Ne6! 18 Bxf5 (not 18 Qe2?? Nf4! and White can
resign) 18 ... gxf5 19 c3 Qxc3 20 Rcl Qa5 21 Rc2 Qb5! 22 Qe2 Qe5! White
is in deep trouble, since 23 0-0? hangs material to 23 ... Nd4!.
16 ... Qc3 17 0-0
White returns the pawn. The problem is Black retains a massive bind.
Instead:
a) 17 Bxf5+?? gxf5 18 Ke2 is hopeless due to 18 ... Rhe8 19 Rcl Nd4+ 20
Kfl Ne6! 21 Ke2 Nf4+ and if 22 Kel then 22 ... Nxg2+ 23 Ke2 Nxe3 forces
mate.
b) 17 Ra2! is the best defensive try; e.g. 17 ... Bxd3 18 cxd3 Rxd3 19 Qc2
Rhd8 20 Qxc3 Rxc3 21 Ke2 (21 0-0? loses a pawn to 21 ... Rcd3! 22 Nf3
Rxb3) 21 ... Ne5 with a difficult yet still playable position for White.
.
17.. Ne5’?
 Nepo sits patiently and retains the tension, as if awaiting a bus. I find it
difficult to reconcile why a player as non-materialistic as Nepo plays the
Najdorf Poisoned Pawn, the epitomic opening of materialism. Here he
probably would be better off taking at once: 17 ... Bxd3! 18 cxd3 Rxd3 19
Ne4 Qxb3 and Black is a pawn up with a winning position. I don’t believe
White, in such a passive state, can utilize the open b-file against Black’s king.
18 e4?
He had to try 18 Ra2 Nxd3 19 cxd3 Rxd3 20 Rc2 Qa5 21 Qcl Rhd8 22
Nc4, when 22 ... Qd5! (threatening ... Rdl) 23 Nb2 (forced) 23 ... Rxb3 24
Rxc7+ Kb8 25 Rdl Qxdl+! 26 Nxdl Rbl 27 Qxbl Bxbl 28 Rcl Bf5 gives
Black a clearly favourable ending for the following reasons:
1. Both the black bishop and rook are more active than their white
counterparts.
2. Black owns a dangerous queenside majority which can move up the
board with the aid of his king.
3. White must worry about the safety of his a3-pawn.
18 ... Be6 19 Ra2
It’s a bit late for this now. 19 Nf3 Nxd3 20 cxd3 Rxd3 21 Qe2 Rhd8! 22
b4 Bg4 is also quite wretched for White.
19 ... Nxd3 20 cxd3 Rxd3 21 Rc2
 Exercise (combination alert): What is Black’s strongest continuation?

Answer: Pinned piece. Black’s queen can chop the b3-pawn and get away with it.
21 ... Qxb3! 22 Qcl
If 22 Nxb3 Rxdl 23 Rxdl Bxb3 24 Rdcl Bxc2 25 Rxc2 Rd8, Black has an
extra pawn and control over the only open file, which equates to an easy win
in the rook endgame.
22 ... Qxa3 23 Rxc7+ Kb8 0-1
White realized he had no chance and resigned here. After 24 Rc2 Qxcl 25
Rfxcl Rhd8 26 Nfl a5 Black’s queenside pawns will easily triumph.

Game 34
Nepomniachtchi-E.Bacrot
I.
Batumi Olympiad 2018
Réti Opening
1 Nf3 d5 2 e3 Nf6 3 c4 e6 4 Nc3 Be7 5 b3 0-0 6 Bb2 c5 7 cxd5 Nxd5
For the isolani lovers out there, 7 ... exd5 8 d4 Nc6 9 Be2 Bg4 10 dxc5
Bxc5 11 0-0 is a typical isolani position, as in I.Nepomniachtchi-I.Nyzhnyk,
PRO League (rapid) 2020. Despite Siegbert Tarrasch’s stem warnings,
stylistically I favour White.
8 h4!?

Question: What is the name of this opening?

Answer: It’s a nameless variation which defies conventional classification. I’ve called it a Réti
because it began with 1 NI3, but I doubt it’s one Réti himself would recognize as such.

Question: Is it my imagination that everyone in the chess world is enamoured

with their h-pawn?

Answer: \ blame AlphaZero. Now everyone from Grandmaster to 1000-rated player wants to push
their h-pawns. It wouldn’t surprise me if Everyman offered me a contract for the potentially best-selling
book 1 h4H: Move by Move. Either 8 Qc2 or 8 Nxd5 would be less belligerent.
8.. . b6
There is nothing particularly wrong with this move, yet it does feel a touch
slow when the opponent declared his intention of delivering checkmate.
Possibly 8 ... Nc6 9 Qbl f5! is better, as Wesley So has played a couple of
times, following the principle: Counter in the centre when attacked on the
wing.
9 Qbl’?
White’s queen takes long distance diagonal aim at the black king.
9.. . h6!?
Somewhere out in the Pacific Ocean is an inconceivably gigantic island of
floating trash churning in the current. By last estimate it’s larger than the state
of Texas. Moral: don’t ignore a small problem, since it is certain to grow.
Bacrot, perhaps not worrying about White’s coming attack, actually invites
g2-g4, hoping Nepo will overextend in his mating efforts. I don’t like the
move since it gives White a biting point on g5. Instead:
a) 9 ... Nf6? 10 Ng5 Nbd7?!, J.K.Duda-S.Karjakin, Chess.com (blitz)
2020, and now 11 Nce4! gives White a powerful attack.
b) 9 ... Nc6 looks best, when 10 g4 can be met by the central counter 10 ...
f5!. If you aren’t able to halt your opponent’s intent, the next best thing is to
slow it down. After 11 gxf5 Nxc3 12 Bxc3 Bf6! Black looks just fine.
10 g4
 Here he comes, full speed ahead. There was a zero percent chance that
Nepomniachtchi would back down and make a quiet developing move.

Question: Isn’t this and his previous h2-h4 a blatant violation of the principle:
Develop your pieces rapidly in the opening -1

Answer: “It is better to die on your feet than to live on your knees,” stated Emiliano Zapata. Bear in
mind that Zapata said this before dying in a government ambush in 1919. Breaking a principle isn’t
always a hanging offense, especially when you have the white pieces and therefore more leeway.
.10.. Bb7 11 Rh3
It’s wise to get the rook off the long diagonal.
11 ... Nd7!
The most accurate square, since the knight may be needed to help out his
king.
12 g5 h5 13 Bd3!
 Multi-purpose:
1. Now any attempt for Black to play ... g7-g6 will invite Bxg6.
2. It clears e2 for the manoeuvre Ne2-g3, attacking Black’s h5-pawn. And
when this happens, keep in mind no.l on the list: that Black is unable to
protect the h-pawn with ... g7-g6, due to the threatened bishop sacrifice on
this square.
13 ... Nb4?’
This doesn’t have the desired intent and so just wastes time. Black should
play 13 ... Nxc3 14 Bxc3 Qc7, when it’s not obvious how White will proceed
on the kingside.
14 Bh7+!
A useful insertion, driving Black’s king to h8, on the same diagonal as
White’s dark-squared bishop.
14 ... Kh8 15 Be4
Challenging Black’s best piece.
 15 ... Nd5
Back again, since 15 ... Bxe4 16 Nxe4 only benefits White; e.g. 16 ... Qc7
17 a3 Nc6 18 Ng3 and Black still can’t play ... g7-g6 because the pawn is
pinned.
16 Ne2
Heading across to the kingside, as indicated above.
.
16.. f5?
This overreaction weakens Black’s king and now his position becomes
totally ungovernable. 16 ... Bd6! 17 Ng3 Bxg3 18 Rxg3 leaves Black worse
but still in the game.

17 gxf6 N7xf6
If 17 ... Bxf6 then 18 Ng5 Bxg5 19 hxg5 Qxg5 20 Rg3 wins; or 17 ... Rxf6
18 Ng5 (threatening Bxd5 and Qh7 mate) 18 ... Qg8 19 Bg6 and Black is
busted.
18 Ng5 Nxe4
 Not 18 ... Qe8?? 19 Bxd5 Bxd5 20 Bxf6. eliminating the defender of h7
and forcing mate.
19 Qxe4!
The once and future queen returns to power, threatening mate on the
move.
19 ... Bxg5
Black has no choice, despite opening the h-file for White’s rook.
20 hxg5 Nf4
Fancy, but it just doesn’t work. On the other hand, 20 ... Qxg5 21 Rg3 and
20 ... Qe8 21 Ng3 are very simple.
21 Qxb7 Nd3+
It soon becomes clear that this counterattack is a sham. 21 ... Nxh3? is
even more inaccurate as it ignores the threat of 22 Qxg7 mate.
22 Kfl Nxb2 23 Rxh5+ Kg8
Hashtagdespair. Black’s king is lost in an uncaring world.
24 g6! 1-0
Threatening Rh8+, Qhl+ and Qh7 mate. Black has no good defence: 24 ... Rf6 (or 24 ... Rf5 25 Rh7
Qf8 26 Rh8+! etc) 25 Nf4 (or just 25 Qhl) 25 ... Nd3 26 Nxe6! (threatening both Qxg7 mate and Nxd8)
26 ... Rxe6 (taking on f2 doesn’t help) 27 Qf7 mate.
Chapter Three
The Road to Magnus: 2019-2021
The key tournament in this chapter is the FIDE Candidates in Yekaterinburg, which Nepomniachtchi
won with 814/14. The slightly unusual circumstance was the length of time between the seventh and
eighth rounds: 390 days, on account of the COVID pandemic. In the intervening period Nepo played a
lot of chess and has 495 games listed in the databases. We’ll examine a tiny fraction of those, and a few
more besides.

Game 35
R.Mamedov-I.Nepomniachtchi
World Team Championship, Astana 2019
Sicilian Defence

1 e4 c5 2 Nf3 d6 3 Bb5+
The Moscow Variation should really be renamed the Low-Expectations-
Kill-Joy Variation. It’s solid and safe, due to being less tactical than a typical
Open Sicilian.
3 ... Nd7
He wants to keep pieces on the board. 3 ... Bd7 is the steady option, while
3 ... Nc6 tends to be sharper.
4 c3
Preparing d2-d4 while making room for the light-squared bishop on c2 if
required. 4 0-0 and 4 d4 are alternatives.
4 ... Ngf6 5 Qe2 a6 6 Ba4 b5 7 Bc2 Bb7
These positions can easily resemble a Closed Ruy Lopez, should Black
elect to play ... e7-e5.
8d4 g6!?
 Nepo indicates that he is up for an adventure. This line is a shade risky
since it allows e4-e5-e6 ideas. 8 ... e6 is a more solid choice.
9 0-0

Question: Why did White refrain from the e4-e5 plan?

Answer: If you wait just a moment you’ll see your question was premature. At this point 9 e5 is
sharp yet doesn’t yield White an advantage. Play can go 9 ... Bxf3! 10 gxf3 (after 10 Qxf3!? dxe5 11
dxc5 Nxc5 it’s debatable whether White gets full compensation for the sacrificed pawn; Black certainly
doesn’t stand worse) 10 ... dxe5 11 dxe5 Nh5 12 e6 fxe6 13 Qxe6 Bg7, which looks dynamically
balanced, A.Delorme-L.Fressinet, French League 2019.
9 ... Bg7 10 e5’?
Mamedov decides to initiate complications before the black king castles to
safety. Nepo has played this position before and since, scoring 2/2/3 against
10 Nbd2, 10 Rel and 10 a4, while
10 Rdl Qb8 also looks okay for Black.
10 ... dxe5 11 dxe5 Bxf3! 12 gxf3
 The pawn sacrifice is even worse here: 12 Qxf3?! Nxe5 13 Qe2 Qc7 14 h3
Nc6 and I don’t see White’s full compensation.
.
12.. Nd5 13 e6 N7f6’?
Another option is 13 ... fxe6 14 Qxe6 N7b6, intending to challenge queens
with ... Qc8 or ... Qd7; e.g. 15 c4?! bxc4 16 Nc3 Qd7 17 Qe2 0-0 18 Nxd5
Qxd5 19 Be4 Qe5 and White is struggling to prove compensation for the
pawn, A.Ginovart-F.Rakotomaharo, Gonfreville 2016.
14 c4!
Stronger here, with the black knight having gone to the kingside.
14 ... bxc4 15 exf7+ Kxf7

Welcome to the I-Don’t-Care-About-My-Pawn-Structure Variation.

Question: Of the two, whose structure looks worse?

Answer: It’s a close call. Black’s structure looks just a shade more wobbly, since after Qxc4
Black’s c5-pawn remains a target, as does his king, whereas White’s double f-pawns are not attacked -
at least not yet.
16 Nd2
I would just take the c4-pawn: 16 Qxc4! e6 17 Nc3 Qc7 18 Rel, when I
prefer White by a microbe.
16 ... Bh6?’
Nepo wants to disable White’s bishop pair with a swap. The trouble with
the move is that it loses time. Black equalizes with 16 ... c3! 17 Nc4 Rf8! 18
Rel Kg8, having castled by hand.
17 Ne4?’
This looks like an inaccuracy. After 17 Rel! Rf8 18 Nxc4 Bxcl 19 Raxcl
Nf4 20 Qe5 Qd4 21 Bb3!, I would be nervous if playing Black here.
17 ... Bxcl 18 Raxcl Nxe4
The black king becomes a touch safer with each exchange.
19 Bxe4 e6 20 Qxc4 Qg5+ 21 Khl Rhd8!
 Now Black looks okay again.
22 Rfdl
If 22 Qxc5 Rdc8 23 Qxc8 Rxc8 24 Rxc8 Nf4 25 Rc7+ Kg8 26 Rgl Qe5,1
don’t believe Black stands worse since White must perpetually keep a rook
near his king as a guard against a potential queen and knight mating attack.
22 ... Rab8 23 b3!?
At this stage Mamedov, perhaps believing himself safe, begins to overplay
his hand. He shouldn’t have allowed the knight to remain on the board, since
from f4 it may in the future cause all kinds of mischief for the white king,
when coupled with the black queen. 23 Bxd5! exd5 24 Qxc5 Rxb2 25 Qd4
Re2 26 Rc7+ Re7 is a likely draw.
23 ... Nf4 24 Rgl
White’s rook has to take up defensive duties.
24 ... Qe5 25 Reel?!
Only Black can win from here. It was wiser to bail out with a line like 25
Qxa6! Rd2 26 Qa7+ Kg8 27 Rxc5 Rxa2! 28 Qe7 Qg7 29 Qd6 Qf8 30 Qe5
Rxf2 31 Rc7 Qh6 32 Rg7+! Qxg7 33 Qxb8+ Qf8 34 Qe5 Qh6 35 Qb8+ Qf8
36 Qe5 with a draw by repetition.
25 ... Qd6
Covering a6.
26 Rg4?
Even worse than a single mistake is a second one, piled upon the first. 26
Rcl should be played.
 Exercise (planning): With his last move White attempted to chase away
Black’s knight from its dominant square. Prove why this was a mistake.

Answer: Activate the b8-rook.

26 ... Rb4!
This move drains all the energy out of White’s position and leaves him
strategically lost.
27 Rxf4+!
A desperate attempt to mix it up. It’s probably his best practical chance,
since a retreat with the queen turns White’s slow-death position into a child’s
toy with weak batteries:
a) 27 Qfl Rd4 28 a4 e5 29 h4 Rd2 30 Qc4+ Kg7 31 Rfl Qd4 and White is
busted.
b) 27 Qcl e5 28 Rggl Rd4 29 Rgfl Rd2 30 a4 Kg7 31 Qc3 Qd4 32 Rcl
Qxc3 33 Rxc3 Rxf2! exploits the weak back rank and Black will win.
 27 ... Qxf4 28 Qxc5?
This was the wrong pawn to capture. White’s only prayer was 28 Qxa6
Qd6 29 Qe2, though Black should win eventually.

Exercise (planning): What is Black’s winning plan?

Answer: Swing the rook over to the kingside for a decisive attack.
28 ... Rb5!
The rook is back, crazier than ever.
29 Qa7+ Kf8! 30 Rgl
Or 30 Qxh7? Rh5 and White must resign.
30 ... Rh5
Threatening mate on h2.
31 Rg3 Qh6! 0-1
White’s king is helpless against the threats at h2 and dl; e.g. 32 Kg2
Rxh2+ 33 Kfl Rdl+ 34 Ke2 Qd2 mate.
 I love those two-for-the-price-of-one specials, where one side has two
ways to win. Also deadly is 31 ... Rdl+! 32 Kg2 Qcl (threatening ... Rgl
mate) 33 Qb8+ Kg7 34 Qa7+ Kf6 and White’s well of checks has run dry.

The next game is a marathon, demonstrating just how difficult it is to beat

Nepo even when he’s on the ropes.

Game 36
Wei Yi-I.Nepomniachtchi
FIDE Grand Prix, Moscow 2019
Sicilian Defence

1 e4 c5 2 Nf3 d6 3 d4 cxd4 4 Nxd4 Nf6 5 Nc3 a6 6 Bg5 e6 7 f4 Qb6 8 Qd2

Qxb2
Once again Nepo senses gold, believing his beloved line to be a
prospector’s paradise.
9 Rbl Qa3 10 f5
 This variation tends to be less wild than its more hot-headed cousin, 10
e5!?, but only slightly!

Question: What is the reason behind the push of White’s f-pawn?

Answer: The idea is to soften up e6 with f5xe6 and then maybe Bc4.

Question: If that’s the case, then why can’t Black just bypass with 10 ... e5 - ?

Answer: Inadvisable, since it leaves a gaping hole on d5.

10 ... Be7
Let’s look at your 10 ... e5 suggestion: after 11 Bxf6 gxf6 12 Nb3 Be7 13
Bc4 I hate Black’s position. Don’t you?
11 fxe6 Bxe6!

Question: Why give away the bishop pair and hang the b7-pawn?
 Answer: First, the move follows the principle: Swap pieces when under attack. Secondly, White
wastes time and loses the initiative if he grabs the b7-pawn. In any case 11 ... fxe6?! is under a cloud:
12 Bc4 scores quite wretchedly for Black.
12 Nxe6 fxe6 13 Bc4
The time-wasting 13 Rxb7? loses the initiative: 13 ... Nbd7 14 Be2 Qc5 15
Be3 Qc6 16 Rbl Bd8!, when ... Ba5 is in the air and White is already
strategically lost.
13 ... Nbd7!
Nepo returns the pawn to catch up in development.
14 Bxe6 Nc5 15 Bf5
White’s bishop has no safe square on the a2-g8 diagonal, which means
Black may still castle kingside. After 15 Bb3 Rc8 16 0-0 Nxb3 17 Rxb3
Qc5+ 18 Be3 Qc4 Black stood no worse, N.Grandelius-M.Vachier Lagrave,
Wijk aan Zee 2021.
.
15.. g6!

Undermining the protector of e4.

 16 Bh3
In another game Wei Yi speculated with the piece sacrifice 16 0-0!? gxf5
17 Rxf5 Rf8 (not 17 ... Nfxe4?? 18 Nxe4 Nxe4 19 Qd4 0-0-0 20 Qxe4 Rd7
21 Bxe7 Re8 22 Rxb7! and Black can resign) 18 Bxf6 Bxf6 19 Nd5 0-0-0!
(19 ... Be5?? loses instantly to 20 Qg5!) 20 Rxf6 Nxe4 21 Ne7+ Kb8 22 Qd5
(threatening Qxb7 mate, Qxe4 and Nc6; Black can cover all of them) 22 ...
Qe3+ 23 Khl Nf2+! 24 Rxf2 Qxe7 with an equal position, Wei Yi-
C.Sandipan, Chengdu 2017.
.
16.. Nfxe4!
Black gets away with the pawn grab.
17 Nxe4 Nxe4 18 Qd4 Qc3+
Forced.
19 Qxc3 Nxc3 20 Bxe7 Kxe7 21 Rb3
Or 21 Rxb7+ Kf6 22 Kd2 Ne4+ 23 Ke3 Nc5 24 Rfl+ Kg5 and Black is
just fine.
21 ... Ne4 22 Rxb7+ Kf6 23 0-0+ Kg5
Forced.
24 g3!
Intending to relocate the bishop to a better diagonal via g2.
24 ... Rab8 25 Ra7
 25 ... Rb2
Another move thrown upon the compost heap of opening theory. This is a
novelty but no improvement. The Najdorf Poisoned Pawn is not for the
memory-challenged. Previous praxis saw 25 ... Nc5 26 Bg2 h5 27 h4+ Kh6
28 Rdl Rb6 29 Bd5 Rh7 30 Ra8 Vi-Vi L.Ovcharenko-A.Guberna,
correspondence 2015.
26 Bg2 d5 27 h4+’?
Wei Yi sees Nepo’s king bouncing around the fourth rank and can’t resist
going after it. The price is that he weakens his own structure. 27 Rd7 is a lot
safer.
27 ... Kg4!
Nepo is not the “Maybe I should back down from this challenge?” type of
player.

Question: Is Black’s wandering king in any danger? Or did White overextend

with his last move?
 Answer. The engine says that the push of the h-pawn is sound, assessing the position at dead even!
White simply lacks the fire-power to endanger the black king, provided that Black finds all the correct
defensive moves. That’s the only problem: Black’s position looks more difficult to play than White’s,
whose intent is obvious. The chickenish 27 ... Kh6!? is less accurate: 28 g4 Rxc2 29 a4 Rc4 30 a5 Ra4
31 Rff7 and I prefer White.
28 Kh2 Nxg3?
After this move Black’s position begins to gurgle erratically. This was
almost certainly a deliberate sacrifice. Nepo probably thought the resulting
position was drawn, and he may have been correct. All the same, there was
no reason for drama when Black can draw easily, and at no risk by taking the
c2-pawn: 28 ... Rxc2! 29 Rf4+ Kh5 30 Rd7 (or 30 Rxa6 Rb8, threatening ...
Rbb2) 30 ... Kh6 31 Rxd5 Re8 32 Rd7 (or 32 g4 Re7) 32 ... Nc5 33 Rc7
Rxg2+! 34 Kxg2 Ne6, and Black regains the exchange with complete
equality.

Exercise: (combination alert): How does White win material?

 Answer: Step 1: Give check on h3.
29 Bh3+! Kxh4 30 Rf4+
Step 2: Give a rook check on f4, disconnecting Black’s king from his
knight.
30 ... Kg5 31 Kxg3 Rxa2
The most logical capture, simultaneously covering a6. The alternative was
31 ... Rxc2!? 32 Rxa6 Rb8, giving up the a-pawn to activate the rooks as
quickly as possible. Play might continue 33 Rg4+ Kh6 34 Bg2 Rc3+ 35 Kh4
Rb2 36 Rg3 Rc4+ 37 Kh3 Rd2 38 Ra5 Rf4 39 Bf3 (not 39 Bxd5? Rf5,
winning the bishop, or 39 Rxd5?! Rxa2 with an easy draw) 39 ... Rfd4 and
it’s not easy to see how White makes progress.
32 Rg4+ Kh6 33 Rc7

Question: You talk as if Nepomniachtchi is in trouble. Aren’t three pawns a

decent bargain for a piece, especially in an ending?
 Answer: Wei Yi clutches his final pawn tightly. As long as White’s c-pawn remains on the board,
he has chances to win, whereas Black’s three pawns are dysfunctional as a threat to promote. So Nepo
has landed in a position which must be tolerated, rather than enjoyed.
33 ... Ra3+ 34 c3 Re8 35 Rh4+ Kg5 36 Rg4+ Kh6 37 Kf3!
Preventing ... Re3+ which drops White’s final pawn.
37 ... Rel 38 Rh4+ Kg5 39 Rg4+ Kh6 40 Rh4+ Kg5 41 Rd4 Re5 42 Rc5
h5! 43 Bd7
43 Rdxd5? falls for Black’s trap: 43 ... Rxc3+! 44 Rxc3 Rxd5 with a likely
draw.
43 ... h4?
Nepo goes wrong, pushing the h-pawn too soon. He needed to advance the
a-pawn first so that he can create threats on both wings; e.g. 43 ... a5! 44 Bc6
Rai 45 Bxd5 (or 45 Rdxd5 Rxd5 46 Rxd5+ Kf6) 35 ... a4 46 Ra5 h4! (now is
the correct moment) 47 Kg2 Rcl 48 c4 Rc2+ 49 Kh3 Rc3+ with perpetual
check.
 Exercise (critical decision): White has three potential pawn targets: a6, d5, and h4.
Which one should he go for?

Answer: The h4-pawn is the most appetizing.

44 Bc6?
Going for the wrong one. The a6-pawn doesn’t signify here, and the d5-
pawn can be targeted at any time, whereas Black can hardly afford to lose his
h-pawn. White should therefore play 44 Rg4+! (it was a mistake to give up
control over g4) 44 ... Kh5 (or 44 ... Kf6 45 Rxh4) 45 Rc8! (threatening Rh8
mate) 45 ... Kh6 46 Rxh4+ Kg5 47 Rg4+ Kf6 48 Rf4+ Kg5 49 Rc6 a5 50
Rg4+ Kh5 51 Rd4 (not 51 Rgxg6?? Rxc3+ and draws) 51 ... Rai 52 Rc5 and
the d5-pawn will soon fall, leaving White with excellent winning chances.
44 ... Rf5+!
White’s king is driven away from the h-file where Black’s pawn will race
forward to the promotion square.
45 Ke3
45 Kg2? allows an immediate perpetual with 45 ... Ra2+ etc.
45 ... h3 46 Rdxd5
Nepo is down to two pawns for his missing piece, but his h-pawn now
provides sufficient distraction to draw.
46 ... h2 47 Rxf5+ gxf5 48 Be4 Kh4!
 49 Bf3
49 Rxf5 Rxc3+ is drawn.
49 ... Kg3 50 Rc8 f4+ 51 Ke4 Ra4+ 52 c4 Ra3! 53 Rg8+ Kh4!
Not 53 ... Kf2?? 54 Rg2+ Kel 55 Re2+ Kfl 56 Rxh2 and White wins.
54 Rh8+
Not 54 Kxf4?? Rxf3+ 55 Ke4 Rfl and Black wins.
54 ... Kg3 55 Rg8+ Kh4 56 Bhl Rai 57 Rh8+ Kg3 58 Rg8+ Kh4 59 Bg2 Ra2 60 Kf3 Rc2!
Reminding White that he can’t win without the c-pawn.
61 Rg4+ Kh5 62 Rxf4 Rcl!
Black has just enough time to save the coming rook ending.
63 Ke3
 Exercise (critical decision): What is Black’s only move to hold the game?

Answer: Delay promotion and move the king to g5.

63 ... Kg5!
This move gains a crucial tempo. The hasty 63 ... hlQ?? 64 Bxhl Rxhl
loses, as the endgame declares mate in 35; e.g. 65 c5 Kg5 66 Rf3 Rh4 67 Kd3
a5 68 Rf2! a4 69 c6 a3 70 Rc2 Rh8 71 Kc4 Rb8 72 Kd5 Ra8 73 Ra2 Rd8+ 74
Ke6 Re 8+ 75 Kd7 Re3 76 Rc2 and so on.
64 Re4 Rgl
He could also promote immediately.
65 Kf3 Rfl+
Nepo extracts some mileage from his little joke.
66 Ke2
GMs rated 2736 don’t fall for 66 Bxfl?? hlQ+ and Black wins with his a-
pawn.
 66 ... Rgl 67 Kf2 hlQ 68 Bxhl Rxhl 69 Re5+ Kf6 70 Ra5 Yz-Yz
Since 70 ... Ke6 71 Rxa6+ Kd7 is a rudimentary draw.

Game 3 7
I.
Nepomniachtchi-Wei Yi
FIDE Grand Prix, Moscow 2019
Caro-Kann Defence

1 e4 c6 2 M3 d5 3 Nc3 g6 4 d4 Bg7 5 h3
This position brings back warm memories since I played the Black side
when I first beat a GM, at age 16. Unfortunately, 99.9% of the glory is
removed when I confess it was in a simul. The mercurial GM Walter Browne
and I went 4-4 in our lifetime score, but I once rudely reminded him that the
score was actually 5-4 in my favour, since I conveniently counted the simul
win against him. The idea behind White’s last move is to prevent ... Bg4,
which allows Black to swap away his most dysfunctional piece.

Question: Why play the passive h2-h3 when White can blunt Black’s
fianchettoed bishop with e4-e5 - ?

Answer: First, preventing ... Bg4 serves a useful purpose. For the second part of your question, it
isn’t clear at all that a white pawn on e5 is going to blunt the g7-bishop, since Black can undermine the
centre with a timely ... c6-c5, swap away White’s d-pawn, and then go after the e5-pawn, which is often
weak in this line. For example: 5 e5 Bg4 6 h3 Bxf3 7 Qxf3 e6 8 Bd3 a6 9 0-0 c5 10 dxc5 Nd7 11 Qg3
Bxe5 and the invincible e5-pawn has gone. I already slightly prefer Black, A.Kosteniuk-S.Sevian,
Titled Tuesday (blitz) 2021.
5 ... dxe4 6 Nxe4 Nf6
I believe this to be a superior line to the one I tried as a 16-year-old. After
6 ... Nd7 7 Bc4 Ngf6 8 Nxf6+ Nxf6 9 0-0 0-0 10 Rel White’s position is
clearly more comfortable at this point, W.Browne-C.Lakdawala, Montreal
(simul) 1977.
7 Nxf6+ exf6
 This resembles the old 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Nf6 5 Nxf6+ exf6
variation.

Question: What factors have altered with the recapture on f6?

Answer: The following:

1. White’s healthy 4:3 queenside majority is capable of creating a passed
pawn in an ending, whereas Black’s majority cannot do the same.
2. White can play for c2-c4 and d4-d5 in the middlegame, when the newly
passed d-pawn may tie Black down.
3. Black’s last move immediately ceded White a central space advantage.
4. By recapturing with the e-pawn, Black’s king becomes a lot safer if he
castles kingside, since the f6-pawn adds further protection.
5. The f6-pawn controls both e5 and g5, which prevent ideas of Ne5 or
Bg5.
6. Black may later play ... f6-f5, ... Nd7, ... Nf6, and then influence the
centre via control over e4.
7. If White castles kingside, Black can actually try a pawn storm with ...
h7-h5 and ... g6-g5, although this is obviously not without risk.

Question: Who do you believe got the better of the bargain?

Answer: White, though only slightly, since it’s close to an actual pawn sacrifice from Black’s side.
8 Bc4
This is a more logical square for the bishop than d3 where it just hits a
wall. White can also develop the bishop to e2 and play for an early c2-c4; e.g.
8 Be2 0-0 9 0-0 Re8 10 c4 Na6 11 Be3 Nc7 12 Qd2 Be6 13 Rfdl and I prefer
White since Black has nothing to do but wait, D.Alsina Leal-D.Gutierrez
Olivares, Seville 2020.
8 ... 0-0 9 0-0 a5
The immediate 9 ... Nd7 is played a bit more often.
10 a4 Nd7
10 ... Na6 is also seen, with the idea of... Nc7, keeping watch over the d5-
square.
11 Bf4 Nb6 12 Bb3 Nd5
Gaining a tempo on the f4-bishop, while if White then attempts to kick the
knight away with c2-c4, it can slide into the newly created hole on b4. This
means that White will likely shift to a plan of c2-c3, Nd2 and Nc4.
13 Bh2 Be6 14 Rel
 .14.. h5!?
This move is based upon the thought: if you seek to be heard, then make
some noise. Black can follow with ... Bh6 and even consider a pawn storm on
the kingside. Such a thrust is potentially injurious, not because it weakens,
but because it may be the precursor to further ambitious pawn pushes on the
kingside.
The previous year Wei Yi tried 14 ... Re8 15 Nd2 f5 16 Nc4 f4 17 c3 g5!
18 Ne5 in Lu Shanglei-Wei Yi, Chinese Blitz Championship, Tianjin 2018,
where I would be happy to play Black after 18 ... Qb6 (threatening ... Nxc3
and ... Qxb3) 19 Nc4 Qc7.
15 Nd2
Intending Ne4-c5, which Black prevents.
15 ... f5 16 c3 Bh6
Seizing control over f4.
17Nc4f4’?
 Risk has a way of following ambition. White can now look to pile up on
the f4-pawn with Ne5-d3 and Qf3.
18 Ne5 Qg5
Threatening ... Bxh3, which is easy to prevent. Black’s queen was better
off on f6.
19 Qf3 Rae8 20 Re2 f6?
Black seeks to take the initiative, but his weakening move has the opposite
effect. 20 ... Re7, followed by ... Rfe8, was preferable.

Exercise (combination alert): Prove why Wei Yi’s last move was a blunder.

Answer: Removal of the guard/knight fork/trapped piece.

21 h4! Qf5
There is no way for Black’s queen to skitter away from the danger zone. If
21 ... Qxh4? 22 Nxg6 Qg5 23 Nxf8 and 23 ... Bg4?, then 24 Rxe8! Bxf3 25
Ne6+ Kf7 26 Nxg5+ Kxe8 27 Nxf3 leaves Black a rook and bishop down.
 22 Bc2
The black queen has fallen on hard times. She is trapped and Black doesn’t
get enough in return.
22 ... Qxc2 23 Rxc2 fxe5 24 dxe5 Bg4!
This wins some of the material back, yet not enough.
25 Qd3 Bf5 26 Qdl Bxc2 27 Qxc2 Re6
Covering g6.
28 Rel

In the combination’s aftermath Black obtained rook and knight for his
queen, which isn’t enough, even if the e5-pawn falls as well.
28 ... Bg7 29 c4’?
Nepo is willing to give up control over b4 in order to drive Black’s knight
from its central post on d5. Here 29 Re4 is a safer alternative, but as we
should know by now Nepo doesn’t do safe very well!
29 ... Nb4 30 Qd2
 More accurate is 30 Qc3! Rf5 31 Qd4 Rfxe5 32 Rxe5 Bxe5 33 Qd8+ Kf7
34 Qd7+ Re7 35 Qd2! Bf6 36 Bxf4 Bxh4 37 Be3, intending Bb6 and Bxa5,
and White is easily winning.
30 ... Rxe5
Or 30 ... Bxe5 31 Rxe5! Rxe5 32 Bxf4 Ref5 33 Be3 c5 34 Qd6 R8f6 35
Qb8+ Kh7 36 Qxb7+ Rf7 37 Qb6 and Black’s queenside pawns begin to fall.
31 Rxe5 Bxe5 32 g3! fxg3
If 32 ... 13? then 33 Qg5! (double attack) 33 ... Bg7 34 g4! hxg4 35 Qxg6
Rf6 36 Qxg4 and White will convert.
33 Bxg3 Bxg3 34 fxg3 c5 35 Qd7
Simpler is 35 Qg5! with a double attack on g6 and c5.
35 ... Rf7
I’ve read that many people distrusted banks during the Great Depression,
preferring instead to stuff their cash under a mattress. Despite Wei Yi’s best
efforts to hang on to his money, he keeps losing it, pawn by pawn.
36 Qd8+ Kg7 37 Qxa5 Na6
With the loss of the a5-pawn, the credibility of Black’s hoped-for fortress
draw is shaken, yet not destroyed completely.
38 Qb6 Kh7 39 Kg2 Kh6 40 a5 Kh7 41 Qd6 Kg7
 Exercise (planning): At first sight it appears as if Black has erected a successful
drawing fortress. It crumbles if you find White’s winning plan.

Answer: Push the g-pawn, which activates White’s king, while stripping Black’s of protection.
42 g4! hxg4 43 Kg3 Kh7 44 Kxg4 Kg7 45 Qe6 1-0
White wins after 45 ... Rf8 (or 45 ... Rf6 46 Qe5 Kf7 47 Qd5+, picking off
the b7-pawn) 46 Qe5+ Kh6 47 Qg5+ Kg7 48 h5 Rf6 49 h6+ Kf7 50 h7 Kg7
51 h8Q+ (overloaded defender) 51 ... Kxh8 52 Qxf6+, followed by Kg5 and a
quick mate.

Game 38
M. V achier Lagrave-I.Nepomniachtchi
Grand Chess Tour, Paris (rapid) 2019
Caro-Kann Defence
1 e4 c6’?
What? Nepomniachtchi on the Black side of a Caro-Kann? Nepo is one of
those players who rarely switches from his normal repertoire. But remember:
we all knew that Bobby Fischer would open with his “best by test” 1 e4 with
the regularity of waves crashing on the beach. Then all of a sudden, in his
World Championship match he stunned Boris Spassky with 1 c4.

Question: Why would a world-class player not just stick to what he knows?

Answer: Because if you do that 100% of the time, your opponent also knows in advance exactly
which positions will be reached, which in turn allows deeply prepared, computer analysed ambushes.
Nepo probably had something new in the opening planned for MVL, which is why he deviated with 1
... c6.
2 d4 d5 3 e5 c5

Question: Isn’t Black just playing an Advance French a tempo down, since he
expended two moves to push the c-pawn to the fifth rank?

Answer: Yes and no. Black did indeed lose a tempo. On the other hand, he still has freedom for his
light-squared bishop, which you don’t get in the French Defence where it is obstructed by a black pawn
on e6. This line is a popular alternative to the more often played 3 ... Bf5.
4 Nf3 cxd4
The immediate capture is Black’s third choice behind 4 ... Nc6 and 4 ...
Bg4.
5 Qxd4!?
 Question: Why would MVL capture with the queen, given that 5 ... Nc6 gains
a tempo?

Answer: Here are some of the consequences of playing 5 Qxd4!?:

1. It is rarely played and MVL smelled a rat when Nepo played 1 ... c6. So
this move is an attempt to dodge Nepo’s preparation, had MVL stuck to his
normal program.
2. When Black plays ... Nc6, White’s queen will shift to f4, which in turn
disables Black’s intent to de-badify the light-squared bishop with either ...
Bf5 or... Bg4.
3. White’s queen on f4 hovers ominously near the black king and may help
participate in a future attack.
4. Moving the queen to f4 leaves White vulnerable to a future ... Nb4.
The alternative capture 5 Nxd4 is played far more often and is probably
stronger.
 5 ... Nc6 6 Qf4
6 Bb5 e6 7 0-0 (or 7 c4 Bd7!) 7 ... Bd7 8 Qf4 is just equal.
. e6
6..

