About the Author

Andrew Martin is a FIDE Senior Trainer and International Master. He is the current Head of the newly-formed ECF Academy, which provides elite training for strong, young players. He teaches in twelve schools, is an experienced chess writer and has produced numerous chess DVDs.

Also by the Author:

Starting Out: The Sicilian Dragon

First Steps; The Queen’s Gambit

Contents

About the Author

Introduction

1) Inspiration

2) Pawn Structures

3) The Classical System

4) The Fianchetto Variation

5) Lines with Bg5

6) The Sämisch Variation

7) The Four Pawns Attack

8) Less Common White Systems

9) King’s Indian Byways

Final words

Index of Complete Games

Introduction

This is my third book on the King’s Indian Defence. The first, Winning with the King’s Indian, was written in 1989 and was more or less a repertoire for Black. The book did pretty well and in the pre-computer era a repertoire book could be just what the doctor ordered. Nowadays, such a book can be cut to ribbons by simply turning on an analysis engine, but it is interesting for me to see how many of my older suggestions are still playable.

King’s Indian Battle Plans, written in 2004, was a sprawling, ambitious project, cooked up by myself and Bob Long – a work full of new ideas, designed to show just how creative the King’s Indian could be. Once again the book sold well, but looking back I can see why it was regarded by some as haphazard. Nevertheless, I am not displeased with it.

Which brings us to the modern day and the work you have before you – First Steps: The King’s Indian. This book will be different again and sets out to provide a good introduction to this wonderful opening. I cannot claim that it will cover everything, as to do so would require at least 1000 pages, but it will get you up and running and ready to play the KID with enthusiasm. How you take it from there will then be up to you.

Thanks are due to Jon Tait who did an excellent job editing this book.

Andrew Martin, Bramley, Surrey, June 2019.

Chapter One

Inspiration

The inspiration needed to play a particular opening has to come from somewhere and for me it was the games of Bobby Fischer and Svetozar Gligoric that provided the initial excitement. The very first game in this book is a Fischer classic and as a young man, after studying and playing it through, I knew that the King’s Indian was the defence for me.

Game 1

V.Korchnoi-R.J.Fischer

Herceg Novi (blitz) 1970

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 Be2 0-0

Here we have the basic King’s Indian set-up. Typically, Black has a well-protected king and plenty of flexibility with his pawns. More on that later.

6 Nf3 e5 7 0-0 Nc6 8 d5 Ne7 9 Nd2

White has employed a Classical formation, has grabbed a bit more space with the advance d4-d5, and now has to choose a way forward. Usually, he advances on the queenside, and so 9 Nd2 is a prelude to these operations. White plans b2-b4, c4-c5 and then Nc4, putting pressure on d6. Black will respond by launching a kingside pawn storm.

9 ... c5!?

Putting the brakes on the intended advance.

10 a3

I think 10 Rb1! is more accurate, intending b2-b4 without delay (see Game 39). Later, White may play a2-a4 in one move, saving a tempo. I think it’s worth pointing out that this was a five-minute game from a very famous blitz tournament. Back then, such events were as rare as hen’s teeth and something to be savoured. The quality of the games was remarkably high.

10 ... Ne8 11 b4 b6 12 Rb1 f5

Enough defending! It’s time to strike out.

13 f3 f4 14 a4 g5 15 a5 Rf6 16 bxc5?!

It was better to open two queenside files instead of one; i.e. 16 axb6 axb6 17 bxc5 bxc5.

16 ... bxc5 17 Nb3 Rg6 18 Bd2 Nf6 19 Kh1 g4

White should be able to defend this position, but Black’s game is much more enjoyable to play and he is going for the king. What more could you want?

20 fxg4 Nxg4 21 Rf3?

Here 21 Bf3! was a better stab at defence.

21 ... Rh6 22 h3 Ng6 23 Kg1 Nf6 24 Be1 Nh8!

A move which made this game memorable for me. The knight heads for the wonderful g5-square, when the rook on f3 will look very clumsy indeed.

25 Rd3 Nf7 26 Bf3 Ng5 27 Qe2 Rg6 28 Kf1?

White can hang on after 28 Kh2 Qd7! (threatening a sacrifice on h3) 29 Qf1 Nxf3+ 30 Rxf3 Qd8 31 Qe2 Rh6 32 Kg1 Bd7, but Black still has much the preferable game. Nevertheless, this would not have lost immediately.

