A very nice quote from Boobby which sums this game up nicely.

Who is the person who said that about this game?

If instead <15...g6>

16.Nh6+ Kg7
17.Bc3

<17...Qb5!>
18.h4 Re8
19.Ng4 Qf5!
20.Nh6! Qc5!
21.Ng4 Qf5 =

<17...Re8?>
18.Ng4 c5
19.Rhe1 Rxe1
20.Rxe1 Kg8
21.Qd5!! Nxd5
22.Re8+ Nf8
23.Nh6++

<9.0-0-0?! "A big advantage would have been given by 9.Nd6+ Bxd6 10.Qxd6 Qe7+ 11.Qxe7 when the two bishops in an open position are not easy to neutralize" (Tal). This would indeed have been simpler and wiser – since also after 10…Ne4 11.Bxd8 Nxd6 12.Bc7 Black is condemned to a difficult struggle for a draw. But Tal, with amazing psychological insight, unusual for one so young, guessed that Smyslov would feel more confident in an inferior endgame than in a double-edged middlegame, and he decided to keep the queens on.>

Btw, after 9.Nd6+ Bxd6 10.Qxd6, Qa5+ followed by Qc5 also looks like a possibility for black.

<13…Bd8! Supporting the knight at f6 – the key piece of the defense. If the bishop, following the queen, had abandoned its king – 13…Bc5?, then Tal's fantastic idea would have appeared in all its glory: 14.Qh4 bxc4 15.Bc3! Believe it or not, black is defenseless against the numerous threats (Rxd7, Nxg7, Qg5, Ng5). For example: 15…Qxa2 16.Rxd7! Bxd7 17.Nh6+ Kh8 18.Qxf6!! with mate.>

Going over the game with Fritz 10, Kasparov's analysis of the last variation appears somewhat superficial. While not denying that 13…Bd8 is better than 13…Bc5, it should be pointed out that after 13…Bc5 14.Qh4 bxc4 15.Bc3 Qxa2 16.Rxd7:

click for larger view

16…Bxd7 is an extremely weak move by black and not forced at all (let's recall that from a material viewpoint white has, for the moment, just balanced his piece sacrifice and nothing more). Instead, black has the resource 16...Bb4(!) with the twofold idea of deflecting the white bishop from the a1-h8 diagonal in case of Bxb4, and making the queen check on a1 a mate, in case of Bxf6. After 17.Nh6+ Kh8 18.Bxb4 Bxd7 19.Bc3 ,black doesn't seem to have anything better than giving up his knight by 19…Ne4 (on 19…Qa1+ 20.Kd2 Qxh1 there's again the pretty 21.Qxf6) 20.Qxe4 Rae8 (20…Qa1+ 21.Kd2 Qxh1 22.Ng5) 21.Qf4 (Qd4/h4 are also possible) f6 (still not 21…Qa1+ 22.Kd2 Qxh1 23.Bxg7+ followed by Qg5+ with mate) 22.Nd2 (now white finally has to defend from Qa1+):

click for larger view

Where white seems to have an advantage, but is it a clearly winning one?

I found a file on the internet which had the corrections that have been made or will be made to the second edition of OMGP.

Supporting the knight at f6 – the key piece of the defense. If the bishop, following the queen, had abandoned its king – 13…Bc5?, then Tal's fantastic idea would have appeared in all its glory:

<14.Qf4!>

(Unclear is 14.Qh4 bxc4 15.Bc3 Qxa2 16.Rxd7?! Bb4!? 0r 16.h5!?, but 16.Nxg7! Bxf2! 17.Qf4! Nd5 18.Nf5!! is decisive)

<14...bxc4

15.Bc3! >

Black is defenseless against the numerous threats (Rxd7, Nxg7, Qg5, Ng5)

I'm still unconvinced by the analysis, however... after 13...Bc5 14.Qf4 bxc4 15.Bc3:

click for larger view

Analyzing 15...Qxa2 here is doing the same "trick" of choosing a crappy move for Black which allows for spectacular wins by White. What about 15...Re8! (which is also good, btw, in case of 14.Qh4) - how can White win then?

(A) 16.Qg5 Bf8 17.Bxf6 Nxf6 18.Nh6+ Kh8 19.Nxf7+ Kg8 and I see nothing better than perpetual for White here.

(B) 16.Bxf6 Nxf6 17.Nxg7 Qxa2 18.Nxe8 Qa1+ 19.Kd2 Qa5+ 20.Ke2 Nxe8 21.Qg5+ Kf8 (21...Kh8?? 22.Rd8) 22.Qh6+ Kg8 (22...Ke7?? 23.Ne5; 22...Ng7 23.Qxh7 looks dangerous for Black - note that if White had played 20.Kc1, though, 23...c3! would be strong here) - and again, I see nothing better than perpetual for White.

