Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Paris WMCCC - were programs better than in Jakarta (1996)?

Author: Andreas Mader

Date: 23:53:01 11/12/97

Go up one level in this thread


On November 12, 1997 at 12:33:37, Amir Ban wrote:

>On November 12, 1997 at 11:43:33, Bas Hamstra wrote:
>
>>>In my opinion there are a couple of programs around who
>>>are able to win such a tournament. You need a strong
>>>program but also a bit of luck to be at the very top.
>>
>>Andreas Mader ran an interesting simulation.
>>
>>- A mediocre program cannot win a tournament by luck (< 5% chance)
>>- The very best program may very well end below the top 3 by pure bad
>>luck
>> (50% chance) or even below the top 5.
>>- Chance that best program wins tournament is only 30%
>>
>>So one of the toppers will certainly win, but one particularly tipped
>>topper willl probably not win.
>>
>>
>>Bas Hamstra.
>
>
>Something wrong with the mathematics here. Surely this depends on what
>the differences in strength are. If the difference between the best and
>worst program is 5 ELO points, then they are all basically the same and
>everyone has equal chances. If the best is 200 ELO points stronger than
>the 2nd best, then he has virtually 100% chances of winning.
>
>Amir

Nothing wrong with the mathematics here. As I wrote in my post I took
the ELO ratings of the top 30 of the current SSDF list and added four
'low' ratings (2200, 2100, 2000 and 1800) to get to 34. Take a look at
the SSDF list and you can see the ratings that were 'involved' in this
simulation.

Best wishes
Andreas



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.