Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Russek -Rebel Match, Game 2

Author: Len Eisner

Date: 09:58:24 01/01/00

Go up one level in this thread


On January 01, 2000 at 10:20:48, James T. Walker wrote:

>Hello Uri,
>You may be right. I assume you think that a doubling of speed is worth LESS than
>60 points.  You must also consider that if you are right then a processor
>running at 1/2 speed does not lose 60 points either.  Are you so sure about
>that??  Doubling speed gives the program 2 times as much information and vs
>computers it is easy to prove that the gain is very close to the 60 points most
>people accept.

*snip*

Larry Kaufman wrote an interesting article several years ago comparing the
effects of increased speed vs. increased time.  He concluded an increase in
program speed was worth more than an increase in thinking time assuming the
human’s time does not vary.

This is difficult to explain so let me give an example.

First Game
Fritz 6 PII 400
Human: 60 seconds a move
Fritz: 30 seconds a move

Second Game
Fritz 6 PII 200
Human: 60 seconds per move
Fritz: 60 seconds per move

You would think Fritz would play equally well in both games but that’s
incorrect.  In game one, Fritz would be stronger because of the effects of the
permanent brain.  If you assume Fritz guesses you move 1/3 of the time, then the
value of the permanent brain is (1/3 X 60 seconds) = 20 seconds per move over
the course of the entire game.

So for game one the total time per move for Fritz including the permanent brain
is 20 + 30 = 50 seconds.  In game two the total time would be 20 + 60 = 80
seconds per move. So Fritz is stroger in game one because 50 seconds per move on
a PII 400 is better than 80 seconds per move on a PII 200.

Len



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.