Author: Robert Hyatt
Date: 08:22:33 10/31/02
Go up one level in this thread
On October 31, 2002 at 08:35:37, Peter Berger wrote:
>Crafty seems to evaluate KRRKRN as a dead draw.
>
>KnightDreamer v3.1b - Crafty-19.0
>[D]5R2/8/8/5n1R/4k1rP/8/5K2/8 b - - 0 1
>
>Analysis by Crafty 19.01:
>
>77...Sxh4 78.Txh4 Txh4
> = (0.00) Tiefe: 12/22 00:00:06 4463kN, tb=5298
>77...Sxh4 78.Txh4 Txh4
> = (0.00) Tiefe: 13/23 00:00:16 11939kN, tb=11852
>
>How about this one? Still dead draw according to Crafty - or is it?
>
>KnightDreamer v3.1b - Crafty-19.0
>[D]8/8/2R5/1R6/4rn2/8/8/k4K2 b - - 0 1
>
>Analysis by Yace after a little training:
>
>92...Ta4 93.Ke1 Ta7 94.Kd2 Ka2 95.Kc2 Se2 96.Td6 Ka3 97.Tb3+ Ka4 98.Tb1 Ka5
>99.Ta1+ Kb5 100.Txa7 Kc5 101.Kd3 Sc1+ 102.Kd2
> +- (7.39) Tiefe: 15/29 00:00:02 1531kN
>
>
>I haven't checked with 6-men statistics or tablebases, but I really doubt that
>this ending is that drawish at all in general - at least a 0.0 can lead to
>serious problems as in this blitzgame:
>
>[Event "Computer chess game"]
>[Site "P422"]
>[Date "2002.10.31"]
>[Round "1"]
>[White "KnightDreamer v3.1b"]
>[Black "Crafty-19.0"]
>[Result "1-0"]
>[TimeControl "240+2"]
>
>1. d4 Nf6 2. c4 e6 3. Nc3 Bb4 4. e3 c5 5. Ne2 d5 6. a3 Bxc3+ 7. Nxc3 cxd4
>8. exd4 dxc4 9. Bxc4 Nc6 10. Be3 O-O 11. O-O b6 12. Qd3 Bb7 13. Rad1 h6 14.
>f3 Rc8 15. Rfe1 Ne7 16. Bd2 Nf5 17. Bf4 Nh5 18. Be5 Nh4 19. Ne4 Bxe4 20.
>Rxe4 Ng6 21. Ba6 Rc6 22. Qb5 Qd5 23. Qa4 f6 24. Kh1 Kh8 25. Bb7 b5 26. Bxc6
>Qxc6 27. Qxa7 Qc2 28. Rde1 fxe5 29. R4e2 Qd3 30. dxe5 Nhf4 31. Qe3 Rd8 32.
>Qxd3 Nxd3 33. Ra1 Kh7 34. b3 Rc8 35. a4 Rc3 36. a5 Ngf4 37. Rd2 Nc5 38.
>Rda2 Na6 39. Rb1 Nd3 40. Kg1 Nxe5 41. Rab2 Nc6 42. Rd2 b4 43. Ra2 Kg6 44.
>Re1 Kf5 45. Rd2 Rxb3 46. Rd6 Nab8 47. Rdxe6 Ra3 48. h4 h5 49. R1e4 Ra1+ 50.
>Kh2 Ra2 51. a6 Nd4 52. a7 Nbc6 53. a8=Q Nxf3+ 54. Kg3 Rxa8 55. gxf3 Ra6 56.
>Rd6 b3 57. Rd5+ Kf6 58. Rb5 Ne7 59. Rf4+ Ke6 60. Rxb3 Nf5+ 61. Kh3 Kf6 62.
>Rc4 Rd6 63. Rc8 g6 64. Rf8+ Kg7 65. Ra8 Rd1 66. f4 Kf7 67. Rb7+ Ke6 68.
>Rb6+ Nd6 69. Rba6 Rd4 70. Kg3 Kf5 71. Rf8+ Ke4 72. Kf2 Kd3 73. Kg1 Nf5 74.
