Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: What are the results of chess programs in correspondence games?

Author: Uri Blass

Date: 03:09:09 04/27/01

Go up one level in this thread


On April 27, 2001 at 06:02:20, Uri Blass wrote:

>I believe that it will be more interesting to get games of chess programs in
>correspondence tournament instead of getting games of chess programs against the
>world.
>
>you can get information about ICCF tournament at http://www.iccf.com
>
>It will be interesting if chess programs can win the email tournament inspite of
>the fact that the opponents can use computers because of the important advantage
>of having more time.
>
>The opponents usually have only one computer and do not use different computers
>against different oponents so if we find enough volunteers to use chess programs
>then they have chances to win the tournament.
>
>In the email ournament it is possible to get average of more than 144 hours per
>move against every opponent(I say more than 144 hours per move because pondering
>is possible).
>
>In the postal tournament (not by email) it is possible to get an average of more
>than 72 hours per move and we only need enough volunteers to do it with good
>hardware.
>
>I guess that winning the second tournament is more easy because people who do
>not use email often also do not use computers.
>
>What is your guess for the rating of this monster when Deep Fritz or Deep Junior
>on a dual or quad is used(it should also calculate when it is in book in order
>to check if the book does not contain errors and play book moves only if the
>score of the best book move is not significantly lower than the score of the
>root position)
>
>My guess is that it can get an ICCF rating of 2600.
>
>Uri

I can add that the 2.5-1.5 against steve ham(2508 ICCF rating) gives performance
of 2608 by the linear formual that I know.

I get 2608 by calculating 2508+(2.5-1.5)/4*400

2.5 is the result of programs
1.5 is the result of Steve Ham
4 is the number of games
400 is a constant that is used to calculate the performance

This formula is not exactly correct but is closed to be correct when the
difference in rating is not very big.

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.