28 ... Nxh3! 29 gxh3 Bxh3+ 30 Kf2 Ng4+ 31 Bxg4 Bxg4 0-1

The twin threats of ... Bxe2 and ... Qh4+ force White to resign.

Back in the mid to late seventies English chess was on the rise. Visionary Leonard Barden used to organize simultaneous exhibitions, whereby teams of very strong juniors were lined up to play the masters, usually the grandmasters at the Hastings Premier. GM Rafael Vaganian came to the London Central YMCA chess club in January 1975, unaware that he was about to face what was, for a simul, some serious opposition.

Game 2

R.Vaganian-A.D.Martin

London (simul) 1975

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 Nf3 0-0 6 Be2 e5 7 0-0 Nc6 8 d5 Ne7 9 Ne1 Nd7 10 Nd3 f5

If it was good enough for Fischer, it was good enough for me. I had visions of battering Vaganian’s king.

11 f3 f4 12 b4 g5 13 c5 h5 14 a4 Nf6 15 b5 g4 16 Ba3 Ng6 17 b6

To go deeply into theory about what has gone before would be quite pointless. Suffice it to say that both players have been ignoring each other and getting on with their own plans. However, there comes a point at which one must pay serious attention to the opponent and this is the moment. I started to feel a bit uneasy. 17 b6 seemed very strong. White was coming round quickly. I decided to follow the advice given before the simul: not to back down and to keep it as complicated as possible.

17 ... axb6 18 cxd6 cxd6 19 Nb5 g3 20 Bxd6?

I think 20 Nxd6! was better, but the poor guy was facing thirty opponents at once. The point is to try to eliminate the light-squared bishop on c8, which is almost always required to sacrifice itself on h3 so that Black’s attack can break through.

20 ... Nxe4!!

Missed or underestimated? Either way, White is now in trouble, as Black’s queen enters the action.

21 fxe4

21 Bxf8 Qh4 22 h3 Bxh3 is instant termination.

21 ... Qh4 22 hxg3 fxg3 23 Rxf8+ Bxf8 24 Bf3 Qh2+

Apparently 24 ... Bd7! is the strongest move, intending to play 25 ... Bxb5 26 axb5 Rxa1 27 Qxa1 Qh2+, followed by ... Qh1+. In my excitement I only saw the check on h2.

25 Kf1 Nh4 26 Ke1?

26 Ne1! was a better try, although I still prefer Black after 26 ... Nxf3 27 Nxf3 Qh1+ 28 Ng1 Bd7.

26 ... Bg4??

A mistake follows a mistake. Instead, 26 ... Bh6!! won straight away, cutting off any escape route for the white king.

27 Bxg4 Nxg2+ 28 Ke2 Ne3+

I saw up to here but simply missed 30 Nf4.

29 Kxe3 Bh6+ 30 Nf4 Bxf4+ 31 Kd3 hxg4 32 Qxg4+ Kh8 33 Nc3?!

White wins easily after 33 Be7!.

33 ... Qd2+ 34 Kc4 b5+ 35 Nxb5?

Here 35 Kb3 was correct. The effect of the simul was clearly taking a toll on Vaganian. We were four hours in and he looked like he would much prefer to go to a restaurant in Soho and drink some wine, rather than grind on against us.

35 ... Qc2+ 36 Nc3 b5+!

Why not chuck another pawn on the bonfire?

37 axb5 Rxa1 38 Qc8+ Kh7 39 Qf5+ Kg7 40 Bxe5+ Bxe5 41 Qxe5+ Kh6?

Seemingly 41 ... Kf8 42 Qxg3 Rc1 holds for Black, who has numerous possible queen checks with the white knight pinned.

42 Qf4+ Kh5 43 Qxg3

White is winning again for sure, but Black can make life tricky for him, as proved the case.

43 ... Ra4+ 44 Kc5 Rxe4 45 Qf3+ Rg4

46 b6?!

It was better to push 46 d6, when White would be on the verge of wrapping it up.

46 ... Qb3 47 d6??

