Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Where can I find what Hardware will be used against GM Gulko?

Author: Sune Fischer

Date: 14:34:59 03/10/02

Go up one level in this thread


On March 10, 2002 at 16:15:04, Uri Blass wrote:

>On March 10, 2002 at 15:48:56, Sune Fischer wrote:
>
>>On March 10, 2002 at 13:01:45, Miguel A. Ballicora wrote:
>>
>>>On March 10, 2002 at 09:39:13, Ian Aston wrote:
>>>
>>>>On March 10, 2002 at 06:32:05, Jonas Cohonas wrote:
>>>>
>>>>>By the way it is a well known fact that conputers of today play better with
>>>>>longer time controls...
>>>>
>>>>I am not sure what you mean by this as obviously computers will generally play
>>>>better given longer time controls. However, if you mean that computers are more
>>>>likely to win against humans at longer time controls than at shorter ones then I
>>>>suggest, with respect, that it is not a well known fact and that you are just
>>>>plain wrong.
>>>
>>>Yes, it is a well known fact.
>>
>>
>>And you can tell by the rating, here on FICS for the top programs:
>>
>>        Blitz                   Standard                 Lightning
>>   1. LittleGoliat 2680     1. vivaverdi    2649     1. ChessTigress 2930
>>   2. Topas        2674     2. Tartessos    2632     2. LittleGoliat 2827
>>   3. ChessTigress 2652     3. ezcape       2580     3. pecable      2784
>>   4. pecable      2646     4. LittleGoliat 2574     4. JahnFeig     2779
>>   5. DeepImpact   2638     5. ChessTigress 2573     5. Tartessos    2756
>>   6. Koibito      2634     6. Gandalph     2514     6. ezcape       2751
>>   7. vippcomputer 2624     7. JahnFeig     2510     7. Artificial   2740
>>   8. baltica      2584     8. JadeiteMech  2509     8. airone       2733
>>   9. Tartessos    2576     9. Hera         2499     9. MrWolf       2731
>>  10. Angledust    2558    10. Kec          2487    10. JadeiteMech  2727
>>
>>
>>Skyhigh lightning rating for the programs, much more realistic ones on standard
>>timecontrols and blitz somewhere in between. I wonder what their correspondence
>>rating would be....
>
>
>It is also possible that after some time more time help the program more than
>the humans because humans may get tired.
>
>The fact that humans perform better at 60 minute per game and not at 1 minute
>per game does not prove that they will continue to perform better at longer time
>control.
>
>Uri

At longer time controls humans don't fall for simple taktics so the positional
play will be more important.
Since programs has a much higher branch factor, I think time works
(exponentially) against programs.

Yes humans tire, but in correspondence chess you are allowed to analyse by
actually moving around pieces, this makes it much easier to avoid tactical traps
IMHO, and noone forbids you to take a break or even a nap if you need it :)

-S.




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.