Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Results of 112 engines in test suite "WM-Test" (100 pos) for download

Author: Albert Silver

Date: 11:48:00 08/20/02

Go up one level in this thread



>>>>it is a bad idea to design a testset for 10-15 seconds now
>>>
>>>From the response you both gave, I can see what I proposed was not well
>>>understood. Think of it this way: suppose you are running a blunder check at 15
>>>seconds a move. It notices that at move 27, with the engine at a mere 9-10 plies
>>>or so, the eval changed strongly. It didn't see anything at move 26. Of course
>>>this is because it is not thinking long enough, but you already know enough:
>>>there was an error done somewhere around move 22-27 probably. If it had been
>>>running at 3 minutes a move, it would have seen the problem at move 23. Now
>>>suppose the game goes 40 moves deep before the losing side resigns. If you do it
>>>at 3 minutes a move that means 80 plies or 240 minutes. In other words it will
>>>take 4 hours. If I take 15 seconds a move it is done in one hour,
>>
>>That should be 20 minutes. Duh. Argument still stands as is though.
>>
>>                                    Albert
>
>the problem is that I try to find the losing mistake in every game to have at
>least one test position from every game but it is not easy and I guess that at
>least in part of the cases I will have to have games that were not drawn with no
>test position.
>
>There are cases when I cannot find moves that blunders more than a pawn before
>the position becomes lost and I cannot be sure what is the losing blunder.

I understand, but this won't be improved by letting the engine run 3 minutes on
every move. You'll still have an idea where the balance was compromised, and
spend 1/12th the time.

                                    Albert

>
>
>Here is the last example from a game against bestia.
>
>[D]2r5/R1p5/P3k1pp/1P1b1p2/1K3P1P/4B1R1/8/2r5 b - - 0 52
>
>Movei played Rc4+ and lost in the 3th division.
>
>52...Rc4+ 53.Ka5 Rg8 54.Bd2 Rd4 55.Re3+ Kd7 56.Bc3 Rxf4 57.b6 Rc4 58.Bb4 59.Rc3
>Rxc3 60.Bxc3 Rc8 61.Be5 g5 62.Bxc7
>
>I suspect that Rc4+ was the losing mistake but I am not sure
>I suspect that finding Rb1+ is a good test position but I am not sure.
>
>
>Here is some analysis of Deep Fritz(Movei can also find Rb1 after enough time
>and I suspect that all programs are going to find Rb1 here but is
>Rb1+ the only drawing move?)
>
>I could not even prove that Rc4+ is losing material.
>
>In a game I would play Rb1+ for positional reasons but without knowing if Rc4 is
>losing.
>
>Bestia 0.88-b2 - Movei 0.0.72h
>2r5/R1p5/P3k1pp/1P1b1p2/1K3P1P/4B1R1/8/2r5 b - - 0 1
>
>Analysis by Deep Fritz:
>
>52...g5--
>  ²  (0.44)   Depth: 1/2   00:00:00
>52...g5-- 53.Bxc1
>  +-  (5.09)   Depth: 1/6   00:00:00
>52...c5+!
>  +-  (5.00)   Depth: 1/6   00:00:00
>52...c5+! 53.bxc6
>  +-  (2.34)   Depth: 1/14   00:00:00
>52...Rb1+!
>  +-  (2.31)   Depth: 1/14   00:00:00
>52...Rb1+! 53.Kc5
>  +-  (1.53)   Depth: 1/14   00:00:00
>52...Rb1+--
>  +-  (1.84)   Depth: 2/8   00:00:00
>52...Rb1+-- 53.Kc5 Re8 54.Rxc7
>  +-  (2.78)   Depth: 2/15   00:00:00
>52...Rc4+!
>  +-  (2.59)   Depth: 2/15   00:00:00
>52...Rc4+! 53.Ka5 Rg8
>  +-  (1.69)   Depth: 2/15   00:00:00
>52...Rc4+ 53.Ka5 Rg8 54.Rg1
>  +-  (1.62)   Depth: 3/7   00:00:00  1kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 Kf6
>  +-  (1.75)   Depth: 4/11   00:00:00  2kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 Kf6 55.hxg6 Rxg6 56.Rxg6+
>  +-  (1.62)   Depth: 5/14   00:00:00  6kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 Kf6 55.hxg6 Rxg6 56.Rxg6+ Kxg6
>  +-  (1.62)   Depth: 6/17   00:00:00  17kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 Kf6 55.Rg1 gxh5 56.Rxg8
>  +-  (1.53)   Depth: 7/21   00:00:00  50kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 Kf6 55.Rg1 gxh5 56.Rxg8 Bxg8 57.Ra8
>  +-  (1.50)   Depth: 8/20   00:00:00  104kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 Kf6 55.Bd2 gxh5 56.Bc3+ Ke7 57.Rxg8 Bxg8 58.Be5
>  +-  (1.50)   Depth: 9/24   00:00:00  288kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 Kf6 55.Bd2 gxh5 56.Bc3+ Ke7 57.Rxg8 Bxg8 58.Be5
>  +-  (1.47)   Depth: 10/25   00:00:01  613kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 g5 55.fxg5 f4 56.Bxf4 Rxf4 57.g6 Rg7 58.Ra3
>  +-  (1.69)   Depth: 11/30   00:00:03  2483kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 g5 55.fxg5 f4 56.Bxf4 Rxf4 57.Rxc7 Rf5 58.g6
>  +-  (1.69)   Depth: 12/30   00:00:09  6425kN
>52...Rc4+ 53.Ka5 Rg8 54.h5 g5 55.fxg5 f4 56.Bxf4 Rxf4 57.Rxc7 Rxg5 58.Rxg5
>  +-  (1.69)   Depth: 13/33   00:00:26  17836kN
>52...Rb1+!
>  +-  (1.66)   Depth: 13/33   00:00:52  35495kN
>52...Rb1+! 53.Kc3 Rg8 54.Rxc7 Rxb5 55.Kd2 Rb3 56.Rh7 Be4 57.Ke1 h5
>  ±  (0.91)   Depth: 13/38   00:03:49  155332kN
>52...Rb1+ 53.Kc3 Rg8 54.Rxc7 Rxb5 55.Kd2 Kd6 56.Rc1 Rb2+ 57.Rc2 Rxc2+ 58.Kxc2
>  ±  (0.81)   Depth: 14/36   00:05:24  222008kN
>52...Rb1+ 53.Kc3 Rg8 54.Rxc7 Rxb5 55.Kd2 Be4 56.Bd4 Ra5 57.a7 Kd5 58.Rd7+
>  ±  (0.72)   Depth: 15/39   00:13:01  548966kN
>
>(blass, tel-aviv 20.08.2002)
>
>
>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.