He should go for 47 Kc6 Qc4+ 48 Kb7 Kh6 or 47 Qh3+ Rh4 48 Qf6+ Kh6, though the win is no longer quite so simple.

47 ... Qc4 mate

This is not the finest game ever seen, but for a young player it was an absolutely thrilling experience. To give you an idea of Vaganian’s treatment: his score on the day was +13, =7, -10. That is tough going.

My 1989 book recommended a queen sacrifice against the Sämisch Variation, developed by the great innovator GM David Bronstein. Not wishing to be caught out by the cliché that it is easy to sacrifice pieces which are not yours, I was anxious to try this idea out whenever I could.

Game 3

J.Levitt-A.D.Martin

Glasgow 1989

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 0-0 5 f3 d6 6 Be3 e5 7 d5 Nh5 8 Qd2 Qh4+

Black isn’t obliged to go in for this. He can also play 8 ... f5 (see Game 21).

9 g3

9 Kd1!? was played against me by GM Chris Ward. The idea of ‘castling long’ slowly looks ex travagant, but White’s central structure is so solid he can afford this slight loss of time. Nonetheless, Black should be okay after the simple 9 ... Qe7 10 Kc2 Na6, intending ... f7-f5.

9 ... Nxg3 10 Qf2 Nxf1 11 Qxh4 Nxe3 12 Ke2 Nxc4

Here it is. For the queen Black has two bishops, two pawns and a connected pawn chain. Is it enough? The sacrifice certainly seemed difficult to play against to me. Perhaps the ultimate endorsement came when, shortly after the publication of the book, Yasser Seirawan decided to risk the idea against none other than Garry Kasparov!

13 Rc1 Na6 14 Nd1 Nb6 15 Ne3 Bd7 16 Nh3 f6

A more recent example saw 16 ... Bb5+ (hoping that the bishop is more active on b5, but the move may push the white king to a safer square) 17 Kd2 c6 18 Rhg1 f6 (if 18 ... cxd5 then 19 Nf5!) 19 f4 Rf7 (possibly Black should try 19 ... cxd5 20 exd5 Rac8 21 f5! Rxc1 22 fxg6 hxg6 23 Kxc1 Be8 24 Nf2 Nc5, but the impression remains that his position is passive) 20 f5 g5 21 Qh5 Nc5 22 Nf2 cxd5 23 exd5 e4 24 h4! (as in our featured game, White finds a way to break through to the black king) 24 ... Raf8 25 hxg5 fxg5 26 Rxg5 Kh8 27 Rcg1 Nbd7 28 Rxg7! Rxg7 29 Rxg7 Kxg7 30 Nfg4 Rf6 (or 30 ... Kh8 31 Qh6 and a capture on d6 will win as the black pieces are in a terrible mess) 31 Nxf6 Nxf6 32 Qh2 Ne8 33 Ng4 Nd7 34 Qh6+ Kg8 35 Qg5+ Kf8 36 f6 1-0 D.Khismatullin-Ju Wenjun, Moscow 2012.

17 Nf2 Nc8

18 Nd3!

Here is the top-level game played earlier that year: 18 Rc3 Ne7 19 Rhc1 Rac8 20 Rb3 Rb8 21 Nd3 Rf7 22 Qe1 Nc8 23 Qa5 Nb6 24 Rxc7!! f5 25 Rc2 (now 25 Rcc3 Na4 26 Rc4 Nb6 27 Rxb6! axb6 28 Qxb6 Bf8 29 b4 makes it very difficult for Black to untangle) 25 ... fxe4 26 fxe4 Rbf8 27 Rxb6 axb6 28 Qxb6 Bh6 29 Qxd6 Rf3 30 Nf5 gxf5 31 Qxh6 fxe4 32 Ne1 ½-½ G.Kasparov-Y.Seirawan, World Cup, Barcelona 1989. Kasparov decides not to push on in what is a rather random position. A lucky escape for Yasser, as according to the machines White holds a decisive advantage.

All the same, I prefer Levitt’s move. It is more direct and allows us to understand the line a bit better. Black’s king can become vulnerable and, while White must play extremely well to prove it, the fact is inescapable.

18 ... c5 19 Qg3!

Preparing to push the h-pawn.