Question: Isn’t this now a tempo-down French Defence after all?

Answer: Yes, but not a very dangerous one, since White’s piece placement looks rather clumsy.
7 Bd3
Here 7 a3 prevents ... Nb4 at the cost of a precious tempo, which Black
might utilize with 7 ... f6!?, removing White’s pawn presence in the centre.
7 ... Nb4!
Consequence #4 above. Black eliminates a potentially dangerous attacker,
weakens White along the light squares and, as a bonus, picks up the bishop
pair.
8 Nc3
White usually castles here, but 8 0-0 Nxd3 9 cxd3 Ne7 10 Nc3 Ng6 is
quite equal. MVL has a different idea in mind.
8 ... Bd7
White’s bishop isn’t going anywhere, so there is no rush to swap it off.
9 h4!?
 This is both aggressive and strategically risky, since the h-pawn can now
be fixed on the same colour squares as White’s remaining dark-squared
bishop (the light-squared bishop won’t be in the picture for long). Instead, 9
0-0 Nxd3 10 cxd3 Ne7 11 Ng5 Nf5 12 g4 Nh6 leads to mutual chances;
Black’s knight is misplaced on h6, while White may later regret playing the
loosening g2-g4.
9 ... h5
Discouraging g2-g4, while enabling ... Nh6.
10Qg3
I would base my strategy around d4 with something like 10 0-0 Rc8 11
Bd2 Nh6 12 Ne2! Nxd3 (not 12 ... Nxc2?? 13 Rael and the knight is trapped)
13 cxd3, when I prefer White’s control of d4 over Black’s control of the light
squares.
.
10.. Ne7 11 Nd4
Pre-empting ... Nf5.
 .11.. Qb6
This way any move by the cl-bishop can be met by ... Nxd3 and ... Qxb2.
12 a3’?
This may not be such a great idea since it allows Nepomniachtchi to make
further gains on the light squares. 12 Nce2 Nec6 looks pretty even.
12 ... Nxd3+
After 12 ... Qxd4 13 axb4 Qxb4 14 Bd2 Qg4 15 Qh2! White has full
compensation for the pawn.
13 Qxd3
Necessary, to defend the loose d4-knight.
13 ... Ng6!
This forces the highly undesirable f2-f4, which further weakens White’s
light squares.
14 f4 Rc8 15 Nce2 Bc5 16 b3
Not 16 b4? Bxd4 17 Nxd4 Rc4 18 c3 (and not 18 Be3?? Nxf4!,
overloading White’s bishop) 18 ... 0-0 19 Bd2 Rfc8 with a clear strategic
advantage for Black, who now has complete control of the light squares.
16 ... Qa6! 17 Qxa6
White would like to keep queens on the board, but not at the cost of
backing down and losing the initiative: after 17 Qdl? 0-0 18 Be3 f6! White is
in deep trouble.
17 ... bxa6
 Question: Black seems to have sacrificed his queenside structure, so shouldn’t
his plan be labelled dubious?

Answer: I like it for the following reasons:

1. White’s main way to play for a win in the position is to attack. By
removing queens, Black also removes any hopes of a white attack.
2. An ending is favourable for Black due to his light square control and
potential pressure down the c-file.
3. Black has a future possibility of posting the light-squared bishop on b5,
inducing Nxb5 a6xb5, after which the once-damaged structure is fixed.
4. Are Black’s doubled a-pawns even weak except as an optical eyesore? I
don’t see any way for White to attack them.
18 Bd2 Bb5
Black can also go for 18 ... 0-0 19 a4 f6 20 Nf3 Ne7, when I prefer his
position.
 19 Nxb5
If 19 Rh3, then 19 ... Bxe2 20 Nxe2 Be7! wins a pawn, since c2 and h4 are
simultaneously attacked.
19 ... axb5 20 b4?’
This is a self-inflicted strategic wound:
1. White places yet more pawns on the same - and therefore wrong -
colour as his remaining bishop, demoting it to real “bad bishop” status.
2. It further weakens his light squares; in particular, White cedes control
over c4 to a black rook.
Instead, White still looks okay after 20 a4 b4 21 a5 Ne7 22 Ra4 Nc6 23
Ncl!Rb8 24 Nd3 Be7 25 a6.
20 ... Bb6 21 a4 bxa4 22 Kdl Rc4 23 Rxa4 0-0

Getting ready to hammer away at White’s backward c2-pawn.

24 Ra2 Rfc8 25 g3 Ne7
Eyeing f5 as a juicy hole.
 26 Rfl Kh7!
Now Black’s king has the option of walking into White’s position via the
sickly light squares.
27 RI3 Kg6! 28 f5+’?
Desperation. MVL didn’t believe passive defence would save him. Let’s
examine and see if his assessment was correct: 28 Rd3 Nf5 (28 ... Kf5?! is a
touch premature and can be met by 29 g4+ hxg4 30 h5!, threatening Ng3
mate, or 30 ... Bf2 31 Rxa7! and White has generated some counterplay) 29
Rb2 d4! 30 Ra2 Nh6! (now is the right time for the king to invade) 31 Rb2
Kf5 32 Ra2 Ng4 (threatening to fork on f2) 33 Rda3 Ke4! (this is the king of
kings!) 34 Rb3 Nf2+ 35 Kcl R8c7 and White doesn’t have a prayer of saving
the game.
28 ... Nxf5 29 Nf4+ Kh7 30 Nxh5 Rg4!
White is now tied down to defence of g3.
31 Raa3 Rcc4 32 Nf4
 Exercise (planning): It’s obvious that White’s hoped-for counterplay is stillborn.
Come up with a winning plan for Black.

Answer: Move the knight to d4 with a double attack on c2 and f3.

32 ... Nd4!
Black wins a pawn and from this point on his pieces pour out joyously like
third graders leaving the classroom at recess.
33 Rfl
33 Rfc3 loses to simply 33 ... Rxb4.
33 ... Nxc2 34 Rd3 Nxb4
Another one bites the dust.
35 Rdf3 Nc6 0-1
After 36 Nd3 Nd4! 37 Rxf7 Rxg3 38 Nf4 Ra4 it is White’s king that gets
fried.

By 2019 Nepomniachtchi’s rating crept close to the stratospheric 2800

range, transitioning into a serious threat to the reigning world champion. I
remember watching the following epic game live and being impressed first
with Nepo’s calculation power and then, later in the game, his dogged
determination to win a drawn position - and against a world-class player.

Game 39
N epomniachtchi-H.N akamura
I.
Sinquefield Cup, St. Louis 2019
Queen’s Gambit Declined

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 Be7 5 Bf4 0-0 6 e3 c5

The most direct response. 6 ... Nbd7 is Black’s second choice.
7 dxc5 Bxc5
 8 cxd5
This move indicates respect for Nakamura’s prowess with opposite wing
attacks. 8 Qc2 Nc6 9 a3 Qa5 10 0-0-0 (10 Rdl is the safer version) 10 ... Be7
11 g4 is a more aggressive option.
8 ... Nxd5
This is the superior capture since it opens ... Bb4+ possibilities later on. If
you look in the database you see myriad Nakamura games from this position,
so possibly Nepo had something prepared, had Nakamura not deviated first.
9 Nxd5 exd5 10 Bd3
Nepo allows the bishop check on b4, refusing to waste a move on 10 a3,
which is White’s main line.
10.. . Bb4+11 Nd2
11 Ke2 is often played too. It’s not as risky as it appears since White has
time for Qc2, Rhdl and Kfl, castling by hand.
11.. . Nc6 12 0-0 Be7
 12 ... d4 doesn’t eliminate the isolani since White would bypass with 13
e4.
13 Rcl Qb6
Slightly unusual just here. 13 ... Bf6 is more common.
14 Qc2 Nb4?’

This is a dubious novelty. Nakamura’s love for a fight gets the better of his
judgment. He ignores White’s threat to chop the h7-pawn, with check, and
counterattacks on the other side. In the coming calculation battle the
complications are in White’s favour, as long as he keeps finding the best
moves. Simply 14 ... g6 15 a3 Bf6 16 b4 Ne5 was equal in R.Fuzishawa-
J.Quinones Borda, correspondence 2015.
15 Bxh7+! Kh8 16 Qc7!
You haven’t fallen for a trap if you have no plans to leave. Only this move
secures a clear advantage. After 16 Qbl?? g6 17 a3 Na6 18 Bxg6 Qxg6
White doesn’t have enough for the sacrificed piece.
 .16.. Qxc7
16 ... Kxh7?? 17 Qxe7 is completely lost for Black.
17 Rxc7 Bd8
Now what? Two white pieces hang simultaneously. Nepo has an answer
ready.
18 Bd6! Bxc7 19 Bxf8 a5’?
After 19 ... Kxh7 20 Bxb4 Be5 21 b3, Black doesn’t have enough
compensation for the sacrificed pawn, despite the two bishops. The text keeps
more pieces on the board, yet also allows White to hang on to his bishops.
20 Bbl Be5 21 b3 Bg4 22 Bc5?’
22 Bxb4! appears counter-intuitive since it weakens the dark squares, yet
the engine shows it to be the best line: 22 ... axb4 23 h3 Be6 24 N13 Bc3 25
Rdl and White’s position is superior to the one he got in the game.
22 ... Rc8 23 Bb6?
He should backtrack with 23 Bxb4! axb4 24 13 Bd7 25 Rdl, when Black
must still prove he can draw a pawn down.
 There is before and there is after, but the most important part is what is in
between. Principle: Never underestimate a desperate opponent, since a
desperate mind is an inventive one, which will do anything to survive. I
played Hikaru a lot of online blitz when he was a 13-year-old IM. He would
routinely get outplayed strategically, yet if there was a tactic or combination
in the position, he saw it with engine-like frequency.

Exercise (combination alert): What did Nepo overlook on his last move?

Answer: Step 1: Sacrifice the knight on a2.

23 ... Nxa2! 24 Bxa2 Rc2
Double attack, which regains the piece with equality.
25 M3 Bxf3 26 gxf3 Rxa2 27 Rdl Rai!
Black has two other routes to a likely draw:
a) 27 ... d4! 28 f4 Bf6 29 Bxd4 Bxd4 30 Rxd4 Rb2 31 Rd5 Rxb3 32 Rxa5
b5 and the ending is drawn, since Black’s passed b-pawn ensures adequate
counterplay.
b) 27 ... Rb2! 28 Bxa5 Rxb3 29 Rxd5 Bf6 and I don’t believe White has
realistic chances to win.
28 Rxal Bxal 29 Bxa5

White should not be able to win the coming bishop ending, despite the
extra pawn.
.
29.. f5!
Now White is forced to swap a pair of pawns if he ever plays e3-e4.
30 Kfl Kg8 31 Ke2 Kf7 32 Kd3 Be5 33 Bc3 Bd6
White remains a pawn up, but this is his only source of solace since Black
should hold the game. Not 33 ... Bxh2?? (we all get flashbacks to Fischer’s
disastrous decision to grab Spassky’s h2-pawn in their world championship
match) 34 f4, when 34 ... g5 35 fxg5 Bd6 36 Kd4 Ke6 37 g6 and Black is
completely lost.
34 e4!
 Question: Isn’t Nepomniachtchi deliberately in violation of the principle:
The pawn-up side should not swap away pawns in an ending -1

Answer: Correct, yet there are extenuating circumstances. In this instance his move is strong, not
because it wins, but because it places upon Black the maximum practical difficulties in holding the
draw. If he obediently follows principle with 34 h3, Black draws with ease; e.g. 34 ... g6 35 Kd4 Ke6
36 f4 Be7 37 Ba5 Bf8 38 Bc7 Be7 39 Be5 Bh4 40 13 Be7 41 Bc7 Bb4 42 e4 (otherwise no progress can
be made) 42 ... fxe4 43 fxe4 dxe4 44 Kxe4 Bc5 and the ending is dead drawn.
.34.. dxe4+
Technically this should still draw, yet Nakamura gives his opponent
practical chances with an inaccurate move. Instead, 34 ... Ke6! 35 Bxg7 Bxh2
is completely drawn; e.g. 36 exf5+ Kxf5 37 Kd4 Ke6 38 Kc5 Bc7 39 b4
Bd6+ 40 Kb5 Kf5 41 Bc3 Kf4 42 Kb6 d4 43 Bd2+ Kxf3 44 b5 Kxf2 45
Kxb7 with a draw.
35 fxe4 Ke6
Hikaru may have originally intended 35 ... fxe4+ 36 Kxe4 Bxh2??, which
fails to 37 Be5! Bxe5 (37 ... Bgl 38 Kf3! traps the bishop) 38 Kxe5 Ke7 39
f4 g6 40 b4 b6 41 b5 (zugzwang) 41 ... Kf7 42 Kd6 Kf6 43 Kc6 Kf5 44 Kxb6
and Black loses the race.
36 Bxg7 Bxh2 37 Bd4
Now comes a test of endurance.
37 ... b5 38 Bb6 Be5 39 Ke3 Bb2 40 Bd4 Bcl+ 41 Kd3 Ba3 42 Bb6 Bb2
43 Bc7 Bf6 44 Ba5 Be5 45 Bb4 Bf6 46 Bc5 Be5 47 Ba7 Bf6 48 Bd4 Be7 49
Ke3 Bg5+ 50 f4
This is a slight concession from Nepo. Now it becomes obvious that the
position is drawn again - except that the game didn’t end in a draw!
50 ... Be7 51 Bb6 Ba3 52 Kd3 Bel 53 Bc7 Bb2 54 Bb8 Bf6 55 Ke3 Bb2
56 Be5!
This guy just never gives up! Nepo’s move is a cheapo attempt. Many
players with White would have given up trying to win and offered a draw by
this point. Nepo’s obstinate determination soon pays off.
56 ... Bcl+
Black must retain the bishops. 56 ... Bxe5?? 57 fxe5 Kxe5 58 exf5 Kxf5
59 Kd4 wins.
57 Kd3
 57 ... b4??
I’m guessing this was a time-pressure error. Principle: Don’t place your
pawns on the same colour as your opponent’s remaining bishop. Now the
position extracts its dues from Nakamura. After 57 ... Ba3! the game is still
drawn; e.g. 58 exf5+ Kxf5 59 Kd4 Bb4! (not 59 ... Bel?? 60 Kc5 Bxf4 61
Bxf4 Kxf4 62 Kxb5 and wins) 60 Kd5 Bel 61 Kc5 b4 62 Kc4 Ke6 63 Bc7!
(intending Ba5) 63 ... Kf5 64 Bd6 Bd2 65 Bxb4 Bxf4 66 Kd5 Bel 67 Bc3
Ba3 and Black holds.
58 Bb8 Kf6
Or 58 ... Bb2 59 Ba7! Bel 60 Be3 Bb2 61 Bd2! Ba3 62 exf5+ Kxf5 63
Kc4 Ke4 64 Bxb4 Bel 65 Bd6 and White wins.
59 Bd6 Ke6 60 Be5 Ba3?’
Making it a little easier. After 60 ... fxe4+ 41 Kxe4 Bd2, the tablebase
declares mate in 49, but the required zugzwang techniques with king and
bishop are quite obscure, and the end result is a bishop endgame with a single
pawn which you’d need to know was won. For example: 62 f5+ Ke7 63 Bf4
Bc3 64 Kd3 Kf6 65 Bd2 Bb2 66 Ke4 Ba3 67 Be3 Kf7 68 Bg5 Bb2 69 Kd5
Bc3 70 Kc4 Bel 71 Kd3 Bg3 72 Bd2 Kf6 73 Ke4 Bd6 74 Bel Bf8 75 Bh4+
Kf7 76 Bg5 Bd6 77 Kd5 Bg3 78 Bd2 Kf6 79 Bxb4 Kxf5 80 Bc3 and White is
winning.
61 Bd4 Bel 62 Be3! Bb2 63 Bd2!
It becomes clear that Black’s position is a sea of grief and regret for his
57th move blunder.
63 ... fxe4+
Or 63 ... Ba3 64 exf5+ Kxf5 65 Kc4 Ke4 66 Bxb4 and wins, as after 48 ...
Bb2 above.
64 Kxe4 Ba3

Exercise (calculation/planning/critical decision): Should White push with 35 f5+

to push the black king back or go straight for the b-pawn with 65 Kd4 etc?
 Answer: White should go after b4.
65 Kd4!
Now Black’s b-pawn falls and his bereaved position can do nothing but
cradle the cold corpse of a beloved pet. There’s no need for 65 f5+?! Kf6,
which only prolongs the game and forces White to find the manoeuvres that
follow 66 ... Ba3 in the note to move 60.
65 ... Kf5 66 Kd5
Or simply 66 Kc4 Ke4 67 Bxb4 Bb2 68 Bd6.
66 ... Kf6
Or if 66 ... Kg4 then 67 Ke6 Bb2 68 f5 Bc3 69 Be3 KI3 70 Bg5 Ke4 71 f6
Kd3 (71 ... Bxf6 fails to 72 Bxf6 Kd3 73 Kd5 Kc2 74 Kc4) 72 17 Bg7 73 Be7
Kc2 74 Bxb4 Kxb3 75 Bd6 Kc4 76 Kf5! Kd5 77 Kg6 wins.
67 Kc5 Kf5 68 Kb5 Ke4 69 Bxb4 1-0

Game 40
E.Tomashevsky-I.Nepomniachtchi
FIDE World Cup, Khanty-Mansiysk 2019
London System

1 d4 Nf6 2 Nf3 g6 3 Bf4

Welcome to my tribe. I started playing the London in the 1990s, long
before it was trendy. My sacred motto: If it’s fun, then I’m against it!
3 ... Bg7 4 e3 0-0 5 c3
 I don’t like this move order for White, since holding the c-pawn back
allows c2-c4 options. I normally play 5 Be2 here.
5 ... d6

Question: Can’t Black pick up the bishop pair here with 5 ... Nh5 and so on?

Answer: Sure, but it comes with the cost to Black’s structure. If 6 Bg5 h6 7 Bh4 g5, O.Dobos-
D.Ngo Tan, Budapest 2008, rather than meekly moving the bishop to g3, White can go for the strong
novelty 8 g4! Nf6 (I also prefer White after 8 ... gxh4 9 gxh5) 9 Bg3 Ne4 10 h4 Nxg3 11 fxg3, when I
like White’s attacking chances.
Instead, 5 ... d5 is Griinfeld versus London, and 5 ... b6 6 Nbd2 Bb7 7 h3
c5 is Réti versus London.
6h3
Preserving the bishop.
6 ... Nbd7
6 ... Nfd7, intending ... e7-e5, is a decent alternative.
 7 Be2
7 Bc4 is also played here, provoking ... d6-d5.
. Qe8
7..
At some point Black plans to toss in ... e7-e5, seizing his fair share of the
centre. 7 ... b6 is the main alternative.
8a4’?
This move is somewhat rare. The idea is to meet... b7-b6 with a4-a5. 8 0-0
is normal.
8 ... a5

A new move. White is denied the space gaining a4-a5, and b2-b4 is also
discouraged since White must then contend with ... a5xb4 and ... Nd5, with a
double attack on the b4-pawn and the f4-bishop. On the other hand, Black
weakens the b5-square, allowing White to bring a knight there, or else to c4,
eventually playing for b2-b4. The immediate 8 ... e5 was played in all six
previous games to reach this position.
 9 0-0 b6
Black can also delay queenside development in favour of 9 ... e5 10 Bh2
Qe7.
10 Na3!
Now Black must watch out for the aforementioned ideas of Nb5, or Nc4
followed by Bh2 and b2-b4.
.
10.. Bb7 11 Bh2 Ne4

Question: What is the point of this move? Black’s knight just seems to hit air
on e4.

Answer: This manoeuvre is standard operating procedure in such London structures. Black’s e4-
knight is annoying and needs to be challenged, usually with Nd2. In that case knights are swapped,
when Black’s position is freed, and his f-pawn is given the go-ahead to move forward.
12 Nd2
I would toss in 12 Nb5 Rc8 first and only then play 13 Nd2.
12 ... Nxd2 13 Qxd2 e5 14 Nb5 Rc8
 15 Rfdl
This is slow and somewhat ineffective since, if White exchanges on e5,
Black will recapture with a piece and the rook doesn’t get a fully opened d-
file. I think White should proceed with his queenside attack: 15 b4! Qe7 (if
15 ... axb4 16 cxb4 then a4-a5 is coming and White stands clearly better) 16
bxa5 bxa5 17 c4 and White’s initiative on the queenside looks more potent
than Black’s on the other flank. (I tell anyone who will listen that the London
is a forced win for White!)
15 ... Qe7 16 Qc2
Slow. He should still go for queenside action with 16 b4!.
16 ... f5
Nepo begins to build for either a central or kingside attack.
17 dxe5
Even now, I would play 17 b4!.
17 ... Bxe5
 He wants to eliminate the h2-defender of White’s king.
18 Bxe5 Nxe5 19 Rd4?
After this blunder Black gets a nasty attack. Chances still look fairly
balanced if White brings a defender closer to his king with 19 Nd4!, and if 19
... f4 then 20 e4.
.
19.. Qg5!
The mate threat is not the point. The main idea is to weaken the white
king’s pawn cover.
20 g3
The unpleasant alternative is to buy his way out with 20 Bfl Nf3+, when
Black wins a clean exchange.
20 ... Qh6! 21 Kh2?
He had to try 21 Rh4 Qg7. Black threatens ... g6-g5, followed by ... Qh6
again, when White’s king is under attack but not as severely as in the game.

Exercise (planning): Find one powerful idea and Black whips up a decisive attack.
 Answer: Push the g-pawn.
21 ... g5!
Tomashevsky faces an apocalyptic development. The threat is ... Rf6,
followed by ... Qxh3+!, forcing mate.
22 Qdl Rf6?
The door of opportunity threatens to slam shut in Nepo’s face. He misses
an immediate knockout with 22 ... Ng4+! which wins easily; e.g. 23 Bxg4 (23
Rxg4 fxg4 is the same story) 23 ... fxg4 leaves f2 and h3 simultaneously
hanging, which 24 Qfl fails to defend because of 24 ... Rxf2+! 25 Qxf2
Qxh3+ and mates.
23 f4!
This is White’s only prayer.
23 ... NÍ7
The Great Romantics may approve but the truth is that the queen sacrifice
23 ... Qxh3+?? is now completely unsound: 24 Kxh3 Rh6+ 25 Bh5 g4+ 26
Kh4! Nf3+ 27 Qxf3 gxf3 28 Rd2 Bc6 (threatening ... Be8) 29 Kg5! Kg7 30
g4 and Black is completely busted.
24 fxg5’?
24 Qfl is an alternative. Black can try 24 ... Kh8, intending to swing the
c8-rook to g8 eventually.
24 ... Qxg5?’
Black’s attack is stronger after 24 ... Nxg5! 25 Rh4 Qg7 26 Qd4 Kh8.
25 Rf4!
Clearing the way for Nd4.
25 ... Kh8 26 Qgl!
Tomashevsky is outplaying Nepo with powerful defensive moves. The
queen adds simultaneous coverage to White’s weak pawns on g3 and e3,
while also making room for Rafi.
26 ... Ne5 27 Rafi Rh6 28 Rxf5??
At this point White could save himself with 28 Qf2! Rg8 (note that 28 ...
Ng4+?? fails miserably to 29 Bxg4 fxg4 30 Rf8+ and White’s attack arrives
first) 29 Rxf5 Ng4+ 30 Bxg4 Qxg4 31 h4 Bc8 32 Rf8 Qh3+ 33 Kgl Rhg6 34
Rxg8+ Rxg8 and White can take perpetual check with 35 Qf6+ Rg7 36 Qf8+
etc.
 Exercise (combination alert): This is a detail-specific position and a guess
isn’t going to cut it. Nepo found a murderous shot on his next move.
How would you continue?

Answer: Sham queen sacrifice.

.28.. Qg4!!
The basketball terminology for such a shot is: nothing but net. The queen
insinuates herself within the white king’s sphere, threatening mate on h3 as
well as the e2-bishop. This was why 28 Qf2 was necessary: to defend the
bishop and give the white king a retreat square at gl.
29 Rf8+
White’s position is in free-fall and nothing works. Black’s queen is not
really hanging, since 29 Bxg4 Nxg4 is mate. 29 h4 doesn’t cut it either, as
Black mates in two with 29 ... Rxh4+! 30 gxh4 Qxh4.
29 ... Rxf8 30 Rxf8+ Kg7
 Swapping rooks hasn’t helped. Black still threatens mate and now the f8-
rook is hanging too.
31RF7+
Super-desperation. 31 Qfl fails to the decoy shot 31 ... Rxh3+! 32 Qxh3
Qxe2+ 33 Kgl Qel+ 34 Rfl Nf3+ 35 Khl Ng5+ 36 Kh2 Qe2+ with mate in
two.
31 ... Kxf7
Not the bone-headed 31 ... Nxf7??, as White can then chop Black’s queen
with 32 Bxg4, winning.
32 Qfl+ 0-1
The idea is to play Bxg4+ next move, since gl is now open for White’s
king. Tomashevsky chose to resign at the same time, seeing that it fails
miserably to 32 Qfl+! Nf3+ (not 32 ... Kg7?? 33 Bxg4 Nxg4+ 34 Kgl) 33
Bxf3 Qxf3, leaving Black a rook up.
This game may well force an admission that my beloved London System
isn’t a forced win for White after all.

Game 41
A.Giri-I.Nepomniachtchi
Candidates Tournament, Yekaterinburg 2020
English Opening

1 Nf3 Nf6 2 c4 c5 3 Nc3 Nc6 4 d4 cxd4 5 Nxd4 e6 6 g3

Here 6 a3 is a major alternative.
. Qb6
6..
Black’s most popular choice, ahead of either 6 ... Bc5 or 6 ... Bb4.
7 Ndb5
7 Nb3 is White’s main reply.
7 ... Ne5
 Covering the d6-square, while attacking c4 in return.
8 Bf4
The alternative is 8 Bg2, offering the c4-pawn.
8 ... Nfg4
Covering the e5-knight while menacing f2. The routine 8 ... d6?! is
considered inferior, since if 9 Bg2 a6 then 10 c5! dxc5 11 Qa4 Bd7 12 Bxe5
Rc8 13 Nc7+ wins material, M.Mulyar-G.Swati, Philadelphia 2006.
9e3
Covering f2.
9 ... a6
Now b2 hangs if the b5-knight moves.
10 h3
10 Qa4 is a major alternative; e.g. 10 ... Ra7!? 11 h3 (not 11 Be2? axb5 12
Qxb5 Qc6! 13 Qxc6 dxc6 14 h3 Nd7! 15 hxg4 e5 and White’s dark-squared
bishop was trapped, N.Meskovs-A.Giri, PRO League rapid 2019) 11 ... axb5
12 Qxb5, V.Topalov-F.Caruana, Moscow Candidates 2016, and now 12 ...
Qxb5 13 Nxb5 Bb4+ 14 Ke2 Ra5 15 hxg4 Nxc4 is probably about equal.
10 ... axb5 11 hxg4 Nxc4 12 Rcl!?
Gambits are the opiate of the masses, and sometimes also of the elite. Giri
tries out a rarely played pawn sacrifice. Instead, 12 Qb3 d5 13 Bxc4 dxc4 14
Qxb5+ Qxb5 15 Nxb5 regains the pawn, but with no advantage for White.
12 ... d5!
A novelty and a slight improvement over 12 ... Qa5, as in P.Sedlacek-
M. Suarez Sedeño, correspondence 2004.
13 b3 Bb4!

Question: Does Black get enough for the sacrificed piece?

Answer: It’s actually a sham sacrifice since White is unable to hang on to his knight. A later game
deviated with 13 ... Nd6 14 g5 Bd7 15 Bd3 Rc8 16 Kfl b4 17 Ne2 Rxcl 18 Qxcl Be7 19 Qb2 f6 20
gxf6 gxf6, A.Giri-P.Svidler, Chess24.com Banter blitz 2020, when 21 Bxd6 Qxd6 22 Nf4 would have
given White a decent position.
 14 bxc4 Ra3! 15 Be5 f6 16 Bd4 Qa5 17 Be2!
Giri correctly returns the piece to give his king room on fl. Not 17 Qc2??
e5 18 Rxh7 Rxh7 19 Qxh7 exd4 20 Qg8+ Kd7 21 Qxg7+ Be7 and White is
busted since he is about to be a piece down.
17 ... Bxc3+ 18 Rxc3 Rxc3 19 Kfl!
Not 19 Bxc3?? Qxc3+ 20 Kfl bxc4 and White is two pawns down for
nothing.
.
19.. b4!?
This move indicates that Nepo is playing for the full point. Instead, 19 ...
Rxc4 20 Bxc4 bxc4 21 g5 e5 22 gxf6! gxf6 23 Qh5+ Ke7 24 Bxe5! Qb6 25
Bd4 Qbl+ 26 Kg2 Qe4+ 27 Kgl Qbl+ is drawn.

Question: Why does Black have to return the exchange when he has 19 ... Ra3 - ?

Answer: That hands White a promising attack after 20 g5! 0-0 (not 20 ... e5?? 21 gxf6! gxf6 22
Bh5+ Kd8 23 Qf3 Qa6 24 Qxd5+ and White wins) 21 gxf6 g6 22 Rh6! with huge compensation for the
exchange. Black is unable to play 22 ... bxc4?? due to 23 Qc2!, when there is no remedy to the coming
rook sacrifice on g6.
20 g5
White’s kingside play grows dangerous, so Nepomniachtchi forces Giri to
reclaim the exchange, eliminating the dangerous dark-squared bishop.
20 ... e5! 21 Bxc3 bxc3 22 gxf6 gxf6
Castling is playable too: 22 ... 0-0!? 23 fxg7 Kxg7 and Black stands no
worse.
 23 Qbl?!
Attacking h7 while covering a2. The simple 23 cxd5 Qxa2 24 Qd3 looks
better, when I would rather take White than Black.
23 ... Qc7 24 Qd3
24 Qb4 c2 25 Qa4+ Qc6 (or 25 ... b5 26 Qxc2 bxc4 27 e4) 26 Qxc2 dxc4
27 Rh4 b5 28 a4 is also even.
24 ... b5! 25 Qxc3
White can’t take on d5, as either 25 Qxd5?? Bb7 or 25 cxd5?? c2 costs
him a rook.
25 ... bxc4 26 e4!
The c4-pawn is undermined.
26 ... dxe4 27 Rh4’?
Once you raise your fist, it pretty much ends all negotiations. Giri begins
to play for a win in a position he probably should be trying to draw. For
instance, 27 Qxc4 Qxc4 28 Bxc4 Ke7 29 Bd5 Ba6+ 30 Kel Bd3 31 Kd2 f5
32 Rh6 Rd8 33 Bb3 Bc4+ 34 Kel Bxb3 35 axb3 Rd3 36 Rxh7+ is a near
certain draw.
27 ... Be6 28 Rxe4 0-0 29 Bxc4?’
Giri proceeds with the untethered false confidence of the true believer. I
doubt that Black could convert if Giri had just left the c4-pawn alone and
played 29 Rh4! Rc8 30 Kgl.

Exercise (calculation): White’s last move isn’t absolutely an error since he

should still hold a draw, but it does unnecessarily offer Black practical
chances.
What should Black play?

Answer: Move the king to g7.

29 ... Kg7!
White is surprisingly unable to extricate himself from the c4-pin. 29 ...
Kh8? allows an escape with 30 Qb4 Rb8 31 Qa4 f5 32 Bxe6! fxe4 33 Qxe4
and White doesn’t stand worse.
30 Qb3?’
A second inaccuracy. The best option was 30 Qb4 Rb8 31 Bb5 Bd5 32
Re2 Qb7 33 Qg4+ Kh8 34 Qh4 Qxb5 35 Qxf6+ Kg8 36 Qg5+ with excellent
drawing chances, in view of the limited material on the board and Black’s
exposed king.
.
30.. Rb8!
About here Giri must have realized he was out-calculated. He is forced to
give up the queen for rook and bishop.
31 Bxe6!
Giri wisely plays for a fortress draw. Keeping the queen is clearly worse:
a) 31 Qc2?? Rc8 wins heavy material.
b) 31 Qd3? Bf5 wins the exchange.
c) 31 Qa4? Bd7 32 Qdl (or 32 Qc2 Bf5) 32 ... Bf5 33 Rg4+ (or 33 Rh4?
Rbl) 33 ... Kh8 34 Bb3 Bxg4 35 Qxg4 Qc6 and Black has decent chances to
convert.
31 ... Rxb3 32 Rg4+ Kf8 33 Bxb3 Qcl+ 34 Kg2 Qc6+ 35 Kgl h5
 Exercise (planning): Come up with White’s correct path to a fortress draw.

36 Rg8+?
Answer: 36 Rh4! Qcl+ (or 36 ... Q13 37 Bc4 f5 38 Bfl) 37 Kg2 Qg5 38 Bdl Qd2 39 Bxh5 Qxa2 40
Re4 Qd5 41 Bg6! Qe6 42 g4! creates an air-tight fortress and White will hold a draw.
36 ... Ke7 37 Rg7+ Kd6 38 Rh7 Qf3!
Preventing Bdl and hanging on to the critical h-pawn, without which
Black can’t win.
39 Rh8
Not 39 Bf7? Qdl+ 40 Kg2 Qbl! 41 Rg7 (41 Rxh5 Qb7+ picks off the
loose bishop) 41 ... Qe4+ 42 Kgl Qel+ 43 Kg2 Qe2 and Black hangs on to
his h-pawn since 44 Rh7?? hangs the rook to 44 ... Qe4+.
.
39.. e4!
Now ... e4-e3 is a serious threat.
40 Rd8+ Ke7 41 Bdl Qc3 42 Rd5
 Exercise (combination alert): How does Black force the win of White’s bishop
for two pawns?

Answer: Step 1: Push the h-pawn.

42 ... h4! 43 gxh4
No choice, since 43 Kg2 hxg3 44 fxg3 e3 45 Kf3 Qel 46 Rd3 Qf2+ wins.
43 ... f5!
Step 2: Push the f-pawn, which also must be captured.
44 Rxf5
If 44 Kg2 Qg7+ 45 Kh2 Ke6 46 Bb3 Qg4! 47 h5 Qxh5+ 48 Kg2 Kf6,
White is busted since Black’s queen, king and pawns will combine for a
winning attack.
44 ... Qel+ 45 Kg2 Qxdl 46 Rg5 Qal 47 Rg4 Qbl 48 Rg3 Qxa2 49 Rh3
 Question: Isn’t this a basic fortress draw, seeing as Black’s king can’t
ever penetrate White’s third rank?

Answer: It appears that way but it isn’t. Black’s king can infiltrate to White’s second and first
ranks. I can’t tell you how, since this would give away the answer to the next exercise.
49 ... Qd5 50 Kfl Qdl+ 51 Kg2 Qg4+ 52 Rg3 Qh5 53 Ra3 Qd5!
The h-pawn isn’t going anywhere. In fact it’s essential not to take it yet,
since 53 ... Qxh4?? 54 Kfl! would allow White his fortress.
54 Kgl Kf6 55 Rg3 Qdl+ 56 Kg2 Kf5 57 Rg5+ Kf4 58 Rg3 Qd5 59 Kfl
Qd2 60 Kg2 Qdl 61 Re3 Kf5
Heading back to pick off the h-pawn with his king.
62 Rg3 Kf6 63 Rh3
If 63 Rg5 then 63 ... Q13+ 64 Kgl e3! breaks through and wins, since
White cannot save the h-pawn. The endgame database declares mate in 26.
63 ... Kg6 64 Rg3+ Kh5 65 Rh3 Qbl 66 Re3 Kxh4 67 Rg3 Kh5 68
Rh3+ Kg4
Come on Nepo, please get on with it. My readers are busy people.
69 Rg3+ Kf4 70 Re3 Qdl 71 Ra3 Ke5 72 Rg3

Watching the game live at this stage I knew that if an over-the-hill IM like
me could see the win, then so had Nepo. I later gave this position to students
and asked them to evaluate it. When the great majority claimed the game was
“dead drawn”, I informed them that I don’t subscribe to the customer-is-
always-right motto.

Exercise (planning): How did Nepo break Giri’s fortress?

Answer: Step 1: Move the king to d4.

72 ... Kd4! 73 Re3 Qd3! 0-1
Step 2: Offer the queen on d3. If it’s taken, Black wins the king and pawn
ending; if it is declined, Black’s shielded king slips by the third rank barrier,
via c3.
 For example: 74 Rg3 (after 74 Rxd3+ exd3 75 Kfl Kc3 76 Kel Kc2 the
newly passed d-pawn promotes) 74 ... Kc3! (from this point Nepo pierces his
opponent’s defences without struggle or resistance) 75 Rh3 Kd2 76 Rg3 Ke2
77 Re3+ (or 77 Kgl Qxg3+! 78 fxg3 e3 79 g4 Kd2 etc) 77 ... Qxe3 (Step 3:
Simplification, into a no-brainer winning king and pawn ending) 78 fxe3
Kxe3 79 Kfl Kd2 forces promotion.

Nepomniachtchi must have incredibly fortunate Petroff s Defence karma,

since it was the Petroff which was instrumental in helping him become
Magnus’ official challenger. In the World Championship Candidates
tournament, in both 2020 and again when it resumed in 2021, he first
defeated Wang Hao on the White side of a Petroff, and then in the 2021
version repeated the feat on the Black side! If Wang Hao had held both
games, maybe Magnus would be facing MVL in the 2021 world
championship match, rather than Nepo.

Game 42
I.
Nepomniachtchi-Wang Hao
Candidates Tournament, Yekaterinburg 2020
PetroffDefence

1 e4 e5 2 Nf3 Nf6 3 Nxe5 d6 4 Nf3 Nxe4 5 d4

Wang Hao opted for 5 d3 in their return game (see Game 60).
5 ... d5 6 Bd3 Bf5
And here he tosses in another sideline, one that Nepo had faced several
time before. 6 ... Nc6, 6 ... Bd6, and 6 ... Be7 are seen far more often.
7 0-0 Be7 8 Rel 0-0
 9Nbd2’?
Oh, the excitement. How are we to tame this wild stallion of a position?
Nepo’s last move essentially says: “Let’s go low tech.” The coming position
seems temperamentally suited only for players with a sedentary style and low
life expectations.

