Knowing, and Following, the Opening’s Correct Plan

| 12 | Amazing Games

IMPORTANT: [At the end of the puzzles, you should click MOVE LIST so you can see my instructive notes and variations. If you are having trouble solving a problem, just click SOLUTION, and then MOVE LIST. Even if you solve everything, DO click MOVE LIST or you might miss an important bit of prose.]

Big_T_Maya (1241) – Thor9 (1567), ICC 2012 (time control, 2/10)

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6

The beauty of the King’s Indian Defense is that you can play it against everything but 1.e4.

5.Nf3 0-0 6.Be2 e5 7.d5

Nothing wrong with this popular move. Most common is 7.0-0 Nc6 8.d5 Ne7 (the Mar del Plata Variation), which leads to mind-numbing mega-theory (Black plays for the thematic …f7-f5 advance followed by a blistering kingside attack, while White plays for b2-b4 and c4-c5 with a queenside roll). In general, endgames favor White (since white’s queenside advantages work well in an endgame, but Black often avoids an endgame by scoring a brutal knockout on the kingside).

Other good lines are 7.dxe5 (seeking a Queen exchange and a tiny endgame edge) and 7.Be3. 

Here is an old example of a typical Mar del Plata Variation attack:


Most popular. It stops white’s b2-b4 advance and prepares to swing the b8-Knight to the c5-square. Kasparov has tried both this, 7…Nbd7, and 7…Na6.


Hard to argue with his obvious move. However, 8.Bg5 is white's main choice (the Petrosian Variation) – it pins the Knight and makes it hard for Black to get in the usual “Knight on f6 retreat followed by …f7-f5 plan”. Other acceptable moves for White are 8.h3 (Kramnik), 8.Be3, and even 8.Nd2. Here are a couple examples of 8.Bg5:

8…Na6 9.a3? 

 White’s best moves are 9.Ne1, 9.Bg5, 9.Be3, or 9.Qc2. Though your desire to play b2-b4 is understandable and well-intentioned, 9.a3 is badly timed. In fact, this move is very instructive since it shows something that White needs to avoid at all costs! 

This position demonstrates a key KID idea: White would love to chase away the well-placed c5-Knight and gain tons of queenside space by a2-a3 followed by b2-b4. However, 1.a3? is a mistake since it allows 1…a4 which stops white’s queenside expansion cold (2.b4 axb3 e.p.). Due to this, White first plays 1.b3! and only then (at some point) a2-a3 and b3-b4. Here’s the whole game which shows this idea (and a general queenside takeover) in crystal clarity:


Here's a problem that will show if you learned anything:

10.Bg5 Nc5 11.Qc2 a4

And there it is – White has absolutely nothing since his dreams of queenside expansion have been axed. 

12.Nd2 Qc8?! 

I guess he’s stepping out of the pin. However, if he intends to play …h6 next move, why not do that right away? Then, if White responds with Be3 (as he did in the game) …Qc8 will be completely unnecessary. 


Not a move that should be played lightly. The reason is that black’s normal plan is a kingside assault, and the h3 move might turn out to be a target later in the game. 

13…h6 14.Be3


Black could also have considered 14…b6 intending to meet 15.Bxc5 with …bxc5.


This is either a horrible move that makes no sense at all, or a very deep move that anticipates some cool stuff on move 16. Since White didn’t see the real point behind 15.Nf3, it has to be labeled a mistake (and with best play it's shown to be useless). Instead, White should generate queenside play before Black gets his kingside stuff rolling. Thus 15.Bxc5 dxc5 16.b4 axb3 17.Nxb3 b6 18.a4 followed by a4-a5.


The thematic move, which gains kingside space and instantly gets the black Rooks into the battle. However, 15…b6! should have been tossed in and only then …f7-f5.


A bit embarrassing, but it gives e4 support and also stops …Nb3. The only way to make 15.Nf3 worthwhile is 16.Nh4! when Black is suddenly forced to come up with some imaginative stuff in order to avoid a serious disadvantage:


I guess …b6 just isn’t on black’s agenda.

17.f3 f4 18.Bf2

By now it’s clear that White has no intention of ever chopping on c5, but if he doesn’t do that, and if he holds off of b2-b4, then where is his play coming from?

