Author: Sune Larsson
Date: 15:37:06 05/26/02
Go up one level in this thread
On May 26, 2002 at 17:02:17, Roy Eassa wrote: >On May 26, 2002 at 08:36:08, Sune Larsson wrote: > >> >> [D]k1r5/1p1b4/p1q1r2p/3pRpp1/PB1P4/1QP3P1/5RPP/6K1 b - - 0 36 >> >> This is a position from the game GM Votava-IM Berg in the 8th round >> of the International Hamburg-ch. With only 4 moves to go before the >> 1st time control, Emanuel decides to set a little trap for his opponent >> with 36.-Tce8! Votava walks right into it with 37.Rfxf5 Rxe5 38.Rxe5 Rxe5 >> 39.dxe5 Qb6+ 40.Kf1 >> >> Test 1: Avoid this move sequence for white, starting with 37.Rfxf5. >> >> > > >Can we assume that, with perfect play, Black wins after 36...Rce8 37.Rfxf5? Hi Roy. Yes, the moves after the taking on f5 are forced. > >If so: > >1) What should White do instead on move 37 to avoid losing? I think white must be very humble in this position (after 36.-Rce8) and back home for a draw. Something like 37.Rxe6 Qxe6 38.Bc5 (some counterplay with a possible Qb6) 38.-Qe1+ 39.Rf1 Qe3+ 40.Kh1 Qf2 41.Qb1. It's late so this is by no means a complete analyse. Point is that 37.Rfxf5 just loses. Alexander is right of course with the little trick 54.-b6! in the pawn ending. > >2) This is a good example of something a human can see and understand that a >computer cannot (it is *much* too deep for any computer) -- do you agree? Yes, completely. After Rce8 the calculation goes like this: " If take, take, take, take, take then I have Qb6+ and a5. Then he must take on d5 which gives me a passer on b2. Yes, that must be winning. He has to put his queen on b1." Patterns of knowledge then play a big role when evaluating the upcoming pawn ending. You first visualize the position you want to reach - then work out a way to get there. > Sune > > > >> [D]k7/1p1b4/pq5p/3pP1p1/PB6/1QP3P1/6PP/5K2 b - - 0 40 >> >> Now comes the point - creating a deadly passer and forcing a won >> pawn ending. 40.-a5! 41.Qxd5 axb4 42.Qxd7 b3 43.Qd3 b2 44.Qb1 Qb3 >> 45.Ke2 Qxc3 46.Kd1 Qd4+ 47.Kc2 Qxa4+! 48.Kxb2 Qb5+ 49.Kc2 Qxb1+ 50.Kxb1 >> >> Test 2: Find 40.-a5! and 41.-axb4 >> >> >> [D]k7/1p6/7p/4P1p1/8/6P1/6PP/1K6 b - - 0 50 >> >> This pawn ending is lost for white. >> >> Test 3: Steer into this ending with 47.-Qxa4+! >> Evaluate the above position as won for black. >> >> >> The whole game below. >> >> >>[Event "IHEM 2002"] >>[Site "Hamburg"] >>[Date "2002.05.25"] >>[Round "8.2"] >>[White "Votava, Jan"] >>[Black "Berg, Emanuel"] >>[Result "0-1"] >>[WhiteElo "2508"] >>[BlackElo "2514"] >>[PlyCount "116"] >>[EventDate "2002.??.??"] >> >>1. e4 e6 2. d4 d5 3. Nc3 Bb4 4. Bd3 Ne7 5. Nge2 c5 6. exd5 exd5 7. a3 Bxc3+ 8. >>Nxc3 c4 9. Be2 Nbc6 10. Bf3 Be6 11. Ne2 Nf5 12. c3 g5 13. Ng3 Nxg3 14. fxg3 h6 >>15. O-O Qd7 16. Bh5 O-O-O 17. Be3 f5 18. Qd2 Qd6 19. Rae1 Rdf8 20. Qe2 Rf6 21. >>Bc1 Rhf8 22. Rf2 Bd7 23. Ref1 Kb8 24. Qd1 Ka8 25. Qd2 Ne7 26. Bd1 Be6 27. Bc2 >>a6 28. b4 cxb3 29. Bxb3 Nc6 30. a4 Na5 31. Qa2 Rc8 32. Ba3 Qc6 33. Bb4 Nxb3 34. >>Qxb3 Bd7 35. Re1 Re6 36. Re5 Rce8 37. Rfxf5 Rxe5 38. Rxe5 Rxe5 39. dxe5 Qb6+ >>40. Kf1 a5 41. Qxd5 axb4 42. Qxd7 b3 43. Qd3 b2 44. Qb1 Qb3 45. Ke2 Qxc3 46. >>Kd1 Qd4+ 47. Kc2 Qxa4+ 48. Kxb2 Qb5+ 49. Kc2 Qxb1+ 50. Kxb1 Kb8 51. Kc2 Kc7 52. >>Kd3 Kd7 53. h4 g4 54. h5 Ke6 55. Ke4 b5 56. Kd4 b4 57. Kc4 Kxe5 58. Kxb4 Ke4 >>0-1 >> >>Sune
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.