Question: So why did Nepo play it then?

Answer: Good question. 9 c4 is White’s only chance for an edge.

9 ... Nd6! 10 Nfl
Deviating from an earlier game, since 10 Nb3 c6 11 Bf4 Bxd3 12 Qxd3
doesn’t give White much of anything, I.Nepomniachtchi-J.K.Duda, Wijk aan
Zee 2019.
10 ... Bxd3 11 Qxd3 c6

Question: Am I correct in concluding that Nepomniachtchi hasn’t handled

 the opening with great energy?

Answer: Is it okay if I close my eyes for a bit? I’m having trouble staying awake. Somehow I don’t
sense a whole lot of adventure coming our way, since the position is essentially a painfully boring
version of an Exchange French, where caffeine in high doses becomes essential to continue playing
though the game. Nepo later makes up for his uninspiring play in the opening with a masterfully
energetic handling of the ending.
12Bf4Na6’?

Question: Why to a6 rather than d7?

Answer: Black almost always puts the knight on a6 here. White’s only real asset is that he can plant
a piece on e5, and if Black later plays ... f7-f6, this would weaken e6. With ... Na6! and ... Nc7, Black
reinforces that square. The downside of transferring his knight to c7 is that it is further away from
Black’s king. Nepo immediately attempts to take advantage of this subtle alteration.
Instead, 12 ... Nd7 13 Qb3 (or 13 Ng3 Re8 14 Re2 Nf6 15 Rael BIB) 13 ...
Nb6 14 a4 Nbc4 15 Ne5 Nxe5 16 dxe5 Ne4 looks both boring and equal,
A.Esipenko-J.K.Duda, Wijk aan Zee 2021.
13 h4!
A novelty and a clever one. Nepo indicates that he isn’t interested in a
draw, making the first aggressive gesture of the game. If White is going to
get anything in the position, it will be on the kingside where the black king is
slightly short on defenders. He may have come up with this idea after his
game with Duda above. Still, it’s difficult to foresee any danger just yet,
since semi-symmetrical Exchange French lines are not known for their
slashing attacking qualities.
 13 ... Nc7 14 Ng5!
Nepo seeks to create an imbalance.
14 ... Bxg5
The natural response. The following year Duda opted for 14 ... g6!? 15
Ne3 Qd7 16 h5, W.So-J.K.Duda, Opera Euro Rapid 2021, when Black’s
game looks fine after 16 ... Bxg5 17 Bxg5 f5! 18 Bf4 Ne6 19 Be5 Nf7.
15 Bxg5 f6 16 Bf4 Qd7 17 Ng3 Rae8?’
Believe it or not, this move, the most natural one on the board, was
inaccurate. 17 ... g6! 18 h5 Kf7! was better; e.g. 19 Qd2 Ne6 20 Bxd6 Qxd6
21 Qh6 Rh8 22 Re3 Qf4 23 Qxf4 Nxf4 24 hxg6+ hxg6 25 Rael Rae8 and
Black stands no worse. In fact a draw is a virtual certainty.
 Exercise (planning): Find a plan for White to gain an edge.

Answer: Step 1: Eliminate the defender of f5.

18 Bxd6! Qxd6 19 Nf5
Step 2: Plant the knight on f5 with tempo.
.
19.. Qd7
If 19 ... Qd8 then 20 Qg3 Qd7 21 Qh3! transposes.
20 Qh3!
Step 3: Transfer the queen to h3, threatening Nh6+ and Qxd7.
20 ... Kh8 21 h5
Step 4: Push on to h5, ensuring that White’s knight gets to remain on its
f5-perch.
21 ... Rxel+
Wang Hao logically swaps down, thus depleting White’s attacking
chances.
 22 Rxel Re8 23 Rxe8+ Nxe8

White stands a touch better for the following reasons:

1. The annoying f5-knight is difficult for Black to challenge (since ...
Nd6?? is unplayable at the moment).
2. White’s h-pawn can later push on to h6, loosening Black’s kingside and
creating potential attacking chances.
Having said that, I still believe it’s easier for Black to draw than it is for
White to win.
24 g4?!
This is a touch inaccurate as it allows Black’s king to move back to g8. If
White is going to make something of his micro edge, it would be by
centralizing his king and expanding in the centre with something like 24 Kfl
a6 (not 24 ... Nd6?? 25 Nxd6 and White wins a piece, since 25 ... Qxd6? is
back-ranked by 26 Qc8+) 25 b3 Qf7 26 Qg4 Qd7 (passing) 27 f3 Qf7 28 Ke2
Qd7 29 c4, when the annexation of territory has improved White’s position
and there are still chances to win.
24 ... a6
Black should still be okay after this. Nonetheless, 24 ... Kg8! 25 g5 Qf7! is
a clearer path to the draw; e.g. 26 Qg4 fxg5 27 Qxg5 Qf6 28 Qg4 Kf7 29 c3
Qe6 30 Qf4 Qf6 and White can’t make progress.
25 b3!
Now c2-c4 is coming.
25 ... Qe6 26 Ne3 Nd6 27 h6
The only chance is to create some kind of weakness around Black’s king.
27 ... g6 28 c4!

This shouldn’t achieve anything, but at least Nepo is trying.

28 ... dxc4 29 bxc4 Kg8
I prefer 29 ... Nf7!, which keeps White tied down to defence of h6. Black
stands no worse.
Note that 29 ... Nxc4?? fails miserably to 30 Nxc4 Qxc4 31 Qh2! and if 31
... Kg8 (or 31 ... Qg8? 32 Qd6 Qf7 33 Qd8+ and mates) 32 Qb8+ Kf7 33
Qxb7+ Ke6 34 Qxh7 and White wins as there is no perpetual check.
30 Qh2! Kf7 31 c5 Nb5 32 Qb8

The weakness of Black’s second rank grows more conspicuous, yet all is
well if Wang Hao plays an accurate move.

Exercise (critical decision): Black can choose between 32 ... Qe7, 32 ... Qd7,
or 32 ... Nxd4. Two of the these are fine, whereas one is a blunder.
Which would you play?

32 ... Qd7?
Evidence trumps the narrative Wang hoped for. This move loses by force
and the roulette wheel lands on Nepo’s number this tournament.
Answer: Either 32 ... Nxd4 or 32 ... Qe7 enables Black to survive. For example:
 a) 32 ... Qe7! 33 Qh8 Ke6 34 Qg8+ Qf7 35 Qc8+ Qd7, when White can’t
make progress and he might as well take the perpetual.
b) 32 ... Nxd4! should also draw after 33 Qxb7+ Qe7 34 Qxa6 Qe4! 35
Qa7+ Ke6 36 Qxh7 Nf3+ 37 Kfl Nd4! (threatening ... Qhl mate) 38 Qg8+
Ke7 39 Kel (or 39 Kgl Ne2+) 39 ... Qd3! (threatening mates on both e2 and
13, which forces White to scramble) 40 Nd5+! cxd5 41 Qg7+ Kd8 42 Qf8+
Kc7 43 Qe7+ Kc6 44 Qe8+! Kxc5 45 Qe7+ Kc4 46 Qc7+ Kb4 47 Qe7+ and
Black must allow the draw since 47 ... Kc3?? loses to 48 Qa3+ Kc2 49
Qxd3+ Kxd3 50 h7 etc.
33 Qh8
Everything looks okay for Black, yet it most certainly isn’t!
33 ... Ke6 34 f4!
Threatening f4-f5+, followed by Qxh7+.This is the move Wang Hao must
have missed. Black is forced into a dead lost knight ending.
34 ... Nxd4
Instead, 34 ... Qxd4?? 35 Qe8 mate makes it easy; while 34 ... f5 fails to 35
gxf5+ gxf5 36 Qe5+ Kf7 37 Nxf5 Kg6 38 Nh4+! Kf7 39 Qg7+ Ke8 (39 ...
Ke6 hangs the queen to 40 f5+) 40 Qg8+ Ke7 41 Qxh7+ and wins.
35 Qg8+! Qf7
If 35 ... Ke7 then 36 Qxh7+ Kd8 37 Qxd7+ Kxd7 38 h7 promotes.
36 Qc8+ Qd7 37 Qg8+ Qf7
 Exercise (combination alert): Find White’s killer move.

Answer: Step 1: Swinging the queen across to d8 is devastating for Black.

38 Qd8!
Threatening Qd6 mate.
.
38.. Qd7
Every attempted defence fails miserably: 38 ... Qe7 39 Qxd4 picks off
Black’s knight, and 38 ... Ne2+ 39 Kf2 doesn’t change anything.
39 f5+!
Step 2: Overloaded defender. White wins a piece.
39 ... gxf5 40 gxf5+ Nxf5 41 Qxd7+ Kxd7 42 Nxf5 Ke6
Black’s solitary king, despite its active position, has no chance against the
knight.
 43 Ne3 1-0

Black can’t do a thing: 43 ... Kf7 (both 43 ... f5 44 Kf2 f4 45 Ng4 and 43
... Ke5 44 Kf2 Kd4 45 Ng4 Kxc5 46 Nxf6 win easily) 44 Kg2 Kg6 45 Ng4 f5
46 Ne3! f4 47 Ng4 Kf5 48 Kf3 Kg5 49 a4 Kf5 50 Nf2 Kg6 51 Kxf4 Kxh6 52
Ne4 and the knight raids the black queenside.

Game 43
I.
Nepomniachtchi-Ding Liren
Candidates Tournament, Yekaterinburg 2020
Ruy Lopez

1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 0-0 Be7 6 d3 b5 7 Bb3 d6 8 a3

When White plays this instead of 8 a4 or 8 c3, Black nearly always
challenges with a future ... Be6, since White’s bishop exerts power along the
a2-g8-diagonal.
 8 ... 0-0 9 Nc3

There is no point in playing c2-c3 when White already expended a tempo

on a2-a3.
9 ... Na5
This way he can expand on the queenside with ... c7-c5.
Nepomniachtchi and Carlsen have squared off in this variation twice: 9 ...
Rb8 10 Nd5 Nxd5 11 Bxd5 Bd7 12 b4 Bf6 13 Be3 Ne7 14 Bb3 Ng6 15 a4
(or 15 c3 a5 16 bxa5 Ra8 17 a4 bxa4 18 Bxa4 Rxa5 19 Bxd7 Rxal 20 Qxal
Qxd7 with an even position, I.Nepomniachtchi-Ma.Carlsen, Carlsen
Invitational rapid 2021) 15 ... bxa4 16 Bxa4 Rxb4 17 Bxd7 Qxd7 18 Rxa6
Ra4 19 Rxa4 Qxa4 20 Qbl Qc6 was equal, I.Nepomniachtchi-Ma.Carlsen,
Carlsen Invitational rapid 2021. Both games were drawn.
10 Ba2 Be6
At some point Black needs to challenge White’s powerful light-squared
bishop. 10 ... c5 11 b4 Nc6 12 Nd5 Be6 is the alternative.
 11 b4
This move disrupts Black’s desire to play ... c7-c5, yet doesn’t gain time
since White’s rook will stand awkwardly on a2.
11 ... Bxa2 12 Rxa2 Nc6 13 Bg5 Qd7
Black also plays 13 ... Ng4 14 Bd2 Nf6, offering a repetition with 15 Bg5
Ng4, which was surprisingly accepted in M.Vachier Lagrave-Ma.Carlsen,
Stavanger 2018.
14 Bxf6!

Question: Why did White swap away his dark-squared bishop unprovoked?

Answer: For two reasons:

1. Nepo is playing for a win and needs some kind of imbalance to do so.
2. Black is stuck with a potentially bad bishop, since his central pawns are
fixed on the same colour squares.
14 ... Bxf6 15 Nd5 a5
 16 Rb2
This was a new move.

Question: Can’t White play 16 Nxf6+ first, inflicting damage to Black’s kingside?

Answer: Black wouldn’t mind that, for two reasons:

1. It opens the g-file which Black may try and exploit with ... Kh8 and ...
Rg8.
2. Black’s kingside pawns are not actually weak after the break... f6-f5.
For example: 16 ... gxf6 17 Rb2 axb4 18 axb4 f5 19 Nh4 f4 20 Nf5
(threatening mate, which is trivially easy to defend against) 20 ... Kh8 21 d4
Rg8 22 Rb3, E.Bacrot-H.Gretarsson, PRO League (rapid) 2018, and I doubt
that White has much after 22 ... Ne7; or if 22 dxe5 then 22 ... Qe6! 23 Rb3
dxe5.
16 ... axb4 17 axb4 Bd8 18 c4 Nd4 19 Nxd4 exd4 20 Qc2 Re8 21 g3
bxc4
The engine is confident Black remains even after 21 ... c6! 22 Nf4 g5!? (a
difficult move for a human to play) 23 Ne2 Bf6 24 f4 g4 25 cxb5 cxb5 26
Qd2 Ra3, when Black generates some counterplay to compensate his
structural deficit.
22 Qxc4 c6 23 Nf4 Bg5 24 Ne2 d5 25 exd5 cxd5 26 Qb3
Obviously 26 Qxd4?? hangs a rook to 26 ... Bf6.
26 ... h5!
 Ding’s position is structurally beaten down, yet remains unbroken.

Question: Isn’t Black in deep trouble, seeing as White is essentially a pawn up?

Answer: I thought exactly that when I watched this game live, believing Black’s attack to be an
imposter. A few moves later the realization dawned that I greatly underestimated Ding’s level of
counterplay against the white king. There is no question of White’s structural superiority, yet the engine
calls it close to even, since White’s king is under-protected and vulnerable to attack, as we see in the
coming moves.
27 b5
Principle: Passed pawns should be pushed. Black’s attack may become
constrained if White’s pawn moves too far up the board.
27 ... h4 28 b6 h3?’
This tempting move may be the spot where Ding’s position goes from
dynamically balanced to slightly inferior. It isn’t easy to piece together a plan
from the position’s scant clues, but 28 ... Rab8 29 Rai g6! is probably
Black’s optimal course, intending ... Kg7! and ... Rh8, when his attack gains
momentum. For example: 30 Qdl Kg7 31 Kg2 Bd8 32 Rabi h3+! (now is the
correct timing) 33 Kgl Qf5 34 b7 (or 34 Nxd4 Qf6 35 Ne2 g5 36 b7 Qf3 37
Qfl Qxd3 and Black is okay) 34 ... Bf6 35 Rb4 g5 36 Nxd4 Bxd4 37 Rxd4
Rxb7! (playing upon White’s weak back rank) 38 Rxb7 Qe5 39 Qfl Qxd4 40
Qxh3 Rel+ 41 Kg2 Qxd3 42 Qd7 Qfl+ 43 Kf3 Qe2+ 44 Kg2 Qfl+ with
perpetual check.
29 Khl! Reb8 30 Rfbl Bd8 31 Qb5?!
Nepo should have applied the principle: Passed pawns should be pushed.
That is 31 b7! Ra7 32 Nf4!, and if 32 ... Raxb7 33 Qxb7 Rxb7 34 Rxb7 Qf5
35 Rb8 Kh7 36 Rxd8 g5 37 Nxh3 Qxh3 38 Rxd5, White’s two rooks and
extra pawn will overcome the queen, since his king is now relatively safe
from checks.
.
31.. Qg4?
A more serious mistake. After 31 ... Qf5! 32 Nxd4 Qg4! 33 b7 Ra7 34
Qe8+ Kh7 35 Qxf7 Bf6! 36 Rb4 Bxd4 37 Qxd5 Ra5! 38 Rxd4 Rxb7! (weak
back rank) 39 Qxb7 Qxd4 40 Qe4+ Qxe4+ 41 dxe4 Ra2! 42 f3 Re2! White is
unable to make progress due to his perpetually weak bank rank. Try playing
this position out. 43 Kgl Rg2+ 44 Khl Re2 merely repeats.
32 Qxd5
The natural capture, occupying the vulnerable long diagonal. The engines
see no danger and prefer 32 Qe8+! Kf7 33 Qxf7, assessing White as
completely winning; e.g. 33 ... Rxb6 34 Rxb6 Bxb6 35 f3 and Black has
nothing.
32 ... Ra5 33 Qc6?
Nepo’s path to the win at this point lay in 33 f3! Rxd5 34 fxg4 Rb7 35
Ngl! Rd6 36 Nxh3 Rdxb6 37 Rxb6 Rxb6 38 Rxb6 Bxb6 and Black won’t
hold the ending for the following reasons:
1. He is a pawn down.
2. He is saddled with the clearly inferior minor piece.
3. White’s king will rush to e4 and then gang up with his knight to pick off
the d4-pawn.
 Exercise (combination alert): Nepo’s last move was a mistake, allowing Black
a route to equality. What did he overlook?

Answer: Weak back rank.

33 ... Rc5?
This hellish position is so complex and confusing that two of the world’s
most brilliant players are unable to find the correct path. Both Nepo and Ding
missed the incredible defensive combination 33 ... Rxb6H (seek and ye shall
find! - after this tactical sleight-of-hand, it is White, not Black, who must
fight for the draw) 34 Rxb6 Qxe2! (zwischenzug) 35 Rb8 Re5! (is Black’s
salvation real or is it a bewildering dream?) 36 Rxd8+ Kh7 (weirdly, despite
being a full rook up, Nepo is unable to overcome the issue of his weak back
rank) 37 Rh8+ Kxh8 38 Qc8+ Kh7 39 Qxh3+ Kg6 40 Qfl Qf3+ 41 Kgl Re2
(intending ... Rd2 to win the d3-pawn) 42 Rdl Rxf2! (discovered
attack/overloaded defender) 43 Qxf2 Qxdl+ 44 Qfl Qxfl+ 45 Kxfl Kf5 46
h3 and White is lucky to be able to hold a draw.
34 Qe8+ Kh7 35 Ngl!
Suddenly, White’s king is safe and Black’s once-promising attack turns to
dust.
.
35.. Rxb6
This desperate piece sacrifice no longer works. But if 35 ... Qf5 36 b7
Qxd3 37 Qxf7 Qe4+ 38 f3 Qd5 39 Qxd5 Rxd5 40 Nxh3, Black is just two
pawns down and losing.
36 Qxd8 Rxb2 37 Rxb2 Rcl

Threatening ... Qf3 mate or ... Rxgl+ and ... Qdl mate. However, there is a
simple solution.
38 Qh4+!
Simplification. If not for this resource, White would be mated.
38 ... Qxh4 39 gxh4 Rdl
Nobody is happy to see their hard earned tax dollars going to support a
useless and expensive government bureaucracy. Black has spent too much on
his attack, which is now officially dead.
40 B! 1-0
White consolidates in the following way: 40 ... Kg6 41 Rb6+ Kh5 (or 41 ...
Kf5 42 Rb7 Kf6 43 Rd7 Rxd3 44 Nxh3 and wins since if 44 ... Rxf3 then 45
Rxf7+! Kxf7 46 Ng5+ forks) 42 Rb5+ Kxh4 43 Rf5 f6 44 Rf4+ Kg5 45 Rxd4
and wins.
One trait both Nepo and Magnus have in common is that, occasionally,
both go clinically insane, give away all their material, and then fail to mate.
The only reason you won’t see much of that from Nepo’s side is that we
aren’t showing his losses in this book!

Game 44
Ma.Carlsen-I.Nepomniachtchi
Carlsen Invitational (rapid) 2020
Sicilian Defence

1 e4 c5 2 NB d6 3 d4 cxd4 4 Nxd4 Nf6 5 Bc4’?

 In psychiatric terms this move would fall under the category of shock
therapy. Magnus attempts to lure his opponent out of mainstream theory as
early as possible.
. Nxe4!
5..
This pawn grab is a reliable equalizer. After 5 ... e6 6 0-0? Be7?! (6 ...
Nxe4! is even better now; see the next note) 7 Bd3 Nc6 8 Nxc6 bxc6 9 c4 0-0
10 Nc3 e5 (10 ... d5 may be more accurate) 11 Be3 Be6 12 Qe2 Nd7 13 Rfdl
Qc7?!, Ma.Carlsen-M.Vachier Lagrave, Carlsen Invitational (rapid) 2020,
Magnus could have gained a strategic advantage with the trick 14 Nd5!, when
Black must lose the bishop pair, no matter how he responds.
6 Qh5
It’s not quite Scholar’s Mate, but it’s a close knock-off.
Note that 6 Bxf7+ isn’t as brilliant as it first appears: 6 ... Kxf7 7 Qh5+ g6
8 Qd5+ Kg7 9 Qxe4 e5 10 Nf3 Nc6 11 0-0 d5 and I already prefer Black,
who owns the centre and bishop pair, J.Cacko-P.Sekowski, correspondence
2015.
The Morphian sacrifice-for-development 6 0-0? isn’t very inspiring either,
scoring a depressing 9.1% for White, who gets little to no compensation after
. e6 7Rel d5.
6..
. e6
6..

Question: Why invite a sacrifice on e6 when Black can just play 6 ... g6 and gain
a tempo?

Answer: You just fell into a trap! 7 Qd5! e6 8 Qxe4 wins a piece.
7 Nxe6?
Magnus would have been wise to assume the attitude of a penitent and
settle for 7 Bb5+ Nd7 8 Nxe6 Nef6 9 Nxd8 Nxh5 10 Bxd7+! (this way
White’s d8-knight isn’t trapped) 10 ... Bxd7 11 Nxb7 Bc6 12 Na5 Bxg2 13
Rgl Be4 14 c4, though Black stands at least even due to the bishop pair,
J.Jordan-We.Müller, correspondence 1995.
 Okey dokey, suicide it is. Like Icarus, who flew too close to the sun and
then came plunging back to Earth with a splat, Magnus succumbs to a case of
temporary insanity with an unsound piece sacrifice. It may be true, as with
Emanuel Lasker, that it’s no simple matter to respond to Magnus’ unsound
ideas, who counts on colossal natural ability to bail him out in the
complications. On the other hand, a piece is still a piece.

Exercise (critical decision): Black can play 7 ... Bxe6 immediately or throw in
7 ... Qe7. One line is level, while the other wins for Black. Which would
you play?

Answer: Chopping the knight at once gives Black a winning position.

7.. . Bxe6!
This is pretty much a refutation of White’s sacrifice. After 7 ... Qe7?, as in
B.Sawyer-M.McCue, Dallas 2002, White had 8 Qb5+! Nc6 9 Bd5! Bd7 10
Bxe4 Qxe6 11 Qd5, when he escapes to an equal position.
8 Bxe6 Qe7!
This move is the problem. White’s bishop is unable to move off the e-file
without allowing Black a lethal discovered check.
9 Bxf7+
Now if 9 Qb5+ Nc6, Black keeps the piece since 10 Bd5?? Nc3+ picks off
White’s queen.
9.. . Qxf7 10 Qe2 Qe7 11 0-0

Question: Doesn’t White still have a considerable lead in development?

Answer: Unfortunately for Magnus, it isn’t considerable enough! Not enough to justify having
given up a piece.
11 ... Nf6 12 Be3
Or 12 Qdl Qf7 13 Rel+ Be7 14 Qxd6 Nc6 15 Nc3 0-0 and White’s two
extra pawns aren’t enough for the missing knight.
12 ... Nc6 13 Nc3 d5!
Threatening ... d5-d4.
14 Qf3 0-0-0
Or 14 ... d4 anyway, since 15 Rfel 0-0-0 16 Bg5 Qf7 17 Ne4 Be7 is
winning for Black.
15 Rfel
Threatening discovered attacks on b6 and a7, which are easily avoided.
 .15.. Qd7 16 Nb5 a6 17 a4’?
If 17 Nd4 then 17 ... Bd6 consolidates.

.17.. Qg4!
White’s chances fall close to zero with queens off the board. The engines
may want to grab the knight: 17 ... axb5!? 18 axb5 Nb8 19 Bb6 Bd6 20 Re3
Ne4 and Black is winning; but no human would pick this line over the one
Nepo played in the game.
18 Qxg4+ Nxg4 19 Bb6
19 Na7+ Kc7 is completely hopeless for White.
19 ... axb5 20 axb5 Nb8 21 Bxd8 Kxd8
Black’s three minor pieces will soon overwhelm White’s extra rook and
two pawns.
22 h3 Nf6 23 Ra7 Kc7 24 Re6 Bc5 25 b6+ Bxb6 26 Re7+ Kd6
Again Nepo opts for the simplest, most forcing solution ahead of anything
like 26 ... Nbd7 27 Ra3 g6.
 27 Raxb7 Nbd7 28 Rxg7

Exercise (combination alert): Calling this a “combination” is a bit of an

exaggeration. How did Nepo force resignation in a single move?

Answer: Trapped piece.

.28.. Kc6! 0-1
White loses another exchange.

Game 45
I.
Nepomniachtchi-Ding Liren
Legends of Chess (rapid) 2020
Scotch Game

1 e4 e5 2 Nf3 Nc6 3 d4 exd4 4 Nxd4 Nf6 5 Nxc6 bxc6 6 e5 Qe7 7 Qe2 Nd5
8 c4 Ba6 9 h4’?
 Apparently, chess principles are only for the ordinary and not for the elite.
Even great players can love a shady line and, for their forbidden love, later
suffer cruelly.

Question: It’s difficult to believe the solution to a lag in development is to

waste more time! Isn’t this move obviously an error?

Answer: I suppose the mind of imagination considers orthodoxy a vice which should be suppressed.
On the other hand, such a blatant violation of principle probably lurches past risky into the realm of
dubious. The ideas behind the move:
1. To meet a future ... g7-g6 with h4-h5.
2. Pushing the h-pawn allows the hl-rook to enter the game via h3
3. Black nearly always challenges the e5-pawn with a future ... d7-d6 or ...
f7-f6. With a rook on the third rank, White can respond with Re3 (assuming
we chase Black’s knight from d5 first).
4. It prevents Black from playing the ... g7-g5 option.
 The trouble is that these don’t make up for White’s massive development
deficit.
9.. . f6
Principle: Create confrontation and open the position when leading in
development. Or perhaps castle first. 9 ... 0-0-0! is a superior response,
backed up by Fat Fritz's approval, after which White scores an alarmingly
low 35%. For example, after 10 Rh3 Re8 11 f4, P.Dachtera-Z.Sanner,
correspondence 2003, and now 11 ... f6, White’s position already looks
difficult.
10 Rh3
Nepo seems intent on bringing out major pieces before minor. Believe it or
not, this is a book move, the idea of Dutch GM John Van der Wiel back in
1987. The rook can later swing across to either a3 or e3, once Black’s knight
is induced away from d5.
10.. . fxe5
10 ... Nb6 is a safer option; e.g. 11 exf6 Qxe2+ 12 Bxe2 Bxc4 13 fxg7 (or
13 Re3+ Kd8) 13 ... Bxg7 14 Bh5+ Kd8 15 Nc3 Kc8, P.Dukaczewski-
I.Seitaj, Moscow Olympiad 1994, which the engine calls even, whereas I
prefer Black due to the central pawn majority.
11 Bg5 Nf6 12 Re3 d6?’
This natural move doesn’t work out well. Instead, 12 ... 0-0-0! 13 Ra3 (not
13 Rxe5? Qb4+ 14 Qd2 Bxc4) 13 ... Bb7 14 Rxa7 Qe6 and if 15 Nc3 Bc5 16
Ra5 Bb6 17 Rxe5?!, R.Vaishali-Lei Tingjie, Women's Speed Chess (blitz)
2021, then 17 ... Bxf2+! 18 Kxf2 Ng4+ 19 Kgl Nxe5 20 Bxd8 Rxd8 is better
for Black.
13 Qf3!
 Unpinning the c-pawn while gaining a tempo on c6.
13 ... e4
Black has also tried:
a) 13 ... Bb7 14 Rb3 0-0-0, D.Szczesny-R.Polaczek, correspondence 1988,
and now 15 c5! ? e4 16 Qc3 dxc5 17 Nd2 looks good for White, since only an
abstractionist artist would approve of Black’s tripled pawns.
b) 13 ... Qd7 14 Bd3! (threatening Bf5 at once; not 14 Nc3? Be7 15 Bd3 0-
0 16 Bf5 Qe8 and Black was better in I.Nepomniachtchi-L.Aronian,
Chess.com blitz 2020) 14 ... Kf7 15 Nc3 Be7 16 Bxf6 gxf6 (16 ... Bxf6 17
g4! is also awful for Black) 17 Qh5+ Kf8 18 Bf5 Qe8 19 Qh6+ Kf7 20 Ne4
and White has a decisive attack.
14 Qdl d5?!
Already Ding has a single solution just to stay in the game, and that is to
play ... Bb7, ... 0-0-0 and ... d6-d5 in some order or other. The correct one
was 14 ... Bb7 15 Nc3 d5 16 cxd5 0-0-0 and even then Black is clearly worse.
 15 Nc3’?
Simply 15 cxd5 Bxfl 16 Kxfl cxd5 17 Bxf6 Qxf6 18 Qxd5 would win a
pawn. It seems Nepo is now after more than that.
15 ... Bb7
Or 15 ... 0-0-0 16 cxd5 Bb7, and not 16 ... Bxfl? 17 Kxfl cxd5 18 Nb5! a6
19 Nd4, which is heavily in White’s favour.
16 cxd5 cxd5?
Ding is no longer ahead in development and therefore shouldn’t be in a
rush to create central confrontation. 16 ... 0-0-0 was obligatory, as indicated
above.
17 Bxf6!
Principle: A lead in development supersedes the bishop pair.
17 ... Qxf6 18 Nxd5 Bxd5

Exercise (combination alert): For the moment White’s plan is an abstraction.

Now comes time for the specifics. What is White’s strongest continuation?
 Answer: Step 1: Zwischenzug.
19 Bb5+!
White can disrupt by first giving check on b5 and only then regaining the
piece.
19 ... Bc6
19 ... c6? runs into 20 Qxd5!.
20 Rxe4+!
Step 2: Pinned piece.
20 ... Be7
Forced. 20 ... Kf7?? 21 Qb3+ Kg6 22 Qg3+ Kh6 23 Qe3+ Kg6 leads to
mate in ten after 24 h5+! Kxh5 25 Be2+ Kg6 26 Rg4+ Kf5 27 Rf4+ Kg6 28
Bd3+ Kf7 29 Bc4+ Kg6 30 Rg4+ etc.
21 Qd5!
Nepo continues to play upon pin tricks.
21 ... Rd8
Since 21 ... Bxb5? 22 Qxa8+ Kf7 23 Qd5+ picks up a rook as well as the
bishop.
22 Bxc6+ Kf8 23 Qc4
 Threatening Rf4. Black is busted due to the principle: Opposite-coloured
bishops favour the attacker, which in this case is clearly White, whose grip
on the light squares is the significant factor.
23 ... Bd6
Preventing Rf4. If 23 ... Qxb2 24 Rcl g6 then 25 Kfl! Rd2 26 Rf4+ Bf6
27 g4! leaves Black busted.
24Rdl!
Threatening Rd3 and Rf3. Castling is pretty strong too.
24 ... g6
Ding hurries to give his king an escape hatch on g7.
25 Rd3 Kg7 26 Rf3 Qxb2 27 Rf7+ Kh6
Going to g8 allows mate in two. At this stage Black’s king is the escaped
prisoner in the swamp, attempting to evade police and tracking dogs by
breathing through a reed.
28 g4?
 When you achieve a winning position, don’t expect smooth conversion
without incident, since your opponent is desperate to ensure that it doesn’t
happen. This is a serious mistake which gives Black the chance to draw.
The correct and necessary move is 28 Kfl!, when all back rank checks can
be met by Rel, while after 28 ... Rhf8 29 g4! Black is soon mated.

Exercise (critical decision): Black must somehow get to White’s king.

He can accomplish this with a queen check, followed by the clearance shot
... Bb4!. The question remains: Should Black’s queen check on al or bl?
One version holds the draw while the other loses. Take your pick.

Answer: Only the check on al holds the game.

.28.. Qbl+?
Ding missed 28 ... Qal+! 29 Ke2 Bb4!! (clearance) 30 g5+ Kh5 31 Bd5
Qel+ 32 Kf3 Qhl+ with perpetual check. Note that White can’t take the
bishop: 31 Qxb4?? loses to 31 ... Qdl+ 32 Ke3 Rd3+ 33 Kf4 Qf3+ 34 Ke5
Qxf7 (threatening ... Qf5 mate) 35 Rf4 Re8+! 36 Bxe8 Qxe8+ 37 Kf6 Rd7!
and, amazingly, White doesn’t have a single check on Black’s hopelessly
placed king, whereas White’s own king is mated in two moves.
29 Ke2 Bb4
Threatening 30 ... Qdl+ 31 Ke3 Bd2 mate, or if 30 Qxb4?? then 30 ...
Qd3+ 31 Kel Qdl mate. Unfortunately, the idea no longer works since White
seizes control over dl.
30 Rd4!

This couldn’t be played with Black’s queen on al. The rook now blocks
his intended ... Qdl+, while vacating e4 enables g4-g5+ and Bf3+ mate.
.
30.. Rhe8+
If 30 ... Qb2+ 31 Kfl Rxd4 then 32 g5+ Kh5 33 Rf4!! Rxf4 (or 33 ... Rd3
34 Qe4!, intending Qf3+ and mates) 34 Qxf4 Qbl+ 35 Kg2 leaves Black
helpless to prevent the white bishop’s transfer to f3; e.g. 35 ... Qdl 36 Bf3+
Qxf3+ 37 Qxf3+ Kxh4 38 Qg3+ Kh5 39 f3! and 40 Qg4 mate.
 31 KB!
White’s bishop is worth a lot more than Black’s rook, and 31 Bxe8??
would allow 31 ... Rxe8+ 32 Kf3 Qhl+ 33 Kg3 Qgl+ with perpetual check.
31 ... Qhl+ 32 Kg3 Qgl+ 33 Kh3!
Threatening g4-g5+ and either Rxh7 mate or Bf3+ and mates next move.
Black has run out of useful checks and must hand over a rook just to keep
checking.
33 ... Re3+
We don’t really worry about financial insolvency when the alternative is
death. It’s either this or get mated even more quickly; e.g. 33 ... g5 34 hxg5+
Kg6 (or 34 ... Kxg5 35 Rd5+ Kg6 36 Rg5+! Kh6 37 Rf6+! Kxg5 38 Qf4
mate) 35 Qd3+! Kxf7 36 Qxh7+ Ke6 37 Qg6+ Ke7 38 Qf6 mate.
34 fxe3 Qxe3+ 35 BB! 1-0

White’s last move not only blocks, it also renews the threat of g4-g5+.
Black resigned in view of 35 ... Be7 (or 35 ... g5 36 Rf6+ Kg7 37 Qf7+ Kh8
38 Rxd8+ etc) 36 g5+ Bxg5 37 hxg5+ Qxg5 38 Rh4+ Qh5 39 Qf4+ and
mates.

Game 46
A.Giri-I.Nepomniachtchi
Legends of Chess (rapid) 2020
Grünfeld Defence

1 d4 Nf6 2 c4 g6 3 Nc3 d5 4 cxd5 Nxd5 5 Bd2

A successful fighter must discover the opposition’s secret weakness in

order to destroy him. Giri most certainly knew that Nepo didn’t have a stellar
score as Black against this line.

Question: Isn’t this move rather passive?

Answer: It’s actually quite an aggressive line for White. The idea is that if Black plays ... Nxc3,
White recaptures with the bishop and, at some point, plays d4-d5 to swap away Black’s powerful dark-
squared bishop and go after his king.
5 ... Bg7 6 e4 Nxc3
6 ... Nb6 is played slightly more often. The idea is to gain a tempo again
the d4-pawn.
7 Bxc3 0-0
Nepo lost a painful miniature after 7 ... c5 8 d5 Bxc3+!? 9 bxc3 0-0 10
Qd2 Qd6 11 f4 e6 12 Nf3! exd5 13 Bc4! Be6? (13 ... d4 looks okay for
Black) 14 0-0 d4? (one move too late; he had to try 14 ... Nd7 15 f5 Ne5 16
Bb3 c4 17 Nxe5 Qxe5 18 Bc2 gxf5 19 exf5 Bd7 20 f6, still with a dangerous
attack brewing for White) 15 f5!! (where did that come from?) 15 ... Bxc4 16
e5! Qd7 17 f6 1-0 V.Anand-I.Nepomniachtchi, Online Nations Cup 2020.
Black is mated in four moves at the most after 17 ... Kh8 18 Qh6 Rg8 19 Ng5
Qh3 (it isn’t a good sign when the engine gives a move as ridiculous as this
as Black’s best option) 20 gxh3 and mate next move.
8 Qd2
More aggressive than 8 Nf3, since White has the option of castling long
and turning the game into opposing wing attacks.
8 ... Nc6 9 Nf3 Bg4
Increasing the pressure on d4.
10 d5 Bxf3 11 Bxg7 Kxg7 12 gxf3 Ne5
 Question: Does Black already stand a shade better, in view of White’s
potential bad bishop?