(C) 16.Rxd7 Bxd7 17.Nh6+ Kf8 18.Bxf6 Re6! 19.Qg5 (a very complicated position - other options for White here are Bxg7+, Ne5 and Ng5, but this looks best)...gxh6 20.Qxc5+ Kg8 (20...Ke8? 21.Re1) 21.Qd4 Rxf6 22.Qxf6 Qxa2 with what seems like equality.

“I chose this move, not thinking of gaining any advantage for White, but to avoid the usual variations.” - Tal

<2…d5 3.Nd2 e5>

A standard move in this position, but also possible are the more elastic 3…Nf6 and 3…g6

<4.Ngf3 Nd7>

Passive play from Black, accepted theory today is the more natural 4…Bd6

<5.d4>

“The opening of the center secures an advantage in development for White” - Tal

<5…dxe4 6.Nxe4 exd4 7.Qxd4 Ngf6 8.Bg5 Be7 9.0–0–0>

“"A big advantage would have been given by 9.Nd6+ Bxd6 10.Qxd6 Qe7+ 11.Qxe7 when the two bishops in an open position are not easy to neutralize" (Tal). This would indeed have been simpler and wiser – since also after 10…Ne4 11.Bxd8 Nxd6 12.Bc7 Black is condemned to a difficult struggle for a draw. But Tal, with amazing psychological insight, unusual for one so young, guessed that Smyslov would feel more confident in an inferior endgame than in a double-edged middlegame, and he decided to keep the queens on.” - Kasparov

<9...0–0 10.Nd6 Qa5 11.Bc4 b5 12.Bd2>

“A very important intermediate move, the main aim of which is to free the square g5 for the knight, or, as we shall see later, to activate the Queen” - Tal

<12…Qa6 13.Nf5 Bd8>

Supporting the key piece of the defense.

13...Bc5 14.Qf4 bxc4 15.Bc3

(A) 15...Qxa2 16.Nxg7 Rd8 17.Nf5 Qa1+ 18.Kd2 Qa4 19.Qg5+ Kf8 20.Qg7+ Ke8 21.Rhe1+ Ne5+ 22.Nd4 Bxd4

(B) 15…Re8 = and White cannot make any progress

(B1) 16.Qg5 Bf8 17.Nh6+ Kh8 18.Nxf7+ Kg8 19.Nh6 = with a draw by repetition.

(B2) 16.Bxf6 Nxf6 17.Nxg7 Qxa2 18.Nxe8 Qa1+ 19.Kd2 Qa5+ 20.Ke2 Nxe8 21.Qg5+ Kf8 22.Qh6+ Kg8 = with a draw by repetition.

(A) 15…Ne8 16.Qxd8 Qxa2 17.Bc3 Ndf6 (17…Nef6 18.Rxd7 ) 18.Ne7+ Kh8 19.Qa5 Qxa5 20.Bxa5 Ng4

(B) 15...g6 16.Nh6+ Kg7 17.Bc3

(B1) 17…Qb5 18.h4 Re8 19.Ng4 Qf5 20.Nh6 Qc5 21.Ng4 Qf5 = with a draw by repetition.

(B2) 17…Re8 18.Nf5+ Kg8 19.Qh6 gxf5 20.Qg5+ Kh8 21.Rhe1

(B3) 17...Qxa2 18.Ng4 h5 19.Rxd7 hxg4 20.Rxd8 Qa1+ 21.Kd2 Rxd8+ 22.Ke3 Re8+ 23.Kf4 Qxh1 24.Bxf6+ Kg8 25.Qh4 Qc1+ 26.Kg3

<16.Nh6+ Kh8 17.Qxh5 Qxa2>

Thought to be an error but is in fact the only move which retains any advantage for Black.

(A) 17…Bf6 18.Nxf7+ Kg8 19.Ng5 f6 20.Ne4 Qxa2 21.Nxf6+ Nxf6 22.Qa5 Qxa5 23.Bxa5 Bg4 =

(B) 17…gxh6 18.Bc3+ f6 19.Qxh6 Rg8 20.Ng5 Rxg5 21.Qxg5

<18.Bc3 Nf6>

The losing move.

(A) 18…Bf6

(A1) 19.Ng5 Bxg5 20.Qxg5 f6 21.Qg3 Ne5 22.Bxe5 fxe5 23.Nf7+ Kg8 24.Nh6+ = with a draw by repetition.