19 ... Bh6 20 h4 Bb5?!

20 ... Bxe3 21 Kxe3 Ne7 seemed unattractive to me, but maybe Black can try to tough it out. The problem will always be the absence of counterplay, as any pawn break, such as ... f6-f5, runs the risk of exposing Black’s king still further.

21 a4! Bxd3+

Levitt also gave 21 ... Bxa4 22 h5 g5 23 Nf5 Bg7 24 h6 Bh8 and 21 ... Bd7 22 h5 as winning for White.

22 Kxd3 Nb4+ 23 Ke2 f5 24 h5!

Pressing on in aggressive style. By now I knew what was coming but could not see any way to avoid it. According to Levitt, Black has more chances to survive after 24 Nxf5 (or 24 Qg2 Ne7 25 h5 g5) 24 ... Bxc1 25 Rxc1 Nxd5!? 26 Nh6+ Kg7 27 exd5 Kxh6 28 Qg5+ (or 28 h5 Ne7) 28 ... Kg7 29 h5 Rf6.

24 ... Bxe3 25 hxg6! h6

Not 25 ... Ne7? 26 Qh4 or 25 ... Bxc1? 26 gxh7+ Kh8 27 Rg1 and mates.

26 Rxh6!!

A bonecrusher. Both the white rooks are available to capture, but this offers no help to Black. His king is too lonely.

26 ... Bxh6

If 26 ... Bxc1 then 27 Rh7 is quickly decisive; but not 27 Qh3? because of 27 ... Ne7! and Black hangs on.

27 Qh3 Kg7 28 Rh1 Bg5

I thought initially that 28 ... Rh8 might bail me out, but White has a decisive queen switchback: 29 Qxf5 Rf8 30 Qh3 Rh8 31 Qd7+! Kxg6 32 Qf5+ Kg7 33 Rg1+ etc.

29 Qh7+ Kf6 30 Qd7!

The final, nice touch. Black is paralysed. Not 30 g7? due to 30 ... Ne7! again.

30 ... Bh6 31 Rxh6 1-0

You can learn a lot from getting crushed. I now understood that if Black wanted to play this line on a regular basis, he needed to gain speedy counterplay before all hell was let loose on his king. To the present day I don’t believe Black has found a satisfactory way to achieve this.

About 25 years ago I got into correspondence chess, hoping to improve my analytical skills and to try out some new ideas. The engines were getting stronger at that time, but they had not yet killed the game dead. Today, I have visions of a little man sitting in a big office, surrounded by banks of computers. Soon his next set of moves will be ready. Later he will claim the games as his own. Is this really what chess is all about? For a few years I really enjoyed playing this type of chess. The last straw came after a crushing defeat in a domestic tournament: a perfectly played Open Sicilian by White, where all I could do at the end was to congratulate my opponent on his excellent win. Upon consulting the grading list afterwards I found his over-the-board rating to be Elo 1000.

The following game is one of my better efforts and, conveniently, it is a King’s Indian.

Game 4

V.Lyukmanov-A.D.Martin

World CC Championship Semi-Final 1994

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 f3 b6!?

Dubbed futuristic by IM Gerald Welling. My simple thought at the time was that the Sämisch pawn structure is inflexible and with good timing Black should be able to obtain a good position full of dynamic chances.

6 Be3 Bb7

Clearly b7 is not the ideal square for this bishop with e4 so securely protected. Furthermore, d4-d5 looks as if it will throttle the bishop completely. Is that the full story? I don’t think so. Black deliberately delays castling, waiting for White to commit himself.

7 Qd2 c5

An important prod, opening up prospects for the bishop on g7. In Game 23 Black opts for 7 ... Nbd7 and 8 ... c6, which gives him a very passive position.

8 d5 Nbd7! 9 h4

Lyukmanov could not resist it. The Sämisch set-up lends itself naturally to a kingside assault – but what if Black doesn’t play ball, leaving his king in the middle? Won’t h2-h4 turn out to be weakening? Well, that was roughly what was going through my mind at the time. I couldn’t assess the position exactly but resolved to carry on making flexible moves.