Answer: First of all it’s not yet a bad bishop since the central pawns remain unfixed. Secondly,
Black’s king looks more exposed than White’s, whose position I slightly prefer.
13 0-0-0 c6
13 ... Nxf3?? is a monumentally dumb decision which hangs a piece to 14
Qc3+.
14 Qc3
Later in the year Magnus crushed Nepo with 14 Bh3 Kg8 15 Qc3 Qb8??
(15 ... Qc7 is necessary) 16 d6! exd6 17 f4 (oops, Black’s knight is trapped)
17 ... Re8 18 Rhel a5 19 Kbl a4 20 fxe5 Rxe5 21 f4 Rc5 22 Qf6 a3 23 b3
Rh5 24 Rxd6! Qe8 (or 24 ... Rxh3 25 Redi and wins, since there is no
remedy to the coming Rd8+) 25 Be6! (interference; not yet 25 Redi? Qxe4+
26 Kai Rd5! and White’s king is suddenly a target) 25 ... fxe6 26 Redi 1-0
Ma.Carlsen-I.Nepomniachtchi, Skilling Open KO (rapid) 2020.
14 ... f6 15 f4 Nf7 16 d6 exd6 17 h4 Qb6’?
A novelty, though it’s not obviously better than just playing 17 ... d5 at
once; e.g. 18 h5 Qe7 19 exd5 Qe4 20 Rh4 cxd5 21 Bd3 Qf3 22 Qd2 Nd6 23
hxg6 hxg6 24 f5, as in T.Biedermann-L.Chesakov, correspondence 2018,
when chances look balanced after 24 ... g5 25 Rgl Nf7.
18Qg3
Protecting the f2-pawn and increasing the pressure on the kingside.
18 ... d5 19 exd5

.19.. Rad8
In the third and final game of their mini-match game Nepo switched to 19
... cxd5 20 h5 Rac8+ 21 Kbl Rfe8 22 hxg6 hxg6 23 Bd3 f5 24 Rdgl Qf6 25
Qf3 Rh8 26 Rxh8 Rxh8 27 Qxd5 Rd8?! 28 Qxf5 Qxf5 29 Bxf5 Rd6 and was
lucky to have held a draw, A.Giri-I.Nepomniachtchi, Legends of Chess
(rapid) 2020.
 20 h5 Qc5+ 21 Kbl Rxd5 22 Rcl! Qd4 23 Bc4 Rc5?
He had to try 23 ... Rdd8 24 Rhgl Qe4+ 25 Kai Rd2 26 hxg6 hxg6 27 Be6
Nd6 28 f3 Qd3 29 f5 g5 30 Qh4! Rf7 31 Bxf7 Nxf7 and hope to survive.
24 hxg6 hxg6 25 Rhdl
This is highly favourable. Even stronger is 25 Rcgl! g5 26 Bxf7 Qe4+ 27
Kai Kxf7 28 Qb3+ Rd5 29 Qxb7+ and Black is completely busted.
25 ... Rxc4
No choice. If 25 ... Qe4+ 26 Bd3, it’s time for Black to resign.
26 Rxd4 Rxd4 27 Rgl Nh8

It’s pretty clear that the opening and middlegame have been a disaster for
Nepo, His rook, knight and pawn for the queen aren’t materially so bad. The
real issues are:
1. It won’t be so easy for Black to hang on to his queenside pawns.
2. Black’s king isn’t terribly safe since White can pry it open with f4-f5
and f2-f4.
 3. His frozen knight on h8 is one only Nimzowitsch could love.
28 f5 g5 29 f4 Nf7 30 fxg5 fxg5 31 Qc7!
Pawns are about to fall and at this point earning a draw would be a miracle
for Black. Then how much more so actually to win from this wretched
position?
31 ... g4
There is no choice except to allow his queenside pawns to fall, since 31 ...
Rb4?? loses on the spot to 32 Qe7!, intending 33 Rxg5+ Kh7 34 Rgl! with
mate to follow.
32 Qxb7 Rfd8 33 Qxc6
Threatening Qg6+ and f5-f6.
33 ... R8d6 34 Qf3?’
From this point on the evaluation keeps dropping for Giri. Better was 34
Qc3! Ne5 35 a3 Kf6 36 Ka2 Kxf5 37 Qc8+ Kg5 38 Qc5 Kf4 39 Qxa7 and
White should win comfortably.
34 ... Ne5 35 Qe3?’
After 35 Qhl! Kf6 36 a3 Kxf5 37 Qh5+ Kf4 38 Qh2+ Kf5 39 Ka2 Black
has no way to advance his g-pawn.
35 ... Kf6
 36 Qh6+?’
The great luxury of the rich is that they can waste without fear of negative
consequence. It is with this feeling that Giri is happy to offer his f-pawn,
probably thinking that he will later be able to exploit the open lines to reach
Black’s king. 36 Rg2 is more accurate.
36 ... Kxf5 37 Qf8+ Kg5 38 Qe7+ Kf4 39 Qf8+ Kg5 40 Qg8+?’
Here 40 a3 keeps winning chances alive.
40 ... Kf4 41 Qb3 Rd3 42 Qb4+ R3d4 43 Qb3 Nf3
At this point the evaluation drops to even. The problem is that Giri
continues to overestimate his position, perhaps believing that he can deliver
perpetual check any time he pleases.
44 Rfl
After 44 Q17+ Ke3 45 Rcl g3 46 Rc3+ Rd3 47 Qxa7+ Ke2 48 Qe7+ K12
49 Rc2+ Nd2+ 50 Kai g2 51 Qh4+ Ke2 52 Qg4+ K12 53 Qh4+ White has to
be satisfied with perpetual check.
 44 ... Kg3 45 Qe3 Rd3 46 Qf2+ Kf4 47 Qxa7 Rdl+!
Now Black’s king is perfectly safe, so perpetual is the best White can hope
for.
48 Rxdl Rxdl+ 49 Kc2 Rd2+ 50 Kc3 g3

Power usually usurps wealth on the chessboard. White’s extra material is

clearly temporary since the passed g-pawn has taken on dangerous
dimensions. It’s clear that only Black can win now, while White is the one
who should be looking to draw.

Exercise (critical decision): White can give a queen check on b8, c7 or f7.
Which one would you play?

Answer: Checks on b8 and c7 draw; the one on f7 loses.

51 Qf7+??
Principle: Ifyou screw up your previously winning position, don 7 dwell on
that fact, live in the moment’s position, not one which only existed in the past.
Giri just can’t bring himself to take the draw, since he keeps remembering the
time when he had a completely won game. But by failing to act, there is not
even a draw for White.
Giri should have taken that with either 51 Qc7+ Kg4 52 Qc8+ Kf4 53
Qc7+ etc, or 51 Qb8+ Kg4 52 Qg8+ Kh3 53 Qe6+ Kh2 54 Qh6+ Kg2 55 Qe3
Rf2 56 a4 Kfl 57 a5 g2 58 Qd3+ Kgl 59 a6 Khl 60 Qh7+ Nh2 61 a7 Rf3+
62 Kc4 glQ 63 a8Q Qcl+ 64 Kd5 Qxb2 and the game should be drawn with
correct play.
51 ... Ke3 52 Qe7+ Kf2 53 Qc5+ Rd4!
White is weirdly helpless against the advance of the g-pawn and,
frustratingly, is unable to deliver perpetual check. Giri may have expected 53
... Kg2? 54 Qe3 Rf2, transposing to the 51 Qb8+ drawing line above.
54 Qf5 g2 55 Qc2+ Rd2 56 Qg6 glQ 57 Qxgl+ Kxgl 58 a4
Let’s just say this promotion attempt is a wee bit optimistic.
58 ... Rd8 59 a5 Ne5 60 b4 Nc6!
This move freezes the pawns.
61 a6 Rb8 0-1
After 62 Kc4 Na7 White’s pawns aren’t going anywhere.

Game 4 7
V.Ivanchuk-I.Nepomniachtchi
Legends of Chess (rapid) 2020
Grünfeld Defence

1 d4 Nf6 2 c4 g6 3 Nc3 d5 4 Bf4 Bg7 5 e3 0-0 6 Qb3 c5!

Nepo is happy to give a pawn away for a lead in development.
7 cxd5 cxd4 8 exd4 a5!
 Principle: Create confrontation when leading in development. The idea is
two-fold:
1. Black wants to disrupt with a future ... a5-a4.
2. If White responds with a2-a4, Black can play ... Na6! and ... Nb4.
I like this move better than the more traditional treatment 8 ... Nbd7 9 g3
Nb6 10 Bg2 Bf5 HNge2.
9 Nf3
After 9 a3 a4 10 Qc4 (or 10 Nxa4?! Nxd5) 10 ... b5! 11 Qxb5 Ba6 12 Qc5
Bb7 13 Qb5 Bxd5 14 Nxd5 Nxd5 15 Bd2 Bxd4 Black has regained the
sacrificed pawns while remaining ahead in development; and 9 Bb5 Na6 10
Nge2 Nb4 11 Bc4 b5! 12 Bxb5 Nfxd5 offers Black loads of compensation for
only one pawn.
9 ... a4! 10 Qc4
A new move. After 10 Nxa4 Nxd5 11 Bd2 Nc6 Black’s development
counts for more than White’s extra pawn, A.Mikhalchishin-V.Bagirov,
Frunze 1979.
10 ... Na6
Covering c7. The quick route 10 ... b5!? is also completely sound: 11
Qxb5 Ba6 12 Qb4 Bxfl 13 Kxfl Nxd5 14 Qb7 Nc6! 15 Nxd5 Qxd5 16 Rcl
Rfc8 with obvious compensation for Black.
11 Be2 Bd7 12 a3
If 12 0-0 then 12 ... Rc8 13 Qd3 Nb4 14 Qd2 Nbxd5 and Black stands
better.
12 ... b5!

It’s only money. Nepo is willing to hand over a second pawn to open
queenside lines and gain more time.

Question: Isn’t Black going overboard, giving too much away for an
initiative which may or may not lead to something?

Answer: There is a wide range in remedies. If you are having trouble sleeping, there is a big
difference between choosing a cup of camomile tea and popping three tablets of Ambien. Black’s last
move was Nepo’s three tablets! The engine says the decision is not only sound, it’s even its first choice.
Which means that Nepo’s move is probably less radical than it actually appears.
13 Nxb5 Rc8!
Have another one!
14 Qd3?
From this point on Ivanchuk gets no rest from Nepo’s unrelenting
initiative. He should have accepted: 14 Qxa4! Nxd5 15 Bd2 Ndc7 16 Qb3
Nxb5 17 Bxb5 Rb8 18 a4 Nc7 19 Ba5 Bxb5 20 axb5 Qd6 21 Bxc7 Qxc7 22
0-0 Qb6 23 Qe3 e6 24 Rfcl Qxb5 25 Rc2 with dynamic equality.
14 ... Nxd5 15 Bg3

Exercise (combination alert): Black is a pawn down. How can he make use of
his temporary lead in development?

Answer: Bang down either knight on b4.

 15 ... Bf5?!
Black still gets a dangerous initiative with this move, yet it’s not a win.
Instead, after 15 ... Nab4! 16 axb4 Nxb4 17 Qbl Nc2+ 18 Kfl Bxb5 19 Bxb5
Qb6 20 Bd3 Nxal 21 Qxal Rfd8, White is too far behind in development to
survive.
16 Qdl Qa5+ 17 Qd2
17 Kfl? Rc2 is even more unpleasant for White.
.
17.. Qxd2+!

Question: Isn’t it counter-intuitive to swap queens when attempting to

exploit a lead in development?

Answer: Not if doing so increases that lead, as in this case. Black’s initiative is sustained even with
queens off the board.
18 Nxd2 Rc2 19 Nc3

.
19.. Nac7’?
 What is Nepo willing to risk? A great deal.

Question: Why did Nepo not only decline to win two pawns but also give
another one away?

Answer: I would have gone for the safe way, but maybe that is why Nepomniachtchi is about to
play for the world championship while I only get to write about it. In the other version Nepo would
have been a pawn up with all the pawns on the same side of the board, enhancing White’s drawing
chances. Amazingly, the engine says the lines are equally strong.
The alternative 19 ... Nxc3 20 bxc3 Nb8 21 Bdl Rxc3 22 Bxa4 Bxd4 23 0-
0 Rd3 24 Ra2 Be6 25 Rc2 Rxa3 leaves Black a pawn up, with the bishop
pair. Even it may not be so easy to win, I enjoy such positions as the pawn-up
side, who plays for two results: a win, or a draw.
20 Nxa4?’
It isn’t easy to process the jungle of data. This unremarkable-looking move
exerts profound influence on events to come. It is better to reduce material at
the cost of the bishop pair with 20 Bxc7! Nxc7 21 Nxa4 Bxd4 22 Bdl Bxf2+!
23 Ke2 Bg4+ 24 Kd3 Bf5+ 25 Ke2, when Black can take the draw or go for
more after 25 ... Rxd2+!? 26 Kxd2 Rd8+ 27 Ke2 Ba7 28 Nc3 Ne6 with a
monster endgame initiative for the exchange.
20 ... Ne6! 21 Nc3?’
Now Black’s initiative spins out of control. Although 21 Bdl is met by 21
... Nxd4!, this is still a better option for White than Ivanchuk’s choice.
21 ... Ndf4! 22 Bxf4
There goes the bishop pair, since White can’t allow either ... Nxe2 or ...
Nd3+.
22 ... Nxf4 23 g3 Nd3+
The engine prefers the unnatural, self-trapping 23 ... Ng2+H 24 Kfl Bh3
25 Nce4 Bxd4 with decisive pressure for Black. Note that 26 Kgl is met by
.Ne3!.
26..
24 Bxd3 Bxd3 25 Nd5
The only move. 25 Ra2? Bxd4 leaves White helpless.
25 ... Rxb2 26 Nb4
26 Nxe7+? not only wastes time, it also opens the e-file for Black’s rook.
After 26 ... Kh8 White is hopelessly lost.
26 ... Bb5 27 Rbl Rc8!
 Black’s b2-rook is immune due to White’s weak back rank. 28 Rxb2??
Rcl mate would be a delightful gift for Black.
28 Kdl Bxd4 29 Rel e6 30 Rxb2 Bxb2 31 Nbl
What a painful move to be forced to endure.
31 ... Bd4 32 f4 Kg7 33 Kd2 Rd8 34 Kcl Ba4! 35 Re2 h5!
This way he leaves open the possibility of a later king raid on the light
squares via ... Kf6 and ... Kf5.
36 Nd2 Rc8+ 37 Kbl Rc3
Threatening ... Rxa3 and ... Rai mate.
38 Re4 Bf6 39 Nc4 RB! 40 Re2?
 This game is one of those movies with an incredibly complicated plot
which only sorts itself out when miraculously unlikely events lead to an
inevitable happily ever after. Ivanchuk’s last move is a time-pressure error in
an already lost position. 40 Ne3 Bb5 41 Kcl g5! also wins for Black.

Exercise (calculation): You lucked out and got an easy one this time.
Black to play and mate in three.

Answer: Weak back rank. There is no way for White’s king to escape the Valley of the Shadow of
Death.
40 ... Rfl+ 41 Ka2 Rai mate

Game 48
I.Nepomniachtchi-Ma.Carlsen
Legends of Chess (rapid) 2020
Sicilian Defence

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

 Question: I understand that White wants to play g2-g4, but why not do so
with the more traditional f2-f3 and then later g2-g4?

Answer: By replacing 12-13 with Rgl, White gets the following options:
1. White’s future g4-g5 comes with more force with a rook behind the g-
pawn.
2. By avoiding f2-f3 White has the option in some cases to go crazy and
play f2-f4 in one go.
6 ... b5 7 g4 Bb7
Note that 7 ... b4?! is premature, as after 8 Nd5 Black can’t play 8 ...
Nxe4? due to 9 Bg2 Nc5 10 Nf6+ (discovered attack) 10 ... gxf6 11 Bxa8 and
Black doesn’t have enough for the exchange.
8g5!?Nxe4!
I like Magnus’ choice of grabbing the e-pawn, even though the game
ended in disaster for him. Declining looks inferior, despite later leading to the
opposite result: 8 ... Nfd7!? 9 a3 (unnecessary; 9 Be3 b4 10 Nd5 e6 11 Nxb4
Bxe4 12 Qd2 d5 13 0-0-0 looks a touch better for White, whose development
is more dangerous than Black’s central control) 9 ... g6! (now White’s rook is
misplaced on gl since he needs to push his h-pawn) 10 h4 Bg7 11 h5 Nc6 12
Be3 Nc5 13 f3 0-0 14 hxg6?! fxg6! 15 Nxc6 Bxc6 16 Bh3 Bxc3+! (brilliant
judgment; Magnus correctly gauges that White’s structural deficit is real
while his attacking chances on the dark squares aren’t) 17 bxc3,
I.Nepomniachtchi-Ma.Carlsen, Legends of Chess (rapid) 2020, and now 17 ...
Qa5! 18 Bd2 (or 18 Bd4 e5) 18 ... Qc7 was the strongest continuation,
planning to double rooks on the f-file with a clear advantage.
9 Nxe4 Bxe4 10 a4 e5
Given what happens in the game, the prudent 10 ... b4 is a strong
consideration.
11 axb5!?

Question: How could this possibly be sound?

 Answer: At first it does seem as if White’s knight is the dumbest student in the class, who doesn’t
get it no matter how many times the teacher tries to explain. Yet the engine evaluation is a shocking
“0.00” - dead even, since Black experiences a terrible time developing if the piece is accepted.
11 ... Be7?’
Despite the drawbacks in doing so, Black should probably accept anyway:
11 ... exd4 12 Qxd4 d5 13 Rg3 Nd7 14 Re3 Be7! 15 Qxg7 Rf8 16 f3 Bxc2 17
Bh3 (or 17 Rxa6 Bf5) 17 ... Nc5! 18 Qc3 Be4! (now ... d5-d4 is a serious
threat) 19 Qd4 Ne6 20 Bxe6 fxe6 21 fxe4 Bxg5 22 bxa6 Bxe3 23 Bxe3 with
even chances.
12 Rg4! axb5!
Magnus correctly avoids 12 ... Bb7? 13 Nf5 with a strong initiative; or 12
... Bg6? 13 Bg2 d5 14 Be3! 0-0 (14 ... exd4? 15 Rxd4 0-0 16 Bxd5 wins
material) 15 Ne2 with a clear advantage for White.
13 Bxb5+ Nd7
Here he was better off playing 13 ... Kf8 14 Bd2 Rxal 15 Qxal d5 16 Ne2
Bxc2 17 Qa7 h5! 18 gxh6 gxh6, when White’s advantage is less than in the
game.
14 Bd2! Bb7
For some mysterious reason the engine wants to move the bishop to hl!.
15 Nf5 0-0
15 ... g6 16 Nh6 Kf8 17 Rxa8 Bxa8 18 Ra4 doesn’t look like much fun for
Black either.
 16 Rxa8?
This lets Magnus back into the game. White should play 16 Rga4! Rxa4
17 Rxa4, when Ra7 is coming; e.g. 17 ... Nb6 (after 17 ... Bxg5 18 Bxg5
Qxg5 19 Bxd7 Qgl+ 20 Kd2 Black doesn’t get enough for the piece) 18 Ra7
Qc7 19 Ba5 Bxg5 20 Nxd6 Rb8 21 Ba6 Be7 22 Rxb7 Rxb7 23 Nxb7 and
wins.
16 ... Bxa8 17 Rh4
Threatening Qh5.
17 ... g6
There is happy, there is unhappy, and then there is that gauzy place in the
middle, which is where both sides live now. The position is dead even,
according to the engines. Not 17 ... Bxg5?? 18 Bxg5 Qxg5 19 Bxd7 Qgl+ 20
Kd2 Qxf2+ 21 Qe2 and White has won a piece.
18 Qg4! Nc5
Of course not 18 ... gxf5?? 19 Qh5 and Black is mated.
 19 Qh3

Exercise (critical decision): White threatens Rxh7. Black’s choices are: a) 19 ... gxf5,
grabbing the knight; b) 19 ... h5, attempting to clog the h-file; c) 19 ...
Bxg5,
eliminating the g5-pawn. Make your choice.

Answer: Only 19 ... Bxg5! works.

19 ... h5??
Psychologist’s Report: “Subject is not adjusting well to his new, enclosed
environment.” Magnus’ sense of danger fails him as he grossly
underestimates the force of Nepo’s coming rook sacrifice. He had to play 19
... Bxg5! (Project: Don’t Screw This Up saves Black; not 19 ... gxf5?? 20
Rxh7 Bf6 21 Qh6! Ne6 22 gxf6 and it’s time for Black to resign) 20 Rxh7
Bxd2+ 21 Kxd2 Qg5+ 22 Ne3 Qf6 23 Kel! (23 Ng4? is met by 23 ... Qf4+
24 Kdl B13+ 25 Be2 Qxg4 26 Rh8+ Kg7 27 Rh7+ Kf6! 28 Qxf3+ Qxf3 29
Bxf3, when Black has all the winning chances) 23 ... Bf3 and it’s anybody’s
game.
20 Rxh5!
Annihilation of defensive barrier. At this point I’m certain that Magnus
visualized a flashing neon sign which read: “You are too late!”
20 ... gxh5
No choice.
21 Qxh5 Ne6 1-0
Magnus resigned before 22 g6 fxg6 23 Qxg6+ Kh8 24 Qh6+ Kg8 25
Qxe6+, when it’s mate in six at the most. 21 ... Be4 22 g6! also leads to mate.
There aren’t many people walking the earth today who can truthfully claim
to have beaten Magnus Carlsen in 21 moves. Just in case you are under the
false impression that Magnus was in awful form in this tournament, keep in
mind that he won it by a mile: a whopping five points ahead of his nearest
rival, Nepomniachtchi!

Game 49
A.Giri-I.Nepomniachtchi
Legends of Chess (rapid) 2020
Grünfeld Defence

1 d4 Nf6 2 c4 g6 3 Nc3 d5 4 cxd5 Nxd5 5 e4 Nxc3 6 bxc3 Bg7 7 Be3 c5 8

Rcl Qa5 9 Qd2
Sometimes I suspect that the knowledge required to play this trendy line of
the Grünfeld surpasses that required for astronauts bound for a Mars landing.
. 0-0 10 Nf3 Bg4 11d5 b5
9..
The lines beginning 11 ... Nd7 and 11 ... Bxf3 are more common.
 12 Bxc5
In the final Carlsen tried 12 Be2 Nd7 13 0-0 Rac8 14 Rfdl Nb6 15 Bh6
Bxf3! 16 gxf3 b4!? 17 Bxg7 Kxg7 18 cxb4 cxb4 19 Rc6 (stronger is 19 Rbl!
Rc3 20 Qd4+ Kg8 21 Rxb4 Rfc8, when Black has some compensation for the
sacrificed pawn, but I’m not sure it’s enough) 19 ... Rxc6 20 dxc6 Rc8 21
Rcl Na4 22 Rc4 Nc3 23 Bfl Qb6 (after 23 ... Nxa2 24 c7! Rxc7 25 Rxc7
Qxc7 26 Qxa2 Qcl Black should hold) 24 a3 Rxc6 25 Rxc6 Qxc6 26 axb4
Nb5 27 Bxb5 Qxb5 28 Qd4+ e5 29 Qxa7 Qxb4 and the game was soon
drawn, Ma.Carlsen-I.Nepomniachtchi, Legends of Chess (rapid) 2020.
12 ... Rc8 13 Bb4 Qc7
Threatening ... a7-a5, chasing away a key defender of c3.
14 Nd4 a5 15 Ba3
Not 15 Nxb5? Qe5 16 Ba3 Qxe4+ 17 Qe3 Qxd5 and Black is better.
15 ... b4 16 Bb2 Qb6 17 c4 Na6 18 f4
In the second set Giri tried 18 Nc6 (if 18 Qe3 then 18 ... b3! 19 axb3 a4!
keeps the balance) 18 ... Bxb2 19 Qxb2 Rxc6! (more or less forced) 20 dxc6
Qxc6 21 Bd3 Nc5 22 Bc2 Nxe4 23 Qd4 Bf5 24 Bxe4 Bxe4 25 f3 Bf5 26 0-0
a4 27 Rfel b3 28 axb3 axb3, when Black's extra passed b-pawn gave him
enough for the exchange, A.Giri-I.Nepomniachtchi, Legends of Chess (rapid)
2020.
.
18.. e5!!

19 h3!
Taking en passant with 19 dxe6? Bxe6 leaves White dangerously behind in
development.
.
19.. Nc5’?
Nepo offers a piece to keep White’s king in the centre.
20 fxe5 Bxe5! 21 hxg4
Otherwise White suffers without payment.
21 ... Nxe4 22 Qe3 Re8! 23 Kdl Bg7 24 Qgl Nc3+??
A serious mistake which should turn the game in White’s favour. Instead,
24 ... a4! 25 Bd3 b3 offers Black enough compensation for the piece. Fat
Fritz even gives Black an edge.
25 Bxc3 bxc3 26 Nc2??
Missing 26 Nb5! Qf6 27 Rxc3 and Black doesn’t have anything much for
the piece.
26 ... Qf6 27 Rh2
27 Qd4 is met by 27 ... Re5!, intending ... Rae8, or 28 Bd3 Qg5.

27 ... Rad8’’
What an incredible concept. Nepo sees that White is unable to evade a
line-opening rook sacrifice on d5. Another way to win is 27 ... Bh6! 28 Rbl
Be3! (interference) 29 Nxe3 Qd4+ 30 Ke2 Qxc4+ 31 Kf2 (31 Kf3 Qe4+ pops
the rook on bl) 31 ... Qf4+ 32 Ke2 Qxg4+ 33 Kf2 Rxe3! 34 Kxe3 Re8+ 35
Kd3 (not 35 Kf2? Qf4 mate) 35 ... Qe4+ 36 Kxc3 Rc8+ 37 Kd2 Qc2+ 38 Ke3
Qxbl and White’s king is unable to survive the attack; Fat Fritz declares
mate in 14.
 28Rbl?’
Not that best defence, but White lacked a good one now:
a) 28 Bd3 Rxd5H 29 cxd5 Qf4 30 g3 Qf3+ 31 Be2 Qxd5+ 32 Kel Qd2+
wins.
b) 28 Rai Qg5 (threatening mate on d2) 29 Qf2 Rxd5+!! 30 cxd5 Qxd5+
31 Kcl h5! (threatening ... Bh6+) 32 g5 Qxg5+ 33 Kdl Qd5+ and Black
wins, since 34 Kcl is duly met by the crushing 34 ... Bh6+.
28.. . Rb8!
This gains a tempo.
29 Rcl
Clearly 29 Rxb8 Rxb8 30 Kcl Qf4+ is hopeless; while 29 Rb3 is met by
29 ... a4 30 Rb4 Rxb4 31 Nxb4 Rb8 32 a3 c2+! 33 Kxc2 Re8! 34 Rh3 Qb2+
35 Kdl Qbl+ 36 Kd2 Rel and White is mated.
29 ... Bh6! 30 Rai
Or 30 Rxh6 Qf4! (double attack) 31 Ne3 Rxe3 32 Rh3 Qd4+ 33 Kc2 Rb2
mate.
30 ... Rb2! 31 g3
31 Bd3 Qf4 32 g3 Qf3+ is similar to the finish in the game.
31.. . QB+ 32 Be2
 Exercise (combination alert): Black has two ways to win. Find one of them.

Answer #1:32 ... Rxe2!

The human move, which wins without fuss.
Answer #2: The optimal path is 32 ... Qe4!, which forces mate after 33 Kel Qxc2 34 Qfl Qd2+ 35
KÍ2 Rxe2+.
33 Rxe2 Rxc2! 0-1
Pinned piece/overloaded defender.

Game 50
Ma.Carlsen-LNepomniachtchi
Legends of Chess (rapid) 2020
Pirc Defence

1 d4 d6!?
 The Najdorf or Griinfeld player’s curse is that they agree to become a
slave to the latest new idea, forever having to keep up with fickle theory’s
whims. Here Nepo takes a theory vacation, possibly to evade Carlsen’s
preparation.
2 e4 g6 3 Nf3 Nf6 4 Bd3
This is one of White’s safest systems. 4 Nc3 leads to the Classical
variation of the Pirc.
4 ... Bg7 5 0-0 0-0 6 Rel c5 7 c3
Alternatives are 7 d5 and 7 dxc5.
7 ... cxd4 8 cxd4 Bg4
Otherwise 8 ... Nc6 9 h3 is slightly in White’s favour.
9 Nbd2 Nc6 10 h3
Normally Black exchanges on fB, so Nepo’s next move must have come as
a shock to Carlsen.
.
10.. Bh5?’?
 Take that, theory! There is the law and order of the zoo and then there is
the law of the jungle. 10 ... Bh5 was a novelty, and a mighty brave one when
played against a world champion!

Question: Isn’t this unsound? It feels as if White has time to defend

and consolidate.

Answer: We all have our goals, yet how much are you willing to risk to achieve them? I agree it’s
closer to unsound than sound, and so does the engine. I would have given this move an unmodified
dubious mark if it were played in a slow, classical time control game. Yet we must toss in the practical
aspect. This was a rapid game and it isn’t so easy to defend such a position with the clock running.
Either 10 ... Bd7 or 10 ... Bxf3 is of course the non-crazy course.
Somehow the normal felt too boring for Nepo, who didn’t want to defend a
slightly inferior position against the world champion.
11 g4
Magnus isn’t going to back down.
11 ... Nxg4 12 hxg4 Bxg4 13 Be2
I like the idea of hanging on to the light-squared bishop, as opposed to 13
d5 Nd4 14 Be2 Nxe2+ 15 Qxe2, when Black maybe gets some chances on
the weakened kingside light squares.
13 ... Rc8
Contemplating ... Nb4. 13 ... d5!? is worth a thought.
14 a3
Perhaps 14 d5! straight away is strongest, and if 14 ... Nb4 15 Nh2 Bxe2
16 Rxe2 Nc2 17 Rbl Nd4 18 Re3 e5 19 Ndf3 f5, Black only gets fishing
chances on the kingside for his sacrificed piece.
.
14.. Qd7?’
Here 14 ... d5 is correct, before White gets another chance himself
15 d5! Bxf3 16 Nxf3?’
This inaccuracy reduces White’s advantage. The bishop capture was
stronger: 16 Bxf3! Nd4 17 Rbl e5 18 Bg2 f5 19 Nf3 and I don’t believe in
Black’s compensation.
16 ... Qg4+ 17 Kh2 Ne5
 18 Ng5! Qh4+ 19 Kg2?!
Another inaccuracy. Magnus should have played 19 Nh3! Qxe4 20 f4 Nc4
21 Bf3 Qd4 22 Qxd4 Bxd4 23 Rbl! (rather than 23 Rxe7?! Rfe8 24 Rxe8+
Rxe8, which leaves White tangled up) 23 ... Rc7 24 b3 with the superior
chances.
19 ... Bf6
Neither side is operating at full strength. Here 19 ... Rxcl! 20 Qxcl Bh6
21 f4 Bxg5 22 fxg5 f5! 23 Qc3 f4! 24 Rfl Qxg5+ 25 Khl Qh4+ is perpetual
check.
20Rhl?’
White should throw in 20 Nxh7! Qxh7 21 Rhl, though after 21 ... Bh4 22
Bg5 Bxg5 23 Rxh7 Kxh7 24 Kg3 Bf6 25 f4 Nc4 26 Bxc4 Rxc4 27 Rcl! Rd4
28 Qe2 b5 a draw is likely.
20 ... Rxcl! 21 Rxh4?!
Magnus had to try 21 Qxcl Qxg5+ 22 Qxg5 Bxg5 23 Kg3 Bf6 24 f4 Nd7
25 Rael Bxb2 26 Rc7 Nc5 27 Kf3 e5 28 dxe6 fxe6 29 Rhxh7 Be5, which the
engine assesses as “0.00”.
21 ... Rxdl 22 Rxdl Bxg5 23 Rh3 Rc8

Advantage Black, who gets:

1. Two extra pawns for the exchange.
2. A powerfully entrenched knight on e5, since White is unable to eject it
with f2-f4.
3. The dark squares ruled by his powerful bishop.
4. Control of the c-file.
Conclusion. Magnus is now fighting for the draw.
24 Rc3’?
This is necessary, despite inflicting heavy damage to White’s structure,
since Black’s rook is too powerful if it is allowed free reign on the c-file.
However, it wasn’t necessary just yet. More accurate is 24 Rb3 b6 25 Kg3
h5! (preventing 12-f4) 26 Rc3 Rxc3+ 27 bxc3 Nd7 28 f4 Bf6 29 c4 and
White’s position looks better than in the game.
24 ... Rxc3 25 bxc3 Nd7 26 Rbl Nc5 27 e5’?
Magnus goes completely crazy and begins to give away all his pawns. I
doubt that he can save himself after the passive 27 Kf3 h5.
27 ... dxe5 28 a4’? Nxa4 29 Rxb7 Nxc3 30 d6 exd6
Amazingly, 30 ... Nxe2 also works: 31 dxe7 Bxe7 32 Rxe7 f6 33 Rxa7
Nf4+ 34 Kg3 h5 and White is losing; the fact that the black king is
temporarily cut off isn’t going to save him.
31 Ba6 e4 32 Rxa7 e3!

Nepo wants all of Magnus’ pawns off the board.

33 fxe3 Bxe3 34 Rc7 Bd4 35 Bc4 d5
Oh no you don’t. Black doesn’t care about losing the d5-pawn, if he can
eliminate White’s bishop for his knight.
36 Bb3 Kg7 37 Rd7 Kf6 38 Kf3 Be5 39 Bxd5 Nxd5 40 Rxd5 Kf5 41
Rd7 Ke6 42 Ra7 h5 43 Ra6+ Kf5
 This is one of those just-a-matter-of-time situations where White can
resign now, since he isn’t going to get three pawns for his rook.
44 Ra7 f6 45 Rh7 Bc3 46 Rh8 Bel 47 Re8 Bc3 48 Rh8 Kg5 49 Rc8 Be5
50 Rh8 Kh4 51 Kg2 g5 52 Ra8 Kg4 53 Ra4+ Bf4 54 Ra3 h4 55 Rb3 f5 56
Rb4
Or 56 Ra3 Bg3 and the pawns march.
56 ... h3+ 57 Kf2 Kh5!
57 ... h2? 58 Kg2 would be much more difficult for Black to win.
58 Rb5 Kg6 59 Rb6+ Kf7 60 Kf3 Be5 0-1
 After something like 61 Rb7+ Kg6 62 Rb6+ Kh5 63 Rb5 g4+ 64 Ke3 h2
65 Rbl Kh4 66 Kf2 Kh3, it is game over.
Unfortunately, this was Nepo’s only success in the final, which Magnus
won by two sets to nil after winning four straight games.

Game 51
Ma.Carlsen-I.Nepomniachtchi
St. Louis (rapid) 2020
English Opening

1 c4 e5 2 g3 Nc6 3 Bg2 h5’?

Isn’t there some principle which advises us to bring out pieces quickly in
the opening? I don’t trust this non-developing move, even though at this point
it’s virtually obligatory to push our h-pawns in an attacking rage.
4Nc3’?
 Magnus provokes, allowing the h-pawn to keep going forward at a cost of
time for Black. Something like 4 h4 Bc5 5 Nc3 d6 6 e3 a6 7 Nge2 Bg4 8 a3
Ba7 9 b4 looks more normal.
4 ... h4 5 e3 Nf6 6 Nge2

.6.. h3!?
A novelty at move six. AlphaZero may have won half a dozen games
against Stockfish by rushing its h-pawn down the board, but unless you can
play like AlphaZero the h-pawn may just shield the enemy king from a future
kingside attack.
Not yet 6 ... hxg3?! anyway, as after 7 hxg3 Rxhl+ 8 Bxhl Black just
wasted a lot of time.
The simple 6 ... d6 looks more logical, retaining the tension, when 7 d4
Bf5!? 8 e4 Bd7 9 d5 Ne7 10 Bg5 Ng6 11 Qd2 Be7 12 f4 hxg3 13 hxg3
Rxhl+ 14 Bxhl exf4 15 gxf4 Nh5 turned out okay for Black in Lu.Radovic-
B.Zlatanovic, Paracin 2013, since if 16 Bf3 Bxg5 17 fxg5 Ne5 18 Bxg5 then
18 ... g6 regains the piece.
7 Bf3 e4’?
Maybe Nepo is trying to mess with Magnus’ head. Black hands over an
important central pawn to eliminate White’s potentially powerful light-
squared bishop, or else win the c4-pawn. Do I believe in the move’s
soundness? Actually, no. Black loses time and swaps an e-pawn for a less
valuable c-pawn. On the other hand, if 7 ... d6 8 d4 g5!?, White can calmly
play 9 0-0 and I like his position since Black is unable to open kingside lines
easily.
8 Nxe4 Ne5
Double attack.
9 N2c3 Nxc4’?
Instead:
a) 9 ... Nxf3+? is unsound: 10 Qxf3 Be7 11 Nxf6+ Bxf6 12 Nd5 Bg5 13
Qe4+ Kf8 14 f4 and Black got nothing for the sacrificed pawn.
b) 9 ... Nxe4 10 Bxe4 Nxc4 11 d4 leaves White ahead in development but
may be preferable to what Nepo played in the game.
10 Nxf6+ Qxf6 11 Nd5’?
I’m not sure this is a tempo-gaining move, since White’s knight will later
be kicked back with ... c7-c6.1 know this may sound crazy, but maybe White
should keep developing with 11 0-0 c6 12 d4 with a clear advantage.
.
11.. Qd6 12 d4 c6 13 Nc3
 13 ... Qb4?’
This wastes more time. 13 ... Qg6 looks correct. Even here I love White’s
position.
14 a3!
Gaining more time.
14 ... Qa5
14 ... Nxb2?? hangs a piece to 15 Qc2.
15 0-0 d5?!
As awful as it sounds, Nepo should perhaps go passive and give Magnus
the centre with 15 ... Be7 16 e4 Qd8 17 Qd3 Nb6.
16 e4!
Principle: Create confrontation and open the game when leading in
development.
16 ... dxe4 17 Bxe4?!
Too slow. Now, miraculously, Black’s position actually looks playable.
Instead, 17 Rel! gives White a big advantage; e.g. 17 ... Be6 (if 17 ... f5? 18
Bh5+ Kd8 19 d5 c5 20 Bg5+ Kc7 21 Be2 Ne5 22 b4!, White has generated a
decisive attack) 18 Rxe4 0-0-0 19 Qe2 (threatening Rxe6, undermining
Black’s knight) 19 ... Nd6 20 Re5 Qc7 21 d5! and Black won’t survive.
17 ... Be6 18 d5
Magnus logically pries open the centre. Unfortunately, Nepo’s resources
are sufficient for survival.
.
18.. cxd5

19Bxd5’?
One trait both Magnus and Nepo have in common is some of their
sacrifices are only semi-sound. Maybe White has better chances to extract
something with 19 b4! Bxb4 20 Nxd5, though Black looks okay after 20 ...
Bc3 21 Nxc3 Qxc3 22 Qa4+ Kf8 23 Bxb7 Rd8 24 Bg5 f6.
19 ... Rd8 20 Qe2 Rxd5 21 Qxc4 Re5 22 Qd3
Magnus banks on his development trumping Nepo’s power on the light
squares. Both engines prefer Black’s position.
22 ... Qc5
Thinking about... Qc6 and ... Qg2 mate.
23 Qf3
Wisely securing control over the hl-a8 diagonal.
23 ... Qc8 24 Bf4 Rc5 25 Ne4 Rc6 26 Ng5
White wants the light-squared bishop.
26 ... Be7!
26 ... Bg4?! is a waste of time, since 27 Qb3 pretty much forces 27 ... Be6
again.

