Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Junior 9 Test (40'/40) after 300 games

Author: Kurt Utzinger

Date: 15:56:35 12/31/04

Go up one level in this thread


>Tests carried out by Volker P have shown that relative results are remarkably
>similar whether you use ponder on or ponder off when using a single cpu.
>As for learning, that is a completely separate issue. However you wouldn't get
>comparable results from different testers because the same program will play
>differently depending on the learning that has occurred and this could differ
>markedly.
>Keep up the good work Kurt. It doesn't matter how much soe of us put into
>testing, there wll always be illjoys.
>
>Graham.


      Hi Graham
      All these issues have been discussed for many times already.
      And the same arguments always come when somebody is not
      satisfied with the results of his beloved engine. As to
      ponder=on/off it is well known that it makes (as far as
      results between two engines are concerned) almost no difference,
      see also at http://www.pittlik.de/winboard/ponder.html
      and as to learning and books used I can only say that it has always
      been our aim to test the real strength of an engine and not the
      influence of the books and learning function. Other people might
      prefer testing the whole package of an engine as it came with
      the CD. This is of course legitimate and also ok. Personally (chess
      friend Rolf Bühler's opinion is not the same) I really prefer
      matches with Nunn2, Noomen select and other opening positions
      and I would almost never use the books delivered.
      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.