Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: New crap statement ? Perpetuum mobile(test positions for Deep Fritz)

Author: Uri Blass

Date: 07:19:33 10/04/01

Go up one level in this thread


On October 04, 2001 at 08:22:31, Sune Fischer wrote:

>On October 04, 2001 at 07:47:40, Uri Blass wrote:
>>
>>Bob can at most say that it is impossible for crafty but
>>I understand that even for Crafty there are positions when parallel search can
>>give programs super linear improvement.
>>
>>The question is what is the reason for the super linear improvement of Crafty in
>>some positions and how it is possible to prove that there cannot be
>>average superlinear improvement for the existing Crafty at long time control
>>that Crafty was not tested(I agree that in this case it is possible to improve
>>Crafty for one processor at long time control but the discussion is about
>>existing programs and not about science that talk about theoretic programs).
>>
>>
>>Another note is that an average super linear improvement does not mean always
>>doing better than linear improvement.
>>
>>
>>If I have a way by 2 processors to be 4 times faster in 1/2 of the cases and 4
>>times slower in 1/2 of the cases then I have a super linear improvement but I
>>think that it is still better to be twice faster in all the cases.
>>
>>Uri
>
>If I ask you to guess a number between 1 and 1000, you could be lucky and guess
>it at the very first guess. You could also be unlucky and guess it last.
>On average we would expect you to need 500 guesses to find the number using
>random guesses.
>
>Move ordering has that bit of randomness about it, sometimes you pridict wrong
>and you get the right move second (or even last).
>Most of the time you would get the right move first though (or else just reverse
>move ordering, so that is a safe assumtion).
>Example:
>
>1. move has 60% chance of being the best.
>2. move has 30% chance of being the best.
>3. and rest 10%
>
>Search move 1 => succes has 60% chance.
>Search half move 1 and half move 2 in the same time => 0.5*(60+30)=45% chance.
>
>It will not pay off on average, only once and a while.

It is not so simple.
programs do not have exact scores for a lot of moves and they search for good
enough moves and not for the best moves.

if 2 processors search for a good enough move then it is possible that they can
find in small part of the cases better good enough move than one processor and
later in the next iterations the computer can remember that better good enough
move to get a better branching factor.

Uri



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.