27 Nxe6 fxe6!?
I doubt the open f-file makes up for the weakness in Black’s structure. I
would give a pawn up with 27 ... Rxe6 28 Rael Rc6 29 Rfel Rxcl 30 Rxcl
Qe6! 31 Qxb7 0-0, when Black has enough counterplay since there will be an
eternal mating threat on g2.
 28 Be5 Bf6 29 Bxf6 0-0 30 Qg4 Rxf6 31 Qxh3
Magnus is willing to give up the initiative for the clamping h3-pawn.
31 ... Rc2 32 b4 Qc6 33 Radi Qb6 34 Qg2 Rc3?’
The best course is to do nothing and make a waiting move like 34 ... a6.
35 Qe4! Rxa3 36 Rd2
After 36 Rd7 Qc6 37 Rd8+ Rf8 38 Qxc6 bxc6 39 Rxf8+ Kxf8 40 Rcl Ra6
41 Kg2 Ke7 42 h4 Kd6 43 g4 Ra4 Black stands no worse in the ending.
36 ... Raf3 37 Kg2 a6 38 Rc2 Rf8 0-1

Question: Huh? Why did Carlsen resign?

Answer: He didn’t. Magnus lost his connection and was unable to re-establish it before losing on
time, in a position where he still stands a shade better in view of Black’s inferior structure. So instead
of having to work for a draw, Nepo was gifted a “technological” win.

I have a hunch that between the years 2023 and 2029,1 may be writing the
book Firouzja: Move by Move. The Iranian-born/French teen phenom is
thought by many, including me, to be a safe bet as a future challenger for the
world championship.

Game 52
Nepomniachtchi-A.Firouzja
I.
St. Louis (blitz) 2020
Nimzo-Larsen Opening

1 b3
Classical chess is the fundamentalist, while hypermodernism is the attempt
at a reform movement. Nepo is at heart a member of the Classical school, but
he does sometimes flirt with hypermodernism, as here.
1 ... e5 2 Bb2 Nc6 3 Nf3
This is White’s third choice, behind 3 e3 and 3 c4. The idea is to provoke
Black into an odd, reversed version of Alekhine’s Defence, should Black
choose to chase the knight with ... e5-e4.
3 ... d6
The alternative is 3 ... e4 4 Nd4 Nxd4 5 Bxd4 d5 6 e3 Nf6 7 c4 dxc4 8
Bxc4 Bd6 9 Nc3 Bf5 10 f3, when I slightly prefer White’s position.
4 d4 e4 5 d5
 All theory so far.
5.. . exf3
This line doesn’t score all that well for Black. Instead, 5 ... Nce7! 6 Ng5
e3! 7 fxe3 Nf5 8 Nf3 Nxe3 9 Qd3 Qe7 10 Bel! Nxfl (10 ... Ng4!? is another
option) 11 Rxfl Nf6 12 Nc3 g6 looks equal, P.Fry-E.Guo, Sydney 2009.
6 dxc6 bxc6 7 gxf3
The dynamic choice, opening the g-file for White’s rook. 7 exf3 develops
faster, at the cost of devaluing White’s kingside pawn majority.
7 ... Nf6 8 Qd4!

Question: Isn’t this a violation of principle?

Answer: Maybe so. Yet I still like it since it prevents Black from playing ... g7-g6 and ... Bg7. I
don’t care for Black’s position, no matter how he responds.
8.. . Be6
After 8 ... Be7 9 Rgl 0-0 10 Nc3 d5 11 0-0-0,1 would be nervous playing
Black, with so much force aimed at the g7-square.
9 Rgl
Now it becomes difficult for Black to develop his kingside.
9 ... c5 10 Qa4+!

A novelty and improvement over 10 Qh4, N.Barth-D.Justen, German

League 2006, where 10 ... Nd5 gives Black a better position than Firouzja
achieves in the game.
10 ... Nd7
If 10 ... Bd7 11 Qf4 Nh5 12 Qe4+ Be7 13 f4 Nf6 14 Qf3 0-0 15 Nc3,
White’s coming attack looks very dangerous; while after 10 ... Qd7 11 Qa5
c6 12 Nd2 Nd5 13 0-0-0 the question remains: how will Black develop his
kingside?
11 Nc3 Rg8
This means that Black’s king will remain in the centre for quite some time.
12 Ne4
 Threatening a dirty cheapo. 12 f4! is even stronger: 12 ... g6 13 e4 Bg7 14
e5! looks dangerous for Black, who shouldn’t play 14 ... dxe5? because of 15
0-0-0, threatening Bb5 and wins.
12 ... f6
Now Black’s g8-rook is protected so White’s Nxf6+ tricks no longer work.
Not 12 ... d5??, which loses to 13 Nf6+ gxf6 14 Rxg8. The engine suggests
the radical 12 ... Be7!?, just giving the g7-pawn away. White may do best to
ignore it and play 13 0-0-0! Kf8 14 f4 with a clear advantage.
13 e3 Kf7 14 0-0-0 Nb6
After 14 ... a5 15 Nc3 Nb6 16 Qh4 White’s attack looks more dangerous
than Black’s.
15 Qa5!
This way Black is denied ... a7-a5-a4.
15.. . Qe7 16Bb5’?
The bishop heads for the cubbyhole on c6.
16.. . h6
He wants to play ... f6-f5 without allowing Ng5+. After 16 ... d5 17 Nc3
Qd6 18 h4 c4 19 bxc4 Qc5 20 Na4 Nxc4 21 Nxc5 Nxa5 22 Nxe6 Kxe6 23
Bd7+! Kd6 24 Ba4, threatening Bxf6 or 24 ... Be7 25 e4, Black is in deep
trouble.
17 Bc6 Rd8 18 Qxa7!?
 Question: Isn’t this crazy, taking an enemy pawn in front of one’s own king?

Answer: I’m pretty certain Nepo’s motivation was other than greed. Maybe he thought that, by
grabbing the a7-pawn, his queen creeps closer to Black’s king. But yes, I agree it’s risky, and perhaps
unnecessary since normal moves offer White a big advantage. For instance, 18 Nc3 d5 19 Nb5 looks
more thematic.
18 ... d5
The idea is to dislodge the e4-knight and then play ... Qd6, hopefully
displacing White’s light-squared bishop as well.
19 Nc3 Qd6 20 Bb7
There was no need to move the bishop, since 20 Qb7! gives White a
winning position; e.g. 20 ... Bc8 (20 ... Rb8!? is well met by 21 Nb5!,
declining the offer) 21 Bxd5+ Nxd5 22 Qxd5+ Qxd5 23 Nxd5 Bd6 24 f4 and
Black is busted, two pawns down in an endgame.
20 ... Be7!
Now ... Ra8 is a serious concern for White.
 21 Qa6!
The engine prefers to give pawns away with 21 Nb5!? Qxh2 22 Kbl. In
any case, it’s a mistake to ignore Black’s intent: 21 f4?? Ra8! 22 Bxa8 Rxa8
23 Qb7 Bc8 (the queen is trapped) 24 Qxa8 (or 24 Nb5? Bxb7 25 Nxd6+
Bxd6) 24 ... Nxa8 25 Nxd5 Qe6 and I doubt White can survive.
21 ... c4
It was probably better to do some pawn grabbing of his own with 21 ...
Qxh2!. This time the engine wants to give away a million pawns with 22
Nb5!? Qxf2 23 Kbl Bd6 24 a4 Qxf3 25 a5 Nc8 26 e4! Qxe4 27 Nc3 Qh4 28
Nxd5, when White has strong attacking compensation and actually stands
better.

22 Nb5!
The military paratrooper looks at the ground below and then leaps into the
wind, praying his parachute will function as it was designed to do. I just
described Nepo’s last move. Nepo once again offers his h-pawn, with more
pawns to follow. His move looks stronger than 22 f4 Qc5.
22 ... Qxh2 23 Nd4 Qxf2
The pressure mounts against Nepo, in that his initiative must produce
something concrete or he bleeds pawns for nothing. After 23 ... Bd7!? 24
Bxd5+ Nxd5 25 Qxc4 Qd6 26 e4 Qc5 27 Qxd5+ Qxd5 28 exd5 h5, Black can
fight on and hope to generate counterplay with the h-pawn, but he certainly
doesn’t get full compensation for two pawns.
24 e4?’
Nepo furiously opens central lines and attempts to ram on through, hoping
to get at Black’s king. The engines say this move gives away a big chunk of
White’s advantage. 24 Kbl! looks more difficult for Black.
24 ... Bc8?’
A suboptimal retreat. After 24 ... dxe4! 25 fxe4 Qf4+ 26 Kbl Qe5! 27
Nxe6 (or 27 Nc6 Rxdl+ 28 Rxdl Qc5) 27 ... Qxe6 28 Qb5 Black doesn’t
look that much worse, if at all.
25 Bxc8 Rxc8 26 exd5?
We all love to pander to our vices. Here Nepomniachtchi displays a
recurring trait: he refuses to take time out for a necessary defensive move
when he has one which fuels his initiative. Objectively, he should protect his
a-pawn with 26 Kbl! and White’s position is still better.
26 ... Ra8 27 Qb7
 Exercise (critical decision): Black has four logical choices: a) 27 ... cxb3 to
open lines; b) 27 ... c3 to open lines; c) cover c7 with 27 ... Bd6 before
attacking;
d) chop White’s a-pawn with 27 ... Rxa2. Two of the these save Black;
the other two lose. Which would you play?

Answer: Both pawn moves save Black, whereas ... Bd6 and ... Rxa2 allow White a decisive
advantage.
27 ... Bd6?
Instead:
a) 27 ... Rxa2?? loses to 28 Qxc7 cxb3 29 Rgel Nxd5 30 Qc4 Rd8 31
Qxb3 Ra6 32 Re2! Qg3 33 Rxe7+! and the rook can’t be taken due to a fork
on f5.
b) 27 ... c3! 28 Kbl (not 28 Bxc3?? Qe3+ and White loses the bishop since
29 Kb2 gets mated by 29 ... Ba3+ and ... Qxc3) 28 ... cxb2 29 d6 Bxd6 30
Qe4 Be5 31 Qg6+ Kf8 32 Ne6+ Ke7 33 Nxg7 is a colossal mess which the
engine unhelpfully says is dead even!.
c) 27 ... cxb3! also saves Black: 28 axb3 Nc4! 29 d6 (not 29 bxc4?? Rgb8
30 Qxc7 Qe3+ and White gets mated eventually) 29 ... Qe3+ 30 Kbl Nd2+
31 Rxd2 Qxgl+ 32 Bel Bxd6 33 Qd5+ Kg6 34 Qf5+ Kf7 35 Qd5+ with
perpetual check.
28 Nf5?
The position is too complex for a rapid game. Here 28 Kbl! Rgb8 29 Qc6
cxb3 30 axb3 Ra5 31 Ne6! (targeting g7) 31 ... g5 32 Rhl gives White a
winning attack.
28 ... Bf4+?
Now 28 ... Qc5! 29 Kbl h5 is okay for Black.
29 Kbl Rxa2?
29 ... Qc5 is a better try. The engine gives the crazy line 30 b4! (aiming to
control d6 and push the d-pawn) 30 ... Qb5 (not 30 ... Qxb4? 31 d6! Qa4 32
a3 Qd7 33 Qe4! and Black is losing since there is no good place for his
bishop) 31 Rg4 Bh2 32 Rhl Rae8 33 a3 and Black is busted.
 Exercise (combination alert): Should White take the rook or decline?
And if the offer is declined, what should White play instead?

Answer: Zwischenzug.
30 Qc6!
Threatening Qe6+ and Qe7 mate. 30 Kxa2? throws away White’s win: 30
... Qxc2 31 Nd4 cxb3+ 32 Kai Qc4 33 Kbl Qa4 (threatening mate on a2) 34
Nxb3 Qxb3 35 d6 Be5 36 Rd2 Ra8 37 Qxc7+ Ke6 38 Rel Qa2+ 39 Kcl
Qc4+ 40 Rc2 Qf4+ 41 Kbl Nd5 and Black is okay.
30 ... Rxb2+
30 ... Qe2 31 Rxg7+! Rxg7 32 Qxf6+ is game over.
31 Kxb2 Be5+
31 ... Qe2 32 Rdel Be5+ 33 Kbl changes nothing.
32 Kbl 1-0
Suddenly, Black’s attack is an engine devoid of fuel and e6 can’t be
defended. Firouzja resigned in view of 32 Kbl Ra8 (or 32 ... Bd6 33 Nxd6+
cxd6 34 Qc7+ Kf8 35 Rdel etc) 33 Qe6+ Kf8 34 Qe7+ Kg8 35 Qxg7 mate.

Game 53
L.Aronian-I.Nepomniachtchi
Skilling Open KO (rapid) 2020
English Opening

1 c4 c5 2 Nc3 Nf6 3 g3 d5 4 cxd5 Nxd5 5 Bg2 Nc7

Principle: The side with more (future) space should avoid swaps. 5 ...
Nxc3 is also playable.
6d3 e5
On 6 ... Nc6,1 always go for 7 Bxc6+ bxc6 8 Qa4 Bd7 9 Nf3, trusting that
White’s superior structure outweighs Black’s power on the light squares.
7 Be3
White is still waiting for ... Nc6 so he can play Bxc6+.
7 ... Be7 8 Rcl Ne6 9 Nh3’?
 Question: Isn’t Aronian taking his I-insist-on-playing-Bxc6 stance a bit too far?
White’s knight on the h3-rim really does look dim, as the rhyme suggests.

Answer: I agree with your assessment; he should probably just give up on the Bxc6 project and
develop with 9 Nf3, blocking the diagonal.
9 ... 0-0 10 0-0 Nc6!
A novelty which comes with the unspoken taunt: “Go ahead, punk, make
my day!” 10 ... Nd4, as in V.Harasta-J.Banas, Rimavska Sobota 1975, is okay
but unnecessary, since Bxc6 isn’t a real strategic threat anymore.
11 Bxc6!?
Aronian perhaps takes it too far. With White’s knight on h3 this swap
doesn’t look right anymore. It was probably wiser to refrain and play the
more thematic 11 f4.
.
11.. bxc6 12 f3 Rb8 13 b3 f5 14 Nf2
I know we aren’t supposed to allow an opening of the position when the
opponent owns the bishop pair, yet maybe here White should consider a
direct violation of the principle and play 14 f4!? to challenge Black’s
kingside expansion; e.g. 14 ... exf4 15 Nxf4 Nxf4 16 Bxf4 Rb7 17 Na4 Re8
18 Qc2 Bf6 19 Nxc5 Bd4+ 20 e3 Rxe3! 21 Qc4+ Re6+! 22 Rf2! Bxc5! 23
Qxc5 Rd7 and I only slightly prefer Black.
14 ... h5 15 Khl g5!

Clearly, patience is a commodity of which Nepo has precious little! This is

an intimidating image if you sit on White’s side of the board.

Question: Isn’t this crazy, seeing as Black already carries the burden of a
pawn weakness on c5. Should he be pushing on the kingside and creating
new weaknesses?

Answer: Sometimes crazy is just fine; the crazy don’t tend to overthink a decision. It’s actually the
correct strategy because of the c5-weakness, which induces Black to proceed energetically, since quiet
play probably leads to slow death.
 16 Na4 Rb5 17 Bd2
Aronian isn’t interested in wasting the white pieces by proposing a draw
with 17 Nc3 Rb7 18 Na4.
17 ... g4 18 e4?’
White seeks to close the opportunity gap before it grows any wider, which
induces radical action.

Question: Why declare this move dubious when White clearly follows the
principle: Counter the opponent’s wing attack with a reaction in the centre
.?

Answer: This is why chess is such a miserably difficult game. Aronian is indeed following
principle, yet he times it badly. It was necessary to exchange on g4 first: 18 fxg4! hxg4 (forced, since
18 ... fxg4? 19 Bh6 is good for White) 19 e4!, when White can count on equal chances since Black
lacks the benefit of an open f-file. For example: 19 ... Nd4 20 Nc3 Nf3!? (or 20 ... Rb7 21 exf5! Rxg5
22 Qxg4 Bg5 23 Bxg5 Rxg5 24 Qe4) 21 Nxb5 cxb5 22 Be3 Qe8 23 Bxc5 Qh5 24 h4 Bxc5 25 Rxc5 f4
26 Rxc8! Rxc8 27 Nxg4! (undermining) 27 ... Qxg4 28 Qxf3 Qh3+ 29 Kgl Rc2 30 Rf2 Rcl+ 31 Rfl
Rc2 with a draw.
18 ... Nd4! 19 f4
Now 19 fxg4 is met by 19 ... fxg4!, when White is in all sorts of trouble
with weaknesses everywhere; e.g. 20 Nb2 Be6 21 Nc4 Bxc4 22 Rxc4 Bg5!
23 Kg2 Rb7! 24 Rxc5 Bxd2 25 Qxd2 Qf6 (threatening ... Q13+, followed by
... Ne2+) 26 Qe3 h4! 27 Rfcl h3+ 28 Kgl Qg7 29 Rxc6 R13 30 Qel Rbf7 31
R6c2 Qf6 and wins.
19 ... Nf3! 20 Be3 h4! 21 Kg2 Qe8!
The queen is transferred to the kingside.
22 Bxc5 Bxc5 23 Nxc5
 White has picked up a pawn, but it’s not as if he pulled off a bank heist
since his spoils are very much outweighed by the dire situation of his king.
23 ... Qh5 24 Rhl
Covering the infiltration point h2.
24 ... exf4 25 gxf4 h3+
25 ... Qh6, threatening the indefensible f4-pawn, looks even more deadly.
26 Kfl
26 Kg3?? Qh4 mate would be highly inadvisable,
.
26.. fxe4
Here too 26 ... Qh6! is very strong.
27 Ncxe4 Rxf4
Now ... Rbf5 is coming.
28 Rxc6 Bb7!
Black’s bishop enters the attack.
29Rf6
 If 29 Rc7 Qe5 30 Qcl then 30 ... Nd2+! (overloaded defenders; taking the
knight drops either rook) 31 Kel Nxe4 wins.
.
29.. Rbf5!
Nepomniachtchi allows a knight fork on g3.
30 Ng3
30 Rxf5 Qxf5 31 Ke2 Ng5 is lost for White.

Exercise (combination alert): What is Black’s strongest continuation of his attack?

Answer: Step 1: Clearance.

30 ... Nd2+! 31 Kel
31 Qxd2 loses to 31 ... Rxf2+ 32 Qxf2 Rxf2+ 33 Rxf2 Qc5.
.
31.. Qe8+! 0-1
Step 2: Zwischenzug. White resigned since 32 Kxd2 Rxf2+ 33 Kel Rxf6
wins a rook.
 The next game was a wild fight, with both players missing wins.

Game 54
I.
Nepomniachtchi-Ma.Carlsen
Skilling Open KO (rapid) 2020
Ruy Lopez

1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Ba4 Nf6 5 0-0 Be7 6 d3 d6 7 c3 0-0 8 Nbd2

Re8 9 Rel Bf8 10 d4 b5 11 Bc2 exd4’?

Question: Why would Black give up the centre like this?

Answer: In order to attack it:

1. White’s e-pawn becomes a touch more vulnerable with the opening of
the e-file.
2. Black can play ... Bg4 next, applying pressure to the d4-pawn.
3. Black also has the option of... Nb4 and ... c7-c5 once White recaptures
with c3xd4.
11 ... Bb7 is a safer alternative.
12 cxd4 Bg4 13 Nfl
After 13 h3 Bh5 14 g4 Bg6 15 a3 Qd7 16 d5 Ne7 17 Nfl c6 18 dxc6 Qxc6
19 Nd4 Qb7 20 Ng3 d5 21 e5 Bxc2 22 Qxc2 Ng6 23 f4, White’s chances of
overextending are about equal to those of whipping up a winning attack,
Au.Caruso-V.Hetzer, correspondence 2018.
13 ... g6
In an earlier game Black inserted 13 ... Bxf3 14 gxf3 before 14 ... g6,
R.Golubev-E.Inarkiev, Moscow 2015, when I wouldn’t mind playing Black’s
position.
14 Ng3 Bg7 15 Be3 Nd7 16 Rcl
A novelty. 16 Qd2 had been played before.
16 ... Bxf3
This creates an interesting imbalance. White gets the bishop pair and
enhanced control over the light squares, at the cost of slight damage to his
structure.
17 gxf3 Qf6!
 Pressure begins to mount upon d4.
18Bb3’?
Threatening Bd5. 18 Ne2 is also possible since 18 ... Qxf3 can be met by
19 Bbl! Rxe4 20 Ng3! Qxdl 21 Rexdl Rxd4 22 Bxd4 Nxd4 23 Rxc7 Nc5
and I prefer White by a shade.
18 ... Nxd4 19 Bxd4
This turns it into an opposite-coloured bishops position, which favours the
side with the more promising attack.
19 ... Qxd4 20 Rxc7 Ra7!
Magnus cleverly activates his sleeping a8-rook and exchanges it for
White’s more active one on c7.
21 Rxa7 Qxa7 22 Qxd6 Bxb2 23 f4

Question: It looks as if White will benefit from the opposite-coloured bishops

since his attack is brewing, while Black has no attack. Is this assessment
correct?
 Answer: I actually prefer Black! With such an open centre, I don’t believe White’s attack on the
light squares is any more potent than Black’s potential to attack. The engine calls it dead even at this
point.
23 ... Nc5 24 Bd5 Bd4’?
Targeting White’s soft f2-spot. I prefer 24 ... Qe7! 25 Qxe7 (otherwise
Black will play ... Nd3!) 25 ... Rxe7 and favour Black’s queenside majority to
White’s on the kingside.
25 Re2
The tempting 25 Bxf7+ Qxf7 26 Qxd4 is at least even for Black after 26 ...
Ne6! 27 Qb6 Nxf4, since 28 Qxa6?? runs into 28 ... Nh3+ 29 Kg2 (or 29
Khl?? Qf3 mate) 29 ... Qxf2+ 30 Kxh3 Qxel with a winning position.
25 ... Nd3 26 Kg2?’
After 26 Rd2! Bxf2+ 27 Kg2 Bxg3 28 hxg3 Nel+ 29 Kfl NI3 30 Rf2 Qe3
31 Qc6 White stands no worse.
26 ... Kg7?’
Both sides make rapidplay-typical, useful-looking, yet non-incisive king
moves. Instead, 26 ... Qb6! (not 26 ... Nxf2?! 27 e5! intending e5-e6, while
26 ... Bxf2 27 Bxf7+ Qxf7 28 Qxd3 Bc5 is only equal) 27 Qxb6 Bxb6 28 Kf3
Rd8, intending ... Nb4, is in Black’s favour.
27 Rd2?
Ironically, another king move, 27 KI3, is necessary here.
 Exercise (planning): Prove why Nepo’s last move was a blunder.

Answer: Removal of the guard.

27 ... Bc5!
White’s queen is driven away from her coverage of f4, so Magnus wins a
clean pawn.
28 Qc6 Nxf4+ 29 Kf3
Nepo must have counted on this move to regain material, since he menaces
dual attacks on e8 and f4.
 Exercise (planning): Black has a sneaky way of hanging on to all his pieces
and retaining the extra pawn. What should he play?

29 ... Rd8?
This piece sacrifice offers sufficient compensation, yet Magnus missed an
opportunity.
Answer: 29 ... Bb4! 30 Qxe8 Bxd2 31 Qe5+ f6 32 Qb2 Nxd5 33 exd5 Ba5 34 Ne4 Bd8 offers Black
excellent chances to convert the extra pawn.
30 Kxf4
Otherwise White will just be a pawn down.
.
30.. Rd6!
It’s getting crowded in there!
31 Qe8?
Due to time pressure both sides managed to surpass the hesitancy level of
Hamlet. Correct was 31 Qc8! Bxf2 32 Rc2 Qe3+ 33 Kg4 Bxg3 34 hxg3
Rxd5! 35 exd5 Qe4+ 36 Kh3 Qhl+ 37 Kg4 Qe4+ with perpetual check.

Exercise (combination alert): White’s last move was a blunder. Prove why.

Answer: Moving the queen to c7 wins heavy material.

31 ... Bxf2?
Black goes from winning to losing! Magnus perhaps thought that without
dramatic measures he had no chance of survival. In fact, he was winning had
he found 31 ... Qc7! (threatening deadly discovered checks) 32 Qe5+ (if 32
Kf3? Rd8!, White loses his queen for a rook, or else a full rook) 32 ... f6! (32
... Rf6+?? drops a piece to 33 Nf5+! etc) 33 Qc3 Rxd5+ 34 Kf3 b4 35 Qb2
Rxd2 36 Qxd2 Qd6, when White is a pawn down and busted.
32 Kg4!
Now Black’s position is on the verge of entering a debtor’s prison, since
he just doesn’t have enough for the sacrificed knight.
32 ... Rf6 33 e5 h5+ 34 Kh3
 Black’s attackers prove to be remarkably useless, mainly because he
doesn’t have enough fire-power on the light squares to endanger the white
king.
34 ... Rf4
Magnus puts up a defiant front, even if he probably didn’t believe it in his
heart.
35 e6!
This line wins, as does the more difficult to see 35 Bxf7H Rxf7 (35 ...
Qxf7 36 Rd7 wins the queen) 36 Rd8 Kh6 37 e6 Rf4 38 Qh8+ Kg5 39 Rd5+
Rf5 40 Qd8+ Kh6 41 Nxf5+ gxf5 42 Qf6+, forcing mate.
35 ... fxe6 36 Bxe6 Bd4
After 36 ... Kh6 37 Rd7 Qc5 38 Qh8+ Kg5 39 Qd8+ Kh6 40 Bg8! Black
must hand over his queen to prevent immediate mate on h7.

Exercise (combination alert): How did Nepo force mate on the reigning
world champion?
 Answer: Annihilation of defensive barrier.
37 Nxh5+!
As bad omens go, this is the equivalent of seeing Odin’s ravens gathering
over Black’s position. His king can’t survive the opening of the g-file.
37 ... gxh5
No choice, since Black’s rook was hanging.
38 Rg2+ Kf6
Or 38 ... Rg4 39 Bxg4 hxg4+ 40 Rxg4+ Kf6 41 Rg6+ Kf5 42 Qe6+ Kf4 43
Rg4+ Kf3 44 Qe4+ KI2 45 Rg2+ Kfl 46 Qe2 mate.
39 Rg6+ Ke5 40 Bb3+ 1-0
In view of 40 ... Kf5 41 Qe6 mate

Game 55
I.
Nepomniachtchi-Dan.Dubov
Airthings Masters (rapid) 2020
King’s Indian Defence

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 Be2 0-0 6 Be3 Nc6’?

Careful consideration about the future ramifications of present
recklessness isn’t Dubov’s strong suit.

Question: Why is Black willing to lose tempi like this?

Answer: 6 ... Nc6 is based upon the same principle of Alekhine’s Defence, where Black plays under
the assumption that giving away free pawn pushes is not such a disadvantage, since White also
becomes more vulnerable to overextension. Having said that, I don’t believe this version is as sound as
Alekhine’s Defence. 6 ... e5 is Black’s main move.
7d5
Trying for a more normal position with 7 NI3 is a bit cowardly and, in any
case, is thwarted by 7 ... Ng4!. For example: 8 Bg5 f6 (or 8 ... h6) 9 Bh4 e5
10 h3 Nh6 11 dxe5?! (11 d5 is correct) 11 ... dxe5 12 Qxd8 Nxd8 13 Nd5
Ne6 and Black stood well, D.Khegay-Dan.Dubov, Titled Tuesday (blitz)
2021.
7 ... Ne5 8f4’?
More energetic than 8 h3 e6 9 f4 Ned7.
. Neg4!
8..
There is no backing down now. If 8 ... Ned7?! 9 Nh3! c6 10 NI2 cxd5 11
cxd5, as in Pe.Schreiner-T.Reich, Aschach 2016, I don’t see where Black’s
counterplay is coming from.
9 Bd2 Nh6 10 h3’?

This is a war both sides are anxious to fight.

Question: I see the idea is g2-g4-g5 winning a piece. Does White have the time?

Answer: It’s going to be close since Black will use the spare time to create chaos in the centre. The
engines say White is favoured, but not by much. 10 Nf3 is a safer, yet probably not superior,
alternative.
.10.. e6!
We already know where this is headed. Black will give up a piece for
attacking chances. Striking from the other side with 10 ... c6?! is inferior, as
after 11 g4 b5 12 cxb5 cxb5 13 a3 e6 14 g5 Nxe4 15 Nxe4 Nf5 16 Bc3 Bb7
17 Bxg7 Kxg7 18 Nf3 Bxd5 19 Nf6 Black doesn’t have enough for the piece.
11 dxe6!
 Now 11 g4?! is met by 11 ... exd5 12 exd5 Re8 and if 13 g5 Nf5 14 gxf6
Qxf6 15 N13 Ne3, Black gets full compensation for the sacrificed material.
.
11.. Bxe6’?
Any time I watch a match-up between Nepo and Dubov, the only certainty
is that there will be gratuitous violence! Unsurprisingly, Dubov is willing to
give up a piece for development and an attack. Safer is 11 ... fxe6 12 g4 Nd7
13 h4!, when the engine says even, though I still like White’s chances.
12 g4 Re8 13 Nf3!
Not yet 13 g5? since 13 ... Nxe4! 14 Nxe4 Nf5 15 Kf2 d5! leaves White in
deep trouble.
13 ... Bd7! 14 g5 Nxe4

The position resembles a Jackson Pollock painting, splotched in random

spray. Already, the position is devoid of familiar strategic landmarks.
15 gxh6
15 Nxe4 Rxe4 16 gxh6 Bxh6 transposes.
 15 ... Bxh6 16 Nxe4 Rxe4 17 f5
After 17 Ng5 Bxg5 18 fxg5 Qe7 19 Kf2 Qe5 20 B13 Qd4+ 21 Kg3 Qe5+
22 Kg2 Re8! 23 Rel Rxel 24 Qxel Qxb2, Black has decent compensation
for the sacrificed piece.
17.. . Qf8
The idea is to make room for ... Rae8. Instead, 17 ... Be3 looks slightly
more accurate; e.g. 18 Bxe3 Rxe3 19 Qd2 Qe7 (threatening ... Rxf3) 20 Kf2
Re8 21 Rael Qf6 22 Bd3 Rxel 23 Rxel Rxel 24 Kxel Bxf5 25 Bxf5 Qxf5
26 Kf2 Qxh3 27 Qa5 Qd7 28 Qxa7 Qc6 and the engine calls it dead even.
18 0-0! Rae8
Again 18 ... Be3+! is stronger; e.g. 19 Bxe3 Rxe3 20 fxg6 hxg6 21 Rf2
Bxh3 22 Qd2 Rae8 23 Bfl Qh6 and chances look balanced.
19 Bd3
Now after 19 Rf2! Be3 20 Bxe3 Rxe3 21 fxg6 hxg6 22 Bfl Qh6 23 c5!
White stands slightly better.
19.. . Re3!?
 Inflation is one of life’s constants. When I was in my 20s, a homeless
person would normally hit me up for spare change. Today they always ask
for a spare dollar, and in the future I may need to give them my credit card.
Dubov feels it necessary to add more material in the deficit column and is
willing to be a full rook down to fuel his attack. Once again 19 ... Be3+! is
safer: 20 Bxe3 Rxe3 21 Rf2 Qh6 and I doubt that White has much of an edge,
if any.
20 Bxe3 Bxe3+ 21 Kg2?’
It was better to put the king in the corner and leave g2 for the queen: 21
Khl Bc6 (or 21 ... Qh6 22 Qc2 Qxh3+?! 23 Qh2) 22 Qc2 Bc5 23 Qg2 Re3 24
Radi, though it still won’t be easy for White to unravel.
21 ... Bc6 22 Qel g5
After 22 ... Qg7! (threatening ... g6-g5-g4) 23 Qg3 Qxb2+ 24 Khl Qc3 25
Radi Bh6, Black’s chances look better, mainly because it now looks even
more difficult for White to unravel.
 23 Qg3 Bf4 24 Qg4 Re3 25 Radi

25 ... Qg7?
Here 25 ... Qe8! 26 h4 Rxd3! 27 Rxd3 h5 28 Qxh5 Qe2+ 29 Rf2 Qxd3
more or less equalizes; e.g. 30 f6 Qh7 31 Qxh7+ Kxh7 32 Kh3 Bd7+ 33 Kg2
Bc6 34 hxg5 Bxg5 35 Kg3 Bxf6 reaches an even ending.
26 Rf2
Inserting 26 f6! Qxf6 before 27 Rf2 is stronger.
26 ... Qh8?
Nimzowitsch may approve, but nobody else does. The threat is ... h7-h5,
trapping White’s queen. The problem is that it simply loses two tempi after
White’s next move. 26 ... f6 keeps Black’s disadvantage to a minimum.
27 Qh5!
Diffusing... h7-h5.
27 ... Qg7
An admission that his last move was incorrect.
 28 Kfl! Bg3
Now 28 ... f6 is no good, since 29 Nd4 heads into e6.
29 f6!

Not yet 29 Qxg5?? Bxf2 30 Qxg7+ Kxg7 and White is busted.

29 ... Qxf6 30 Qxg5+ Qxg5 31 Nxg5
Black’s initiative is running dry and he remains heavily behind in material.
31 ... h6 32 Nxf7! Kg7
32 ... Bxf2 33 Nxh6+ Kg7 34 Nf5+ Kf6 35 Nxe3 Bxe3 leaves White a
clean exchange up.
33 Nd8!
Forcing simplification.
33 ... Bxf2
Dubov finally decides that poverty doesn’t agree with him.
34 Kxf2 Rf3+ 35 Ke2 Rxh3 36 Rfl! Rh2+ 37 Ke3!
Stronger than blocking with the rook.
 37 ... Rxb2

Exercise (combination alert): Black just blundered in an already lost position.

What is White’s best continuation?

Answer: It’s mate in two.

38 Ne6+ 1-0
A nasty message arrives in the black king’s inbox: 39 Rf8 is mate, no
matter which square he moves to.

The following game is one of the most amazing in the book. It displays
Nepomniachtchi’s almost superhuman power when in possession of the
initiative.

Game 56
N epomniachtchi-H.N akamura
I.
Carlsen Invitational (rapid) 2021
Ruy Lopez

1 e4 e5 2 M3 Nc6 3 Bb5 Nf6 4 0-0

In recent months Nepo seems to have abandoned 4 d3 (as in Game 24) in
favour of castling.
4 ... Nxe4
This move is the entryway to the long, theoretical, opening/ending
variations of the main line of the Berlin Lopez.
5d4
5 Rel leads to a boring Exchange French(ish) position after 5 ... Nd6 6
Nxe5 Be7 7 Bfl Nxe5 8 Rxe5 0-0 9 d4 Bf6 10 Rel Re8 11 c3 Rxel 12 Qxel
Ne8 13 Bf4 d5.
5 ... Nd6 6 Bxc6 dxc6 7 dxe5 Nf5 8 Qxd8+ Kxd8

This position comes from noble ancestry. It dates back to the late 19th
century and is generally believed to have made Vladimir Kramnik world
champion, since he drew every game from here in his match against Garry
Kasparov, thereby negating Kasparov’s white pieces advantage.

Question: Why would anyone be attracted to this position for Black, who
can’t castle, is behind in development, and is stuck with a non-fimctional
queenside pawn majority against White’s healthy 4:3 on the kingside?

Answer: Black has hidden advantages as well:

1. It’s an ending so it’s not always a disadvantage for Black’s king to be in
the middle.
2. Black owns the bishop pair.
3. White is missing his light-squared bishop, which could mean potential
weakness on that colour complex later on.
4. The white e-pawn being pushed to the fifth rank is significant since this
makes it difficult for White to utilize the kingside majority; in particular f2-
f4-f5 will be difficult to achieve without the inclusion of the double-edged
g2-g4.
9 Nc3 Ke8 10 h3 Be6
Two games later Nakamura switched to 10 ... h5, discouraging g2-g4, and
made a draw.
11 g4!? Ne7 12 Nd4! Bd7
12 ... Nd5 13 Nxe6 fxe4 14 Ne4 is good for White.
13 f4 h5
Striking at the kingside pawn phalanx and activating his h8-rook.
14 f5 hxg4 15 e6!?
 Polite formalities are dispensed from the get-go. Nepo’s last move
declares: “I can do whatever I want!”

Question: Isn’t White risking overextension?

Answer: Intuition is the basis from which miracles are performed. In real life, whenever there is a
battle between public good and corporate profit, the latter holds sway since it has an almost unlimited
ability to spend. Not so on the chessboard. I would rather have White’s initiative over Black’s extra
money. White’s nasty offensive ideas have come a long way since the mellow Berlin Wall days of the
Kasparov-Kramnik match. Sure, he risks overextension, yet it’s a sensible risk for two reasons:
1. After the d7-bishop is sent home to c8, Black lags dangerously behind
in development.
2. White’s last move opened attacking lines against the black king stuck in
the middle.