If White had played 9 Bh6 instead, Black should certainly castle long. An effective example of this strategy was seen in the following game: 9 ... Bxh6 10 Qxh6 a6 11 Qd2 Qc7 12 b3 0-0-0 13 Bd3 e6 14 Nge2 Ne5 15 Rb1 exd5 16 cxd5 Kb8 17 b4, A.Shenin-V.Gavrilov, Dagomys 2004, and now Black’s best move is probably the flexible 17 ... Nfd7!.

9 ... a6! 10 Nh3 Ne5! 11 Be2 h5

This had to be played now. White was threatening to swamp Black with g2-g4-g5 and f3-f4. Lyukmanov probably thought he was clearly better. Black continues to wait.

12 Nf2 Bc8!?

A strange and remarkable move. Black covers g4 necessarily and it is becoming clear that White must take some further action soon, otherwise Black will find a way to play ... b6-b5 somewhere. Perhaps 13 a4 is best now, intending Ra3 and Kd1-c2, though it’s difficult to say.

13 Bf4 Ra7

Provocative!

14 0-0-0?!

Naturally White could have taken on e5. I couldn’t call the position but felt that the dark-squared bishop would see me through; e.g. 14 Bxe5 dxe5 15 Nd3 Bh6 16 Qd1 Nd7 17 Qa4 Qc7.

But now, after having waited for so long, Black’s attack is surprisingly immediate and deadly.

14 ... Nfd7 15 Kb1 b5!

The Benko Gambit is quite good when White castles long! For a measly pawn Black blasts open lines.

16 cxb5 Qa5 17 Rc1 axb5 18 Bxb5 Ba6

Black’s initiative proceeds very smoothly.

19 Bxa6 Qxa6 20 b3 0-0!!

The finest move of the game, all the more powerful for having been delayed. At this point I had calculated a lovely combination.

21 Bh6

Playing right into it. But with ... Rb8 etc imminent, what could he have done?

21 ... Nc4!

This felt very good to play.

22 bxc4 Rb8+ 23 Ka1 Bh8!

A full piece up White is powerless to stop Black invading, either with ... Ne5 or ... Nb6-a4.

24 Nfd1 Rab7 25 Re1

Or 25 Bf4 Nb6 26 e5 Nxc4 27 Qe2 dxe5 and wins.

25 ... Ne5 0-1

Lyukmanov gave up the ghost. Enigmatic resignation or zugzwang on a full board? The more you look, the more difficult it becomes to see a good move for White – or indeed any move. One brief example might be 26 Qc2 (or 26 Bf4 Nxc4 27 Qe2 Qa3) 26 ... Nxc4 27 e5 Na3 28 Qd2 Qb6 29 Nb2 (or 29 Ne3 Bxe5) 29 ... Qxb2+ 30 Qxb2 Rxb2 31 B4 Nc2+ and Black wins.

Gligoric, one of the most famous of post-war grandmasters and certainly one of Yugoslavia’s finest ever players, was a genius in King’s Indian positions. Just look at this game against a former world champion.

Game 5

T.V.Petrosian-S.Gligoric

Rovinj/Zagreb 1970

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 Be2 0-0 6 Nf3 e5 7 0-0 Nc6 8 d5 Ne7 9 b4

This is the Bayonet Variation. White wastes no time pushing forward and gaining space on the queenside.

9 ... Nh5

Black sees a nice square for the knight on f4 and gets ready to play ... f7-f5. Today, 9 ... a5 is equally popular.

10 Nd2!?

10 Re1 (for which see the notes to Game 32) and 10 c5 are the main alternatives. Petrosian’s move is interesting and intrigued me enough to research the idea for my 1990 book, Secret Weapons. White is allowing Black to play ... Nf4 and then ... Nxe2 if he wishes. The light-squared bishop is nothing special in this position.

10 ... Nf4 11 a4 f5

Clearly Black can play 11 ... Nxe2+ 12 Qxe2 f5, but then White is definitely a little better after 13 f3 with c4-c5 to come.

12 Bf3

Setting up the positional idea of g2-g3.

12 ... g5!

12 ... fxe4 13 Ndxe4 Nf5 is not as good. White has the upper hand after 14 g3! Nh3+ 15 Kg2, and if 15 ... Nd4 then 16 Bg4.

13 exf5 Nxf5 14 g3

14 ... Nd4!?

