Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: 8 game match 34 game total Crafty vs CT13

Author: Paul Doire

Date: 21:06:56 12/30/00

Go up one level in this thread


On December 30, 2000 at 23:50:13, Peter Berger wrote:

>On December 30, 2000 at 18:22:55, Paul Doire wrote:
>
>>
>>Chess Tiger 13 won 67.64% of the match,this equates to approximately
>>140 ELO stronger if my math is correct.
>>Feedback?Friendly Comments?
>
>Well done :-) !! Neither Crafty nor Tiger should have been weakened by your last
>settings . The trial with near 0 hashp and 0 cash for Crafty : I would forget
>about those games if I were you .
>
>Equal is not always equal btw ; I think 24 MB hash and 5 MB hashp will let
>crafty perform just as well :-) ( and then you don't have to worry about paging
>at all ) .
>
>I think 24MB Hash for Tiger as opponent would have been fairest ; and I also
>think this wouldn't change your results very much .
>
>pete
>>
>>Regards,
>>Paul

Thanks Pete,
There are a million opinions out there about proper allocation of Crafty hash.
Thats why I posted the excerpt from Crafty.doc.
There was no paging at all.
If you check ChessBits Rating List 12/19/00(NewsTicker) You will find
Chess Tiger13 rated at 2712 Crafty at 2598.
A difference of 114 Elo,26 less than my estimate.
This is not to say that list is the ultimate guide, but it helps in overall
assessment.
I am done testing these 2 engines, after over 2 days of solid testing I have
formed my opinion and posted my results.
CONCLUSION:
Don't test Crafty unless you want to here how this or that was wrong.
I will stick with engines that have one hash setting.

Thanks for your input and Happy New Year,
Paul



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.