Author: Tobias Lagemann
Date: 16:33:23 01/12/06
Go up one level in this thread
Hello Eduard, >Interesting analyses! > >Firs the game: > >[Event "Thessaloniki ol (Women)"] >[Site "Thessaloniki"] >[Date "1988.??.??"] >[Round "?"] >[White "Xie Jun"] >[Black "Polgar, Judit"] >[Result "0-1"] >[ECO "B47"] >[BlackElo "2320"] >[PlyCount "106"] >[EventDate "1988.11.??"] >[EventType "team"] >[EventRounds "14"] >[EventCountry "GRE"] >[Source "ChessBase"] >[SourceDate "1998.??.??"] > >1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 Nc6 5. Nc3 Qc7 6. Be2 a6 7. O-O Nf6 8. >Kh1 Nxd4 9. Qxd4 Bc5 10. Qd3 h5 11. f4 Ng4 12. Qg3 b5 13. Bxg4 hxg4 14. Qxg4 >Bb7 15. Qe2 b4 16. Nd1 Ke7 17. Be3 Bxe3 18. Qxe3 Rh4 19. h3 Rah8 20. Kg1 g5 21. >fxg5 Bxe4 22. Qf2 f5 23. gxf6+ Kf7 24. Qe3 Bxg2 25. Qg5 R4h7 26. Kxg2 Rg8 27. >Qxg8+ Kxg8 28. Ne3 Qc6+ 29. Kg3 Qd6+ 30. Rf4 Rxh3+ 31. Kxh3 Qxf4 32. Ng4 e5 33. >Rg1 Kf7 34. Rg2 Qf3+ 35. Kh2 d6 36. Rg3 Qf4 37. c3 bxc3 38. bxc3 Qd2+ 39. Kh3 >Qxa2 40. Rf3 Qd2 41. Kh4 e4 42. Re3 Qg2 43. Kg5 Ke6 44. c4 a5 45. Kf4 Qf1+ 46. >Kxe4 Qxc4+ 47. Kf3+ Kf5 48. Re5+ Kg6 49. Re7 Qd3+ 50. Kf4 Qf5+ 51. Kg3 Kh5 52. >Rg7 a4 53. Rg8 Qd3+ 0-1 > >We see here (please load for the analysis the full game), that black is in >advanced after 22...f5! (like the BS-Test-Set). But in my opinion the game is >very tactical and after 22...f5! For an amateur player not easy to win. Or? > >Ok, when not easy to win for me, what is then the next best move? Interesting is >22...Bxg2! With a draw. > >But if I let run the test (FEN from BS-Test) I see that many engines thinks >about 22...Th7 +=, 22…Kd8 += or 22...Bg6 += in the way to finds 22...f5 -+! But >only little engines says me that Bxg2 is here an draw. Have You know this? > >Why it is also? Why can not every engine find the move Bxg2, with a draw line? > >Example: > >Xie Jun - Polgar,J, Thessaloniki ol (Women) 1988 > >[D]7r/2qpkp2/p3p3/6P1/1p2b2r/7P/PPP2QP1/R2N1RK1 b - - 0 1 > >Analysis by Rybka 1.01 Beta 9 32-bit: > >22...Lxg2 23.Kxg2 Txh3 24.Dxf7+ Kd8 25.Df6+ Ke8 26.Tf4 --->Rybka see the draw! > = (0.00) Tiefe: 9 00:00:15 147kN >22...f5 23.Te1 Dxc2 24.Dxc2 Lxc2 25.Tc1 Le4 26.Tc5 ---> and then the win ! > ³ (-0.30) Tiefe: 9 00:00:23 249kN >22...f5 23.Te1 Lxg2 24.Dxg2 Txh3 25.Se3 Ke8 26.Sxf5 > -+ (-1.99) Tiefe: 10 00:01:13 836kN > >(Nemeth, Privat 12.01.2006) > >7r/2qpkp2/p3p3/6P1/1p2b2r/7P/PPP2QP1/R2N1RK1 b - - 0 1 > >Analysis by Shredder 9.1 UCI: > >22...Kd8 23.Se3 T8h7 24.a3 b3 25.Df6+ Kc8 26.cxb3 --->Shredder 9 UCI += > ² (0.48) Tiefe: 8/25 00:00:02 159kN >22...T8h7 > ² (0.47) Tiefe: 8/25 00:00:02 172kN >22...T8h7 23.Se3 Ke8 24.a3 b3 25.Tac1 De5 > ² (0.30) Tiefe: 8/28 00:00:03 208kN >22...T8h7 23.a3 Dxc2 24.axb4 Dxf2+ 25.Sxf2 Lb7 26.Sg4 d6 27.Kf2 > ² (0.29) Tiefe: 9/30 00:00:05 360kN >22...T8h7 23.a3 Dxc2 24.axb4 Dxf2+ 25.Sxf2 Lb7 26.Sg4 d6 27.Kf2 T4h5 += > ² (0.30) Tiefe: 10/33 00:00:09 683kN >22...f5 ---> and now, Shredder 9 UCI see abrupt a line for the win. Draw-Line >see S9 not! > ² (0.29) Tiefe: 10/33 00:00:11 923kN >22...f5 > = (-0.06) Tiefe: 