This complicated and unusual game seems to have confused the participants (judging from their errors) and annotator Lev Aronin, whose notes, published while the match was in progress, are republished in the book Igor Botvinnik published about the rematch. Below a selection of Aronin’s notes are in brackets, while my/Shredder’s comments are in plain text. <Talisman> was right to query Tal’s 22nd move, but there were several other turning points. Anyway, errors aside, it was an amazing game!

1.e4 c6 2.d4 d5 3.e5

The first of eight times Tal played this move in the rematch, scoring +1-2=5. In a post-match press conference, Botvinnik criticized Tal for preparing just one variation. Botvinnik, by contrast, repeatedly varied his responses.

3….c5

<Botvinnik thought over this move for more than ten minutes and, as we see, he avoided the ‘most theoretical’ 3….Bf5. The reason why a particular move was made or was not made remain, up to a certain time, known only to the participants themselves. After the conclusion of the match both grandmasters will probably report on much that for the moment cannot be fully disclosed.

However, the move 3….c5, aiming for an immediate attack on the center, has, along with 3….Bf5, every right to exist. Moreover, one can recall games where this move led, in the end, to success for Black. This is what happened, for example, in the game Spassky-Kotov (22nd USSR Championship).> Spassky vs Kotov, 1955

4.dxc5 e6 5.Nc3

In games 6 and 8 Tal played the sharper 5.Qg4. After losing the latter game, Botvinnik switched to 3….Bf5 for the remainder of the match.

5….Nc6 6.Bf4 Nge7

Shredder thinks Black is a little better after 6….Bxc5 7.Qg4 Kf8: if 8.Nf3 Bxf2+ 9.Kxf2 Qb6+ 10.Be3 Qxb2 11.Bd3 Qxc3. But Botvinnik didn’t tend to play this way against Tal….

7.Nf3 Ng6 8.Be3

8….Ncxe5 9.Nxe5 Nxe5 10.Qh5

<A useful active move, putting the opponent’s kingside under fire. >

10….Nc6 11.0-0-0 Be7

<Perhaps the immediate 11….g6 would have been better here, and in reply, say, to 12.Qe2 (or 12.Qf3) – 12….Bg7. After 13.Nb5 it is true that Black has to castle queenside (but this is also not bad). In the mutual sharp attacks on the flanks, that are possible in this case, Black’s chances look fully equivalent. In playing 11….Be7, Botvinnik evidently thought it advisable to retain control of d6 (in the event of a possible Nb5) and to create a possible threat to the c5 pawn.>

12.f4

<Not only impeding the Black pawns in the center, but creating the threat of f4-f5.>

12….g6 13.Qh6 Bf8 14.Qg5<!>

<Again an interesting decision. It is true that there is no longer any other acceptable queen move, but it is obvious that all this was planned by Tal beforehand. White goes in for the removal of yet another one of his pawns from the center, after which it soon becomes inevitable that Black will create two connected passed pawns. In return, however, White gains new possibilities.

14….Qxg5

<Would it not have been better to continue developing with 14….Bd7, leaving White himself to decide what to do with his queen?> Botvinnik also could have played 14….Be7, inviting a repetition after 15.Qh6 Bf8; Shredder likes 14….Bg7 and thinks Black has slightly the better of a sharp position. Tal clearly hasn’t gotten anything out of the opening. 15.fxg5

<Now the threat of Nb5 becomes very serious, together with a possible c2-c4. Therefore Botvinnik makes the following move, after which Black is practically forced to sacrifice the exchange. However, as will become clear from what follows, Tal should not have been in too much of a hurry with the winning of the exchange.>

15….a6.

Igor Botvinnik’s book gives 15….h6. Since that doesn’t prevent Nb5, I assume it is a typo and 15….a6 as in the cg score is what Botvinnik actually played. Shredder comes up with a seemingly simple and strong alternative: 15….d4. Now if 16.Nb5 (16.Bxd4 Nxd4 17.Rxd4 Bxc5 ) 16….Bxc5 17.Nxd4 Bd7 (or 17….Bxd4 18.Bxd4 Nxd4 19.Rxd4 Bd7 20.Be2 Bc6 21.Bf3 Bxf3 21.gxf3 Rd8 =). But it turns out White has an strong exchange sacrifice of his own: 16.Rxd4! Nxd4 17.Bxd4 Rg8 18.Nb5 (or Ne4).

After Botvinnik’s 15….a6, things get very complicated.

