Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Crafty ponder experiment in CSS forum: ponder ON is worse

Author: Harald Faber

Date: 08:13:48 09/26/02

Go up one level in this thread


On September 26, 2002 at 08:07:00, Volker Pittlik wrote:

>On September 26, 2002 at 07:40:26, Harald Faber wrote:
>
>...
>>So the programmers should consider playing without permanent brain, this should
>>reduce the chances of hash collisions etc. And it does not weaken the engine.
>>Only pros, no cons, think about it. :-)
>
>Weakening the engine by setting ponder off and playing tournaments where all
>opponents use ponder=on or ponder=off are two complete different circumstances.


Partly agreed.
a) Why is it weakening? The results show almost identical numbers, so where is
the weakening?
b) The opponents played with pb=on, so they played at maximum strength according
to your statement above. Do you expect the opponents perform better with pb=off?


>The test tried to analyse if results of tournaments with different settings are
>comparable. No word about both settings are playing at equal strengths.


No? Crafty/pb=on vs. programs a-g/pb=on got almost the same results as
Crafty/pb=off vs. programs a-g/pb=on. This usually leads to the thesis that
Crafty/pb=on is as strong as Crafty/pb=off. If no, why not?


>How long will it take until the last will stop to cunfuse this issues?
>Volker


First clarify so that I can understand this simple experiment.



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.