10/33 00:00:13 1013kN >22...f5 23.gxf6+ Kf7 24.De3 > = (-0.06) Tiefe: 10/33 00:00:16 1276kN >22...f5 23.gxf6+ > ³ (-0.31) Tiefe: 11/29 00:00:21 1685kN >22...f5 23.Te1 Dxc2 24.a3 b3 25.Sc3 Dxf2+ 26.Kxf2 d5 27.Tac1 Kd6 > ³ (-0.68) Tiefe: 11/31 00:00:27 2232kN >22...f5 23.Te1 T4h5 24.g6 > ³ (-0.59) Tiefe: 12/33 00:00:44 3736kN > >(Nemeth, Privat 12.01.2006) > >We see here, that Rybka saw us one much better analysis about this position than >Shredder 9 UCI. > >Where I am wrong? > >Eduard. please test the position after 22... Bxg2! I'm sure Shredder would be say draw, too. Testing the position you have posted only shows the way of "thinking" of engines... For Rybka "22... Bxg2 with = 0.00" is the best move for one moment in time... (Tiefe: 9, Time: 15sek), a few seconds later Rybka find 22... f5 with (-0.30) In "Tiefe 10" 22... f5 rules.... -+ (-1.99) To find the best move is a kind of step by step.... The engines all used user steps .... Okay, here's a sidestep: Take a look at SlowChess Blitz WV... 7r/2qpkp2/p3p3/6P1/1p2b2r/7P/PPP2QP1/R2N1RK1 b - - 0 1 Analysis by SlowChess Blitz WV: 2...T4h7 3.Se3 ± (0.76) Tiefe: 2 00:00:00 2...T4h7 3.Se3 d5 ² (0.64) Tiefe: 3 00:00:00 2...T4h7 3.Se3 De5 4.Df6+ Dxf6 5.Txf6 ± (0.76) Tiefe: 4 00:00:00 2...T4h7 3.Se3 De5 4.a3 b3 ² (0.68) Tiefe: 5 00:00:00 2...T4h7 3.Se3 De5 4.Df6+ Dxf6 5.gxf6+ Kd6 6.Kf2 ± (0.88) Tiefe: 6 00:00:00 128kN 2...T4h7 3.c3 Ld3 4.Df6+ Ke8 5.Tf4 Dc5+ 6.Kh2 Tg8 7.cxb4 Dxg5 ± (0.92) Tiefe: 7 00:00:01 557kN 2...Lg6 ± (0.92) Tiefe: 7 00:00:04 941kN 2...Lg6 3.Df6+ Ke8 4.Tf2 Dg3 5.Tf3 De1+ 6.Kh2 Lf5 ² (0.68) Tiefe: 7 00:00:04 1043kN 2...T8h7 ² (0.68) Tiefe: 7 00:00:05 1218kN 2...T8h7 3.c3 Ld3 4.Df6+ Ke8 5.Tf2 Dg3 6.Tc1 ² (0.64) Tiefe: 7 00:00:06 1432kN 2...T8h7 3.Tc1 De5 4.Df6+ Dxf6 5.gxf6+ Kd6 6.Se3 Ld5 7.c4 ± (0.80) Tiefe: 8 00:00:08 2464kN 2...Lg6 ± (0.80) Tiefe: 8 00:00:11 3763kN 2...Lg6 3.Se3 Dc5 4.Tad1 Dxg5 5.Dd2 Db5 6.Sg4 ² (0.68) Tiefe: 8 00:00:11 4119kN 2...Lg6 3.a3 Dxc2 4.Df6+ Ke8 5.axb4 De2 6.Sf2 Lf5 7.Tac1 ² (0.68) Tiefe: 9 00:00:15 5787kN 2...Lxg2 ² (0.68) Tiefe: 9 00:00:16 6797kN 2...Lxg2 3.Dxf7+ Kd8 4.Kxg2 Txh3 5.Df6+ Ke8 6.Kf2 Dxc2+ 7.Ke1 De4+ 8.Kd2 Dd3+ 9.Kc1 Dc4+ 10.Kb1 De4+ 11.Kc1 Dc4+ ² (0.43) Tiefe: 9 00:00:16 7856kN 2...f5 ² (0.43) Tiefe: 9 00:00:22 10547kN 2...f5 3.gxf6+ Kf7 4.Se3 Txh3 5.gxh3 Txh3 6.Dd2 Dg3+ 7.Sg2 Dh2+ 8.Kf2 Tf3+ 9.Ke1 Txf1+ 10.Kxf1 Dh1+ 11.Kf2 Dxa1 = (0.00) Tiefe: 9 00:00:27 13682kN 2...f5 3.Te1 Dxc2 4.Dxc2 Lxc2 5.Tc1 Le4 6.Kf2 Tg8 7.Kg3 = (-0.16) Tiefe: 10 00:00:58 29388kN 2...f5 3.Te1 Dxc2 4.Te2 Dc7 5.Se3 Kf7 6.a3 bxa3 7.bxa3 d5 = (-0.24) Tiefe: 11 00:01:36 49226kN 2...f5 3.Te1 Dxc2 4.Te2 Dc7 5.De3 Df4 6.Sf2 Ld5 7.g6 Dxe3 8.Txe3 = (-0.24) Tiefe: 12 00:03:10 97672kN 22... Bxg2 looks not so good (+0.43, Tiefe: 9, 16 sec) 22... f5 is draw! (Tiefe: 9, 27 sec.) Uhps... 22...f5, +0.43 (Tiefe: 9 00:00:22) seems as good as 22... Bxg2 Okay, 22... Bxg2 is draw..., but for Slow Chess "one moment in time" 22... f5 looks like draw.... Best regards Tobias
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.