<Draws Don't Count

I am often asked which of my games is the most memorable and usually I refer to this one - my game against E.Geller in the USSR Championship when I first became USSR Champion.

Every grandmaster has played many intresting games , but the memory is not stirred by subtle opening ideas, solid middlegame play, or intricate endgames. No, the games that you remember are the games with the most sporting significance. As a rule, the more mistakes there are in the game, the more memorable it remains, because you have suffered and worried over each mistake at the board. In my life as a chess player there have been many such battles; the outcome of them usually had adecisive influence on the final result of a strong tournament. But even amongst games of this importance, the game with Geller is pre-eminent and the tension of the battle is evident.

1... ♘f6
<It is generally thought that the sharpest and most active defence to 1.e4 is the Sicilian. However is there not a resemblance between the Sicilian defender and the wrestler who begins the fight before he has climbed into the ring?>

11.♗f4?!
<As Black has chosen an unusual, 'incorrect', move order, White does not need to play his queen's bishop to e3, and therefore Geller tries to improve its position compared with the normal line. Nevertheless, playing 'according to theory' with 11.exf6 ♗xf6 12.♗e3 ♘c6 13.♕d2 created more chances>

17... ♘e3!
<Clearly after 17... ♔xg7 18.♕d4+ White has the the advantage - but what about 17... ♘xb2 ? Yes during the game I seriously thought of this possibilit, altough it is quite clear to me now - and everuy master will agree with me without any concrete analysis of lines - that the weakening of Black's king position is more than sufficent compensation the pawn. I played the remainder of the game on the following principle: in view of the weakening of your own king position the greatest harm that you can do to your opponent is to make a breach in the defences of the enemy king!>

29.♘e2?
<White could have still drawn by 29.♕d8, but now he is faced with insurmountable difficulties.>

30.♘g1
<if 30.♕g1 then 30...♕h4 (with threats of ...♘h2+ and ...♗d3) and if 31.{Qd4 then 31...b6 32.♗a3 c5 e.g. 33.♕d6+ ♔f7 ♕d5+ ♔xe7! and wins ; or 32.♘f4+ ♔h6 33.♕d5 ♘h2+ and 34...♕xf4.>

31.♕d8
<If immidiately 31.♗a3 then 31...♕h4 was very strong.>

Fischer vs Geller, 1962

where Geller built up a winning advantage, wrote down the winning move on his scoresheet, then erased it and substituted a lemon, and went down to his only loss of the tournament. Had Geller won this game, he'd have played Botvinnik in 1963.

<This was the destruction of all Geller's hopes. A destruction which Petrosian had foreseen, and which could in no way have been accidental. Disastrous reversals of form had accompanied Geller in many tournaments, and here it had come again, a definite psychological instability, which revealed itself in the most dramatic moments of the struggle. Twenty-two rounds without defeat, and now this failure practically reduced to nought his previous efforts.>

I thought this was a slightly indelicate comment, considering that Vasiliev's book was published around 1974, and Geller was still a very active player (he won the Soviet Championship again in 1979).

Another example that the same book gives is Geller's performance in his first Soviet Championship in 1949. Petrosian had an awful tournament, starting off with 5 straight losses, but recovering somewhat to go +1 the rest of the way. Geller, on the other hand, did so well that he nearly won the tournament. Here's what Vasiliev says about that one:

Geller vs Kholmov, 1949

The Geller-Portisch game you mentioned is a remarkable one too.

Geller vs Portisch, 1973

After fighting for 30+ moves in a dead drawn, pawn up ending, Geller suddenly oversteps in a still-dead-drawn position. I find this game almost impossible to explain, even with psychological theories. He could have played almost ANY move and still drawn this. The only way to lose was to play no moves. Incredible.