Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: So we can conclude: Shredder 9 UCI isn“t stronger than Shredder 8 UC

Author: Drexel,Michael

Date: 23:08:40 07/31/05

Go up one level in this thread


On July 31, 2005 at 22:47:13, Andrew Walker wrote:

>On July 30, 2005 at 06:01:06, Drexel,Michael wrote:
>
>>
>>...at all with this time controls and hardware used?
>>
>>http://chessprogramming.org/cccsearch/ccc.php?art_id=364117
>>
>>Sandro claimed Shredder8.bkt and more effective learning should increase the
>>rating by about 35 points over Shredder 8 CB.
>>
>>Michael
>
>Well, if you look at the size of the error bars the two ratings overlap so you
>cannot conclude anything yet. A lot more games would be needed to settle a 35
>poiint difference.
>
>Andrew

So if the error bars do not overlap with 95% confidence interval but overlap
with a 97% confidence interval I am allowed to make conclusions? :)

There are anyway no games at all from Shredder 8 UCI, so we will never know it's
SSDF rating.
It seems highly unlikely that it would be considerably lower than 2821 though.

Anand has currently a rating of 2788 and Svidler 2738. Both play usually against
the same opponents.
So almost anyone concludes Anand is the stronger chess player.

Maybe Svidler just had constantly bad luck during his career and is in reality
as strong as Anand?
Doesn't matter, we don't care.

You have to accept some compromise otherwise you will never come to any
conclusions.

Michael












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.