Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Computer or Human in a REAL match??

Author: martin fierz

Date: 17:18:16 08/19/02

Go up one level in this thread


On August 19, 2002 at 15:08:00, Dann Corbit wrote:

>On August 19, 2002 at 12:58:08, Jeff White wrote:
>
>>In a 24 game match with the time control of 2 1/2 hours for 40 moves, who wins?
>>The best human player or the best computer player? In my opinion this would be
>>the only true way to solve this. How come it's never been done like this before?
>
>Why do you imagine that 24 games will create an answer to the question?
>
>It will only raise more questions.
>
>There have been many man/machine matches.  None of 24 games that I know of, but
>there is nothing magical about the number 24.

there sure is nothing magical to "24", but i think it's clear that longer
matches are more desirable from a statistical point of view.
i guess they should favor the human, who can try to adapt to his opponent -
something the computer can't do very well (change openings is probably all you
can do as a computer). the short matches we have seen so far (van wely - rebel,
4 games; deep blue - kasparov, 6 games; fritz - huebner, 6 games etc.) tend to
mask the fact that the machines can't learn, since the humans don't get a second
chance to do it right.

aloha
  martin



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.