Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Fritz SSS TPR after 8 Games - 2643

Author: blass uri

Date: 00:33:48 05/17/00

Go up one level in this thread


On May 16, 2000 at 23:17:41, Robert Hyatt wrote:

>On May 16, 2000 at 22:35:51, Chris Carson wrote:
>
>>Peter,
>>
>>I am leaning towards not counting the forfeit (check spelling).  I
>>think there might be a rule on that.  I am checking, but please take
>>a look also.
>>
>>The Bosboom game is more difficult.  It would count for a human
>>player and give an accurate measure of strength.  This game looses
>>significance as more games are played.  I am reconsidering, however,
>>Bosboom chose to play and chose to quit.  That is a measure of
>>strength.  I would like more opinions on this, Bosboom has a lot
>>of short games in this tournament, I am sure the human opponents will
>>get all the points they are entitled to.  I have had opponents (rated
>>higher than myself) that quit after a few moves because they were upset
>>playing a lower rated person.  I got the points.  I have had opponents
>>that drug the game out to checkmate even when the game was over long
>>before that.
>>
>>Please let me know your rational for not including the game.  Also,
>>should losses not fair to the programs also be excluded since they
>>also distort the ratings.  It seems to me to equal out in the wash.
>>Just my two cents.
>>
>>I try not to set policy, I think the tournament directors do that,
>>also, you can always recalc without the game(s) that you feel are
>>not fair.  I want to be consistant and fair, what do the rules say?
>>Do the rules apply only when we think they should and only to the
>>opponents we choose?
>>
>>I am still thinking on this.  :)
>>
>>Here is the TPR without the Bosboom game, still very impressive:
>>
>>                            FIDE
>>                            Opp
>>Program     HW         TPR  Avg  + = - Tot
>>Fritz  SSS  4x500 PIII 2612 2554 2 4 1  7  - Dutch Championship only
>>Pritz 6/SSS 4x500 PIII 2663 2563 3 4 1  8
>>
>>Best Regards,
>>Chris Carson
>
>
>In this case, it is pretty clear.  Bosboom refused to play.  He played a few
>moves to meet the letter of the rule that said "if you don't play the machine,
>you lose the chance to win the bigger prize."  IMHO he simply "refused to show
>up" but it was a "practical refusal to play" and not a "technical refusal to
>play"...
>
>I'm much more interested in seeing how the program _really_ does, without all
>the silliness of the humans factored in.  Although now it is impossible to do.
>I would count the last Fritz draw as a win, IMHO.  It couldn't lose with the
>opponent having so little time.

It also could not win if the GM decide to repeat the position and I expect the
GM to know enough about computers to know that they are happy to repeat in a
drawn position so he could get an easy draw if he wanted to do it.

I think that practically it could not win so there is no justification to count
it as a win.

Fritz could win in the case that the GM was claiming a draw and even in this
case it is not clear but dependent on the arbiter.

I do not think that Fritz could win if the GM decide to try to do a draw on the
board by repetition and I think the GM knows enough about computer to understand
that programs are usually happy to do a draw with a pawn down.

Maybe Fritz could win the first game and not this game but we will never know
because the operator offered the opponent a draw when the opponent was in time
trouble.

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.