Opening surveys in CBM 132

King's Indian with 6.h3 (part I)

by Michal Krasenkow

At the beginning of the 90s I had serious problems dealing with the King's Indian. The "rare" classical system with 9.Nd2 I had started playing several years earlier became extremely popular after the Kasparov-Karpov match in Seville, and its theory started growing like Godzilla. Therefore I had to look for something less hackneyed. In 1991, during the C'an Picafort open tournament, I was sharing a room with future GM Igor Khenkin. Before my game against GM Jeroen Piket I told Igor about my problems. "Why don't you play the Bagirov system?", he asked. "What Bagirov system?" - "6.h3".

1.d4 Nf6 2.c4 g6 3.Nc3 Bg7 4.e4 d6 5.Nf3 0-0 6.h3

Igor showed me some games by GM Vladimir Bagirov from recent "Informants" (we didn't have computers yet), which White won quite convincingly. I liked the general idea of the system - restriction of the opponent's active possibilities on both sides, with a chance to create an attack on the black king. According to the database, it was first played in... 1855 by John Cochrane in several games in India (well, are those games authentic?), then occasionally applied by players like Reti, Saemisch, Gruenfeld, Teichmann, Tarrasch, Bogoljubow, Spielmann, Makogonov, Lilienthal, Tolush, Larsen, Kavalek, Olafsson, etc. - there is even one game by Kasparov (Bagirov's influence?)! However, in fact, it was Bagirov who laid the foundations of the strategic concept of the system. So I decided to give it a try. Against Jeroen I didn't play very confidently (the game ended in a draw) but further experience was much more successful - apparently, the 6.h3 system has brought me one of the highest percentage scores among all the openings of my repertoire.

The point is that, compared to the other systems in the King's Indian, it is quite rarely seen in practice and therefore paid little attention by Black players. At the same time the system is not strategically simple. It is very difficult (especially for a poorly prepared player) to find a "hole" in White's strategy of restriction. Moreover, White players who regularly apply and analyse this system will discover those holes and patch them in advance. I'll tell you more about all these strategical niceties when we examine the main lines of the system in parts II and III. Here I'll mention only one thing, which is more or less common: the white dark-squared bishop should (if allowed) go to g5 rather than e3, in order to provoke the weakening ...h7-h6. Therefore White usually plays Bc1-g5 at once and plans further development depending on Black's reaction.

Black's main reply to 6.h3 is 6...e5. We'll deal with it in parts II and III (planned for CBM 133 and 134). Now let's take a look at the numerous alternatives.

A) ...c7-c6, ...a7-a6, ...b7-b5. A well-known plan. If Black starts it at once, White simply develops his pieces, advances in the centre (e4-e5) and obtains a comfortable position, as shown in Manolache,M - Vajda,A ½-½. In some games Black tried a more cunning move order, first taking the e5-square under control: 6...Nbd7 7.Bg5 a6, keeping an option to play ...c7-c5. In Krasenkow,M - Schmidt,W 1-0 and several other games I successfully replied 8.Nd2, controlling the important c4 and e4 squares.

B) 6...a6 7.Bg5 c5. The idea is that White can't play 8.d5 in view of 8...b5! However, central play starting with 8.dxc5 brought White a pleasant position in Kazhgaleyev,M - Evdokimov,A 1-0 and other games.

C) 6...Na6. A flexible move keeping two options: ...e7-e5 and ...c7-c5. It is typical for several systems of the King's Indian. However, here White's reply is obvious: 7.Bg5,

and now Black must decide what to do. To 7...c6 I like 8.Nd2. White is ready to meet 8...e5 with 9.d5, and if Black tries to avoid that structure then he can't create any pressure on White's centre. Krasenkow,M - Kempinski,R 1-0 is an interesting illustration of manoeuvring by both sides.

7...Qe8 is probably more appropriate, with a positional threat 8...e5 9.d5 Nh5. To prevent that, White should apparently play 8.g4. Now 8...e5 9.d5 leads to the 6...e5 7.d5 Na6 line, which we'll discuss in part II. The alternative 8...c5 9.Bg2 (9.d5?! e6 is now favourable for Black) 9...cxd4 leads to a good Maroczy structure for White, with the black queen absurdly placed on e8. GM Andrey Kovalev's attempt 9...h5!? proved insufficient in Nielsen,P - Michelakis,G 1-0.

D) 6...c5 7.d5. In this structure (often arising when Black starts with 2...c5 3.d5 and then plays 3...g6 etc.) White can safely develop his bishop to d3. However, there is an important thing: he should not hurry with short castling! If Black closes the centre completely (...e7-e5), it will be more appropriate to play g2-g4 (preventing ...f7-f5) and castle queenside or leave the white king in the centre. Therefore White should first make other useful moves: Bc1-g5, Qd1-d2 etc. If Black decides on ...e7-e6xd5 then White's short castling will be appropriate. Bareev,E - Damljanovic,B 1-0 is the classic example: Black made his own useful moves like ...Nb8-a6-c7 but eventually took on d5. Please note that Black played ...h7-h6, and White later won a tempo playing Qd1-d2. In another remarkable game Psakhis,L - Chatalbashev,B 1-0 Black refrained from ...h7-h6 and even managed to push ...b7-b5 but then the pin along the h4-d8 diagonal proved too annoying.

The immediate 7...e6 8.Bd3 exd5 is a separate story. 9.cxd5 now leads to a popular system of the Modern Benoni Defence (A70), which is outwith the scope of this article. Alternatively, White can play 9.exd5. If Black now allows White's castling, White keeps a small but clear edge due to his space advantage (Maric,A - Prudnikova,S ½-½ is an example).

9...Qe7+ is not so good either as the position of the queen is quite unfortunate here (see Tregubov,P - Evdokimov,A 1-0).

Therefore Black usually plays 9...Re8+ 10.Be3 Bh6 (again White is slightly better if Black refrains from this active move, as shown in annotations to Krasenkow,M - Mikrut,D 1-0) 11.0-0! Bxe3 12.fxe3.

Now he can hardly take the e3-pawn (White obtains a strong attack using his big advantage in development, after 13.Qd2 Re8 14.Qh6 etc.); therefore, Black tries to create a blockade on the e5-square. However, he still suffers from a lack of space while a lot of pieces remain on the board. White's plans are: doubling rooks along the f-file; the queenside attack by means of a2-a3 and b2-b4; sometimes he transfers his queen to g3. It is interesting that the queen exchange doesn't necessarily help Black as his queen is an important defending piece.

Marin,M - Rausis,I 1-0 (where Black played 12...Nbd7) and Belous,V - Kokarev,D ½-½ (with 12...Qe7) illustrate the ideas of the position.

Of course, this survey concerning the side lines (especially 6...c5) is quite brief. They mostly give White a pleasant position with a play "for two results". 6...e5 is a much more important and principled continuation.