Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: SSDF, Fritz5 games

Author: Robert Hyatt

Date: 11:07:23 02/28/98

Go up one level in this thread


On February 28, 1998 at 12:03:03, Enrique Irazoqui wrote:

>On February 28, 1998 at 10:55:31, Robert Hyatt wrote:
>
>>On February 28, 1998 at 08:00:35, Enrique Irazoqui wrote:
>>
>>>On February 28, 1998 at 05:40:01, Moritz Berger wrote:
>>>
>>>>On February 27, 1998 at 17:52:53, Thorsten Czub wrote:
>>>>
>>>>>>That looked quite amazing until I noticed the machines.  Fritz had a
>>>>>>3X speed advantage, maybe more, so that is not quite so surprising as
>>>>>>if they were run on equal hardware...
>>>>>
>>>>>Confirm again !
>>>>>Are this official ssdf-games ???
>>>>
>>>>SSDF has been doing this with *ALL* over programs for a long time ...
>>>>
>>>>Now Thorsten finds out that they treat Fritz 5 in the same way (i.e.
>>>>playing 200MMX vs. P90).
>>>>
>>>>*Now* he's beginning to wonder ...
>>>>
>>>>Moritz
>>>
>>>And I wonder why.
>>>
>>>By the way, SSDF is not the only one to run matches this way. Even in
>>>Paris we saw programs running on platforms 2 or 3 times faster than
>>>others.
>>>
>>>Something else: a 3x increase in speed is worth roughly 100 points,
>>>equivalent to scoring 62%. In both matches reported by Tony, Fritz 5
>>>scored almost 80%, which represents quite a bit more than the expected
>>>increase in strength due to a 3x speed up.
>>>
>>>Enrique
>>
>>
>>
>>I'm not sure I agree with the "100 Elo" idea.  IE for years, I have seen
>>this "doubling speed produces a 70 point rating increase."  I would
>>guess
>>that *may* be right in normal games.  But I often run matches of Crafty
>>vs
>>Crafty, and when I try a 2x time handicap, I get *huge* changes in the
>>win
>>loss ratio.  I haven't done this recently, but the last time I tried it,
>>I
>>saw the one getting a 2x handicap win about 3 of every 4... which was
>>more
>>than I expected...
>>
>>But in any case, it would be nice to see only 200mmx vs 200mmx or
>>whatever,
>>and stop this unequal platform competition, since it doesn't provide any
>>useful information, basically...
>
>Of course it would be better to run all matches on equal platforms, but
>this does not disqualify results Fritz 5 got in matches on uneven
>computers. SSDF always did this with the 59,000 games they played so
>far. Fritz 5 played both, matches on two P200MMX and on P200MMX vs.
>P90s. Like all other top programs recently: Rebel 8 and 9, Hiarcs 6,
>Nimzo98, Mchess 6 and 7, Genius 5  Also these programs played matches on
>P200MMX against P90s.
>
>In the use of different platforms I don t see anything atypical that
>allows singling out Fritz s case.
>
>Enrique


I wasn't implying anything wrong at all.  Just the huge one-sided wins
by fritz looks amazing until I noticed the 3X speed handicap.  That
makes
those particular win/lose numbers mean something different than if they
were
posted for equal hardware matches...



This page took 0.01 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.