Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Ruffian 2.0.0 strength - another results

Author: Ulrich Tuerke

Date: 08:12:13 01/02/04

Go up one level in this thread


On January 01, 2004 at 16:07:33, Robert Hyatt wrote:

>On January 01, 2004 at 15:15:16, Luis Smith wrote:
>
>>On January 01, 2004 at 14:56:22, Robert Hyatt wrote:
>>
>>>On January 01, 2004 at 12:32:00, Luis Smith wrote:
>>>
>>>>>So I suggest using PonderOff for the WCCC, because poff leads to the same
>>>>>results as many "experts" have proven.
>>>>
>>>>I don't see why, at WCCC all the games are not played on the same computer.
>>>>Ponder on = higher quality of games.
>>>
>>>In a tournament, "quality" counts for _nothing_.  The _result_ is all that
>>>counts...
>>
>>While very true Bob, I was just remarking about how its the people who watch and
>>go over these games to want quality.
>
>I was just responding to the pondering/no pondering point.  I happen to
>think pondering is better, because that is the way engines generally are
>tested, and that means they are potentially going to play better.  While
>it is possible, as some claim, that there is no difference, there is also
>a reasonable probability that there is a significant difference.  Much older
>versions of Crafty were certainly worse with ponder=off.  I've tuned it from
>time to time, but I have no idea if it plays as well with ponder=off as with
>ponder=on.  And since ponder=on is more interesting since that is the way I
>play all the time, I _know_ that setting will work as well as any for my
>program.

I think that with pondering option enabled, we will observe a less number of
draws. In games, you have often the situation that either side has a certain,
but still not decisive advantage; as a result the replies of the other side may
be forced. The side standing better will have permanently ponder hits in this
situation which may help a lot to realize the advantage. This may be more
difficult when pondering is disabled.

Uli


>
>
>>
>>>
>>>>
>>>>Another suggestion could be using shorter
>>>>>time controls. This way the WCCC could be finished in one day.
>>>>>Many "experts" have proven that shorter time controls lead to the same results
>>>>>as long time controls.
>>>>
>>>>It could be done, but longer time control = better quality of games.
>>>>
>>>>>kind regards
>>>>>Bernhard



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.