Author: Robert Hyatt
Date: 20:56:35 04/17/01
Go up one level in this thread
On April 17, 2001 at 13:51:20, Sandro Necchi wrote: >On April 17, 2001 at 13:42:31, Christophe Theron wrote: > >>On April 17, 2001 at 13:19:19, Sandro Necchi wrote: >> >>>We are the World Champion until the ICCA new tournament will be held and another >>>program win the title. >>>If we do not recognize such a title and event there is no reason to make such >>>event anymore!!!!!!!!!!!!!!! >>> >>>Sandro Necchi >> >> >>The event is recognized for what it is: a nice event with a lot of randomness in >>the result, but which is fine because it is the only opportunity for programmers >>and computer chess businessmen to meet each other. > >I do not agre. >> >>I personally enjoyed a lot participating in person to the WMCCC 1997 (Paris) and >>the WCCC 1999 (Paderborn). I have nice memories of everybody, and for me that >>was a dream come true. >> >>But you should not assume people are so stupid that they do not understand the >>low reliability of such events. > >That's your opinion! >A World champion title is a title! > >> >>If the event was so reliable, why would the SSDF results be so interesting for >>everybody? Why would testers play home tournaments? > >That's their choice. I think this argument is amusing. Use the SSDF rating. Ignore the current title "World computer chess champion". But for the human, ignore the FIDE rating list and grab the human with the title "world champion." Consistent, wouldn't you say? :) So which human should play the computer? Highest rated (FIDE) vs highest rated (SSDF)? World champion vs world champion? Have a playoff with the top 4 FIDE-rated humans? And a playoff between the top 4 SSDF-rated computers? Winners meet? Be consistent guys. No apples and oranges at the same time.. sheesh... It isn't about what is the _right_ thing to do. It is about the selfish thing to do... >> >>What you are doing here is trying to disinform people about the reliability of >>chess events. > >I am not trying to disinform anyone. I did not say that winning such event >automatically means that the program is the strongest. >I say that it becomes the World Champion! > >>But it won't work. People are more and more aware that a high >>number of games is necessary to evaluate the relative strength of chess >>programs, and that a 7 or 9 or even 11 rounds event means close to nothing. > >Look, I am involved in computer chess before you did, so I know this better than >you. I did not say the contrary. >I am not trying to convince anyone that you! >> Your point is understood. But a bit of twisting is necessary. "The ends justify the means" comes to mind... >>In the tournaments organized by the ICCA, all you can do is say that there must >>be a stength difference between the bottom and the top of the final rankings. >>But between, say, the 5 first programs it is impossible to say which is the >>best. Even if the same program wins time after time? that's simply baloney. Luck only stretches so far... > >if these tournaments means nothing then lets cancel them! >> >>20 years ago these events were significant because there were significant >>differences in the strength of chess programs. That's why the "Chess" program >>was reliably leading, and that's also why Richard Lang's programs have been able >>to do the same in the eighties. >> >>Nowadays the difference in playing strength is less obvious, and the reliability >>of the ICCA tournaments is close to nil. > >You are offending ICCA! >> >>This is not a critic against the ICCA. Given the time and money constraints it >>is impossible to do any better, and anyway the events are very enjoyable for the >>participants. > >Sandro Necchi >> >> >> >> Christophe
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.