Computer Chess Club Archives

Search

Terms

Messages

Subject: Re: SSDF(Pro Deo 1.1 - Junior 9)A1200,0-3, now 1-9

Author: Ed Schröder

Date: 15:35:46 06/01/05

Go up one level in this thread

On June 01, 2005 at 18:26:02, Kurt Utzinger wrote:

>On June 01, 2005 at 11:29:55, Ed Schröder wrote:
>
>[...]
>>Conclusion - You must be playing with some weird settings, in any case you are
>>not playing with the default settings. I checked all learner settings, no
>>setting can create the 2 moves (and scores) in question.
>
>      I agree with this. Both moves played by Pro Deo 1.1 at SSDF
>      can't be reproduced on my P4 1.8/60 MB hash. I have tested
>      under Shredder 9 Classic GUI and are going to repeat the
>      test under Arena-GUI.
>      Kurt

It's not so simple because of the (possible) effect of position learning.
Position learning slightly increases the value of the [Attractiveness] parameter
to escape from previous lost games. But the silly moves and too low iteration
depths as shown in the games can not be blamed to these (unseen) internal
changes.

Ed

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.