Larsen surprised Spassky in the opening with the rare move 10....e5. Larsen says that, in home analysis, he had seen the position through to move 17, but when it appeared on the board during the game he worried that he had evaluated it wrongly. "I was rather nervous during the game." But Spassky used up an hour in the opening and Larsen got an advantage in Spassky's time trouble.

Spassky's plan with 24.Bb3 turns out not to work. It was better to retreat the bishop to f3. In his notes to the game, Spassky is very impressed with Larsen's play in the middlegame. Larsen's scattered, 'weak' pawns turn out to be very effective - they're not easy to attack and they control key squares. That kind of paradox (weak pawns that turn out to be strong) is, I think, central to Larsen's style.

White's critical chance was 29.Rc6, but after 29....Re4, black is close to winning in all variations.

Larsen played 37....h5 to take advantage of Spassky's time pressure - although says that it was also probably the best move.