Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Can humans teach computers to be less than 2500 or 2400 strength?

Author: James T. Walker

Date: 09:57:51 06/30/01

Go up one level in this thread


On June 30, 2001 at 12:29:34, Uri Blass wrote:

>Suppose that humans who play against programs get 10000$ for winning and 5000$
>for drawing so they are going to be motivated to learn how to play against the
>programs.
>
>Suppose also that the programs and the book  are not available to the humans
>
>They may only buy the best case something similiar like commercial Deep Fritz
>when they play against Deep Fritz7(beta).
>
>Suppose the time control is 40 moves/2 hours+20 moves/1 hour.
>
>What is your guess for the rating of the programs under these conditions.
>
>What is your guess of the rating drop of humans under similiar conditions when
>the opponents are going to be motivated to prepare against him(her) more than
>preparing against other humans?
>
>Uri


Hello Uri,
Maybe that much money will make the humans "choke".  I understand your
"incentive" comment but I don't think that kind of money will make humans play
that much better.  Most high ranked humans have a high elo because they hate to
lose/love to win and have an inner drive to do the best they can all the time.
Just my opinion of course.  So I say the money will not make that much
difference in the human performance.  I think the key to your question is that
the humans will not be able to practice vs the actual programs they play.  I
think this will be a bigger factor in the performance.
Regards,
Jim



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.