Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: An idea to generate a new test suite that is not only tactics

Author: Uri Blass

Date: 10:32:00 01/15/01

Go up one level in this thread


On January 15, 2001 at 13:26:45, Uri Blass wrote:

>On January 15, 2001 at 13:15:59, Uri Blass wrote:
>
>>On January 15, 2001 at 13:06:38, Dann Corbit wrote:
>>
>>>On January 15, 2001 at 13:01:39, Uri Blass wrote:
>>>
>>>>On January 15, 2001 at 12:56:14, Dann Corbit wrote:
>>>>
>>>>>On January 15, 2001 at 12:24:10, Uri Blass wrote:
>>>>>
>>>>>>It is possible to use some games and a lot of computer time to genrate a test
>>>>>>suite.
>>>>>>
>>>>>>The idea is to give some top programs to analyze all the positions in the games
>>>>>>and to give them 24 hours per position.
>>>>>>
>>>>>>The test suite can include all the positions that there was an agreement between
>>>>>>all the top programs about the best move after 24 hours of search when part of
>>>>>>them needed more than 3 minutes to converge to the best move(otherwise the
>>>>>>position is too easy).
>>>>>>
>>>>>>People can claim that the fact that all of the program converge to the same move
>>>>>>is not a proof that it is the best move and I agree but I think that in
>>>>>>more than 90% of the cases the solution will be correct.
>>>>>
>>>>>On the SSDF is a positional test suite similar to the one that you propose.
>>>>
>>>>I remember that the test suite of the ssdf give some options for every move when
>>>>there is not only one move.
>>>>
>>>>It is impossible to give more than one option in my test.
>>>>
>>>>I also have doubts about the question if the solutions are correct.
>>>>
>>>>If I know that all top programs converge after 24 hours to the solution then I
>>>>believe more in the test.
>>>
>>>The reason that I mention that particular test is because it is very thoroughly
>>>documented.
>>>
>>>I refer (of course) to this one:
>>>
>>>Given this search:
>>>http://www.google.com/search?q=berkeley+sis
>>>
>>>This is the 3rd listing:
>>>http://home.interact.se/~w100107/fentest.htm
>>
>>I have doubts if the solutions are correct.
>>
>>No program solve position number 4 so I have doubts if the solution is correct
>>
>>[D]r1b1r1k1/1p1n1pbp/2p1n1p1/q1P1p3/4P3/1PN1BNPP/2Q2PB1/1R1R2K1 w - - 0 1
>>
>>Position number 3 is a pawn endgame and I suspect that programs with tablebases
>>can find a forced mate.
>>
>>Uri
>
>I can add that there was another position that no program solved.
>
>It seems to me that the test was composed by humans and I trust more long search
>of computers.
>
>I also suspect that the composers of this test are not strong GM's.
>
>Uri

I see that 2 positions are from games of kasparov when kasparov is responsible
to it but there are other games when the annotator is not a known player.

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.