It takes real guts to play like this against one of the greatest ever tacticians. Gligoric cheerfully sacrifices a knight, leaving the white king wide open. As a young player this style of play had me thirsting for more and wanting to play the King’s Indian at every opportunity. By contrast 14 ... Ng6? 15 Nde4 Nd4 16 Bxg5! Nxf3+ 17 Qxf3 Rxf3 18 Bxd8 leaves White well on top.

15 gxf4 Nxf3+ 16 Qxf3

We will see shortly that 16 Nxf3! is an improvement.

16 ... g4 17 Qh1?!

When the great Petrosian buries his own queen in the corner, he has to be pretty concerned. Petrosian instinctively feels that he has to cover g2, but 17 Qh1 was a mistaken judgment call. 17 Qd3 exf4 18 Ra2 has to be a better way to go, even if Black’s attack looks frightening after 18 ... f3 and ... Qh4.

17 ... exf4 18 Bb2 Bf5 19 Rfe1 f3

I suppose White thought that he could get his queen out with a subsequent h2-h3. Gligoric makes sure that this plan will not succeed.

20 Nde4 Qh4! 21 h3 Be5! 22 Re3

There is no way to break free: 22 hxg4 Qxg4+ 23 Kf1 Bxc3 24 Bxc3 (or 24 Nxc3 Bd3+ 25 Ne2 fxe2+) 24 ... Bxe4 25 Rxe4 Qxe4 26 Qg1+ Qg6 leaves Black with a decisive material advantage.

22 ... gxh3 23 Qxf3 Bg4! 24 Qh1 h2+ 25 Kg2

Since 25 Kf1 Rf3! 26 Rxf3 Bxf3 27 Qxf3 h1Q+ wins immediately.

25 ... Qh5 26 Nd2 Bd4 27 Qe1

27 Rae1 Bh3+! 28 Rxh3 Rxf2+ 29 Kg3 Qg5 mate is a very nice line.

27 ... Rae8

White’s scattered forces are no match for the total coordination of the black army.

28 Nce4

If 28 Nd1 then 28 ... Bxe3! 29 Nxe3 (or 29 fxe3 Bh3+) 29 ... Rf4!, threatening ... Bf3+, is the quickest way to win.

28 ... Bxb2 29 Rg3 Be5 30 Raa3 Kh8

If we are to split hairs then 30 ... Bxg3! 31 Rxg3 Rxf2+! 32 Kxf2 h1Q 33 Rxg4+ Kh8 was stronger.

31 Kh1 Rg8 32 Qf1 Bxg3 33 Rxg3

Taking with the knight might have allowed more resistance. Even so, Black wins relatively easily: 33 Nxg3 Qh6 34 Nf3 Rgf8! 35 Nxh2 Qd2 36 Qa1+ Kg8 37 Qg1 Rxf2 38 Nxg4 Re1.

33 ... Rxe4 0-1

An attacking masterpiece.

The last game had me thinking and, after considerable time spent burning the midnight oil, I managed to come up with an improvement for White. I think the original version of Fritz was the only engine around at that time, but I don’t recall using it to help me. I felt confident enough to write a chapter in Secret Weapons (Tournament Chess 1990) about this very line and was ready to try out my analysis at the earliest opportunity.

Game 6

A.D.Martin-R.Britton

British Championship, Eastbourne 1991

1 d4 Nf6 2 c4 g6 3 Nc3 Bg7 4 e4 d6 5 Nf3 0-0 6 Be2 e5 7 0-0 Nc6 8 d5 Ne7 9 b4 Nh5 10 Nd2 Nf4 11 a4

I decided to follow the main line. 11 Bf3!? leads to an unclear position, not even allowing Black to take on e2. The whole point is to strand the knight on f4, and g2-g3 follows unless Black does something aggressive such as 11 ... Nd3 12 Ba3 a5 13 bxa5 Rxa5. Then White has the odd move 14 Ncb1!? with ideas of Nb3, Qc2, or Bb4 and a2-a4-a5; for example, 14 ... f5 15 Nb3 or 14 ... Nc5 15 Bb4.

11 ... f5 12 Bf3 g5 13 exf5 Nxf5 14 g3 Nd4

All as before; Britton