Chess

THE LEAD changed hands at Hastings this week thanks to a bad misjudgment of a King and Pawn endgame by Yevgeny Bareyev against Judit Polgar. In the diagram position after the 40- move time control, Bareyev (White), with weak pawns on e5 and h5 stood worse, but after 41. Rf3 a draw still looks the likely result. Instead, he allowed an exchange of rooks with 41. Rh3? Rxh3 42. gxh3.

The next few moves were forced: 42 . . . Kc6 43. a4 Kd5 44. a5 (otherwise the e-pawn is lost for nothing) Kc6 45. Kd3 Kb5 46. b3 Kxa5 47. Kc4 Kb6 and this is the position that White needed to calculate before exchanging rooks.

ch07out-harts-nws

White's three pawns seem to hold Black's four on the K-side, and his king keeps Black's out of d5, but a little logic shows that White is lost.

It is easy to see that 48. b4 loses: after 48 . . . cxb4 49. Kxb4 Kc6 50. Kc4 Kd7 51. Kc5 Ke7 and Black wins with f6, so the pawn must stay on b3.

The rest is a study of related squares. When Black plays Kc6, White must reply Kc4 (to stop Kd5); when Black plays Kb5, White must play Kc3 (to prevent Kb4 and c4). So after Black's Kb6, White must stay in contact with c3 and c4, so must play Kd3. Continuing the process, after Kc7 (connecting with c6 and b6), White must be next to d3 and c4, so must play Kc3, but he needs the same move in reply to Kb7. QED

After 48. Kd3 Kc7 49. Kc3 Kb7] White gave up. 50. Kc4 Kc6] or 50. Kd3 Kb6] end it.