Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: I plan to settle this.

Author: John Hatcher

Date: 20:44:53 06/12/01

Go up one level in this thread


On June 12, 2001 at 22:52:22, Dann Corbit wrote:

>On June 12, 2001 at 22:18:31, Slater Wold wrote:
>
>>I am not sure about the other CCC readers, but the recent "Are computers GM
>>strength?" has gotten me curious.  Are they?  One would think.  But then you see
>>a computer play a bad game, and think, "HA!  A 2100 COULD PLAY BETTER THAN
>>THAT!"
>>
>>So let's finalize this.  In a full rules, simulated FIDE event.  I am talking
>>about the 4 top programs vs 4 2500+ GM's.  Swiss tournament, standard time
>>controls.  All the good stuff.  Simulate ratings, and get a decent GUESS at
>>what, in this tournament, the computers would achieve.
>>
>>I would rather a 3 round swiss tournament @ 120/0.  3 games against a GM is a
>>good gathering of performance.  12 games.  See how the computers do.
>>
>>The summer is arriving, and a lot of GM's will be playing a lot of tournaments
>>soon.  If you'd like to help, please contact me.  If you can help, PLEASE
>>contact me.  If you want to tell me what a worthless waste of time this is,
>>DON'T CONTACT ME!  NOT EVEN ON THE CCC BOARD!  I don't think anyone wants to
>>hear it.
>>
>>I doubt ANYONE could argue the point of a 3 round swiss.  If the computers walk
>>away undefeated, then we'll know.  It could be close, it could be a killing.
>>
>>If ANYONE has a better idea.  PLEASE SPEAK NOW!
>
>I think it is a good idea, but I don't think it settles anything.
>
>Joe is a boxer.  Joe goes undefeated in 12 fights.  Then, in fight 13, the first
>time he really gets pasted in the jaw, we discover Joe has a glass jaw.  How
>will Joe fare in his next 12 bouts?
>
>Fred is a boxer.  Fred wins 12 bouts in a row.  In the 13th bout, Fred gets
>hammered every time his opponent uses an uppercut.  How will Fred fare in the
>next 13 bouts?
>
>On the other hand, if Kramnik gets handed his hat after careful preparation, I
>think that will be very interesting.  Not conclusive of course, but it would
>definitely alter the way that I think about computer chess engines.

The question may be interesting in the abstract, and of interest to chess
programmers, but since both Grandmasters and chess programs are stronger than
99% of the chess players on the planet, I think what interests the average
program user is how well they can be made to play in a way that mimmics a
human's understanding of the game.

Of course, computers don't play chess; they crunch numbers and process
algorithims.  They don't "know" whether they're playing chess, "writing" a
letter, or playing a Midi tune.  It's the programmer's skill at making a
computer mimmic human activity that makes it so fascinatng.  It's an illusion; a
clever and successful one; but just an illusion.

John





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.