Question: Isn’t opening the position also helping Black, who owns the
bishop pair?
 Answer: Not when the bishops lie dormant on their original squares.
The unadorned version 15 hxg4 is also worth a thought; e.g. 15 ... Rh4 16
Rf4 Nd5 17 Nxd5 cxd5 18 NI3 Rh8 19 Be3 c5 20 c3 and I slightly prefer
White, S.Karjakin-S.Mamedyarov, FIDE World Cup, Baku 2015.
.
15.. fxe6
This natural move opens the f-file for White’s rooks. 15 ... Bc8 is a serious
option, yet after 16 exf7+ Kxf7 17 hxg4 Rh4 18 Rf4 Nd5 19 Nxd5 cxd5 20
NI3 Rh8 21 Kg2 Bd6 22 Rd4 g6! 23 Bf4 gxf5 24 Bxd6 cxd6 25 g5 Black still
stands worse.
16 fxe6 Bc8 17Bg5’’

It requires an astoundingly high level of attacking intuition to perceive that

White isn’t just giving away pawns without compensation. Nepo correctly
gauges that his attack is at least equal to Black’s material. Instead, 17 hxg4
Rh4 18 Rf4 Nd5 19 Nxd5 cxd5 20 g5 Rxf4 21 Bxf4 Bc5 22 c3 Bxd4+ 23
cxd4 Bxe6 is a near certain draw.
 17 ... gxh3
Having learned from a lifetime of sales offers: If it’s “free”, then it’s either
worthless, or there is some irritating catch. After 17 ... Rxh3? 18 Rael Ng6
19 Ne4 Black can barely move.
18 Kh2 Ng6?
Black may yet survive with 18 ... b6! 19 Ne4 a6! (intending ... c6-c5
without allowing Nb5) 20 Rf7! Bxe6! 21 Rxf8+ Kxf8 22 Nxe6+ Kf7 23 Nxc7
Ra7 24 Rfl+ Nf5! 25 Rxf5+ Kg6 26 Rf4 Rxc7 27 Bh4 Rd7 28 Kxh3 Rdl and
Black should probably hold the draw.
19 Ne4 Rh5
If 19 ... Bd6+ 20 Nxd6+ cxd6 21 Rf7 Rh5, then 22 Rgl Bxe6 23 Rxg7 Bf7
24 Rfl! Rxg5 25 Rgxf7! Rb8 26 Rh7 Re5 27 Rf6! Nf8 28 Rxf8+! Kxf8 29
Rh8+ Kf7 30 Rxb8 and Black won’t survive.
20 Rgl! a6?!
Black doesn’t really have time for this, especially as it doesn’t achieve its
aim. He should try 20 ... c5 21 Nb5 Bxe6!, though after 22 Nxc7+ Kd7 23
Nxa8 Bd5 24 Rael b5 25 Bel Ne5 26 Rgfl (covering the fork threat on 13),
Black’s long-term survival is unlikely. Certainly not 20 ... Be7??, which loses
quickly to 21 Bxe7 Nxe7 22 Rxg7 Rh6 23 Rfl (threatening Rff7) 23 ... Bxe6
24Rf6!.
21 Radi!
Threatening Nb5! anyway and mate on d8 or a fork on c7.
21 ... Ne7
Not 21 ... c5? 22 Nb5! Bxe6 23 Nxc7+ etc, or 21 ... Be7? 22 Bxe7 Nxe7
23 Rxg7 Rh6 24 Rfl Bxe6 25 Rf6! and wins.
22 Rdfl! c5?
He had to try 22 ... Rh8 23 Rf7 Bxe6 24 Rxf8+ Kxf8 25 Nxe6+ Kf7 26
Nxc7 Raf8 27 Nd6+ Kg8 28 Bxe7 RI2+ 29 Khl Rh6, although I doubt Black
can hold the game even here.
 Exercise (combination alert): Do you see White’s shot?

Answer: Annihilation of defensive barrier.

23 Rxf8+! 1-0
The exchange sacrifice allows White’s pieces to swoop in: 23 ... Kxf8 24
Bxe7+ Kxe7 (24 ... Ke8 25 Rxg7 and 24 ... Kg8 25 Nf6+ are equally
hopeless) 25 Rxg7+ Kf8 26 Rf7+ Kg8 27 Nf6+ Kh8 28 Nxh5 cxd4 29 Nf6
Bxe6 30 Rh7 mate.

Game 5 7
Ma.Carlsen-I.Nepomniachtchi
Carlsen Invitational (rapid) 2021
Queen’s Gambit Declined

1 d4 Nf6 2 c4 e6 3 Nf3 d5 4 Nc3 dxc4 5 e4

 This turns it into the Vienna Variation of the Queen’s Gambit Declined. 5
e3 steers the game into Queen’s Gambit Accepted territory after 5 ... a6 6
Bxc4 b5 7 Bd3 Bb7 8 0-0 c5.
5 ... b5
This variation is only for those who are unafraid of the dark. 5 ... Bb4 is
Black’s main line.
6 e5
You can train your parrot to say “I love you”, but you can’t train it to mean
it. Have you ever played a line where you keep getting promising positions
and then always manage to lose? Well, this line is that situation for Magnus,
who at the time of this writing has a dismal 0-3, including two losses to
Nakamura in their epic match in the Carlsen Tour Final. After failing to get
anywhere with the immediate 6 Nxb5, Magnus now returns to 6 e5 and loses
for the third time.
The other game went 6 Nxb5 Nxe4 7 Qa4 c6 8 Nc3 Nxc3 9 bxc3 Ba6 10
Bxc4 Bxc4 11 Qxc4 Bd6 12 0-0 0-0 13 Rbl Qc7 14 Qd3 Nd7 15 c4 c5 and
Black looked okay, Ma.Carlsen-I.Nepomniachtchi, Carlsen Invitational
(rapid) 2021. Magnus later spent 32 moves probing for a blunder in a queen
and h-pawn vs. queen ending, until finally conceding a draw on move 100.
6 ... Nd5 7 Nxb5
This is a strange sort of Slav, Geller Gambit, except without the gambit or
Slav part of the equation.

Question: Shouldn’t Black be worried with such a wobbly structure?

Answer: The potential future loss of Black’s c4-pawn is an occupational hazard in this line. He
should be a little worried but not overly so, since even if the pawn does drop off, Black has the d5-
square, bishop pair, and light square control as compensation.
7 ... Nb6
 8 Be2
Magnus varies from his Nakamura games, in both of which he pushed his
a-pawn, once to the third rank and once to the fourth.
a) 8 a3 (preventing a future ... Nc6-b4-d5, as occurs in the main game) 8 ...
Nc6 9 Be3 Na5 10 Qc2 a6 11 Nc3 Be7 12 Rdl Bb7 13 Be2 Qd7,
Ma.Carlsen-H.Nakamura, Carlsen Tour Final (rapid) 2020, and now I prefer
White after 14 0-0 0-0 15 Nd2.
b) 8 a4 (threatening a4-a5, followed by Bxc4) 8 ... Qd7 (preventing a4-a5,
since the b5-knight would drop) 9 Be2 Nc6 10 0-0 Na5 11 Ng5! h6 12 Ne4
Bb7 13 Nc5! Bxc5 14 dxc5 Nd5 15 Ra3 a6 16 Nd4 Ne7 17 Qd2 Qd5 18 Rg3
Qxc5, Ma.Carlsen-H.Nakamura, Carlsen Tour Final (rapid) 2020, and here I
prefer White after 19 Qf4.
8 ... Nc6 9 0-0 Be7 10 Qd2
The idea is (sometimes) to follow with Rdl and Qf4, transferring the
queen to the kingside. 10 Be3 is played slightly more often.
 .
10.. Bb7 11 Rdl Qd7

Question: I take it Nepo plans to castle queenside. Isn’t the black king safer
on the kingside?

Answer: Kingside short is certainly possible but I don’t believe Black’s king is safer there, since
White’s e5-pawn and potential for piece massing mean a greater chance of attack on that wing than on
the queenside, which is flooded with black defenders.
12 Nc3
12 Qf4 looks more consistent with his 10th move.
12 ... Nb4
Preparing a transfer to d5.
13 a3 N4d5 14 Ne4 Qc6 15 Rel
We are in new territory. 15 Qc2 has been played before and since; e.g. 15
... Nb4 16 axb4 Qxe4 17 Qxe4 Bxe4 18 b5 c6 19 bxc6 Bxc6 20 Nd2 Bd5 21
Nfl 0-0 22 Ne3 Rfc8 23 Bd2 and I slightly prefer White’s position, Ding
Liren-R.Praggnanandhaa, Ding vs. Challengers (blitz) 2021.
.
15.. 0-0-0
It isn’t so easy to deliver mate to Black’s king, who is surrounded by four
defenders.
16 Bfl f5’?
 The default assumption of the natural optimist is that only good can arise
from change. I have never understood the mystery of perception, where two
insanely strong players can look at the same position and react with opposing
evaluations. Magnus seems to like White’s positions which arise from this
line (as do I) due to Black’s loose pawns. But I bet Nepo was happy here as
well, since his pieces are active and he has an open attacking lane on the g-
file.
17 exf6!

Question: Doesn’t he help Black by opening the g-file for an attack?

Answer: The open file does benefit Black, while the loosening of Black’s central pawns helps
White. I prefer Magnus’ chances at this point. Instead, 17 Nc3? g5! is dangerous for White.
17 ... gxf6 18 Qe2
Triple purpose:
1. Getting the queen off the d-file.
 2. Opening the cl-bishop’s diagonal.
3. Threatening to put heat on the e6-pawn with either Nc5 or Nc3.
18 ... Kb8 19 Bd2
He wants to play Rcl, eyeing the black queen on c6. Another option is 19
Nc5 Bxc5 20 dxc5 Qxc5 21 Qxe6 Rhg8 22 Bd2; whereas 19 Nc3?! Nxc3 20
bxc3 Rhg8 looks better for Black.
19 ... Rhg8 20 Rael Bd6 21 g3! f5 22 Nxd6’?
I don’t much like the idea of straightening out Black’s once deformed
pawns, although the engine likes Magnus’ move just fine. I would go for 22
Nc5 Bxc5 23 Ne5! (zwischenzug) 23 ... Qd6 24 dxc5 Qxc5 25 Nxc4 Nxc4 26
Rxc4 with the bishop pair and a superior structure.
22 ... cxd6

Exercise (critical decision): Should White play 23 Qxe6 or should he refrain?

Answer: At first sight the e6-pawn looks poisoned. Deeper analysis shows that White can and
should take it.
23 Bg2
This is a natural reaction, yet not the best.
Answer:Neither side may have realized that 23 Qxe6! is playable, since 23 ... Nf4? fails to 24 d5!
(this zwischenzug messes up Black’s plan to deliver mate on hl) 24 ... Nbxd5 25 Nd4! Nxe6 26 Nxc6+
Bxc6 27 Rxe6 Nb6 28 Bxc4 Nxc4 29 Rxc4 Bd5 30 Rb4+ Kc7 31 Re7+ Rd7 32 Re3, when White is a
pawn up and Black must defend four separate isolanis. I don’t think the opposite-coloured bishops will
save him.
23 ... Rde8 24 Nh4!

Question: A knight on the rim ... ?

Answer: ... is dim. But not here. This strategically brilliant move puts on display Magnus’
inhumanly deep understanding of the position, for the following reasons:
1. It pretty much stops Black’s future threats on the hl-a8 diagonal.
2. Black’s d5-knight is now pinned, which means White threatens b2-b3.
3. With a knight on h4 it becomes a lot harder for Black to play ... e6-e5,
since d4xe5 d6xe5 hangs a pawn to Nxf5, so Black would have to recapture
with the rook, weakening his structure.
24 ... Qb5 25 Rbl
Again b2-b3 is a possibility.
25 ... Qb3 26 Bc3?’
26 Bh6! keeps White’s advantage intact, since it prevents Nepo’s next
move shot.
26 ... Nf4!
 Engine-like tactical alertness is a Nepo trait. This is no more than a
temporary piece sacrifice which weakens White’s structure. Amazingly,
Black only equalizes with it, rather than wins as it first appears to do.
27 gxf4
Forced. 27 Qdl? Nxg2 would leave White suffering on the light squares
for the remainder of the game.
27 ... Rg4!
This is the point. Black regains the sacrificed piece since 28 Nf3?? loses to
the simple 28 ... Reg8.
28 h3! Rxh4 29 Bxb7 Kxb7 30 QB+ Ka6
 31 d5?
Now we see a distorted reflexion of Carlsen’s original intent. Magnus
grows tired of his bishop’s role as second-class citizen and sacrifices a pawn
to activate it. This is a case of right idea, wrong move order.

Question: Isn’t White just dead lost, with or without this move?

Answer: I thought so too and White’s repulsive structure seems to confirm it. However, the
overlord engines say otherwise. Magnus should stay calm and play 31 Kh2! Nd5 (I’m the first to admit
that there isn’t much glamour going on for White, but just wait for his next move; if 31 ... Qb5 32 b3!
Qd5 33 Qg3 Rh6 34 bxc4 Nxc4 35 Qd3 Rg8 36 13, covering the mate threat on g2, then 36 ... Qc6 37
d5! Qxd5 38 Qxd5 exd5 39 Re7 Rhg6 40 Bd4 Rg2+ 41 Khl Nb6 42 Rxh7 and White is fine, according
to the engines) 32 Rxe6! (now is the correct moment to crack open his piggy bank) 32 ... Rxe6 33 Qxd5
Reh6 34 Qc6+ Qb6 35 Qc8+ Qb7 36 Qxc4+ Kb6 37 Qb3+ Kc6 38 Qc4+ isn’t what Magnus wanted a
few moves ago; on the other hand, he can’t complain either. Black’s king must return to b6 and take
perpetual check, since 38 ... Kd7?? loses to 39 QI7+ Kc6 40 Qe8+ Kb6 (or 40 ... Kd5 41 Qg8+) 41
Qd8+ Kc6 (or 41 ... Ka6?? 42 Qa5 mate) 42 d5+! Kxd5 43 Rdl+ Kc6 44 Rd3! with a decisive attack.
31 ... Nxd5 32 Rxe6
 32 Qg3 Rh6 doesn’t look so appealing for White either.
32 ... Rxe6 33 Qxd5 Rg6+
Do you see the problem? In the other line White’s king was on h2 so ...
Rg6 doesn’t come with check. In this version White doesn’t get time to play
Qc6+.
34 Kfl Qb5! 35 Rdl
Or 35 QI3 Qb7! and White must swap queens anyway.
35 ... Qxd5 36 Rxd5 Rxf4

White is down too much material; even being one of the greatest endgame
players of all time can’t save Magnus when Nepo is given such a huge head
start.
37 Bd4 Rf3 38 Be3 f4 39 Rf5 Re6!
40 Bd4
Since 40 Bxf4? Re4 wins the bishop, while 40 Rxf4? Rexe3 leaves Black a
rook up.
 40 ... Re4! 41 Kg2 Rd3 42 Bc3 d5
Once the passed d-pawn rolls, it’s all over.
43 Rf6+ Kb5 44 Rf7 d4 45 Bb4 Re2 0-1

Game 58
Ma.Carlsen-I.Nepomniachtchi
Carlsen Invitational (rapid) 2021
Queen’s Gambit Declined

1 d4 Nf6 2 c4 e6 3 Nc3 d5 4 cxd5 exd5 5 Bg5 Bb4

We reach a Ragozin Variation of the Queen’s Gambit Declined.

6 Qa4+

Question: Isn’t White losing time with this check?

Answer: There is a plus side too. White lures the b8-knight to c6 where it is misplaced and hinders
both ... c7-c5 and ... c7-c6.
The main move is 6 e3, which Magnus tried four games later: 6 ... h6 7
Bh4 g5 8 Bg3 Ne4 9 Nge2 h5 10 h3 (10 h4! is also played) 10 ... Nxg3 11
Nxg3 h4 12 Nh5!? (novelty) 12 ... c6 13 Be2 Nd7 14 Qb3 Bxc3+ 15 Qxc3
Nf6 16 Nxf6+ Qxf6 17 b4 with even chances, Ma.Carlsen-I.Nepomniachtchi,
Carlsen Invitational (rapid) 2021. Nepo went on to win this one as well.
6 ... Nc6 7 e3 h6 8 Bxf6
Magnus is in a mellow mood and prefers the calm, medicated version to 8
Bh4 g5 9 Bg3 Ne4 10 Bb5 Bd7 11 Nge2 h5 12 13 Nxg3 13 hxg3.
8 ... Qxf6 9 Nge2 0-0 10 a3
A small novelty. After 10 Nf4 Rd8 11 Rcl Qd6 (threatening ... Nxd4) 12
a3 Bxc3+ 13 Rxc3 I slightly prefer White, V.Veremechik-B.Okner,
correspondence 1988.
10 ... Bxc3+ 11 Nxc3 Ne7 12 Be2 c6 13 0-0 Nf5 14 b4 Nh4

Black’s chances to pester White on the kingside balance out White’s

attempted queenside minority attack.
15Khl!
 Not yet 15 b5?, as that walks into Black’s trap: 15 ... Bh3! 16 gxh3 Qg6+
17 Bg4 f5 and if 18 Qdl h5 19 13 fxg4 20 hxg4 Rf6 with a very strong attack.
15 ... Qg6 16 g3
Reigning world champions tend to see threats of mate in one.
16 ... Nf5 17 b5 h5!
The engine says even, but I like Black a shade better since White’s threats
on the queenside feel far off when juxtaposed with the immediacy of Black’s
kingside attack.
18 bxc6 bxc6 19 Rgl
Discouraging ... h5-h4. After 19 Rael Nh6 20 h4 Re8 21 Kg2 Bf5 22 B13
Rab8 (intending ... Rb2 and ... Rxe3) 23 Rfel Bd3 (making room for ... Nf5)
24 Qdl! Ng4 25 Na4, chances remain balanced.
.
19.. h4!?
Nepo isn’t discouraged!

Question: Should Black be playing this when White’s rook is already on the g-fde?

Answer: It does seem risky, but that is Nepo’s nature. Objectively, he should perhaps hold back the
push back and go for 19 ... Nh6 20 Rael Ng4 21 Ndl Bd7 22 Qa5 Rfe8, when I still slightly prefer
Black’s chances.
20 gxh4! Qf6 21 h5 c5?
This move looks so natural; after all White would be crazy to play Nxd5,
allowing ... Bb7, pinning the knight on the long diagonal.
 22 Nxd5!
Nobody goes 125 classical time control games unbeaten, without being
one of the greatest defenders of all time. This move is yet another example of
Magnus’ inhuman defensive intuition.
22 ... Bb7 23 Bf3
Threatening Nxf6+. The engine comes up with a stronger, impossible-to-
see variation: 23 Qd7!! Bc6 24 BfB! Bxd5 25 Qxd5 cxd4 26 Be4 Nh4 27
exd4! Rad8 28 Qe5 Rd6 29 Rg4 Nf3 30 Qg3 Nd2 31 Rgl! Nxe4 32 Rxe4
Rxd4 33 Qxg7+! (removal of the guard) 33 ... Qxg7 34 Rxd4 Qxgl+ 35
Kxgl and White should convert, despite the fact that one of his extra pawns
is doubled.
23 ... Bxd5 24 Bxd5 Rad8 25 Qc6! Qxc6 26 Bxc6 cxd4 27 Rg5
If 27 e4 then 27 ... Rd6 28 Bd5 Rc8 also looks okay for Black.
27 ... Ne7 28 Bf3 dxe3 29 fxe3 Rd3 30 Re5 Nc8
 Question: Does White have any winning chances?

Answer: It’s a long shot, but yes, for two reasons:

1. Magnus may have treated his pawn structure with cruel indifference, but
at least he got payment for it in the form of an extra pawn.
2. White’s bishop has the potential to be the superior minor piece.
3. White has a plan of pushing his a-pawn down the board and later
winning Black’s a-pawn. Of course this is a long shot as well.
From this position we can only imagine two results: either the game ends
in a draw or Magnus wins. The shocker is how Nepo actually won an inferior
ending against perhaps the greatest endgame player of all time.
31 a4 Nd6 32 Rcl Rb8 33 Kg2!
Magnus hurries to get his king out of the comer before it is cut off by ...
Rb2 or... Rd2.
33 ... Rbb3 34 Rc6 Kf8
 The engine likes the inhuman, decentralizing move 34 ... Kh7!?.
35 a5
His best shot was 35 Rc7! Rxe3 36 Rxe3 Rxe3 37 Rxa7 Ra3. Magnus
probably concluded that White doesn’t have enough to win here.
35 ... f6! 36 Re6 Nf5! 37 h6’?

Dreams must be based in at least a sliver of reality. You may dream of

building a spaceship and visiting planets outside our solar system; but if you
have no training in science and just $37.50 in your bank account, your dream
is unlikely to come true. Magnus returns the pawn with the long-shot hope to
go after Nepo’s king.
I remember with a shudder of distaste a rapid game played in 2002, when I
was rated around 2550 and a pawn up in a rook ending against an 1890
player. To my surprise he defended perfectly as I played on and on, refusing
to cede the draw. I kept taking risks, one more dangerous than the last, and
ended up losing. The lesson: follow the position’s dictates. Your will alone
isn’t enough to alter reality.
Magnus’ play in this game reminded me of that nightmare loss, since he
loses an essentially unloseable position.
37 ... Nxh6 38 Rc8+?’
Principle: In endings, don’t give check to the enemy king if doing so
improves the opponent’s king position. Now White is the one fighting for the
draw. Correct was 38 Bh5 (threatening Re8 mate) 38 ... Rb8 39 Bg6 Rd5 40
h4 Rxa5 41 Rc7 Re5 42 Rxe5 fxe5 43 Rxa7 Rb6 44 h5 Rf6 and Black will
hold.
38 ... Kf7 39 Ra6?
It was Jimmy Buffett who ominously warned that it’s our own damned
fault. White’s position goes from difficult to lost. He can still save the game
with 39 Rec6! Rd2+ 40 Kfl Kg6! (after 40 ... Rxe3 41 Bh5+ g6 42 R8c7+,
White won’t lose) 41 Rc3 Rbl+ 42 Rcl Rbb2 43 Rlc2!, and if 43 ... Rbxc2
44 Be4+ f5 45 Rxc2 Rxc2 46 Bxc2 Ng4 47 Ke2 Nxh2 48 e4! f4 49 e5+ Kg5
50 Be4 Ng4 51 e6 Kf6 52 Kf3 Ne5+ 53 Kxf4 Kxe6 54 a6 will draw.
39 ... Rd2+ 40 Kg3
Not 40 Kfl?? Rbl+ and mate in two. This was the problem for White
when he moved his rook to a6, rather than c6. He denied himself a defence
with Rcl.
40 ... Nf5+ 41 Kf4
 It’s difficult to imagine that Carlsen was once on the cusp of a winning
position. Unfortunately “once” no longer exists.

Exercise (planning): Come up with a way for Black to win material.

Answer: Step 1: Go after the bishop with ... Rf2! and ... Nh4.
41 ... Rf2! 42 Rxa7+?’
Even after 42 Kxf5 Rxf3+ 43 Kg4 Rbxe3! 44 Rxa7+ Kg6 White problems
look fatal, given his weak h-pawn, his a7-rook out of the game, and his king
in a potential mating net.
42 ... Kg6 43 Rcc7 Kh6!
Step 2: Move the king to h6, “hanging” the knight, so that Rxg7 doesn’t
come with check.
44 Kxf5 Rxe3 45 Rc5
Or 45 Rxg7? Rfxf3+ 46 Kg4 f5+ 47 Kh4 Rh3 mate, which is one of the
amazing points behind 43 ... Kh6.
 .45.. g6+!
Even stronger than taking the bishop straight away.
46 Kg4 Rfxf3 0-1
White’s position drowns in death. 47 Rcc7 is met by 47 ... f5+ 48 Kh4 g5
mate or 48 ... Rh3 mate; while 47 Ra6 only evades mate at the cost of a rook
after 47 ... f5+ 48 Rxf5 Rxf5.

We now return at last to the Candidates tournament. The next game below
was the third after the resumption and was a must-win for Nepo, who had the
white pieces against the lowest rated participant. Since when is a 2698-rated
player considered the tournament’s easy-to-beat cupcake!? GM Kiril
Alekseenko, the tournament’s wildcard entry, performed quite credibly,
scoring 5/2/14 and actually gained rating points. He also defeated Anish Giri,
thereby essentially ruining the Dutch GM’s chances to be Magnus’
challenger.

Game 59
N epomniachtchi-K.Alekseenko
I.
Candidates Tournament, Yekaterinburg 2021
English Opening

1 c4 Nf6 2 g3 e6 3 Bg2 d5 4 M3 dxc4 5 Qa4+ Nbd7 6 Qxc4 a6 7 Qc2!

Question: Why is the queen running away before being chased by ... b7-b5 - ?

Answer: By moving away early, White actually prevents ... b7-b5, since it no longer comes with
tempo.
7 ... c5
Note that 7 ... b5?? is a blunder which costs Black material after 8 Nd4.
8 Nc3 Be7
This move doesn’t score well for Black, logging in a dismal 14.9%. The
main line in 8 ... Qc7, after which Black’s score moves back into the normal
range.
9 0-0 0-0 10 d4
 .10.. cxd4?!
A plan partially formed, only partially correct, is not one you want to
invest in. After this second inaccuracy, Black is unable to complete queenside
development easily. This game resembles early Catalans from the 1950s and
60s where Black just couldn’t figure out how to develop the queenside pieces
harmoniously.
Black keeps his disadvantage to a minimum with 10 ... b5! (a 100%
effective cure is never to contract the disease in the first place) 11 Ne5 Nxe5
12 dxe5 (12 Bxa8?! cxd4 is at least even for Black) 12 ... Nd5, as in
G.Oparin-V.Vetoshko, Titled Tuesday (blitz) 2020. Even after 13 Nxd5 exd5
14 Be3! Be6 15 Bxc5 Rc8 16 b4 Bxc5 17 bxc5 Rc6 18 Rael Qe7 19 Rfdl
Rfc8 Black stands only an iota worse.
11 Nxd4 Qc7
The immediate 11 ... Nb6 is more accurate.
12 Rdl Rd8
 A not-so-great novelty. Here too 12 ... Nb6 looks preferable, though White
keeps the advantage after 13 Be3 (or 13 Bg5 h6, but not 13 ... e5? 14 Nf5
Bxf5 15 Qxf5 Rad8 16 Bxf6 1-0 L.Galego-M.De Souza, Rio de Janeiro 1999,
in view of 16 ... Bxf6 17 Be4, winning a piece) 13 ... Ng4 14 Bd2 Rd8 15
Bel!?Nf6 16 Rael Bd7 17NM!.
13 Be3 Nb6
Black is definitely struggling. After 13 ... Ne5 14 Na4 Qxc2 15 Nxc2 Ned7
16 Rael White is clearly better; while 13 ... Ng4? is met by 14 Nd5!
(discovered attack) 14 ... Qxc2 15 Nxe7+ Kf8 16 Nxc2 Nxe3 17 Nxe3 Kxe7
18 Rael with a difficult position for Black, who just can’t unravel.
14 Rael e5!

Conquering your fear doesn’t mean you automatically overcome death.

This desperate attempt to cleanse the strategic wound may be Black’s best try
in a difficult situation. 14 ... Bd7 15 Qb3! is terrible for Black, who is still
unable to unravel.
 15 Nf5 Bxf5

Question: Black’s last move gave up the bishop pair in an open position, while
on the previous move he weakened his central light squares by pushing
the e-pawn to the fifth rank. Didn’t he have anything better?

Answer: Not really, no. Just to give you a glimpse of Black’s troubles, let’s consider the natural 15
... Bf8. White has the shocking riposte 16 Nxg7!! Kxg7 (if 16 ... Bxg7 then 17 Ne4! Rxdl+ 18 Qxdl
Qd7 19 Bxb6 wins) 17 Qb3 Nc4 (or 17 ... Nbd7 18 Nd5 Qa5 19 Bd2 Qb5 20 Qe3! Nxd5 21 Qg5+ Kh8
22 Qxd8 and wins) 18 Rxd8 Nxe3 19 Rxf8 Be6 20 Qxb7 Qxb7 21 Bxb7 Rxf8 22 fxe3 Rb8 23 Bxa6
Rxb2 24 a4, emerging with two extra pawns.
16 Qxf5 Nc4
Or 16 ... g6 17 Qbl Rxdl+ 18 Nxdl Qd8 19 Bxb7 and Black is busted.
17Bg5
Now Bxf6, followed by Nd5, is in the air.
17 ... Rxdl+ 18 Nxdl Rd8
If 18 ... Qd8 19 Bxf6 Bxf6 20 Bxb7 Nd6 21 Qf3 Nxb7 22 Qxb7, Black
won’t save the game.
 Exercise (combination alert): To describe Black’s position as awful, would
be excessively complimentary. How did White force the win of a pawn,
while maintaining a decisive bind?

Answer: Step 1: Start with a swap on f6, which weakens both h7 and d5.
19 Bxf6! Bxf6 20 Be4!
Step 2: Transfer the bishop to e4, taking aim at h7.
20 ... Qa5
If 20 ... h6? 21 Qh7+ Kf8 22 Ne3 b5 23 Nf5 (threatening Qh8 mate) 23 ...
g6 24 Nxh6, there is no remedy to the coming Qg8+ and Qxf7+.
21 Nc3! Kf8
Nothing else is any better:
a) 21 ... h6 22 Nd5 Rxd5 23 Bxd5 Qxd5 24 Qc8+ Kh7 25 Qxc4 wins.
b) 21 ... Nxb2 22 Qxh7+ Kf8 23 Nd5! (threatening mate on the move) 23
... Rxd5 24 Bxd5 Qxd5 25 Qh8+ Ke7 26 Rc7+ Kd6 27 Qb8! e4 28 Rxb7+
Kc5 (or 28 ... Ke6 29 Qc8+ Ke5 30 Rxb2) 29 Rc7+ Kd6 30 Qb6+ Ke5 31
Rc5 and White wins.
22 Nd5! b5
If 22 ... Rxd5 then 23 Bxd5 Qxd5 24 Qc8+ picks up the loose c4-knight,
winning.
23 Qxh7
Threatening mate in one.
23 ... Rxd5
Or 23 ... Ke8 24 Qg8+ Kd7 25 Qxf7+ Kc8 26 Nxf6 and wins.
24 Bxd5 Qd2 25 Rxc4! bxc4 26 e4!

Black’s king can’t survive the attack by the enemy queen and opposite­
coloured bishop. 26 Qh8+! Ke7 27 Qa8 is also winning for White.
26 ... Qxb2 27 Qh8+ Ke7 28 Qc8 Qb6
Or 28 ... c3 29 Qc7+ Ke8 30 Bc6+ Kf8 31 Qc8+ Ke7 32 Qd7+ Kf8 33 Qe8
mate.
 29 Qxc4 Qb5 30 Qc7+ Qd7 31 Qc5+ 1-0
Since 31 ... Qd6 32 Qa7+ Kd8 33 Qxf7 wins easily.

I keep hearing from ever-so-wise social media pundits about how dull
positions are Nepo’s weakness. So we’ll look at the driest game in the book,
just to examine Nepo’s play when he is outside of his dynamic element.
Chess players are divided into two groups: 1. Legitimists. 2. Anarchists. In
this game we see Nepo the anarchist outplay Wang Hao the legitimist, in a
position completely devoid of tactics or complications, showing that Nepo is
more than capable of playing such positions if they are forced upon him.

Game 60
Wang Hao-I.Nepomniachtchi
Candidates Tournament, Yekaterinburg 2021
PetroffDefence

1 e4 e5 2 Nf3 Nf6

Question: Why didn’t Nepomniachtchi play his signature Najdorf?

Answer: Nepo was leading the tournament with three rounds left to play and might have been
content to draw with the black pieces, to which end the Petroff lends itself very well. On the other hand,
it could have been a psychological ploy. Having lost their fifth round game on the Black side of a
Petroff (see Game 42), Wang Hao unexpectedly finds himself in the same opening with the white
pieces, almost as if Nepo is taunting him.
3 Nxe5 d6 4 Nf3 Nxe4 5 d3
Perhaps worried about preparation against his usual, more dynamic 5 d4 or
5 Nc3, Wang Hao opts for a quieter, non-theoretical line.
5 ... Nf6 6 d4 d5
We can verify from the DNA that, rather than being a Petroff, the position
is actually a genetic mutation of the Exchange French. Specifically, the
extremely dull version 1 e4 e6 2 d4 d5 3 exd5 exd5 4 Nf3 Nf6.

Question: Why is White playing so unambitiously?

Answer: Perhaps because he was disturbed by the choice of opening; and perhaps because
Alexander Grischuk had caused him a few problems with this line earlier in the tournament.
7 Bd3 Bd6 8 Qe2+
 White continues by offering an exchange of queens. The symmetrical
option is 8 0-0 0-0 9 Bg5 Bg4 10 Nbd2 Nbd7 11 c3 c6 12 Qc2 Qc7 13 Rfel
Rfe8, when the position, as you may have guessed, is both sterile and equal.
. Qe7
8..
Nepo is happy for the queens to leave the board. Thirteen months earlier 8
... Be6 9 Ng5 Qe7 10 Nxe6 fxe6 11 c3 c5 12 Nd2 Nc6 13 Nf3 0-0 14 0-0
cxd4 15 cxd4 e5 16 dxe5 Nxe5 17 Nxe5 Qxe5 18 Qxe5 Bxe5 19 f3 had given
White an endgame pull in A.Grischuk-Wang Hao, Yekaterinburg Candidates
2020, due to the two bishops and isolated d5-pawn, though Wang made a
draw without too much trouble.
9 Qxe7+ Kxe7
Damn, this is super-exciting chess!
10 0-0 Nc6 11 c3
Preventing ... Nb4. White more often plays 11 Rel+ first, though it hardly
matters.
 11 ... h6 12 Nh4
Aiming either to prompt a weakness with ... g7-g6 or get Black to hand
over the bishop pair. A few days later the game T.Danada-M.Pacher,
Budapest 2021, ended thematically 12 Rel+ Be6 13 Nbd2 1/2-1/2.
12 ... Re8 13 Nf5+ Bxf5 14 Bxf5

Question: Is the loss of the bishop pair such a tragedy for Black?

Answer: Actually not. My normal strategy on the Black side of an Exchange French is to find a way
to create an imbalance with knight versus bishop and then hope to win the ending in 100 moves.
14 ... Kf8 15 g3 Ne7 16 Bh3
There’s nothing wrong with this, as long as White is careful not to leave
the bishop on the sidelines. Instead, 16 Bd3 can be met by the boring 16 ... c6
or the more enterprising 16 ... c5.
.
16.. Nc8
The knight may later move to d6, its optimal post.
17 Nd2 a5!
 Nepo begins to expand on both sides of the board.
18 a4 c6 19 Rdl
White is a bit tangled, so he clears the fl-square for his knight.
19 ... h5 20 Nfl g6
Principle: Place your pawns on the opposite colour as your remaining
bishop.
21 13
This is not a mistake either, but White certainly has reasons to regret it
later in the game.
21 ... Nb6 22 b3
To keep the black knight out of c4.
22 ... Kg7 23 Kf2 Nbd7

Question: This knight has had an uneventful tour, moving to c6, e7, c8, b6,
and now d7. Maybe it would have been better to play ... Nd7 in one move?
 Answer: Time operates at a leisurely pace in rigid structures. Nimzowitsch called such manoeuvres
“tacking”. The knight has done its job on b6 and now returns to the centre. It later goes on to f8, e6, c7,
e8, d6 and f5.
24 Bg5 Nf8 25 Rel
White finally challenges the e-file, and Black blocks it with his knight.
25 ... Ne6 26 Be3 Rac8 27 Bd2 c5!

Nepo grows tired of putzing around. Black’s activity easily makes up for
the potential weakness. This reminds me of the sixth game in the Carlsen-
Caruana match, where Fabio outplayed the world champion from a similar
Petroff position, boldly taking on an isolani, just as we see here, only later to
miss a study-like win.
28 dxc5?!
Opening up helps Black’s cause. 28 Ne3 cxd4 (or 28 ... Rc7 29 Bxe6 Rxe6
30 Nc2) 29 cxd4 b6 30 Bxe6 fxe6 31 Rael would have been much more
solid.
28 ... Bxc5+ 29 Kg2 Bb6 30 Rabi Rc6 31 b4 Nc7!
 Offering up the b7-pawn, since Nepo sees he will regain the material on c3
and can then look to hassle the white king, which is slightly exposed due to
earlier f2-f3.
32 Rxe8 Ncxe8 33 bxa5 Bxa5 34 Rxb7 Nd6 35 Ra7 Bxc3 36 Bd7’?
Wang Hao returns the now offside bishop for Nepo’s knight in order to get
his rook to an optimal spot on d7.
.
36.. Nxd7
Another option was 36 ... Rc4 37 Bxc3 Rxc3 38 Kf2 Nc4 39 Be6 Ne5 40
Ne3 Kh6 41 Bxd5 (not 41 Bxf7?? d4 and wins) 41 ... Nxd5 42 Nxd5 Rxf3+
43 Ke2 Ra3 44 a5 Ra2+ 45 Kfl Rxh2, but the chances of Black converting
the extra pawn are negligible.
37 Bxc3+ Rxc3 38 Rxd7 Rc6

Normally an outside passed pawn would suggest an edge. Not here, since
Black’s rook will be annoying on White’s second rank.
39 Re7?’
 I don’t know whether Wang was in time trouble or if he simply
underestimated Black’s chances. Either 39 Nd2 Kf6 40 Nb3 or 39 Ne3 d4
(not 39 ... Kf6? due to 40 Rc7! and Nxd5+) 40 Nd5 Rc2+ 41 Kh3 Nc4 42 Nf4
looks like a better defence, although Nepo could continue pressing here with
42 ... Ra2 43 Ne6+ Kf6 44 Nxd4 g5!? and White is not completely out of the
woods.
39 ... Rc2+ 40 Kgl d4
Possibly 40 ... Nf5! was more accurate; e.g. 41 Rd7 Nd4 42 Rxd5 Nxf3+
43 Khl h4 44 gxh4 Nxh4 45 Rd3 Ra2 46 Kgl Rxa4. Normally two versus
one on the same side is a no-brainer draw. Here, however, the remaining
minor pieces are knights which give Black some chances, especially since
White’s king position is still insecure.
41 Rd7 Nf5 42 a5
Not 42 g4?? Nh4 43 gxh5 Nxf3+ 44 Khl gxh5 45 a5 Ra2 46 Rd5 h4 47
Rf5 Rf2 48 h3 d3! 49 a6 Rxfl+ 50 Kg2 d2 and Black wins.
42 ... Ra2 43 Ra7?’
Wang Hao, mesmerized by a shiny and useless object, his a-pawn, fails to
sense the danger. Putting the rook here looks very unappealing. White should
hold the draw if he gives up the pawn to activate his passive knight with 43
a6! Rxa6 44 Nd2!.
 White now deals with a backlog of defensive issues:
1. His rook is awkwardly placed in front of the passed a-pawn, where it
has trouble participating.
2. White’s knight, now demoted to the status of babysitter, is passively
placed on fl, unable to move.
3. White’s king is in some danger, especially if Black’s knight can
manoeuvre to a spot where it attacks f3. Black can also push the h-pawn
forward.
4. Black’s king is mobile and able to participate in the action, whereas
White’s is cut off on gl, destined to serve out his life sentence on the first
rank.
Magnus would be deadly with Black’s position and I would give him 80%
chances to win as Black against any player in the world. Let’s see how
Nepo’s technique compares.
43 ... Kf6?’
 With a faulty idea in mind. The immediate 43 ... h4! was better: 44 g4
(after 44 a6? h3 45 Ra8 Rg2+ 46 Khl Ne3 Black wins quickly) 44 ... Nd6 45
Rd7 Nb5 46 g5 Nc3! 47 a6 Rxa6 48 Nd2 (not 48 Rxd4?? Ne2+) 48 ... Kf8 49
Kf2 Ra4 and Black has serious winning chances.
44 a6 h4
Another possibility is 44 ... Nd6!, intending ... Nc4.
45 Ra8 h3
Now White’s back rank is weak, but not as fatally as after 43 ... h4 44 a5
h3 above.
46 a7 Ke7
At this point Nepo noticed that his intended 46 ... Rg2+ 47 Khl Ne3!?
fails to 48 Rh8! Ndl 49 Rxh3 Ra2 50 Rh4 d3 51 Rd4 Rxa7 52 Kgl and
White survives, so he had to play something else.
47 g4
White might also take the chance to activate the knight with 47 Nd2, given
that 47 ... Rxd2? allows 48 Re8+ Kxe8 49 a8Q+.
47 ... Nd6 48 Rb8! Rxa7 49 Rb4! d3 50 Rb3 Ra2 51 Rxd3
 Wang Hao defended pretty well and should hold the draw, even if his
position remains unpleasant due to the weaknesses at f3 and h2.
51 ... Nc4 52 Ng3 Rg2+ 53 Khl Kf8
Avoiding defensive ideas of Re3+ and Re2.
54 Rc3?’
The rook was better placed on the d-file. After 54 g5 or even 54 Rd8+ Kg7
55 Rd3, White should hold; e.g. 55 ... Kf6 56 Ne4+ Ke5 57 Ng5 Rd2 58
Rxd2 Nxd2 59 Nxh3 Nxf3 60 Kg2, when White’s damage-control measures
have worked and the game is clearly drawn.
54 ... Nb2!
Nepo’s subterfuge is revealed. The threat is ... Ndl and ... Nf2 mate.
55 Rc8+ Kg7 56 Rd8 Rf2
Now the f-pawn falls and White’s task becomes harder.
57Kgl
Not 57 Rb8? Rc2! 58 Kgl Rg2+ 59 Kfl (or 59 Khl? Ndl and mates) 59 ...
Rxh2 and Black wins easily.
57 ... Rxf3 58 Ne4 Re3
It is always possible to blunder. Here 58 ... Nd3? runs into 59 Nd2! Re3 60
Nfl Rf3 61 Nd2 with a draw by repetition.
59 Ng3?
In the movies, love always finds a way. Not so in chess. With this final
mistake, Wang’s hopes are cruelly dashed. After 59 Rd4! Nd3 60 Nd2 Rel+
61 Nfl Nc5 62 KI2 Rai 63 Nd2, White may yet hold the draw, albeit with
difficulty.
59 ... Ra3! 0-1

Wang Hao was clearly demoralized and resigned a touch prematurely. He

might have made Nepo prove the win, but he didn’t want to play on any
more. The position is lost in any case. For example: 60 g5 (or 60 Re8 f6,
fixing the g4-pawn as a target) 60 ... Nd3 61 Kfl Nf4 62 Ne4 Ne6 63 Rd5
Ra2 64 Rd2 Ra5 65 NI2 Nxg5 66 Rd3 Rf5 67 Ke2 Kh6! and Black remains
two pawns up, since 68 Nxh3? Nxh3 69 Rxh3+ Rh5 70 Rxh5+ Kxh5 is a
winning pawn endgame.

