Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How to draw a won pawn endgame

Author: Martin Giepmans

Date: 13:51:09 05/13/03

Go up one level in this thread


On May 13, 2003 at 16:28:07, John Merlino wrote:

>On May 13, 2003 at 15:18:22, Martin Giepmans wrote:
>
>>
>>From a game Ruffian - SpiderChess
>>
>>[D]8/7p/p2k4/1p4p1/4K1P1/P6P/8/8 b - - 0 50
>>
>>
>>
>>In this position SpiderChess played 50 .. Kc5? (it should have played
>>Ke6 Kd4 a5 -+) and thought it was winning.
>>
>>The game continued:
>>51. Kf5 Kc4 52. Kxg5 Kb3 53. Kh6 Kxa3 54. Kxh7 b4 55. h4 b3 56. g5 b2
>>57. h5 b1=Q+ 58. g6 Qe4 59. h6 a5 60. Kg7 Qe7+ 61. Kh8 Qf8+ 62. Kh7 Qf5 63.
>>Kg7 Kb3 64. h7 Qe5+ 65. Kg8 Qe8+ 66. Kg7 Qe7+ 67. Kg8 Qd8+ 68. Kg7 Qc7+ 69.
>>Kg8 Qc8+ 70. Kg7 Qc3+ 71. Kg8 Qc4+ 72. Kg7 Qd4+ 73. Kg8 Qd5+ 74. Kg7 Qb7+
>>75. Kg8 Qa8+ 76. Kg7 Qa7+ 77. Kg8 Qb6 78. Kg7 Kc2 79. h8=Q Qd4+ 80. Kg8
>>Qd5+ 81. Kh7 Qh5+ 82. Kg7 Qe5+ 83. Kh7 Qh5+ 84. Kg7 Qe5+ 85. Kh7 Qh2+ 86.
>>Kg8 Qb8+ 87. Kh7 Qh2+ 88. Kg7 Qc7+ 89. Kh6 Qf4+ 90. Kh7 Qh4+ 91. Kg7 Qe7+
>>92. Kh6 Qe3+ 93. Kh7 Qh3+ 94. Kg8 Qb3+ 95. Kf8 Qb8+ 96. Kg7 Qb7+ 97. Kh6
>>Qh1+ 98. Kg7 Qb7+ 99. Kh6 Qh1+ 100. Kg7 Qb7+ 101. Kh6 Qh1+ 102. Kg7 Qb7+
>>{Drawn by repetition} 1/2-1/2.
>>
>>How do other engines fare?
>>
>>Martin
>
>Chessmaster 9000 finds nothing wrong with 50...Kc5:
>
>Time	Depth	Score	Positions	Moves
>0:00	4/8	-2.69	6235		50...a5 51.Kd4 h6 52.Kc3 Ke5 53.Kd2
>					Kd4 54.Ke2
>0:00	5/9	-2.64	12849		50...a5 51.Kd4 h6 52.Kc3 Kd5 53.Kc2
>					b4 54.axb4 axb4 55.Kb3
>0:00	6/10	-2.65	24484		50...a5 51.Kd4 h6 52.Ke4 Kc5 53.Kd3
>					b4 54.axb4+ axb4 55.Kc2 Kc4
>0:00	6/10	-2.80	41681		50...Kc5 51.Kd3 a5 52.Kc3 a4 53.Kd3
>					b4 54.axb4+ Kxb4 55.Kd4 a3
>0:00	7/11	-2.87	53163		50...Kc5 51.Kd3 a5 52.Kc3 b4+ 53.axb4+
>					axb4+ 54.Kb3 Kb5 55.Kb2 Kc4 56.Kc2
>					b3+ 57.Kb2 h6
>0:00	8/12	-3.26	74726		50...Kc5 51.Kd3 a5 52.Kc3 b4+ 53.axb4+
>					axb4+ 54.Kb3 Kb5 55.Kb2 Kc4 56.Kc2
>					b3+ 57.Kb2 h6 58.Kc1 Kd4
>0:00	9/13	-3.17	120989		50...Kc5 51.Kd3 a5 52.Kc3 b4+ 53.axb4+
>					axb4+ 54.Kb3 Kb5 55.Kb2 Kc4 56.Kc2
>					b3+ 57.Kb2 h6 58.Kc1 Kc3
>0:01	10/14	-3.36	182816		50...Kc5 51.Kd3 a5 52.Kc3 b4+ 53.axb4+
>					axb4+ 54.Kb3 Kb5 55.Kb2 Kc4 56.Kc2
>					b3+ 57.Kb2 h6 58.Kb1 Kc3 59.Kc1
>					b2+ 60.Kb1
>0:01	11/15	-3.29	317502		50...Kc5 51.Kd3 a5 52.Kc3 b4+ 53.axb4+
>					axb4+ 54.Kb3 Kb5 55.Kb2 Kc4 56.Kc2
>					b3+ 57.Kb2 Kb4 58.Ka1 Kc3 59.Kb1 b2
>0:02	12/16	-3.78	514692		50...Kc5 51.Kd3 a5 52.Kc3 Kd5 53.Kd3
>					Ke5 54.Ke3 a4 55.Kd2 Kf4 56.Kd3
>					Kg3 57.Ke4 Kxh3 58.Kf5 b4
>0:04	13/17	-3.78	804669		50...Kc5 51.Kd3 a5 52.Kc3 Kd5 53.Kd3
>					Ke5 54.Ke3 a4 55.Kd2 Kf4 56.Kd3
>					Kg3 57.Ke4 Kxh3 58.Kf5 b4
>0:10	14/18	-4.75	1578501		50...Kc5 51.Kd3 a5 52.Kc3 Kd5 53.a4
>					b4+ 54.Kd3 h6 55.Kc2 Kc4 56.Kb2
>					b3 57.Kc1 Kb4 58.Kb1 Kxa4 59.Kb2
>					Kb4
>0:23	15/19	-4.97	3134673		50...Kc5 51.Kd3 a5 52.Kc3 Kd5 53.Kd3
>					Ke5 54.Ke3 a4 55.Kd2 Kf4 56.Kd3
>					Kg3 57.Ke4 Kxh3 58.Kf5 b4 59.axb4
>					a3 60.Kxg5
>5:21	16/20	-7.34	41907384	50...Kc5 51.Kf5 a5 52.Kxg5 b4 53.axb4+
>					axb4 54.Kh6 b3 55.Kxh7 b2 56.g5
>					b1=Q+ 57.Kh6 Qh1 58.g6 Qxh3+ 59.Kg7
>					Kd6 60.Kg8 Ke7 61.Kg7 Qc8 62.Kh7
>					Qf8
>
>It appears that 51...Kc4 (using the king to capture the pawn) instead of 51...a5
>(to trade off pawns) was the error.
>
>After 50...Kc5 51.Kf5 a5! 52.Kxg5 b4 53.axb4+ axb4 54.Kh6 b3 55.Kxh7, Black has
>a tablebase Mate in 10 with EITHER 55...b2 or 55...Kd5.
>
>The only other option for White (after 50...Kc5) that I can find is 51.Kd3,
>which also appears to lose after 51...a5. White can respond with 52.Kc3, but
>after 52...a4! White is in zugzwang. White's only other option is 52.Ke4 which
>is met by 52...b4, trading off pawns to a won game for Black.
>
>jm

You're right. The mistake came later.
Tablebases would probably have helped here.
The game was played without tablebases.

Martin



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.