Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Different Hydra personalities against Rybka

Author: Kurt Utzinger

Date: 07:22:34 12/13/05

Go up one level in this thread


On December 13, 2005 at 07:05:49, Joachim Rang wrote:

>On December 13, 2005 at 05:43:45, Kurt Utzinger wrote:
>
>>On December 13, 2005 at 05:04:08, Joachim Rang wrote:
>>
>>>On December 13, 2005 at 03:58:50, Kurt Utzinger wrote:
>>>
>>>>On December 13, 2005 at 02:16:50, Chrilly Donninger wrote:
>>>>
>>>>[...]
>>>>>The experiment shows also, that it is fairly easy to tune against one programm.
>>>>>The problem is to find a solution which works against all.
>>>>>
>>>>>Chrilly
>>>>
>>>>      Hi Chrilly
>>>>      This conforms exactly with my own experience when
>>>>      trying different CM-personalities: the best CM-setting
>>>>      vs Shredder had no chance vs Fritz and so on.
>>>>      Kind regards
>>>>      Kurt
>>>
>>>
>>>Hi Kurt,
>>>
>>>I'm surprised by this. I had the impression tuning against a specific opponent
>>>yields only little. Do you have any data on your experience for this?
>>>
>>>Joachim
>>>
>>>Joachim
>>
>>      Hi Joachim
>>      We have tried hundreds of different CM9-settings
>>      without tuning them vs a specific engine. But we
>>      noticed that setting X played excellent vs for
>>      example Shredder but lost badly vs Chess Tiger
>>      and/or other combinations. Such things always
>>      happened with settings where we tried to increase/decrease
>>      some parameters to the extreme. All such settings did never
>>      well and were unsuitable as a good overall setting.
>>      Regards
>>      Kurt
>
>
>Did you have such "anomalities" with matches for more than 50 games? I mean if
>you play 10 matches of 50 games chances are high that one match will have a
>"curious" result, such as your result of Fritz 9 against Fruit-Toga 1.0.
>
>Joachim


      Hi Joachim
      The problem was not that CM X got for example a score
      of 85 % vs Shredder but that the same setting did only
      got an overall score of 30-35 % vs all other engines.
      If a CM9-setting achieved an overall score between
      48-55 % with the exception of 70 % (positive) or 30 %
      (negative) vs a specific engine, this was OK for us. There
      may always be a dreaded/easy opponent for a chess program.
      Best regards
      Kurt



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.