Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: CM8K & Fritz - What am I testing.

Author: stuart taylor

Date: 12:46:13 04/12/01

Go up one level in this thread


On April 12, 2001 at 13:23:26, Scott Woods wrote:

>On April 12, 2001 at 04:20:07, stuart taylor wrote:
>
>>On April 11, 2001 at 08:49:21, Scott Woods wrote:
>>
>>>Stuart,
>>>
>>>Very simple really.
>>>
>>>There have been a lot of posts on this site saying that CM8K is weak.
>>>    -"A larger opening book is the problem"
>>>    -"Weak in the endgame- needs table bases"
>>>
>>>There have been questions over which patch has been tested etc etc.
>>>
>>>Then there are posts of how well "The King" does when "Johan de Koenig" operates
>>>the machine.
>>>
>>>A bit confusing all round.
>>>
>>>The proof of the pudding as they say.....
>>>So I am volunteering myself to play 100 games of CM8K v1.02 against Fritz 6e.
>>>because they are the 2 chess programs I currently own and I have to start
>>>somewhere.
>>>
>>>My budget doesn't run to identical hardware so I have given Fritz the better
>>>machine. It has the benefit 64Mg more memory, twice the amount of hash tables -
>>>a faster processor and tablebases.
>>>
>>> - if the games look a bit one sided in favour of the faster hardware - I'll
>>>swap the hardware at game 50.
>>>
>>>When we get to 100 games - we should have a fair idea of how well CM8K plays
>>>against Fritz 6e.
>>>
>>>I will publish on this site every 10 games the results, settings, hardware specs
>>>etc etc
>>>
>>>Then I'll go out and buy "Deep Something or other" and/or some of the other
>>>engines that are well respected on this site and repeat the exercise.
>>>
>>>Will this prove which is the strongest chess engine in the market?
>>>    - Of course it won't.
>>>
>>>What it will illustrate is how well CM8K plays against those other engines,
>>>
>>>
>>>
>>>As for playing CM6K against CM8K - I will leave that for someone else as CM6K
>>>does not allow winboard232.exe autoplay
>>
>>
>>This certainly sounds very worthwhile. But only if the nature of Fritz 6e is
>>known, and/or other programs will/have also be(en) compared to it.
>>(nb when I said earlier about 6e being weaker than 6a, I meaned 6b is weaker,
>>not 6e. 6e maybe too, but I don't know).
>>  "Will it prove which is the strongest chess engine in the market?" Of course,
>>after a few tests with CM8K it will strongly indicate where CM8K stands, even if
>>does not play all of them.
>>S.Taylor
>
>Hi Stuart,
>Really bizarre 24 hours this.
>
>SSDF has found that many of the CM8K games are invalid because of the "Thinking
>lines switches" and are re-doing the affected CM8K games.
>Hat's off to the guys for doing this - because it's a fair bit of their time to
>
>Also I think the SSDF posting the games will have a lot more clout ( and
>respectability ) than "little old me" but I will continue with them in the
>meantime.
>
>On a by note,
>One thing I find fascinating is how the different engines are scheming and
>plotting - ( and the difference in the points analysis too )
>
>Now if I can only get the Missus to agree that I need Deep Fritz/ Deep Junior/
>Shredder and Gambit Tiger.....etc...
>
>;-p
>
>Scott

You mean they should send them to you as a tester? Good idea!
Anyway, I think what is sorely missing from the ssdf is a willingness to compare
things by testing CM6K.  But I agree it is very good of them to be redoing those
many games of CM8K. Necesary too!
S.Taylor



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.