Game 61
I.
Nepomniachtchi-T.Radjabov
FTX Crypto Cup (rapid) 2021
Nimzo-Larsen Opening

1 b3
Good man, Nepo. I want to write Opening Repertoire: Nimzo-Larsen
Attack and this game will help sales!
1 ... d5 2 Bb2 Nf6 3 g3
3 e3 and 3 Nf3 are played much more often; whereas 3 Bxf6?! gives away
the bishop pair, dark squares and even strengthens Black’s centre after 3 ...
gxf6.
3 ... Bf5
Radjabov goes for the solid reversed London set-up.
4 Bg2 e6 5 d3
Nepo plays it in reversed King’s Indian versus London fashion.
5 ... Nbd7
Slightly inaccurate since, if White plays for e2-e4-e5, Black’s f6-knight is
denied the d7-square. 5 ... Be7 6 Nd2 h6 7 e4 Bh7 is better.
6 Nd2
 Intending e2-e4, with tempo.
6 ... Bc5?’
I don’t like this move since Black’s light-squared bishop is now sent to the
unfavourable g6-square, after which White can gain more time with a future
h2-h4-h5 or Nh3-f4. Instead, 6 ... h6! 7 e4 Bh7 8 e5 Ng8 is similar to the
game’s continuation without the above drawback.
7 e4 Bg6 8 e5!
White gains two tempi with this push.
. Ng8
8..
He can also try and complicate with 8 ... Bh5!? 9 Qcl Ng4 10 Nh3 Bg6
(the e5-pawn is tactically defended: 10 ... Ngxe5?? 11 d4 wins a piece) 11 0-0
Bf5 12 Nf3 h5 13 d4 Be7 14 c4 c6 15 Ba3, though White stands better here
too.
9h4
Nepo takes advantage of Black’s failure to play ... h7-h6. The threat is h4-
h5 and g3-g4, trapping Black’s light-squared bishop. 9 Nh3! and Nf4 may be
even stronger.
9.. . h6 10Nh3!
Planning Nf4 and Nh5.
10 ... Ne7 11 Nf4 Bh7

Question: White’s knight looks powerful. Can Black consider just allowing a
swap on g6, playing 11 ... c6, followed by ... Qc7 and ... 0-0-0 - ?

Answer: It isn’t so simple. Black’s king isn’t safe on the queenside either; e.g. 12 Nxg6 Nxg6 13 d4
Bb6 14 c4 and there is no way Black should castle long.
12 Nh5?’
This is premature and loses White’s edge. Nepo should play 12 d4 Bb6 13
c4! with a clear advantage.
12.. . 0-0’?
It’s a brave decision to castle into White’s potential attack. 12 ... Nf5!? is a
safer alternative, after which White may have a hard time playing for g3-g4,
since the h4-pawn then hangs.
13 g4!?
 Well, that was predictable. We all knew that Nepo wasn’t going to take the
quieter route with 13 Qg4 Nf5 14 0-0. This intimidating stab is a precursor to
a kingside pawn storm which looks stronger than it actually is, since it:
1. Creates holes, potentially on f4.
2. Weakens the f-file.
3. Even endangers White’s king a touch, due to the loosening of structural
integrity.
Nevertheless, the engine considers the game even.
13 ... Ng6! 14 Nf3
14 g5!? is an attempt to open lines against Black’s king. The trouble is that
it also opens lines against White’s king! After 14 ... hxg5 15 hxg5 Qxg5 16
Bfl f6! 17 Nf3 Bxf2+! 18 Kxf2 fxe5 19 Bel Qf5 20 Bh3 Qf7 the engine still
says it’s equal, but a human would have a terribly difficult time defending
White here, perhaps even finding the first move: 21 Bb2!.
.
14.. f6!
 Principle: Meet the opponent’s wing attack with a central counter. I’m
starting to like Black’s chances.
15 exf6 Nxf6 16 Nxf6+ gxf6?
Believe it or not, this most natural of moves is an error. Black should go
for a promising exchange sacrifice with 16 ... Rxf6! 17 g5! (not yet 17 Bxf6?
Qxf6, as I don’t believe White can survive the punctured dark square
weaknesses) 17 ... hxg5 18 Bxf6! (now is the correct moment to accept) 18 ...
Qxf6 19 hxg5 Qc3+ 20 Kfl Nf4 with obvious compensation for the
exchange. The engine again calls it even, whereas I would love to play Black.
17 Qd2!
Not only is ... Nf4 halted, Black has no good way to protect h6.
.
17.. e5
He might try to muck it up with 17 ... Qd6!? 18 Qxh6 Qf4 19 Bel! Qxg4
20 Rgl Ne5! 21 Nh2 Qg6 22 h5 Qxh6 23 Bxh6 Kf7! 24 Bxf8 Rxf8, when it
won’t be easy to mobilize White’s forces, so Black can fight on, despite
being down an exchange.
Note that 17 ... Kg7? is suicidal: 18 h5 Ne7 19 g5! Nf5 20 Bh3 hxg5 21
Nxg5 Qe7 22 0-0-0 d4 23 Nxh7 Kxh7 24 Rdgl and Black is busted, since 24
... Rg8 25 Bxf5+ exf5 26 Rg6! gives White a winning attack.
18 Qxh6?’
Nepo throws away his advantage which he could have sealed with 18 d4!
Nf4 19 0-0-0! Nxg2 20 dxc5 Be4 21 Nel Nf4 22 f3 Bg6 23 Ng2!
(eliminating Black’s best piece) 23 ... Nxg2 24 Qxg2, when White has a
strong attack, enhanced by:
1. The presence of opposite-coloured bishops.
2. The potential break with g4-g5.
3. The option to try and create light square weakness in Black’s camp with
a future f3-f4.
18 ... Nf4
Threatening to take the g2-bishop with check.
19 g5!
 Which Nepo ignores!
19 ... Rf7!
Radjabov finds a strong defensive move. Nepo may have expected 19 ...
Nxg2+? 20 Kfl and if 20 ... Nf4 then 21 Rgl with a winning attack for
White.
20 gxf6! Nxg2+ 21 Kfl Qxf6!
Not 21 ... Nf4?? 22 Rgl + Kh8 23 Nxe5 and Black can resign.
22 Qxf6 Rxf6 23 Kxg2 Raf8 24 Rh3
The only move, but good enough. Attack and defence are in perfect
balance. Black is compensated for his missing pawn with the following
strategic pluses:
1. The bishop pair in an open position.
2. White is potentially weak on the light squares.
3. Black exerts tremendous pressure on 12, with ... e5-e4 in the air.
24 ... d4?
 Black is overly worried about his hanging e5-pawn and is willing to
smother his dark-squared bishop’s power to ensure its safety.
Pushing the e-pawn is unfavourable too: 24 ... e4? 25 Bxf6 Rxf6 26 Rg3+
Kf8 27 Ng5 Rxf2+ 28 Khl (not 28 Kh3?? Bf5+ 29 Rg4 Bd6! 30 Rhl e3 and
White is paralysed; the engine calls mate in 11 for Black) 28 ... Bf5 29 dxe4
dxe4 30 Rel e3 31 h5 with winning chances for White, since the passed h-
pawn is dangerous.
Radjabov could probably save the game with 24 ... Rg6+! 25 Khl! e4 26
h5 Rg4 27 Ne5 Rg5 28 Nd7! Bf5! 29 Nf6+! Kf7! 30 Rg3 Rxg3 31 fxg3 Ke6
32 g4 Rxf6! 33 Bxf6 Bxg4 34 Bg5 Bxh5 and I don’t believe Black can lose.
25 Rg3+ Kh8 26 Rfl
Now Nxe5 is threatened, but this is a slightly inferior way to protect f2.
White holds a larger advantage after 26 Rel! Bd6 27 Re2!.
26 ... Bd6 27 Ng5
Preventing ... e5-e4.
27 ... c5?’
Here 27 ... Bb4!, intending ... Bd2, keeps White’s edge to a minimum.
28 Nxh7 Kxh7 29 Rg4!
 While Black’s position may not yet be lost, it is certainly in the vicinity.
White is a pawn up, with the superior minor piece, and has decent winning
chances.
.
29.. b5!
The ... c5-c4 push is Black’s best shot at counterplay.
30 Bel
The bishop was hitting a wall on b2.
30 ... c4 31 Bg5?’
White should return the pawn to unravel with 31 f4! exf4 32 Kf3! Re 8 33
Rfgl, which leaves Black struggling.
31 ... Rf3 32 Rg3 R3f7
Now White is tied down to defence of f2, at least until Black’s next move.
33 Rh3 Rc8?
Natural, yet incorrect since it allows White to challenge for control of the
f-file. Correct was 33 ... cxd3 34 cxd3 a5!, when Black should have enough
counterplay to hold the draw.
34 Rdl
White should ignore the infiltration at c2 and play 34 Rf3!, since if 34 ...
Rxf3 35 Kxf3 cxd3 36 cxd3 Rc2 then 37 Rcl! Rxa2 38 h5! Bf8 39 Rc7+ Kg8
40 h6 Ra6 41 h7+ Kh8 42 Rc8 wins a piece.
34.. . cxd3?’
It was probably better to admit his mistake and return the rook with 34 ...
Rcf8, and never mind that 35 Rd2 has improved White’s rook somewhat.
35 cxd3 Rc2 36 Rd2
No problem. White’s rook challenges Black’s control over the second
rank.
36.. . Rxd2?’
Why swap an active rook for a passive one? 36 ... Rcl 37 Rf3 Kg7 puts up
greater resistance.
37 Bxd2 Rc7
 Now ... Rc2 appears to be a dangerous threat. The reality is that Nepo has
seen deeper into the position and realizes he can ignore Black’s queenside
play and launch an endgame attack/pawn promotion attempt with rook,
bishop and h-pawn.
38 RB! Rc2 39 Bg5 Rxa2 40 Rf7+! Kg8
A necessary concession, since 40 ... Kg6?? hangs a piece to 41 Rf6+. The
problem is that Black’s king is also in dire trouble on the back rank.
41 Rd7 Bb4 42 h5
There is no good defence to the push of the h-pawn.
42 ... Bel
After 42 ... Bd2 43 Bf6 Bf4 44 Kf3 a5 45 Rg7+ Kf8 46 Ra7 a4 47 bxa4
Rxa4 48 Rb7 b4 49 Bg7+ Kg8 50 h6 (threatening 51 h7+ Kxh7 52 Bxe5+)
Black loses a piece, no matter what he plays.
43 Bh4!

Black’s counterplay just went dead.

 43 ... a5
If 43 ... Bd2 44 KI3! (White’s king joins the attack) 44 ... a5 45 Kg4 a4 46
bxa4 bxa4 47 Kf5 a3 then 48 Kg6 Bh6!? 49 Rd8+ Bf8 50 h6 forces mate in
three.
44 h6 1-0
If white rook and bishop are the undertakers, then the h-pawn is the hearse. There is no reasonable
remedy to the threat of h6-h7+ and Bf6 mate.
Chapter Four
What Would Nepo Do?
Are you ready to take the Nepo exam? This chapter is mostly a tactics quiz from Nepomniachtchi’s
games, although sometimes you will be asked to find the correct plan or make a critical decision. Nepo
will always be on the move and you take his side. When writing the book I gave this exam to some
students, giving them five minutes per problem, and discovered that it isn’t easy even to achieve a plus
score, so take your time with each one.

1. H.Jonkman-I.Nepomniachtchi
Wijk aan Zee 2007

Work out Black’s forced mate in five.

Solution

2. P.Anisimov-I.Nepomniachtchi
Krasnoyarsk 2007
 Both the d2- and f2- are covered for now. How did Nepomniachtchi overcome these obstacles and
force the win?

Solution

3.1.Nepomniachtchi-G.Sargissian
Wijk aan Zee 2008
 How does White regain the sacrificed material with interest?

Solution

Nepomniachtchi-B.Grachev
4.1.
Aeroflot Open, Moscow 2008
 Nepo found one crushing move which forced a winning position.

Solution

5.1.
Nepomniachtchi-I.Salgado Lopez
Pamplona 2008
 Prove that Black’s king is far more vulnerable than White’s.

Solution

6. H.Melkumyan-I.Nepomniachtchi
Aeroflot Open, Moscow 2009
 What is the optimal way to go after White’s king?

Solution

7.1.
Nepomniachtchi-H.Melkumyan
European Championship, Rijeka 2010
 White has several ways to the win. Find the cleanest.

Solution

8.1.
Nepomniachtchi-K.Sasikiran
Khanty-Mansiysk Olympiad 2010
 Simply 52 bxa4+ should win. Do you see something far stronger?

Solution

9.1.
Nepomniachtchi-M.Karttunen
European Cup, Plovdiv 2010
 How did White win material?

Solution

10. D.Khismatullin-I.Nepomniachtchi
Russian Championship, Moscow 2010
 A powerful idea leads to White’s strategic collapse.

Solution

11.1.
Nepomniachtchi-D.Bocharov
Russian Rapid Grand Prix, Apatity 2011
 30 Nxf6+!! works. Your job is to calculate the lines to White’s forced win.

Solution

Nepomniachtchi-D.Shengelia
12.I.
European Championship, Legnica 2013
 Find one powerful idea and Black’s foundation crumbles.

Solution

Nepomniachtchi-N.Vitiugov
13.1.
Russian Championship, Nizhny Novgorod 2013
 Do you see Nepo’s knockout punch?

Solution

14.1.
Nepomniachtchi-S.Sjugirov
Russian Team Championship, Sochi 2015
 Nepo found a killing shot which overwhelmed the defence.

Solution

15. D.Svetushkin-I.Nepomniachtchi
Baku Olympiad 2016
 White’s position isn’t as well protected as it looks. How would you
proceed with Black?

Solution

Nepomniachtchi-V.Kramnik
16.1.
Tai Memorial, Moscow 2016
 Nepo found a brilliant, short-circuiting combination.

Solution

17.1.
Nepomniachtchi-B.Gelfand
Zürich (rapid) 2017
 What is White’s winning plan?

Solution

18.1.Nepomniachtchi-J.Geller
Russian Team Championship, Sochi 2017
 Should White sacrifice on f5? If so, with which piece?

Solution

19. E.Alekseev-I.Nepomniachtchi
Russian Team Championship, Sochi 2017
 Black appears tied down to defence of f7. How should he continue?

Solution

20. Iv.Popov-I.Nepomniachtchi
European Cup, Antalya 2017
 White’s back rank doesn’t just look weak. It is! Prove how.

Solution

21. J.K.Duda-I.Nepomniachtchi
World Rapid Championship, Riyadh 2017
 How should Black continue his attack?

Solution

22.1.
Nepomniachtchi-Ba.Jobava
World Rapid Championship, Riyadh 2017
 The formidable-looking barrier in front of Black’s king is an illusion. How
did Nepo force mate?

Solution

Nepomniachtchi-A.Rakhmanov
23.1.
World Rapid Championship, Riyadh 2017
 It’s obvious that White is winning. But can you find the optimal finish of
mate in three?

Solution

24.I.
Nepomniachtchi-Wang Hao
World Rapid Championship, Riyadh 2017
 White can win material here. How?

Solution

Nepomniachtchi-A.Korobov
25.1.
Poikovsky Karpov 2018
 White forces mate in mate. Try it without moving the pieces, as a
calculation exercise.

Solution

26.I.
Nepomniachtchi-V.Bologan
Poikovsky Karpov 2018
 What is White’s best prosecution of the attack?

Solution

Nepomniachtchi-V.Ivanchuk
27.1.
Jerusalem (rapid) 2018
 Demonstrate why Black’s king is in grave danger.

Solution

Nepomniachtchi-L.D.Nisipeanu
28.1.
Dortmund 2018
 One deceptively quiet move places Black in zugzwang.

Solution

Nepomniachtchi-D.Khismatullin
29.1.
Russian Championship, Satka 2018
 What is White’s winning plan?

Solution

30. S.Zhigalko-I.Nepomniachtchi
Batumi Olympiad 2018
 Black’s big fear is perpetual check. Find a clean line which eliminates any
counterplay.

Solution

31.1.
Nepomniachtchi-E.Najer
Dvoretsky Memorial, Moscow (rapid) 2018
 How did Nepo successfully launch a mating attack against the black king?

Solution

32. G.Sargissian-I.Nepomniachtchi
World Rapid Championship, St. Petersburg 2018
 How can Black exploit the deeply entrenched b3-pawn and White’s
insecure king?

Solution

33.1.
Nepomniachtchi-Iv.Saric
World Rapid Championship, St. Petersburg 2018
 Saric has just attempted to seize the initiative with 21 ... g5. Show why this
was an error.

Solution

34.1.
Nepomniachtchi-S.Vidit
Wijk aan Zee 2019
 What is White’s winning plan?

Solution

35.1.Nepomniachtchi-V.Fedoseev
Wijk aan Zee 2019
 Prove why Nepo’s passed b-pawn is more dangerous than Black’s passed
d-pawn.

Solution

36.1.
Nepomniachtchi-B.Amin
Grand Chess Tour, Abidjan (rapid) 2019
 Nepo found a deadly sequence to end the game.

Solution

Nepomniachtchi-L.D.Nisipeanu
37.1.
Dortmund 2019
 Nepo found a poisonously quiet move which forced immediate
resignation.

Solution

38. S.Ganerdene-I.Nepomniachtchi
FIDE World Cup, Khanty-Mansiysk 2019
 How did Nepo force a winning position?

Solution

39.1.Nepomniachtchi-E.Postny
World Rapid Ch., Moscow 2019
 Black’s bad bishop is not his only problem. One strong move forced his
resignation.

Solution

40.1.
Nepomniachtchi-Ding Liren
Carlsen Invitational (rapid) 2020
 Ding has just moved his rook from f6 to take a pawn on f5. Show why this
was a huge blunder.

Solution

41. M.Vachier Lagrave-I.Nepomniachtchi

Chessable Masters (rapid) 2020
 What is Black’s winning plan?

Solution

42.1.Nepomniachtchi-V.Artemiev
Chessable Masters (rapid) 2020
 White’s h4-knight is attacked. Should he retreat to f3 or is the sacrifice on
f5 sound?

Solution

43. V.Kramnik-I.Nepomniachtchi
Legends of Chess (rapid) 2020
 In this position Nepo forced mate in six.

Solution

44. P.Svidler-I.Nepomniachtchi
Legends of Chess (rapid) 2020
 Black to play and win.

Solution

45. A.Giri-I.Nepomniachtchi
Legends of Chess (rapid) 2020
 Without moving the pieces, calculate Black’s forced mate in seven.

Solution

46. P.Svidler-I.Nepomniachtchi
Skilling Open (rapid) 2020
 White appears to have excellent drawing chances. Prove why this
assessment is incorrect.

Solution

Nepomniachtchi-M.Matlakov
47.1.
Russian Championship, Moscow 2020
 Black’s attack looks dangerous with ... Be5 in the air. Fortunately for
Nepo, it’s his move.

Solution

48. M.Antipov-I.Nepomniachtchi
Russian Championship, Moscow 2020
 Find the correct idea and White is mated.

Solution

49. A.Goganov-I.Nepomniachtchi
Russian Championship, Moscow 2020
 Black’s c-pawn looks lost. What is Nepo’s strongest continuation?

Solution

50. J.K.Duda-I.Nepomniachtchi
Opera Euro Rapid 2021
 Black has two separate mates in three. You need only find one of them.

Solution

51.1.
Nepomniachtchi-A.Grischuk
Opera Euro Rapid 2021
 White’s only chance to win is to pick off the h6-pawn. How can this be
accomplished?

Solution

52. W.So-I.Nepomniachtchi
Carlsen Invitational (rapid) 2021
 Find a path where Black forces resignation.

Solution
 Chapter Five
Solutions
1. Jonkman-Nepomniachtchi

Step 1: Toss the h-pawn forward, after which Black’s pieces grow
vindictive.
47 ... h5+! 48 Kxh5
Or 48 Kg5 Qe5+ 49 Qf5 Qxf5 mate.
.
48.. Qe5+
Step 2: Give check on e5.
49 Kg4 f5+! 0-1
Step 3: Add the f-pawn into the attack. The finish would be 50 Kh5 (or 50 Kh3 Qh2 mate) 50 ...
f4+! (Step 4: Clearance) 51 Kg4 Qf5 mate.

2. Anisimov-Nepomniachtchi
 Queen sacrifice/attraction/discovered attack/pawn promotion.
93 ... Qxfl+! 0-1
White resigned in view of 94 Kxfl d2+ 95 Kg2 Bc6+! 96 Kgl dlR mate.

3. Nepomniachtchi-Sargissian
 Queen sacrifice.
25 Qxg5!
The majority of my students incorrectly chose 25 Bxg5? Rxg5! 26 f6 Rg8
27 Qh8 Bh3+! (not 27 ... Kf8?? 28 Ke2 Bg4+ 29 Kd2 dxe4 30 Ragl Qxd3+
31 Kcl and White wins; e.g. 31 ... Qd7 32 Qh7 Bd8 33 Rxg4! Rxg4 34 Qh6+
Ke8 35 Qh8+ Rg8 36 Qxg8 mate) 28 Rxh3 Rxh8 29 Rxh8+ Kd7 30 Rxd8+
Bxd8 31 exd5 c5. Black should hold the draw.
25 ... Rxg5
25 ... Qxg5 26 Bxg5 is much the same, since 26 ... Rxg5? 27 Rh8+ is even
worse.
26 Rh8+ Ke7 27 Bxg5+ f6 28 Rxd8 Rxd8 29 Be3
Nepo had bagged a pawn and converted efficiently.

4. Nepomniachtchi-Grachev
 Move the queen to e5, after which Black’s king isn’t so safe in his twilight
corner.
35 Qe5!
Not 35 Rb8? Qd5 36 Qf2 e5 and Black is fine.
.
35.. Qcl+
After 35 ... Qc8 36 Rb8! or 35 ... Rg7 36 Rxe6+ Black must hand over his
queen or get mated; e.g. 36 ... Kh7 37 Qxf5+ Kh8 38 Qxh5+ Rh7 39 Qe8+
Kg7 40 Rg6 mate.
36 Kh2 Rxg2+
This desperate attempt at perpetual check fails. But otherwise 36 ... Qc8 37
Rb8 wins the queen again.
37 Kxg2 Qd2+ 38 Kfl Qd3+
Or 38 ... Qcl+ 39 Ke2 Qc2+ 40 Ke3 Qcl+ 41 Kd4 Qgl+ 42 Qe3 Qg7+ 43
Kd3 Qd7+ 44 Qd4 Black’s checks ran out.
39 Kf2 Qd2+ 40 Qe2
 Nepo sacrifices a pawn to clear the cl-h6 diagonal. 40 Kf3 Qdl+ 41 Qe2
Qhl+ 42 Qg2 Qdl+ 43 Kg3 Qel+ 44 Kh2 was another solution.
40 ... Qxf4+ 41 Kg2 Kg5
41 ... Qg5+ 42 Kf3! leaves Black checkless.
42 Rxe6 h4 43 Qe3 Kh5

44 Qf3+! 1-0
I hope nobody fell for 44 Qxf4?? stalemate.

5. Nepomniachtchi-Salgado Lopez
 It’s mate in four, starting with a check on c6.
44 Nc6+ Kb7 45 Qxc8+! 1-0
Queen sacrifice/weak back rank. Black is mated after 45 ... Kxc8 46 Re8+
Kd7 47 Rd8 or 46 ... Kb7 47 Rb8.

6. Melkumyan-Nepomniachtchi
 Weak back rank.
42 ... Rf2! 43 Re4
Nothing saves White:
a) 43 Rxf2 gxf2+ 44 Kxf2 (or 44 Khl Qh5+ etc) 44 ... Qel mate.
b) 43 d6 Rxfl+ 44 Qxfl Qe3+ 45 Khl Qh6+ 46 Kgl Qh2 mate.
c) 43 Rdl Qel+ 44 Rxel Rxel+ etc.
d) 43 Qdl Rxfl+ 44 Kxfl (or 44 Qxfl Qe3+ 45 Khl Qh6+ etc) 44 ... Qh5!
45 Rxg3 Qhl+ picks off White’s queen.
43 ... Qb2! 0-1
It’s mate in four as g2 can’t be defended.

7. Nepomniachtchi-Melkumyan
 Overloaded defenders.
27 Ne4!
This is the simplest and strongest line.
The engine also found 27 Rael Rd6 28 Ne4! Nxe4 29 Qxf7+ Kh7 30 Rxe4
Nd8 31 Qf4 Qxd5 32 Re5! Qg8 33 Bxd7 and wins, since 33 ... Rxd7 34 Qf5+
forks king and rook.
27 ... Nxe4
Black is lost no matter which knight he takes or how:
a) 27 ... Rxe4 28 Nxf6+ gxf6 29 Qxe4.
b) 27 ... Nxd5 28 Qxf7+ Kh8 29 Qxd7 Rxe4 30 Qxd5.
28 Qxf7+ Kh8 29 Qxd7 Rd6 30 Bxc6
The clearer path. 30 Qg4 also wins.
30 ... Rxc6
30 ... Rxd7 31 Bxb7 Rxb7 32 Nxb4 leaves White two pawns up.
31 Qe7 Rd6 32 Rf8+ Rxf8 33 Qxf8+ Kh7
 How do you force Black’s resignation?

Step 1: Move the knight to e7, threatening mate on g8.

34 Ne7! Nf6 35 Rfl! 1-0
Step 2: Removal of the guard. After 35 Rfl! there is no remedy to the
coming sacrifice on f6, removing Black’s knight which defends the g8-
mating square; e.g. 35 ... h5 36 Rxf6 Rxf6 37 Qg8+ Kh6 38 Qh8+ Kg5 39
Qxg7+ Kf4 40 Qxf6+ wins.

8. Nepomniachtchi-Sasikiran
 Overloaded defender.
52 Rc8! Ne2+
If Black takes the rook then 52 ... Rxc8 53 Rxb6+ Ka5 54 b4 is mate;
while 52 ... Rb7 53 bxa4+ Kxa4 54 Rxg8 leaves White a rook up.
53 Kh2 Rxc8
Walking into mate after all. Black has no good moves anyway.
54 Rxb6+ 1-0
In view of 54 ... Ka5 55 b4 mate again.

9. Nepomniachtchi-Karttunen
 Trapped piece.
24 Be5! Bd6
24 ... Qe7 25 b4 f6 26 bxa5 fxe5 27 Ne4 exf4 28 Nxc5 Qxc5 29 Qxf4
leaves White a full exchange up.
25 Bxd6 Qxd6 26 b4!
A rook on the rim is dim! Black’s rook ran out of safe squares.
26 ... Rd5 27 Nxd5 exd5 28 Radi
Nobody can give Nepo a clean exchange and expect to survive.
28 ... a5 29 Bc4! axb4
29 ... Be6 30 f5 wins.
30 Bxd5
Threatening Bxf7+.
30 ... Qc5 31 axb4 Qxb4 32 f5!
Preventing any of... Be6, ... Ne6 and ... Ng6.
32 ... Ra3?
 Show why Black’s last move was an error.

33 Bxf7+!
Annihilation of defensive barrier.
33 ... Kxf7
33 ... Kh8 34 Bb3, threatening both Qd5 and f5-f6, leaves Black with no
chance of survival.
34 Qd5+ 1-0
Since 34 ... Kf6 35 Qd8+ Qe7 36 Qxc8 wins easily.

10. Khismatullin-Nepomniachtchi
 Strategic exchange sacrifice.
26 ... Rxa4’’27f5
The sacrifice comes with a proviso after 27 Qxa4. Black’s queen must now
move to the single correct square: 27 ... Qe3! 28 Rfl (or 28 g4 Qg3+ 29 Khl
Nxf4 30 Rd2 Nxg2 31 Rxg2 Qxh3+ 32 Kgl Bxg4 and White has no chances
to survive) 28 ... Qxg3+ 29 Kgl Bxh3 30 Qc2 Nxf4 31 Rxf4 Bxg2 and White
is busted. Several of my students found the first move of the combination but
then failed to follow through correctly, choosing 27 ... QÍ2?, which fails to 28
Rd3! (covering g3) 28 ... Qxb2 29 Radi and White should convert with the
extra exchange.
27 ... Rc4 28 Qe2
 What is Black’s strongest continuation?
.
28.. Bd5! 0-1
Exploiting the pin on the e4-pawn. Black wins after 29 Rfl Qb8!
(targeting g3) 30 Qf2 Bxe4 or 30 Qd3 Bxe4!

11. Nepomniachtchi-Bocharov
 This is a detail-specific exercise.
30 Nxf6+’’
“Superb calculation for a rapid game,” writes IM John Bartholomew.
30 ... Kf8
If 30 ... gxf6 then 31 Qh8+ Kf7 32 Rh7+ Kg6 33 Rh6+! Kg5 (or 33 ... Kf7
34 Rxf6 mate) 34 Qxf6+ Kg4 35 Rh4 mate.
31 Rf4!
It was important to see this move when you started the combination.
31 ... Qcl+ 32 Kg2 gxf6 33 Rxf6+ Kg8 34 Rg6+! Kf8
Or 34 ... Rg7 35 Rxg7+ Kxg7 36 Nxe6+ Kf6 37 Nxd8, when 37 ... Qd2+
38 Kh3 Qxd8 gets skewered by 39 Qh4+.
35 Qh8+ Kf7 36 Rf6 mate
 Did you see this position in your mind’s eye back at move 30? Not many
of my students succeeded.

12. Nepomniachtchi-Shengelia
 Queen sacrifice/pawn promotion.
31 Qxb8+! Qxb8 32 a7 Qa8 33 Rb2 1-0
There is no remedy to the coming transfer of the rook to b8.

13. Nepomniachtchi-Vitiugov
 Weak back rank.
37 Qg2!
Threatening Qa8+.
.
37.. Re8
Nothing else saves Black either:
a) 37 ... Bxdl 38 Qa8+ is mate in three.
b) 37 ... Qc6 38 Qxc6 dxc6 39 Rd8+ wins the house.
c) 37 ... Qb8 38 Qd5+ Kf8 39 Qd4 (threatening Qh8 mate) 39 ... Qe5 40
Qf2+! Ke8 41 Bxe5 wins the queen.
d) 37 ... Kf8 38 Qa8+ Re8 39 Nxh7+ Ke7 40 Bf6+! Qxf6 41 Rxd7+! Kxd7
(or 41 ... Ke6 42 Qd5 mate) 42 Nxf6+ wins.
e) 37 ... Qa7 38 Rd4 Bc5 39 Qd5+ Kf8 (39 ... Re6 40 Rf4 also forces
mate) 40 Rf4+ Ke8 41 Qg8 mate.
38 Qd5+ 1-0
In view of 38 ... Re6 (or 38 ... Kf8 39 Qf7 mate) 39 Qxd7 (threatening Qg7
mate) 39 ... Rf6 40 Qxh7+ Kf8 41 Qh8+ Ke7 42 Bxf6+ Qxf6 43 Qd8 mate.
The engine gives 38 Rxd7, threatening Rg7+ or Qd5+, as even faster.

14. Nepomniachtchi-Sjugirov

Queen sacrifice/overloaded defenders/weak back rank.

35 Qxd6! 1-0
Since all captures lead to mate:
a) 35 ... Bxd6 36 fxg7 mate.
b) 35 ... Rdc8 36 Qxf8+! (removal of the guard) 36 ... Rxf8 37 fxg7 mate.
c) 35 ... Rxd6 36 fxg7+ Bxg7 37 Rf8 mate.

15. Svetushkin-Nepomniachtchi
Step 1: Transfer the queen to el, attacking f2.
42 ... Qel! 43 Nf3 Bxh3+!
Step 2: Decoy.
44 Kxh3 Qxfl+ 45 Kh2 Bxf2 46 Qb2 Be3! 0-1
 White’s position is hopeless, since the e5-pawn is taboo:
a) 47 Nxe5 Bgl+ 48 Khl Bf2+ 49 Kh2 Qgl+ 50 Kh3 Qxg3 mate.
b) 47 Qxe5+ Kh7 48 Qf6 (or 48 Nh4 Bgl+ etc) 48 ... Bgl+! 49 Khl (or 49
Nxgl Qxf6) 49 ... Bd4+ wins the queen and mates.
c) 47 Qg2 Qdl! 48 Nxe5 Qh5+ wins the knight.

16. Nepomniachtchi-Kramnik
 Overloaded defenders.
38 e3+!!
Not 38 f6? Bh4 39 Nf5 (and especially not 39 17?? Ne3+ 40 Kh2 Bg3+ 41
Kgl Rcl mate) 39 ... Bxf6 40 Rf7 Kxe4 41 Ng3+ Kd4 42 Rxf6 Ne3+ and
Black is the one with all the winning chances.
.
38.. Kxe3
Instead, 38 ... Nxe3+ hangs a piece to 39 Kxf2; or if 38 ... Bxe3 then 39 f6
Ra5 40 17 Rxa2+ 41 Kfl Rf2+ 42 Kel Kg3 43 Nf5+! wins.
39 Rxg4
Black’s dangerous passed pawn is eliminated, whereas White’s f-pawn is
ready to move forward.
39 ... Kd2
Threatening ... Ne3+.
40 Bf3!
Ignoring Black’s threat. Simply 40 f6 also wins.
 40 ... Ne3+ 41 Kxf2 Nxg4+ 42 Bxg4 Rd5 43 Ne4+ Kd3 44 f6!
White’s knight is tactically protected.
44 ... Ra5
Unfortunately for Black, after 44 ... Kxe4 45 Bf3+ Kf5 46 Bxd5 Kxf6
White’s bishop is the correct colour to promote the rook’s pawn.
45 Be2+!

That annoying knight is held up by invisible forces.

45 ... Kd4
If 45 ... Kxe4 then 46 f7 Ra8 47 Bf3+ skewers the rook and wins.
46 Í7 Ra8 47 Ng5 1-0
Since 47 ... Rf8 gets forked on e6, while 47 ... Kd5 is met by 48 Nh7.

17. Nepomniachtchi-Gelfand
 Pry open the kingside with g3-g4!, after which Gelfand’s position
collapsed quickly.
32 g4! Nd7
32 ... e4 33 fxe4 fxg4 34 e5 wins.
33 gxf5 gxf5
33 ... Rxf5 34 Qe3 Rf6 35 Qg5 threatens the h5-pawn and if 35 ... Kg7
then 36 Rgl Nf8 37 f4 wins.
34 f4! e4 35 Qh3 Kh6 36 Rg2 Bxf4
Or 36 ... Rg8 37 Bdl! Rg4 38 Bxg4 fxg4 39 Rxg4 and wins.
37 Bdl! 1-0
Since 37 ... Nf6 38 Bxf6 only delays the end by one move.

