Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF Rating list 00-10-27

Author: Frank Quisinsky

Date: 11:47:22 10/27/00

Go up one level in this thread


On October 27, 2000 at 12:59:00, Bertil Eklund wrote:

>On October 27, 2000 at 12:43:59, Uri Blass wrote:
>
>>On October 27, 2000 at 12:35:28, Enrique Irazoqui wrote:
>>
>>>On October 27, 2000 at 12:23:35, Marcus Kaestner wrote:
>>>
>>>>>Gandalf suffered from a learning bug and lost the same opening twice against
>>>>>Junior because the learning is destroyed.
>>>>>I believe that the same happened against other programs.
>>>>>
>>>>>I think that the real ssdf rating of gandalf should be better because the bug
>>>>>was corrected and I understood that the bug appears only when you stop games
>>>>>played by the autoplayer and is not relevant for the customer.
>>>>>
>>>>>I do not say that it is better than Fritz6a but I believe that it is better than
>>>>>the rating that it got.
>>>>
>>>>
>>>>more important: gandalf wins very much from fast computers. ALL the testers i
>>>>know have the same results: on a 1 ghz machine with game in one hour, gandalf is
>>>>slightly behind fritz6a.
>>>
>>>The 40:2 of the SSDF on AMD 450s is slower than Game/60 on a 1 Ghz machine, and
>>>their rating for Gandalf was much lower than "slightly behind F6a."
>>>
>>>Enrique
>>
>>Their results are not convincing because gandalf suffered from a learning bug.
>>
>>I am interested to know the rating if you do not include losses of gandalf
>>because of the learning bug.
>>
>>It is possible that it is only slightly behind F6a without the learning bug.
>>
>>Uri
>Hi!
>
>I guess this could have affected the outcome in 1-3 games say 1,5 points of over
>300 games. BTW before the corrected bug (about 50-60 games) Gandalf had 2572.
>
>Don't forget that Gandalf is probably the best Winboard-engine, number six in
>the list and last but not least a very nice program to follow. Very sharp play,
>maybee without direct speculative play but very exciting attacks. Also remember
>that it probably plays better with todays cpus and likes longer time-controls.
>
>Bertil

Yes, yes I think so :-)

Best
Frank



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.