Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hiarcs 10 first results - Harry Schnapp test suite 440 used

Author: Heinz van Kempen

Date: 06:16:11 12/23/05

Go up one level in this thread


On December 23, 2005 at 08:00:43, Uri Blass wrote:

>On December 22, 2005 at 19:14:25, George Tsavdaris wrote:
>
>>
>> Indeed as your examples have shown there are many dubious(not necessarily bad)
>>lines. I agree that this is not a good test suite but only if the engines don't
>>play a position one as black and one as white. If they get at random some
>>positions with white and some with black then there is a problem.......
>>
>> BTW what test suite is used: 440 or 220 ? Uri said 220 while Heinz said
>>440.....
>
>I said that 220 positions are very big test suite.
>I claimed nothing about the test suite that was used.
>
>I downloaded the 220 test suite so I thought about it.
>
>440 positions is even bigger and you may need 880 games to finish nunn match
>between 2 programs.
>
>Uri

Hi Uri,

based on the results with the Harry Schnapp suite and statistical investigations
of the CEGT database one could pick the seemingly best 40 positions from it to
run something comparable to a Nunn match. Anyway wait and see, there will come
something even better from Harry one day.

Best Regards
Heinz



This page took 0.07 seconds to execute

Last modified: Thu, 07 Jul 11 08:48:38 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.