Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Funny pos

Author: Uri Blass

Date: 02:19:56 08/10/03

Go up one level in this thread


On August 10, 2003 at 04:23:56, Drexel,Michael wrote:

>On August 09, 2003 at 20:45:47, John Merlino wrote:
>
>>On August 09, 2003 at 17:25:47, Vincent Lejeune wrote:
>>
>>>On August 09, 2003 at 15:38:47, Yar wrote:
>>>
>>>>On August 09, 2003 at 15:16:21, John Merlino wrote:
>>>>
>>>>>On August 09, 2003 at 15:05:39, Yar wrote:
>>>>>
>>>>>>Hello !
>>>>>>
>>>>>>What is your engine says about this pos ?
>>>>>>[D] r1r3kb/pp2pp2/1q1p1npB/6Q1/2nNPNP1/PP3P2/2P5/2KR3R b - - 0 20
>>>>>>
>>>>>>Have a nice day
>>>>>>
>>>>>>Yar
>>>>>
>>>>>Not sure what the best move is, but Chessmaster 9000, on a P4-2.4, seems to
>>>>>think that White has a winning position, and Black's best move is Qa5:
>>>>>
>>>>>Time	Depth	Score	Positions	Moves
>>>>>0:00	1/3	-0.09	10262		20...Qa6 21.a4 Rc5 22.Nf5
>>>>>0:00	1/4	0.05	13991		20...Qa6 21.bxc4 Qxa3+ 22.Kb1 Rxc4
>>>>>					23.Qb5
>>>>>0:00	1/4	-0.04	15233		20...Ne5 21.Qh4 Qa5 22.Bg5 Qxa3+
>>>>>					23.Kb1
>>>>>0:00	1/5	-0.48	59871		20...Ne5 21.Kb1 Rc3 22.Nd5 Nxd5
>>>>>					23.exd5
>>>>>0:01	1/6	1.23	171470		20...Ne5 21.Bf8 Rxf8 22.Nf5 Nh5
>>>>>					23.Nxe7+ Kh7 24.Nxh5
>>>>>0:02	1/6	0.66	341051		20...Qa6 21.a4 Ne5 22.Qh4 Nxf3
>>>>>					23.Nxf3 Qc6 24.Rh2 Qxe4
>>>>>0:04	1/6	0.40	721414		20...Qc5 21.Qxc5 dxc5 22.Nde2 Ne5
>>>>>					23.g5 Nfd7 24.Nd5
>>>>>0:04	1/7	0.40	767801		20...Qc5 21.Qxc5 dxc5 22.Nde2 Ne5
>>>>>					23.g5 Nfd7 24.Nd5
>>>>>0:10	1/8	0.51	1681398		20...Qc5 21.Qxc5 dxc5 22.Nde2 Ne5
>>>>>					23.g5 Nfd7 24.Nd5 Re8 25.Rhf1
>>>>>0:27	1/8	0.43	4813745		20...Qa5 21.Qxa5 Nxa5 22.Nfe2 e5
>>>>>					23.Bd2 Rc5 24.b4 Rcc8 25.g5 Nc4
>>>>>					26.gxf6 exd4
>>>>>0:51	1/9	1.34	8887359		20...Qa5 21.bxc4 Qxa3+ 22.Kd2 Nxe4+
>>>>>					23.fxe4 Bf6 24.Qd5 e5 25.Nb5 Qb4+
>>>>>					26.c3 Qb2+ 27.Kd3 exf4 28.Qxb7
>>>>>4:05	1/10	1.56	43122326	20...Qa5 21.bxc4 Qxa3+ 22.Kd2 Nxe4+
>>>>>					23.fxe4 Bf6 24.Qd5 e5 25.Nb5 Qb4+
>>>>>					26.Kd3 exf4 27.Bxf4 Rc5 28.Qxb7
>>>>>					Qxc4+ 29.Ke3
>>>>>
>>>>>jm
>>>>John,
>>>>
>>>>This is mate i wonder just how many moves ?
>>>
>>>The eval continue to rise, may be it will find the mate after some hours
>>>
>>>CM9000-SKR (Barton 2800+)
>>>
>>>Time	Depth	Score	Positions	Moves
>>>0:33	1/9	0.90	8528166		20...Qa5 21.bxc4 Qxa3+ 22.Kd2 Nxe4+
>>>					23.fxe4 Bf6 24.Qd5 e5 25.Nb5 Qb4+
>>>					26.Ke2 exf4 27.Nxd6
>>>1:38	1/10	1.02	25344838	20...Qa5 21.bxc4 Qxa3+ 22.Kd2 Nxe4+
>>>					23.fxe4 Bf6 24.Qd5 e5 25.Nb5 Qb4+
>>>					26.Kd3 exf4 27.Bxf4 Rc5 28.Qxb7
>>>					Qxc4+ 29.Ke3
>>>4:55	1/11	1.78	79407086	20...Qa5 21.bxc4 Qxa3+ 22.Kd2 Nxe4+
>>>					23.fxe4 Bxd4 24.Nd5 Re8 25.Nxe7+
>>>					Rxe7 26.Qxe7 Qc3+ 27.Ke2 Qxc2+
>>>					28.Rd2 Qxc4+ 29.Rd3 Qb4 30.Bd2 Qb6
>>>22:43	1/12	2.18	377487007	20...Qa5 21.bxc4 Qxa3+ 22.Kd2 Nxe4+
>>>					23.fxe4 Bf6 24.Qd5 Qb4+ 25.c3 Qb2+
>>>					26.Kd3 Rc5 27.Rb1 Qa2 28.Ra1 Qb2
>>
>>There may be some confusion. Yar seems to be claiming that Shredder stated that
>>BLACK has the mate in 58 moves. At least, that's what I'm assuming because it is
>>Black to move and Yar said "mate in 58".
>>
>>However, Chessmaster is saying that WHITE is winning (positive scores always
>>favor White in the CM GUI).
>>
>>I am not at all convinced that Shredder's announcement of mate is accurate. In
>>particular, I would like to see the PV (all 115 plies of it!).
>
>Of course you are right and it is not accurate. I saw this quite often in
>Analysis mode.
>#62,#68,...
>This is just some flaw in the program.
>
>Michael

I see no reason for programs who say mate in 58 to see everything.

A program can say it when it knows based on its chess knowledge that it is mate
in at most 58.

For example a program may know that if the position is no stalemate and it has a
queen and the opponent has only king then it is mate in at most 20 and claims
mate in 20 even if it does not see a forced mate by search.

I know of no program that does it but I see nothing wrong with doing it.

Uri



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.