Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Pairings round 6 : swiss "perfect" makes very strange pairings

Author: Uri Blass

Date: 13:56:05 08/23/00

Go up one level in this thread


On August 23, 2000 at 16:46:05, Sylvain Renard wrote:

>>A quick note.  When you have 9 rounds, and 14 players, it is _not_ a Swiss
>>by any measure.  It is nearly a round robin, in fact.  I don't think pairings
>>matter at all now, other than that the top programs should play each other
>>one time.  Mistakes in the pairings will work themselves out.  It will screw
>>up a color here and there, but eventually all the top programs have to play.
>
>  Playing 9 rounds with 14 players is really unusual. 13 rounds would be
>clearly better. This year in Massy (France) we played a 6 rounds tournament
>with 10 programs. It was not satisfying because one program finished
>with 4/6 playing all weaker programs and only 2 strong programs, and
>another on had 3/6 but played against all strong programs.
>Something similar could happen in London.

I think that it is possible to prevent this problem by better ranking system.

The place of the program should not be decided by the number of points it got
but by the rating that it is going to get assuming that the result of the
tournament repeats again and again.(it is possible to calculate the rating by a
program if we assume that the results repeat 1000000 times in a row).

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.