18. Nepomniachtchi-Geller
 Taking with the bishop is the correct path. Black begins to drop pawns, left
and right.
21 Bxf5!
a) 21 Nxf5? gxf5 22 Bxf5 Nd4! 23 Bxd4 cxd4 24 Nxe4 Bb4! 25 Nd2 d3!
leaves White tied up due to his weak f2-pawn and Black’s passed d-pawn.
b) 21 Ne6? Rf7 22 Ng5 Rg7! improves things for Black, who should now
survive after 23 Bxf5 gxf5 24 Nxf5 Nd4!.
21 ... gxf5 22 Nxf5 Bc8
22 ... Rad8?? hangs material to 23 Nxe7+ Nxe7 24 Rxd8 Rxd8 25 Bxf6.
23 Nxe7+ Nxe7 24 Rd6! h6
24 ... Ne8 25 Rb6 Nf5 26 Rxe4 is also in White’s favour.
25 Nxe4 Nxe4 26 Rxe4 Ra7?
26 ... Nf5! was a better try, though 27 Rg6+ Kh7 28 Rc6 Rf7 29 Rf4 Bd7
30 Rxc5 is still good for White.
27 Rxh6
 White has picked up four pawns for the piece, including three connected
passed pawns on the kingside, and still has a strong initiative.
27 ... Bf5 28 Rf4
Threatening the deadly 29 g4, followed by 30 Rh8+ if the bishop moves.
28 ... Raa8 29 h3! Ng6
If 29 ... Bh7 then 30 Rg4+ wins; or 29 ... Rad8 30 Rh8+ Kf7 31 Rxf8+
Rxf8 32 g4, regaining the piece with a decisive advantage in pawns.

This one is easy. White to play and win material.

Answer: Chop the f5-bishop, undermining Black’s knight.

30 Rxf5!
White gets two pieces for the rook.
30 ... Rxf5 31 Rxg6+
Four pawns is way too much for the exchange.
31 ... Kf7 32 Rg7+ Kf8 33 f4 Rd8 34 Kg2 Rd3 35 Rc7 Ke8 36 Be5! Rf7 37 Rc8+ Kd7 38 Rxc5
Make that five pawns for the exchange.
 38 ... Ke6 39 Bc3 1-0

19. Alekseev-Nepomniachtchi

Who cares about f7? Transfer the black rook to g3, the black king to f4,
and White’s king is caught in a mating attack.
38 ... Rg3! 39 Rxf7+ Kg5
Now both the f3-pawn and b3-bishop are under attack.
40 Bdl Rxh3 41 Ra5
Perhaps looking for a desperate exchange sacrifice on e5.
41 ... Rh2+ 42 Kfl
If 42 Kgl then 42 ... Rd2 43 Rd5 (or 43 Rai Re3! with dual threats of...
Rel mate and ... Bxal) 43 ... Rxd5 44 exd5 Rcl wins a piece.
42 ... Rhl+ 43 Ke2 Kf4! 0-1
 Black’s king turns into an attacker. White has no defence: 44 Rd5 (or 44
Rxe5 Rh2+ 45 Kfl dxe5 46 Rd7 Ra3 47 f6 Raa2 48 Kgl h4, intending ... h4-
h3, ... Rag2+ etc, or 49 f7 Rag2+ 50 Kfl Kg3 and mates) 44 ... Re3+ 45 Kd2
(or 45 Kf2 Rh2+ 46 Kgl Kg3! again) 45 ... Rh2+ 46 Kcl Bb2+ 47 Kbl Ra3
mates.

20. Popov-Nepomniachtchi
 Overloaded defender.
32 ... Rxc2!
White’s knight is unable simultaneously to cover c2 and the back rank
mating square on dl.
33 Qe5
If 33 Nxc2? then 33 ... Qdl+ mates as below.
33 ... Bg7 34 Qe8+ Kh7!
No thanks! This is Black’s most accurate reply. White is mated in five
moves at the most.
35 Nxc2
White elects to acquiesce after all. The engine insists on 35 Qg8+! ? Kxg8
36 Re8+ (spite check number two!) 36 ... Kh7! 37 Nc4 (the mean-spirited
engine is so full of spite; here we see a spite give-away!) 37 ... Rxb2+ 38 Kai
Rxa2+ 39 Kbl Rai mate.
.
35.. Qdl+!
 Weak back rank.
36 Rxdl Rxdl mate

21. Duda-Nepomniachtchi

44 ... g5!
Threatening mate on h4.
45 hxg5
Not 45 g3?? Qhl mate. And 45 Kh2 only delays the inevitable; e.g. 45 ...
g4! 46 fxg4 hxg4 47 Qa3 Qxh4+ 48 Kgl g3 49 Qe7+ Kh6 50 Qf8+ Kg5 51
Rxe5+ Kf4! 52 Rf5+ Ke3 53 Rf3+ Ke2 and wins.
45 ... h4!
Zwischenzug, threatening mate on g3. Not 45 ... fxg5? 46 Kh2 and Black
has nothing better than to take perpetual check.
46 Rd7+
46 Kg4 is met by 46 ... Kg6!, when White must hand over his queen to
evade immediate mate.
 46 ... Kg6 47 Kh2 Qg3+ 48 Kgl

What is Black’s winning plan?

Push the h-pawn, threatening mate on g2 and luring White’s rook to d2.
48 ... h3 49 Rd2
White is forced to hand over his rook.
49 ... Qel+ 50 Kh2 Qxd2 51 Kxh3 Kxg5 52 Qa3
He isn’t going to have time to play Qf8 and then deliver perpetual check.
52 ... Qf4
Threatening ... Qh4 mate.
53 g4 Qxf3+
The engine now declares mate in five. Although Nepo didn’t finish quite
so accurately, his opponent nonetheless resigned on move 60.

22. Nepomniachtchi-Jobava
 Step 1: Slide the bishop to c6.
31 Bc6!
Threatening Qxa6+.
31 ... Na7
31 ... bxc6 32 Qxa6+ Na7 comes to the same thing.
32 Qxa6!
Step 2: Exploit the pin of Black’s b-pawn.
32 ... bxc6 33 Nxa7 1-0
Step 3: Swap on a7, destroying the defender of c6. Black resigned in view
of 33 ... Qxa7 34 Rxb8+ Kxb8 35 Rbl+ with mate in three at the most.

23. Nepomniachtchi-Rakhmanov
 Clearance.
42 Qh8+! Ke7 43 Nxg6+! 1-0
Mate follows after 43 ... Nxg6 44 Rxf7 or 43 ... fxg6 44 Qf8.

24. Nepomniachtchi-Wang Hao

 Step 1: Exchange knights on f6.
15 Nxf6!
White can also reverse the move order with 15 Bxh6+ and 16 Nxf6.
15 ... Bxf6 16 Bxh6+!
Step 2: Deflection.
16 ... Kxh6 17 Qxf6
Nepo has won a clear pawn and converted on move 53.

25. Nepomniachtchi-Korobov
 White’s queen takes an MC Escher-like path to the win.
60 Qg8+ Kf6 61 Qf8+! 1-0
The mating path runs 61 ... Ke5 (or 61 ... Kg6 62 Bd3 mate; 61 ... Qf7!?
prolongs the game by two moves) 62 Qf5+ Kd6 (Black’s king has no choice
but to zigzag his way through the boredom and pain, occasionally glancing
up through the rain) 63 Qe6+ Kc5 64 Qd5+ Kb6 65 Qb5 mate.

26. Nepomniachtchi-Bologan
 Principle: When hunting the enemy king, don’t chase him. Instead, cut off
flight squares.
33Rdl!
Maybe Bologan expected 33 Qg7+? Ke7 34 Rdl Red8! 35 Qe5+ Kf8 with
a draw.
33 ... Qxe4
Instead, 33 ... Re6 34 Qh8+ Ke7 35 Qxa8 leaves White a piece up; while
33 ... Qe6 loses control over c5, so that 34 Bb4+ mates.
34 Qg7+
Black’s king is forced to the e-file, after which his queen is pinned and
won.
34 ... Ke7 35 Ret 1-0

27. Nepomniachtchi-Ivanchuk
 Start with a rook check on h3.
52 Rh3+! Kg7
If 52 ... Kg6 then 53 f5+ Kg7 54 Bh6+ Kh7 55 Rd8 leads to mate; e.g. 55
... Rc7 56 Bf8+ Kg8 57 Bc5+ Kg7 58 Rdh8! and there is no stopping R3h7
mate.
53 Rg3+ Kh6
53 ... Kf8 54 Bb4+ forces 54 ... Rfe7, when White wins the exchange and
the game.
54 Rd8!
Despite reduced material on the board, the creep of threat suddenly arises
against Black’s king.
54 ... Kh7
54 ... Rxd2 is similarly met by 55 f5!, threatening Rg6+ followed by Rh8+,
and if 55 ... Rh7 then 56 Rdg8! leads to mate.
55 f5! 1-0
 In view of 55 ... Rxd2 56 Rh3+ Kg7 57 Rdh8! with R3h7 mate to follow.

28. Nepomniachtchi-Nisipeanu
 50 g6! 1-0
After 50 ... Rc7 51 Kd3! Black is in zugzwang. His knight remains pinned,
while any rook move fatally drops a pawn: 51 ... Rcc8 52 Rxa7, 51 ... Rb7 52
Rxc6, or 51 ... Rb8 52 Rxg7, threatening Nxc6! as well.
Alternatively, 50 Ra2! is also deadly: 50 ... Rc7 (or 50 ... Rbb8 51 Rxg7
Rb3+ 52 Kc2) 51 Raf2! and there is no reasonable defence to a rook sacrifice
on f6, forcing mate.

29. Nepomniachtchi-Khismatullin
 Simplification.
50 c7! Bxc7 51 Nxc7 Rxc7 52 Rxc7 Kxc7 53 h4! 1-0
It was crucial in your calculations to find this move. 53 K13?? is too slow
due to 53 ... Kd6 54 Ke4 Ke6, and 53 f4?? gxf4+ 54 Kxf4 Kd6 55 Kg5 Ke6
56 Kxh5 Kf7 is also a trivial draw.
After 53 h4! Black resigned in view of 53 ... g4 (or 53 ... gxh4+ 54 Kxh4
Kd6 55 Kxh5 Ke5 56 Kg5 etc) 54 Kf4 Kd6 55 Kg5 Ke5 56 Kxh5 Kf5 57
Kh6 Kf4 58 Kg6 K13 59 Kf5! and wins.

30. Zhigalko-Nepomniachtchi
 Giving check on el simplifies to an easily won queen ending.
74 ... Bel+! 75 Kdl Qxd3+ 76 Kxel
The tablebases declare mate in 31 at this point. Nepo doesn’t quite match
that accuracy, but he comes very close to it.
76 ... b3 77 Qb4 Ke8 78 h4 Qbl+ 79 Kf2 Qc2+ 80 Ke3 b2
 All that remains is for Black to dodge the checks.
81 Qb5+ Kf8 82 Qb4+ Kf7 83 Qb7+ Kg6 84 h5+ Kh6 85 Qb6 Qcl+ 86
Ke2 Qc2+ 87 Ke3 blQ!
Nepo correctly calculated that White doesn’t achieve perpetual check. 87
... Kxh5 88 Qxf6 Qc5+ 89 Kf3 Qd5+ 90 Kf4 blQ! 91 Qh8+ Kg6 92 Qe8+
Qf7+ was another way to win.
88 Qxf6+ Kxh5 89 Qf3+ Kh4!
 Black’s king will be shielded by his two wives once he reaches the first
rank.
90 Qf4+ Kh3 91 Qh6+
Or 91 Qf3+ Kh2 92 Qf4+ Khl 93 Qh4+ Qh2 and wins.
91 ... Kg2 92 Qg5+ Khl 93 Qd5+ Kgl 94 Qg5+ Qg2 0-1
White’s checks ran out, as did his luck.

31. Nepomniachtchi-Najer
 In such positions the player on the move nearly always wins. We begin
with a pawn check on e5.
43 e5+! Rxe5
43 ... Kxe5 allows mate in five; e.g. 44 Qg7+ f6 45 f4+ Kf5 46 Rxh5+!
gxh5 47 Qh7+ Kg4 48 h3 mate.
44 Qd6+ Re6
If 44 ... Kf5 then 45 Qd7+ Kf6 46 Rd6+ Re6 47 Rxe6+ fxe6 48 Qd4+ e5
49 Qd6+ Kf5 50 Rf8+ Ke4 51 Qxg6+, followed by 52 Rd8+ and 53 Rc8+
wins.
45 Qd4+ Re5
Other moves allow mate in one: 45 ... Ke7 46 Qd8 mate; 45 ... Kf5 46 Qf4
mate; or 45 ... Kg5 46 Qf4 mate.
46 Qf4+! Rf5 47 Qd6+! Kg7
Or 47 ... Kg5 48 f4+ Kg4 49 h3 mate.
48 Qf8+ Kf6 49 Rd6+ 1-0
 In view of 49 ... Ke5 50 Qe7 mate, or 49 ... Kg5 50 Qh6+ Kg4 51 h3 - a
sweet David and Goliath mate.

32. Sargissian-Nepomniachtchi
 Step 1: It’s crucial to move the bishop to a5 before the exchange sacrifice
on a3.
67 ... Ba5!
The immediate 67 ... Rxa3?? loses to 68 bxa3 b2 69 Rbl Kxa3 70 Kc3
Ba5+ 71 Kc2; while
67 ... Bc7? is too slow. 68 Rc3 Ka5 69 Rxb3 Ka6 70 Bb4! Kb7 71 Bc3+
Kc8 72 Ra3 Rxa3 73 bxa3 Kb7 74 a4 Ka6 is drawn.
68 Rbl
Or 68 Rdl Rxa3! 69 bxa3 b2 70 Rbl Kxa3 and wins.
.
68.. Bd2!
Step 2: Zugzwang.
69 Rdl Rxa3! 0-1
Step 3: Exchange sacrifice, which can’t be accepted.
White resigned in view of 70 Rxd2 (or 70 bxa3 b2 71 Rbl Kxa3 etc) 70 ...
Rai (threatening mate with ... Rcl+) 71 d5 Rcl+ (anyway) 72 Kd4 exd5 73
Ke5 (or 73 Re2 Rc2) 73 ... Rc2 74 Rdl Rxf2 with an easy win.

33. Nepomniachtchi-Saric

Interference, exposing Black’s king.

22 Ne6! fxe6
22 ... Bxd5 23 Rxd5 Rdxe6 24 Rxg5+ Kf8 25 Rxc5 leaves Black two
pawns down with a terrible position.
23 Qxf6 exd5
23 ... Qd8 is little better. White is still completely in command after 24
Qb2 exd5 25 Rxc5, a pawn up with a winning position.
24 Qxg5+ Kh8 25 Qh5 Red8 26 Qe5+ Kg8 27 Rd2 Re6 28 Qc3 b4
28 ... c4 29 bxc4 bxc4 30 Redi Redó 31 Rd4 leaves Black paralysed.
 29 Qc2’?
The c5-pawn isn’t running away. White can actually fall for Black’s trap
and give the queen away: 29 Qxc5! Rc8 30 Qxc8+ Bxc8 31 Rxc8+ Kg7 32
Rxd5 with a crushing position.
.
29.. Qd6
29 ... Rc8 loses to the simple 30 Bxd5.
30 Bg4!
This way Black is deprived of... Rc8 when White’s queen chops the c5-
pawn.
30 ... Rg6 31 Qxc5 1-0

34. Nepomniachtchi-Vidit
 37 Ba5! 1-0
The threat of Rb6+ is decisive. For example: 37 ... Ra8 (37 ... Ba8? 38 Rb8
ends the game) 38 Rb6+ Kc7 39 Rf6+ Kc8 40 Rd6! and Black must hand
over his bishop to evade mate.

35. Nepomniachtchi-Fedoseev
 Decoy.
45 Rxd3!
Not 45 Rc7? Rxf6 46 Rxd7+ Kg6 47 b7 Rb4 48 R2xd3 Rfb6 49 Rg3+ Kf6
50 Rgg7 Rb2 and White is too tied down to make progress.
45 ... Rxf6
Not 45 ... Qxd3? 46 Rc7+ and Black is mated in three moves.
46 Rxd7+ Kg6 47 Rbl!
Principle: Place your rooks behind your passed pawns. After 47 b7? Rb4
Black would hold the draw.
.
47.. Rfe6
 We come to a point of critical decision: should White play 48 b7 at once or
take time out for 48 Rddl to prevent a rook check on el?

White should postpone the push of the b-pawn and make a defensive
move.
48 Rddl!
After 48 b7? Rel+ 49 Rxel Rxel+ 50 Kh2 Rbl 51 Kg3 Rb4 White is
again unable to make progress; e.g. 52 Kf3 Kf6 53 Ke3 Ke6 54 Rh7 Ke5 55
Kf3 Rb3+ 56 Kf2 f4 and so on.
48 ... Re8 49 b7 Rb8 50 Rd7 1-0
The combination of passed b-pawn, more active rooks, and vulnerable
black king mean that White is winning; e.g. 50 ... Kg5 51 Rb6 Kf4 52 Kh2 h5
53 Rh7 Kg5 54 Rc7 (threatening Rc8) 54 ... Ree8 55 g3! (threatening Rg7
mate) 55 ... Rg8 (or 55 ... h4 56 Rg7+ Kh5 57 g4+ fxg4 58 hxg4 mate) 56
h4+ Kg4 57 Kg2! (threatening mate via either Rc4+ or Rb4+) 57 ... f4 58
Rf7! Rbf8 (if 58 ... 13+ then 59 Kh2 and 60 Rf4 mate, or 58 ... fxg3 59 Rb4
mate) 59 Rg6+ Rxg6 60 Rxf8 and the b-pawn promotes.

36. Nepomniachtchi-Amin

Go after g6, via e7.

31 Ne7!
31 Nf4? is the wrong way: 31 ... fxe4 32 Nxg6+ Kg8 33 Rxe4 Re8 34
Rxe5 Qxf3 35 Rxe8+ Kf7 36 Re7+ Kxg6 37 gxf3 Nxb2 with a likely draw.
.
31.. fxe4
31 ... Qc7 32 Nxg6+ Kg8 33 Bd5+ wins the queen to start with.
32 Qf7! 1-0
Black must give up his queen to fend off mate.

37. Nepomniachtchi-Nisipeanu
 Shift the rook to e3, after which Black is unable to defend everything.
19 Re3! 1-0
Threatening Qf5 and wins; e.g. 19 ... Kg8 (19 ... Rxe3? allows 20 Qxh6+
Kg8 21 Qh7 mate) 20 Qf5! g6 21 Rxe6 Qxe6 (21 ... gxf5 22 Rxe8 is
hopeless, and 21 ... Bxe6 22 Qf6 and mate next move) 22 Qf3! (attacking c6)
22 ... gxh5 (or similarly 22 ... Bd7 23 Rel Qg4 24 Qf6 and Black must hand
over his queen to evade mate) 23 Rel Qg4 24 Qf6 Qg7 25 Qd8 Qg4 26 f3
Qg3 27 Re8 and Black can resign, which he had already in fact done.

38. Ganerdene-Nepomniachtchi
 Step 1: Chop the knight on d6.
25 ... Bxd6!
Not 25 ... Rcd8? 26 Bc5 and White defends.
26 Qxd6 Nb4+!!
Step 2: Sacrifice the knight on b4, forcing open the a-file.
27 cxb4
Declining doesn’t work: 27 Kbl Qc4! 28 Qxe5 (28 cxb4 transposes to the
next note) 28 ... Qa2+ 29 Kcl Qal+ 30 Kd2 Qxb2+ 31 Kel Rfe8 wins.
27 ... Qc4+ 28 Kai
Instead:
a) 28 Kbl Qc2+ 29 Kai axb4 30 Ba7 Rc6! 31 Qd3 Qa4+ 32 Kbl b3 wins.
b) 28 b3 Qc2+ 29 Ka3 axb4+ 30 Ka4 Rc3 41 Kb5 Qxb3 and it doesn’t
look like White will last long.
28 ... axb4
Black threatens a deadly rook check on the a-file.
 29 Qd5 Qc2!
Renewing the threat of... Ra8+. Not 29 ... Ra8+? 30 Kbl Qa6 31 Bc5 and
White may yet survive.
30 Qa5
There is nothing better. 30 Bb6 Ra8+ 31 Ba5 Qc7! wins.
30 ... Ra8 31 Ba7 b3!

Increasing the pressure by trapping the white king.

32 Qa3
Or 32 Rd6 Rfc8 (threatening back rank mate) 33 Rbl h6! 34 Ra6 Rd8!,
threatening ... Rd 1, to which there is no good defence.
32 ... Rfc8
White is paralysed.
33 Rgel h6
This move is almost mockingly nonchalant. There is nothing to be done
about the coming ... Rc7. The immediate 33 ... Rc7 also wins.
 34 13 Rc7 35 Bc5 Rxa3+ 36 Bxa3

Black to play and force mate in five.

Simply transfer the rook to the a-file, after which there is no defence to the
coming ... Rxa3+ and ... Qa2 mate.
36 ... Ra7! 37 Rd8+
One spite check before resigning is always gratifying.
37... Kh7 0-1

39. Nepomniachtchi-Postny
 Discovered attack/overloaded defender.
35 Nxh6! 1-0
Black resigned since his position falls apart: 35 ... Rxf3 36 Rxf3 Rxf3 37
Kxf3 Ke7 38 Kg4 Kf6 39 Nf5 Kf7 40 Ng3 Kf6
 As a planning exercise, come up with a win for White.

Answer: Two simple steps:

1. Transfer the knight to c6 and chop Black’s a7-pawn.
2. Create an unstoppable passed a-pawn with b3-b4 and a4-a5.
For example: 41 Ne2 Be7 42 Nel Bd8 43 Nd3 Be7 44 Nb4 and so on.

40. Nepomniachtchi-Ding Liren

 Pinned piece/knight fork.
23 Nf6+! 1-0
Oops. Black loses his queen whether he takes the knight or not.

41. Vachier Lagrave-Nepomniachtchi

 White’s king is helpless if Black goes with ... Qel+, ... Ne2-f4+ and ...
Rd6!-f6!.
30 ... Qel+ 31 Kg2
If 31 Kh2 then 31 ... Bc7+ 32 Ne5 Ne2 33 Qg2 Nxd4! undermines and
wins.
.
31.. Ne2!
Threatening ... Qgl mate.
32 Qe4 Nf4+
Only nominally a knight fork.
33 Kf3 Qhl+!
Black is after bigger game than the exchange. Not 33 ... Nxd3? 34 Nxd3
and the conversion won’t be so simple.
34 Ke3 Qcl+! 35 Kf3 Rd6! 0-1
 Threatening ... Rf6!. White has no good defence; e.g. 36 Qe8+ Kh7 37
Qe4+ g6 38 Qe5 Qhl+ 39 Ke3 (or 39 Kg3 Qg2 mate) 39 ... Re6 wins the
queen.

42. Nepomniachtchi-Artemiev
 The knight sacrifice is both sound and winning, although it’s not easy to
see that at this point.
37 Nhf5+’’
Here comes Nepo, angry as a wasp on a bad day.
37 ... gxf5
No choice, since 37 ... Kf8?? loses without a fight after 38 Nxd6.
38 Nxf5+ Kg8?
Allowing White to capture the g5-pawn with check. Instead:
a) 38 ... Kg6? 39 Nxd6 (threatening Qxe5) 39 ... Re7 is met by 40 Qh3!
and wins; e.g. 40 ... Rai 41 Qf5+ Kg7 42 Qxg5+ Kf8 43 Qf6+ Kg8 44 Nf5
and Black is mated.
b) 38 ... Kf8! puts up the most resistance: 39 Qe3! Rai 40 Qxg5 (not
check) 40 ... Rxdl+ 41 Rxdl Qxe4 and now White has to find 42 Qf6+ Ke8
43 Qh8+ Kf7 44 Qxh7+ Kf6 45 Qh6+! Kxf5 46 f3! Qd4+ (forced; otherwise
57 g4+ follows) 47 Rxd4 exd4 48 g4+ Ke5 49 Kf2 Ne7 50 Qg7+ Ke6 51 f4,
which the engine assesses at +5 for White, and even then it would require
some work for a human to prove the win.
39 Qg3 Qxe4
Giving an exchange back is no good either: 39 ... Rg7 40 Nxg7 Kxg7 (40
... Qxe4 41 Ne8 is clearly hopeless) 41 Qxg5+ Kf7 42 Qf5+ Kg7 43 Rxd6!
and White has a winning attack.
40 Qxg5+ Kf7

Exercise (critical decision): Should White build up with 41 Rd3, intending to

swing across to f3 or g3, or just play 41 Qg7+ - ?

Answer: Sometimes the threat is not stronger than the execution. The queen check to g7 wins.
41 Rd3?
We, the ordinary, are not the only ones who throw away winning
positions. World-class players blunder as well, albeit less often. This move
throws away Nepo’s win.
 He had to find the same idea as in the 38 ... Kf8 line above: 41 Qg7+ Ke6
(not 41 ... Ke8? 42 Qg8 mate) 42 Qh6+! Kxf5 43 43!, threatening 44 g4+
again, and this time Black has no good defence at all, since 43 ... Qd4+ loses
to 44 Rxd4 exd4 45 g4+ Ke5 36 Rel mate.
41 ... Rai!
Nepo may have missed this resource.
42 Rxal
Or 42 g4 Rxdl+ 43 Rxdl Ra7 44 Qg7+ Ke8 45 Qg8+ Kd7 46 Qxh7+ Kd8
and Black’s king threatens to escape via c7. White now has only one path to
the draw: 47 Rxd6+! Nxd6 48 Qh4+! Kc7 (forced) 49 Qe7+ Nd7 50 Qxd6+
Kd8 51 Qe7+ and perpetual check.
42 ... Qxd3 43 Ra3! Qxa3
Now the game should end in perpetual.
44 Nh6+ Ke8 45 Qg8+ Ke7 46 Qg5+ Ke6 47 Qf5+ 1-0

In this completely drawn position, every source I found on this game has
Artemiev losing, so the assumption is that he lost on time.

43. Kramnik-Nepomniachtchi

Step 1: Annihilation of defensive barrier.

32 ... Rxh2+!
This move is an example of the “keep your enemies closer” philosophy.
33 Kgl
Or 33 Kxh2 Qf4+ 34 Kgl (or 34 Kg2 Rf2+ 35 Kh3 Qh4 mate) 34 ...
Qxe3+ 35 Kh2 Rh3+ 36 Kg2 Qf3+ 37 Kgl Rhl mate.
33 ... Rg3+!
Step 2: Clearance, allowing Black’s queen entry to f2.
34 Kxh2 Qf2+ 35 Khl Rh3 mate

44. Svidler-Nepomniachtchi
 Discovered attack/zwischenzug/double attack.
.
36.. Nf2!
Now White’s pieces at h3, d3 and el are all under attack. 36 ... Ng5! is
merely a variation on the same theme.
37 Rde3
Or 37 Rxe7 Nxh3+ (zwischenzug) 38 Kg2 Qxe7 and if 39 Kxh3 then 39 ...
Qe6+! 40 Kg2 Qe2+ wins the rook.
37 ... Nxh3+ 38 Kg2 Rxe3 39 Rxe3 Rxe3 40 Qxe3 Qd7 0-1
Black’s knight is trapped and yet not trapped. Here is an example of the
winning technique: 41 a3 Kf7 42 Qd4 Qe6 43 Kfl g5 44 Qh8 (there is no
perpetual check) 44 ... Qxc4+ 45 Kel (or 45 Kg2 Qe2+! 46 Kxh3 Qfl mate)
45 ... Qcl+ 46 Ke2 Qc2+ 47 Kel Q12+ 48 Kdl Qgl+ 49 Kd2 Qxh2+ 50 Kcl
Qxg3 51 Qxh5+ Kg7 and wins.

45. Giri-Nepomniachtchi
 50 ... Qhl+ 51 Kg4
Not 51 Kg3? Rg2 mate.
51 ... Rg2+
This is quite sufficient to win. To solve the mate in seven, you needed to
find 51 ... Rd4+! 52 Kg3 Rxf3+ 53 Rxf3 Qxh4+ 54 Kg2 Rd2+ 55 Kgl Qh2+
56 Kfl Qhl mate.
52 Rg3 Rh2! 53 Rh3 Qg2+
Here 53 ... Rxh3 54 Rxh3 Qe4+ 55 Kh5 Rf4! is one move faster.
54 Reg3
54 Kh5 Rxh3 55 Rxh3 Qe2+ forces mate in four after 56 R13 Rf4! (even
stronger than taking the rook) 57 Qd3 Qe8+ 58 g6 Qe5+.
54 ... Qe4+ 0-1
In view of 55 Kh5 Qg6+ 56 Kg4 h5 with a David and Goliath pawn mate!

46. Svidler-Nepomniachtchi
 A second pawn falls after:
43 ... Ra3! 44 Bd2
Here is the problem: 44 a5?? hangs a piece to 44 ... Ne3+! 45 Kb4 (or 45
Kd4 Nc2+) 45 ... Nc2+ forking king and bishop, while annoyingly covering
the once loose a3-rook.
44 ... Rxa4+
And Nepo converted his two extra pawns.

47. Nepomniachtchi-Matlakov
 Attraction/double attack.
39 Rxf4! 1-0
Since 39 ... Rxf4 40 Qd6+ (double attack) 40 ... Kg7 41 Qxf4 wins.

48. Antipov-Nepomniachtchi
 Simply shift the rook to g6, threatening ... Qxg3+.
.
30.. Rg6! 0-1
There is no defence, since 31 Kfl Qf3 also leads to a very quick mate.

49. Goganov-Nepomniachtchi
 Slide the bishop to a3, after which White must hand over the exchange.
22 ... Ba3!
Threatening ... Bb2 and ... c3-c2.
23 Bxc3
23 Rxc3 Nxc3 24 Bxc3 c5 is also lost for White.
23 ... Nxc3 24 Rxc3 Bb2
Double attack. White loses the exchange.
25 Rael c5!
Nepo first jettisons a pawn to activate his pieces. 25 ... Bxc3 26 Rxc3 c5!
27 Rxc5 Rc8 was equally effective.
26 Rxc5
Not 26 dxc5?? Bxc3 and White is unable to recapture due to his weak back
rank.
26 ... Bxcl 27 Rxcl Qa5!
White’s queenside falls and he lacks the fire-power to go after the black
king.
28 h4
Or 28 Rc2 Rfc8 29 Rxc8+ Rxc8 30 Qf4 h6! 31 g4 Kg8 32 g5 Qxa2 33
gxh6 Qbl+ 34 Kg2 Qg6+ and Black consolidates.
28 ... Qxa2 29 Qf4 Qxb3 30 Ng5 h6!

This confident move displays why Nepo is one of the premiere calculators
of his era. There is no threat on f7. The nervous type would have gone for the
safe 30 ... Qb7, which also wins.
31 Ne4
Although 31 Nxf7+ looks dangerous, exact calculation proves that it fails:
31 ... Kg8 32 Nxh6+ gxh6 33 Qg4+ Kh8 34 Rc7 Rg8 35 Qf4 Qdl+ 36 Kh2
Rxg2+! 37 Kxg2 Rg8+ wins White’s queen or mates.
.
31.. Rbc8
And Nepo won without much difficulty.
50. Duda-Nepomniachtchi

58 ... Qd2+!
The alternative was 58 ... g5+! 59 Kg3 (or 59 Ke3 Qe2 mate) 59 ... Qf2+
60 Kh3 Qh4 mate.
59 Kg3 Qf2+ 0-1
In view of 60 Kh3 Qh4 mate or 60 Kf4 g5 mate.

51. Nepomniachtchi-Grischuk
 Give up the precious b7-pawn to win the even more precious h6-pawn.
75 Bf5! Kxb7
If Black declines the pawn with 75 ... Bf4, say, he creates a path for
White’s king with 76 Kd4 Bb8 (or 76 ... Kxb7 77 Kd5) 77 Bc8 (back again)
77 ... Kd6 78 Ke4 (Black’s problem is that his king is denied access to e6) 78
... Ke7 79 Kf5 Kf7 80 Be6+ Kg7 81 Bb3! (clearing e6 for his king) 81 ... Bc7
82 Ke6 and the king heads off to c8 and wins.
76 Kd5
Gaining a tempo.
76 ... Bb2 77 Ke6 Kc7 78 Kf7 Kd6 79 Kg6 Ke7 80 Kxh6 Kf6 81 Bc2
Ba3 82 Kh7 Bf8 83 Bb3 Bg7 84 h6 Bf8 85 Ba2! 1-0
 And Black is in zugzwang.

52. So-Nepomniachtchi
 Start with a check on a6.
58... Qa6+! 0-1
Since 59 Kb2 Qb7+! wins the white rook.
Index Complete Games
Anand.V-Nepomniachtchi.I, Mainz (rapid) 2009
Andriasian.Z-Nepomniachtchi.I, World Youth Stars, Kirishi 2007
Aronian.L-Nepomniachtchi.I, Skilling Open KO (rapid) 2020
Astaneh Lopez.A-Nepomniachtchi.I, Batumi Olympiad 2018
Carlsen.M-Nepomniachtchi.I, Carlsen Invitational (rapid) 2020
Carlsen.M-Nepomniachtchi.I, Carlsen Invitational (rapid) 2021
Carlsen.M-Nepomniachtchi.I, Carlsen Invitational (rapid) 2021
Carlsen.M-Nepomniachtchi.I, Legends of Chess (rapid) 2020
Carlsen.M-Nepomniachtchi.I, London Classic 2017
Carlsen.M-Nepomniachtchi.I, St, Louis (rapid) 2020
Carlsen.M-Nepomniachtchi.I, Wijk aan Zee 2011
Dominguez Perez.L-Nepomniachtchi.I, Russian Team Championship, Sochi 2016
Giri.A-Nepomniachtchi.I, Candidates Tournament, Yekaterinburg 2020
Giri.A-Nepomniachtchi.I, Legends of Chess (rapid) 2020
Giri.A-Nepomniachtchi.I, Legends of Chess (rapid) 2020
Giri.A-Nepomniachtchi.I, Wijk aan Zee 2011
Ivanchuk.V-Nepomniachtchi.I, Legends of Chess (rapid) 2020
Kasparov.G-Nepomniachtchi.I, St, Louis (rapid) 2017
Le Ouang Liem-Nepomniachtchi.I, Aeroflot Open, Moscow 2008
Mamedov.R-Nepomniachtchi.I, World Team Championship, Astana 2019
Najer.E-Nepomniachtchi.I, Mainz (rapid) 2008
Nepomniachtchi,I-Alekseenko.K, Candidates Tournament, Yekaterinburg 2021
Nepomniachtchi.I-Aronian.L, FIDE Grand Prix, Geneva 2017
Nepomniachtchi.I-Bacrot.E, Batumi Olympiad 2018
Nepomniachtchi.I-Bosboom.M, Wijk aan Zee 2007
Nepomniachtchi.I-Bu Xiangzhi, Danzhou 2016
Nepomniachtchi,I-Carlsen.M, Legends of Chess (rapid) 2020
Nepomniachtchi,I-Carlsen.M, Skilling Open KO (rapid) 2020
Nepomniachtchi,I-Carlsen.M, World Junior Championships, Halkidiki 2003
Nepomniachtchi.I-Ding Liren, Candidates Tournament, Yekaterinburg 2020
Nepomniachtchi.I-Ding Liren, Legends of Chess (rapid) 2020
Nepomniachtchi.I-Dubov.D, Airthings Masters (rapid) 2020
Nepomniachtchi,I-Firouzja.A, St, Louis (blitz) 2020
Nepomniachtchi,I-Grischuk.A, Russian Team Championship, Dagomys 2010
Nepomniachtchi.I-Hou Yifan, FIDE Grand Prix, Geneva 2017
NepomniachtchiJ-HowelLD, World Rapid Championship, Riyadh 2017
Nepomniachtchi.I-Inarkiev.E, European Championship, Rijeka 2010
Nepomniachtchi.I-Inarkiev.E, FIDE Grand Prix, Geneva 2017
Nepomniachtchi.I-Ivanchuk.V, SportAccord Rapid, Beijing 2013
Nepomniachtchi.I-Jobava.B, European Championship, Rijeka 2010
Nepomniachtchi,I-Kamsky.G, FIDE World Cup, Khanty-Mansiysk 2011
Nepomniachtchi.I-Kramnik.V, Dortmund 2018
Nepomniachtchi.I-Li Chao, FIDE Grand Prix, Sharjah 2017
Nepomniachtchi.I-Meier,G, Dortmund 2018
Nepomniachtchi.I-Najer.E, Russian Rapid Championship, Olginka 2011
Nepomniachtchi.I-Nakamura.H, Carlsen Invitational (rapid) 2021
Nepomniachtchi,I-Nakamura.H, Sinquefield Cup, St, Louis 2019
Nepomniachtchi.I-Ponomariov.R, ACP Cup, Riga (rapid) 2013
Nepomniachtchi.I-Potkin.V, Yaroslavl (rapid) 2014
Nepomniachtchi.I-Radjabov.T, FTX Crypto Cup (rapid) 2021
Nepomniachtchi.I-Rublevsky.S, Russian Championship Superfinal, Moscow 2006
Nepomniachtchi.I-S¡ugirov.S, Russian Team Championship, Sochi 2016
Nepomniachtchi.I-Sutovsky.E, Poikovsky Karpov 2018
Nepomniachtchi.I-Tomashevsky.E, Tai Memorial, Moscow 2016
Nepomniachtchi.I-Wang Hao, Candidates Tournament, Yekaterinburg 2020
Nepomniachtchi.I-Wei Yi, FIDE Grand Prix, Moscow 2019
Smirnov.P-Nepomniachtchi.I, Novokuznetsk 2008
Tomashevsky.E-Nepomniachtchi.I, FIDE World Cup, Khanty-Mansiysk 2019
Vachier Lagrave.M-Nepomniachtchi.I, Grand Chess Tour, Paris (rapid) 2019
Wang Hao-Nepomniachtchi.I, Candidates Tournament, Yekaterinburg 2021
Wei Yi-Nepomniachtchi.I, FIDE Grand Prix, Moscow 2019