Author: Sandro Necchi
Date: 07:54:23 05/15/04
Go up one level in this thread
On May 15, 2004 at 10:32:48, Uri Blass wrote: >On May 15, 2004 at 10:00:15, Sandro Necchi wrote: > >>On May 15, 2004 at 09:42:57, Uri Blass wrote: >> >>>On May 15, 2004 at 04:59:19, Jorge Pichard wrote: >>> >>>>On May 15, 2004 at 04:20:13, Uri Blass wrote: >>>> >>>>>On May 15, 2004 at 00:32:40, ERIQ wrote: >>>>> >>>>>>I would think the ratings could be closer to fide's if the programs ran on 3ghz+ >>>>>>machines. >>>>>> >>>>>>I also think engines could do well in tourneys if competing w/ humans not >>>>>>knowing which program it was or what book program is using. Any thoughts? >>>>> >>>>>Why do you look only at the top of the list? >>>>>I think that the rating of the programs at the bottom of the list is too low >>>>>if you compare it with humans. >>>>> >>>>>Uri >>>> >>>> >>>>The longest the list of programs participating the lowest the rating of the >>>>programs on the bottom of the list will be. The SSDF should not be testing more >>>>than 20 programs, in reality who care about programs with lower rating than >>>>2200. >>>> >>>>Jorge >>> >>>The opposite. >>> >>>Who cares for the rating against humans of the strong programs except some GM's? >>> >>>Maybe it is more important to have correct rating for the lowest part because >>>people care about rating of things at similiar strength to them. >>> >>>For me this is also not important but I see no reason that humans will care >>>about rating of programs against humans when they beat them easily. >>> >>>Uri >> >>I do not agree. >> >>Most people use the chess programs to improve their play and to check >>games/positions. >> >>The weaker programs are the worst on this matter and only the best ones may be >>good enough (I am personally still not satisfied), so for most of the people >>(chess players) weaker programs are simply "museum toys" or something not >>interesting at all... >> >>Computer chess fans (collectors) are very very few. >> >>Sandro > >I did not claim that strength is not important for buyers but I doubt if the >rating against humans is important. Well, it is difficult to have a rating against humans as the strongest players do not play enough and we do not know if when they play they do it at their best or simply try not to lose to play more matches...and get more money. I also think that a stronger engine against other chess programs will perform better against human chess players too, even if not in the same way. To have engines stronger against humans and weaker against computers is not valid anymore (to me). > >The only important thing is that the new thing is better and the ssdf gives that >information. Yes, but why buy a cheaper program if a "much stronger" one would cost about the same? > >Uri Sandro
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.