Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: do not see dishonest marketing and chessbase has the right to do it

Author: Mig Greengard

Date: 22:10:04 08/24/01

Go up one level in this thread


On August 24, 2001 at 20:42:33, Uri Blass wrote:

>
>This was practically an open tournament in regards to the pc.
>more than one title does not mean more than one tournwment.

Quite. More to the point, if they really wanted to create separate titles of
more than a fictive nature, they have to play separate tournaments. This is
because of the separate fields different programs faced, as is the nature of a
Swiss-system tournament. Usually when separate titles are given in the same
tournament (best <2000, best woman, best junior (no pun intended), etc.), there
is no assumption of superiority, only distinction of merit. The winner of the
tournament is still the winner of the tournament; the best <2000 is just that.
(The amateur title here, for example. It does not assume the amateurs are
weaker, only that they deserve special recognition.)

In this case the split-title decision seemed to be based on the assumption that
the duals had an advantage and that it was unfair for them to compete for the
same title as the singles. If that is the case, checking to see who faced whom
is interesting. Gromit finished +1 against duals, facing four of the six.
Goliath scored event against four duals. More indicative of place are the
strengths of the respective fields. Gromit's wins came against #s 11, 13, 18,
and 7. (A far weaker field than Rebel, which faced #s 1,2,3, and 5.) XiNiX lost
so many games that it ran out of weak opponents, and Gromit was lucky enough to
face it in the 8th round. Those are the vagaries of the Swiss, which are more
relevant to the standings than the CPUs.

As expected, the programs decided the scores, not the processors. This
crosstable would not surprised anyone had all the machines been on identical
platforms. If a few machines had been using new, single, 2GHz P4s, would the
ICCA have created a separate title? Doubtful, even though there would have been
an "NPS advantage" for those with the faster CPUs. Adding SMP support is just a
programming advantage, like hand-tuning your code for a certain processor or OS,
or having tablebase access. If some don't take advantage of it, and duals are
allowed, then there is nothing more to say. SMP is not a magic wand, it just
became an excuse to create better odds for everyone to get a title. Either have
a uniform platform or stop whining about platform advantages.

Saludos, Mig




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.