16.Na4 Bd7 17.Bf4<!> (Shredder thinks this is a mistake, as we’ll see) 17….h6

<Since ….e6-e5, weakening the black pawns in the center, is obviously unacceptable, Botvinnik ensures the win of the g5 pawn in return for the exchange, which he is ready to give up. All of his calculations for the future are associated, of course, with the creation of two connected passed pawns in the center – a factor which almost decides the outcome in his favor.>

Instead Shredder thinks Black is better after 17….Rd8 with the positional threat of 18….e5 19.Re1 Bg7. (Black can’t play …e5 immediately because of 18.Re1 Bg7? 19.Nb6 Rd8 20.Nxd5.) 18.Re1 Be7 19.Nb6 0-0 20.Bc7 Bxg5+ 21.Kb1 Be7 is a stronger version of the exchange sacrifice in the game. Instead of 17.Bf4, Shredder thinks White could have kept equality with 17.Nb6 Rd8 18.c3.

18.Nb6 Rd8 19.Bc7 hxg5 20.c4 d4

Instead Shredder thinks it’s about equal after 20….Bxc5 21.Bxd8 Nxd8 22.Nxd7 Kxd7 23.cxd5 e5, though it’s very hard to evaluate the position. Now Tal gets his queenside marjority rolling.

21.b4 Bg7

click for larger view

22.Bxd8

<In the heat of the battle, almost without thinking, the world champion snatches the rook. As shown by Botvinnik’s subsequent play, which up to a point is impeccable, this natural capture was a mistake, leading to a decisive strategic advantage for Black. Meanwhile, by contining 22.b5 Nb8 23.a4 e5 24.Re1! f6 25.g3 and then 26.Bg2 White could have set his opponent very serious problems.>

22….Kxd8 23.b5 Nb8

<Now it is hard for the four white pawns on the queenside to make any progress, whereas the advance of Black’s passed pawns is irresistible.>

24.Be2 f5 25.Bf3

<It is hard to offer White any better advice. He is aiming to clear lines on the queenside, since he has two rooks in reserve, but with accurate play Botvinnik achieves a winning position. >

25….axb5 26.cxb5 (26.Bxb7 at once would have been a little better) Bxb5 27.Bxb7 Kc7 28.a4

<After 28.Bf3 the reply 28….g4 would have been immediately decisive, while if 28.Ba8, then 28…..Na6.>

28….Bxa4 29.Nxa4 Kxb7

<The position has simplified. The black king has acquired the excellent square c6, the c5 pawn is sooner or later doomed, and the black pawns are still enormously strong. Despite the world champion’s exceptional mastery in creating complications and finding tactical resources, it seemed improbable that White would be able to save the game.>

30.Kd2 Nd7 31.Rb1+ Kc6 32.Rhc1

click for larger view

32…Be5

Shredder thinks Black is completely winning after the cold-blooded 32….Rxh2 33.Rb6+ Nxb6 34.cxb6+ Kb5 35.Rc7 Rxg2+ 36. Kc1 Rg1+ 37.Kc2 d3+ 38.Kxd3 Kxa4 39.Rxg7 Ka4 40.b7 Rb1.

Instead Aronin recommends 32….Ra8, which is probably also good enough, but concludes that 32…..Be5 also wins.

33.Kd3
<Not only defending against 33….Bf4+, but also intending to advance the king.>

33…..Ra8

<But now after this move Black is no longer winning. Meanwhile, by continuing 33….Bc7, he would still have had every justification for counting on a win, since 34.Kxd4, immediately or after 34.Rb6+ Bxb6 35.cxb6+ Kd6 36.Kxd4, is not possible on account of …Rh4+.>

34.Rb6+<!> Nxb6 35.cxb6+ Kd7 36.Nc5+ Ke7 37.Re1 Ra3+

<Of course, not 37….Kf6 on account of 38.b7 Rb8 39.Rxe5.> As translator Ken Neat points out, 38….Ra3+ is much better than 38….Rb8, but in any event 38.Nd7+ wins instantly. Being an annotator was a risky trade in the days before engines.

38.Kc4 Rc3+

<If immediately 38….Re3, then 39.Rxe3 dxe3 40.Kd3 is evidently sufficient for a draw.>

Amazingly, it appears that White can hold after 40….Bxh2 41.Na6 e5 42.b7 Bf4 43.b8/Q e4+ 44.Ke2 Bxb8 45.Nxb8 f4 46.Nc6+ Kd6 (or 46.Kf6 47.g4! fxg3 48.Kxe3) 47.Nd8 Kd5 48.Nf7.

39.Kb5 Re3 40.Ra1 Bxh2 41.Ra7+

<The game was adjourned in this position and Botvinnik sealed the move 41.Ke8. The king could not go to f6 since after Rh7 White would even win.

After 41….Kf8 the simplest is 42.Nd7+ (42....Kg7 43.g4!) with a draw. But after 41….Ke8, the move sealed by Botvinnik, White forces a draw by 42.Rh7 Bf4 43.Nd7 d3 44.Nf6+ when he gives perpetual check. Therefore, without resuming the game, the grandmasters agreed a draw.>

After 41….Kf8 Shredder doesn’t care for the immediate 42.Nd7+, but 42.Rh7 Bg3 43.Nd7+ is a simple draw.