>Rxg6 Rxf4 75. Rg5 Ke4 76. Rxh5 Rg4+ 77. Kf2 Nxh4 78. Re8+ Kd3 79. Re3+ Kd2
>80. Rd5+ Kc1 81. Rc3+ Kb2 82. Rh3 Rg2+ 83. Kf1 Rg4 84. Rd2+ Kc1 85. Rf2 Re4
>86. Rhh2 Ng6 87. Ra2 Kb1 88. Rhb2+ Kc1 89. Rc2+ Kb1 90. Rab2+ Ka1 91. Rb5
>Nf4 92. Rc6 Ra4 93. Ke1 Nd3+ 94. Kd2 Nf4 95. Re5 Ng2 96. Rc2 Rf4 97. Re2
>Rh4 98. Rxg2 Rh1 99. Kd3 Rh3+ 100. Ke2 Rh1 101. Rd2 Kb1 102. Rd1+ Kc2 103.
>Rxh1 Kc3 104. Rh4 Kb2 105. Kd1+ Kc3 106. Rg3+ Kb2 107. Ra4 Kb1 108. Rb3#
>{KnightDreamer checkmates!} 1-0
Hard to answer, but I can't imagine that KRR vs KRN is winnable very often,
based on the
.tbs file on my web site. Almost all are draws. Of course, without the KRRKRN
egtb,
a wrong move can result in a loss, but in gereral it is drawn, almost all the
time.
The position you gave is an egtb mate in 19:
1. ... Ra4! 2. Ke1! Ka2! 3. Kd2 Ka3! 4. Rc3+ Ka2 5. Rc8 Ne6 6. Kc1!
Nd4 7. Rb2+! Ka1! 8. Rc7 Ra6! 9. Rc4! Rc6! 10. Rxc6! Nxc6! 11. Kc2!
Nd4+! 12. Kc3! Nf5 13. Rb5! Ne3! 14. Re5! Nc4! 15. Kxc4! Kb2! 16.
Re2+! Kc1! 17. Kd3 Kb1 18. Kc3! Ka1! 19. Kb3! Kb1 20. Re1#
after white's move 78, black should have played Kf4 and that is a draw according
to the
endgame tables. Kd3 leads to a mate in 32. Here is the output:
({11:+2.19} 70. ... Kf5 71. Rf8+ Ke6 72. Kf2 Ke7 73. Raa8 Rd3
74. Kg2 Rd2+ 75. Kh3 Ke6 76. Rfd8 Ne4 77. Rxd2 Nxd2 $18)
71. Rf8+ Ke4
({13:+1.87} 71. ... Ke4 72. Kf2 Kd3 73. Ra3+ Kd2 74. Ra2+ Kd3
75. Kf3 Nf5 76. Ra3+ Kc2 77. Rc8+ Kb2 78. Ra5 Nxh4+ 79. Ke3 Rd1 $18)
72. Kf2 Kd3
({12:+1.94} 72. ... Kd3 73. Ra3+ Kd2 74. Ra2+ Kd3 75. Kf3 Nf5
76. Rh2 Re4 77. Rh3 Re3+ 78. Kg2 Ke4 79. Rxe3+ Kxe3 $18)
73. Kg1 Nf5
({12:+0.01} 73. ... Nf5 74. Rxg6 Rxf4 75. Rg2 Ke4 76. Rg5 Nxh4
77. Rxh5 $10)
74. Rxg6 Rxf4
({12:+0.01} 74. ... Rxf4 75. Rxf5 Rxf5 76. Rg5 Ke4 77. Rxf5
$10)
75. Rg5 Ke4
({13:+0.01} 75. ... Ke4 76. Rfxf5 Rxf5 77. Rxf5 $10)
76. Rxh5 Rg4+
({16:+0.01} 76. ... Rg4+ 77. Kh2 Rxh4+ $10)
77. Kf2 Nxh4
({16:+0.01} 77. ... Nxh4 $10)
78. Re8+ Kd3
({4:Mat32} 78. ... Kd3 $18)
({4:+0.01} 78. ... Kf4 $10)
79. Re3+ Kd2
({4:Mat31} 79. ... Kd2 $18)
({4:Mat32} 79. ... Kc2 $18)
80. Rd5+ Kc1
({4:Mat26} 80. ... Kc1 $18)
({4:Mat30} 80. ... Kc2 $18)
So it is an easy ending to draw with tables, based on the .tbs summary, but it
might well be
a _very_ difficult one to draw for humans...
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.