(A2) 19.Nxh7+ Kg8 20.N3g5 Qa1+ (20…Bxg5 21.Nxg5 Nf6 22.Bxf6 Bf5 ) 21.Kd2 Bxc3 22.bxc3 Nf6 23.Qxh7+ (23.Rxa1 Nxh5 24.Ra5 h6 25.Ne4 Rxf7 26.Rxh5 Bf5 ) 23…Nxh7 24.Rxa1 Nxg5 25.Nxg5 Rxf2+ 26.Ke3

(B) 18…Bc7

(B1) 19.Nxf7+ Kg8 20.Nh6+ Kh8 21.Nf7+ = with a draw by repetition.

(B2) 19.g3 Nf6 20.Qh4 Bg4 21.Qxf6 Qa1+ 22.Kd2 Rad8+ 23.Ke3 Rde8+ (23…Rfe8+ 24.Qe5 Qxd1 25.Rxd1 Rxe5+ 26.Nxe5 Re8 ) 24.Kd2 Red8+ = with a draw by repetition.

<19.Qxf7 Qa1+>

A necessary move to prevent White mating with Rxd8+ after …Rxf7

White might be able to make progress, actually:

(Thanks!)<Jim Bartle>

After 1 e4 c6, Tal chose 2 d3 as an option to take the Caro-Kann opening out of its usual “book” lines. At move 9, Tal has the opportunity for a positional advantage. In “The Battle of Chess Ideas,” Anthony Saidy writes, “Better is 9 Nd6 Bxd6; Qxd6 Qe7+; Qxe7 Kxe7; o-o-o with a superior endgame, but Tal prefers something more exciting. Is this bias a lack of objectivity? No? a warrior is entitled to use his own favorite weapons.” Rybka’s evaluation of white’s alternatives at move 9 is:

Nd6 (.29)
Bxd6; 10 Qxd6 Ne4; 11 Bxd8 Nxd6; 12 Bc7 Ne4; 13 Nd2 Nxd2; 14 Kxd2 Nc5

o-o-o (.17)
o-o; 10 Nxf6 Bxf6; 11 Bxf6 Qxf6; 12 Bd3 Qxd4; 13 Nxd4 Nc5; 14 Rhe1 Be6

Nxf6+ (.15)
Nxf6; 10 o-o-o Qxd4; 11 Nxd4 h6; 12 Bh4 o-o; 13 Bc4 Bd7; 14 Rhe1 Rae8

Tal elected to castle queenside at move 9, and at that move had probably planned the Bc4 line instead of the positional knight exchange Rybka gives in the o-o-o continuation. The fireworks start with 11 Bc4 and black’s reply b5. At move 11, Rybka scores White’s options as:

Nxc8 (.00)
Rfxc8; 12 Qh4 Nc5; 13 Re1 Re8; 14 Bc4 b5; 15 Rxe7 Rxe7; 16 Bxf6 gxf6; 17 Qxf6 Qc7

Bd3 (-.03)
Qxa2; 12 Rhe1 Bxd6; 13 Qxd6 Qa1+; 14 Kd2 Qxb2; 15 Ke2 a5; 16 Kf1 h6; 17 Bd2 Nd5

Kb1 (-.06)
Bxd6; 12 Qxd6 Ne4; 13 Bd2 Qd8; 14 Qd4 Ndf6; 15 Bd3 Qxd4; 16 Rxd4 c5; 17 Ra4 Bd7

a3 (-.11)
Bxd6; 12 Qxd6 Ne4; 13 Bd2 Qh5; 14 Qd4 Ndf6; 15 Bd3 Re8; 16 Bxe4 Rxe4; 17 Qd6 Bg4

Bc4 (-.16)
b5; Bxf7 Rxf7; Nxf7

Tal’s choices at moves 9 and 11 are a clear definition of his chess style. Instead of the positionally “best” continuation, Tal chooses an option for complications which he is betting will result in Tal getting to his opponent’s King before his opponent can solve the defensive problems. Psychologically, most players do not enjoy being attacked, and the pressure Tal creates will lead to calculation mistakes. Tal’s real genius is that more often than not, he creates his pressure while playing with “house money.” Correct calculation by the opponent will lead to a basically level position, while miscalculation creates large advantages for Tal.

In response to Black’s 11 … b5, Tal writes, “Accepting White’s invitation to the “dance of death.” Better was Nb6.”

Rybka scores black’s options at move 11 as:
b5 (-.06)
12 Bxf7 Rxf7; 13 Nxf7

Nb6 (.16)
12 Bb3 Be6; 13 Bxe6 fxe6; 14 Kb1 Bxd6; 15 Qxd6 Qf5; 16 Bh4 Nc4; 17 Qd4 b5;

c5 (.26)

h6 (.27)