Author: Dann Corbit
Date: 02:30:07 06/16/00
Go up one level in this thread
On June 16, 2000 at 04:11:59, Masciulli Gianluigi wrote:
>On June 15, 2000 at 20:56:03, Dann Corbit wrote:
>
>>On June 15, 2000 at 20:44:19, Peter Kappler wrote:
>>
>>>
>
>snip
>
>>The winner of each division will move up to the next higher classification.
>>The loser of each division will move down to the next lower division.
>>The very top and very bottom programs have nowhere to go, so they will be
>>crowned as "World champion" and "Most playable" {to be charitable}.
>
>Its very easy to imagine who will be the "most playable" because all the program
>with force similar to golem has not qualified!!
>In your list golem is the last one with a gap of about 300 points.
>
>Another thing: if you do a calibration too precise may be you will not need
>the four crawns turnament at all;
>It will be just like tennis. the one that win the tounament is good ... but to
>be considered best you have to top the rating list.
They still make Agassi, Rafter, and Sampras play the low seeds. Just because.
Once in a while there is an upset, but not often.
>gianluigi
>
>ps. what program do you use to produce the ratings?
I use ELOSTAT, and crosscheck with ChessBase to ensure approximate agreement.
>How do you assess the
>initial elo for the programs (the whole list would remain consistent if you
>shift all up or down by the same elo)
Intial guess is 2300 for the pool. It makes no difference if it is not even
close, since the delta ELO values will still be consistent.
>ciao
>
>
>
>
>>
>>You can read a lot more about it over on the WinBoard forum:
>>http://f11.parsimony.net/forum16635/
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.