Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: LOL!

Author: Uri Blass

Date: 08:49:02 02/24/06

Go up one level in this thread


On February 24, 2006 at 11:41:23, Martin Andersen2 wrote:

>On February 24, 2006 at 11:20:38, Martin Andersen2 wrote:
>
>>On February 24, 2006 at 10:53:57, George Tsavdaris wrote:
>>
>> The top leader's ELO rating may be wrongly calculated if we match them
>>>with very much weaker opponents....
>>
>>Why ?
>
>To add some more: Fritz 5.32 P200 has a SSDF ELO of 2497, Fruit 2.2.1 A1200
>an ELO of 2853. Difference is 356 so according to the ELO table, Fruit is
>expected to score about 89%, that is 37.5 points in the match.
>
>Fruit scores 40.5 points, a bit higher than expected, but it's not many games.
>
>What would you say is the problem with these matches ?
>
>Martin.

The problem is that rating may not give correct estimate of the expected result.

Note that Fritz5.32 was famous for his learning and if I remember correctly it
beated Rebel8(P90) again and again in the same line.

Unfortunately fritz5.32 did not find a single win to repeat in this match
because it could get only 3 draws and it is not designed to repeat drawing
lines.

Uri



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.