Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: ELO of the chess programs in the 2010

Author: Graham Laight

Date: 02:25:40 08/17/01

Go up one level in this thread


On August 16, 2001 at 16:13:31, Olaf Jenkner wrote:

>On August 16, 2001 at 03:36:19, Graham Laight wrote:
>
>>In 1992, against humans, they were about 2300.
>>
>>Now, they're about 2650.
>>
>>If this trend continues, they'll be about 3000 in 2010.
>>
>>Of course - there is plenty of scope to argue with these numbers. It is my
>>intention to just simply call it as I see it to answer Federico's question - not
>>start a flame war.
>>
>>-g
>The ELO-Rating is not a linear function ...
>OJe

This may well be more true than most people realise.

I believe that, with "optimal" play from both sides, chess is a draw.

The secretary of the USCF has produced the graph on this link:

http://math.bu.edu/people/mg/ratings/Draws.jpg

I think that this indicates that above, say, 3500, you'll only ever get drawn
games.

It's also the experience in computer chess - as their standard of play rises, so
does the proportion of draws (based on a quick, unscientific study of the SSDF
games by myself).

-g



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.