18…g5 19.Kh2? 

If White doesn’t get his queenside play going, he’s going to get smashed on the kingside. It’s do or die, leap or perish, eat or starve.

19…h5 20.g4?

White’s trying to stop an unstoppable attack. I say “unstoppable” because a kingside attack is black’s KID birthright. Perhaps he hasn’t played accurately, but at least Black is following the correct general idea (kingside attack). White on the other hand is wasting much too much time, and soon the position will take a huge bite out of him.


The capture on g4 is bad for a completely different reason than the winning solution found in the above problem. This kingside structure is a famous one, and a clear strategy for Black is known here: 

Dominate the h-file! Since Black has access to 3 squares on the h-file (h6, h7, and h8) while White has only 2 (h1 and h2), black should leave the pawn tension alone, play …Kf7 (you can toss in …b6 here), followed by …Rh8-h6, and then place the other Rook and Queen on the h-file. White won’t be able to deal with that buildup (or if he does, he will be very passively placed) , and when Black finally chops on g4 White will be toast. Note that by capturing on g4 so early (there was zero reason to do so since White is certainly not going to take on h5) he suddenly gives White 3 squares on the h-file too, thereby negating black’s plan.

Here's a classic example:


As mentioned above, thanks to the extremely bad 20…hxg4 the h-file is now everyone’s property. While without that capture the h-file was black’s.

21…Kf7 22.Rh1 Qe8 23.b4

Since Black missed a clear win on move 21 and also ravaged his advantageous pawn structure, black’s kingside attack seems stillborn, while white’s queenside counterplay is finally making little noises.

23…axb3 24.Nxb3 Nxb3 25.Qxb3 b6 26.Kg2 Ba4 27.Qb4

27.Nxa4 Qxa4 (27…Rxa4?? 28.c5! threatening Bb5) 28.Qxa4 Rxa4 is okay for Black. 

27…Kg6 28.Bd1?

28.Nb5 was comfortable for White (It seems to make sense to exchange white’s bad Bishop for black’s good one, but it turns out that the “bad” Bishop was actually serving a key function), but 28.Bd1 actually weakens white’s kingside structure.

28…Bxd1 29.Raxd1 Qc8! 

This is a good move with deadly intentions, but Black doesn’t know why it’s a good move, and either does White!



Going after the solidly defended pawns on a3 and c4. Poor Black starts to play on the wrong side of the board.

31.Nb5 Qb7? 

A sad home for black’s Queen. 

32.Qb1 Rh8 33.a4 Rxh1 34.Qxh1 Rh8 35.Qb1 Rh7 

To his credit, white’s been actively playing on the queenside, while Black has been frozen in doubt and confusion. Now 36.a5 would be the usual way to crash through on his chosen wing (as usual, Black would still have counterplay on the other side of the board), but instead White is overwhelmed with creativity and comes up with some otherworldly stuff! 


Good or bad, wise or insane, White has just earned some serious respect! Clearly White is going whole hog on the queenside.


I’m sure Black feels like he’s been overrun by a tsunami. However, a calm look at the position would help him find the truth: We are somehow back to the basic KID doctrine, which has White rolling through the queenside and Black doing this thing on the other wing. Since White is indeed doing a queenside bum rush, Black needs to remember he’s playing a King’s Indian and turn his attention to the White King. With that in mind, 36…Qc8! (threatening both 37…Qh8 and 37…Nxg4) should have been tried: 

In the game Black chose 36…dxc5 which seems reasonable enough. I would now expect White to continue his queenside roughhousing with 37.a5, but instead we get something that blew my mind.


Both doomed to fail and wonderful at the same time. We know in our hearts that this can’t work, but we stand amazed and awestruck at white’s glorious audacity. 


This is an understandable reaction. Black says, “What?” and chops the Bishop off. Another way was 37…Qc8 (with the usual threats of 38…Nxg4 and 38…Qh8) when Black would have a powerful kingside attack. 

38.a5 c6?!

Poor old Thor9 keeps playing on white’s side. Even though all his pieces (but the Queen) are on the kingside, it never occurs to him to play there. Black wins hands down after 38...Qc8 39.Qg1 Nxg4 40.fxg4 Qxg4+ 41.Kf2 Qh4+ and the attack crashes through.

39.a6 Qxb5?? 

Thor9 continues to walk down white’s fascinating path. Instead, he should have blazed his own trail to white’s King by, you guessed it, 39…Qc8! 40.Nd6 Qh8 41.Qg1 Nxg4! 42.fxg4 Rh4 and wins.

40.Qxb5 cxb5 41.a7 Rh8 42.a8Q Rxa8 43.Rxa8

A remarkable position! Black has two minor pieces and a pawn vs. white’s Rook, which leads one to believe that Black should be doing really well. But the following factors get in the way of that assessment: the minor pieces aren’t doing much, the Rook is active, White has a nice passed d-pawn, black’s queenside pawns are vulnerable to attack by white’s Rook, and the black King is too far away from the queenside. All this somehow creates a more or less even game.

The rest could fill many more pages, but I’ll go through it with minimal notes (see the final game board at the end for the whole shebang):


43…c4 44.Rb8 followed by Rxb5; 43…b4 44.Rc8 Nd7 45.Rc7.


44.Rd8! forces the win of one of the pawns and probably draws. 


Now white’s in trouble.

45.Ra7 Ke7 46.Ke2 Bf8 47.Kd2 Kd8 48.Kc3 Bd6 49.Ra6 Ke7 50.Ra7 Bb8 51.Rb7 Kd8??

51…b4+ suggested itself: 52.Kc4 Kd6 53.Rb5 (53.Kb3 Bc7 54.Ra7 Nb6 55.Ra6 Kd7 56.Ra7 Kc8 47.Ra1 Bd6 followed by …c4+) 53…Bc7 54.Rb7 Bb6 55.Kb3 Ke7 followed by …Ke7-d8-c8, winning white’s Rook.

52.Rxb5 Kc7 53.Kc4 Ba7 54.Ra5 Bb6 55.Ra6 Nf8 56.Kb5 Nd7 57.Kc4 Kb7 58.Kb5 c4 59.Ra3 Bd4 60.Kxc4 Nc5 61.Ra5 Na6 62.d6 Nb8 63.Kd5 Nd7 64.Rb5+ Kc8 65.Rb4 Nb6+ 66.Kc6 Nd7 67.Rb7 Nb8+ 68.Rxb8+ Kxb8 69.d7, 1-0.



* Having an opening that can be played against almost everything (in this case the King’s Indian Defense) really cuts down study time.

* The idea of “freezing” a set of pawns (as shown in the notes to 9.a3) is a very important one. You should be able to avoid this happening to you, and punish others that allow it.

* A move like h2-h3 (on move 13 in this game) has its uses, but can also turn into a weakening of your kingside pawn structure.

* When you make a move, you need to be able to explain why you made it. If you can’t explain the logic of your move, then don’t play it! In the case of white’s 15.Nf3, it’s hard to understand what his idea was.

* If your opponent is attacking your King and your only hope is to overwhelm him with a queenside assault, it’s death if you waste moves defending. You need to get your stuff going as quickly as possible!

* We’ve been taught to exchange bad Bishops for enemy good Bishops. But sometimes a “bad” Bishop is a key defensive piece. One can’t allow oneself to get trapped by general rules – they are guidelines, not ultimate truth. 

* Understanding pawn structures is extremely important. The comments about black's 20th move (and the game that follows it) is well worth a long, detailed look.


If you want me to look over your game, send it to

I need your name (real or handle), your OPPONENT’S name (real or handle), both players’ ratings, where the game was played, and date. If you don’t give me this information, I won’t use your game! BTW: I’ve noticed that many people are reluctant to give me their opponent’s name. This is very strange! Showing the names of both players is the way chess games are presented in databases, books, magazines… everywhere! Permission from the opponent isn’t necessary. If permission was necessary, everyone who ever lost a game wouldn’t allow their name to be on it!

More from IM Silman
The Downs And Ups Of GM Elmars Zemgalis (Silman's Last Article)

The Downs And Ups Of GM Elmars Zemgalis (Silman's Last Article)

How To Build Winning Chess Positions

How To Build Winning Chess Positions