Author: Robert Hyatt
Date: 07:16:44 07/18/99
Go up one level in this thread
On July 18, 1999 at 02:20:38, Hans Christian Lykke wrote: >On July 17, 1999 at 11:39:47, Robert Hyatt wrote: > >>On July 17, 1999 at 10:49:03, KarinsDad wrote: >> >>>On July 16, 1999 at 23:18:36, Robert Hyatt wrote: >>> >>>>On July 16, 1999 at 11:18:06, Aloisio Ponti Lopes wrote: >>>> >>>>>What is the ELO performance of Schredder in the last WCC tournament? >>>>>I have some old chess programs (Sargon 3, Sargon 5), does anyone know their >>>>>estimated ELO ratings?(could not see them in any listing...) >>>>>A. Ponti >>>> >>>> >>>> >>>>the question is impossible to answer. To produce an Elo rating (which >>>>would imply FIDE to most of us) we would need ratings for at least the >>>>simple majority of the programs that participated. No such figures are >>>>available, since none have FIDE ratings... >>> >>>Couldn't you approximate by taking the SSDF ratings for the closest software / >>>hardware combinations and calculating an Elo rating from that? Then >>>approximating the maximum increase in rating due to improved hardware at WCC, >>>recalculate the ratings, and then come up with a range (such as 2425-2530) or >>>somesuch? >> >>The SSDF has nothing to do with FIDE ratings. Years ago they had some human >>games included to 'seed' the initial rating pool. But that has been washed out >>and now the ratings are highly inflated. And have no correlation with FIDE or >>any other 'human' type standard at all. > >It´s not correct that the human games were included earlier, and now removed. > >This is from the SSDF FAQ > >"SSDF has played several hundred games between computers and human players in >serious tournaments and used these results to set a "correct" absolute level for >the rating list according to Swedish conditions. Different national rating >systems are not completely in accordance though, and that has to be taken into >account when reading our list. For instance, US ratings seems to lie >approximately 50 points above the corresponding Swedish ratings (maybe more when >below 2000 and less on the other side of the scale). For ourselves we obviously >use the Swedish scale. > >We firmly believe that our ratings are correct in the sense that if a computer >were to play a sufficient number of games against Swedish humans, it would end >up with a rating close to what it has on our list. Unfortunately, as programs >get better it becomes increasingly difficult to arrange meaningful games against >human players. Reassuringly, we've noted that our ratings are fairly consistent >with the results from the yearly Aegon tournament in Holland. " > >Venlig hilsen > >Hans Chr. Lykke - SSDF please try again. +when+ were those games played? Any recently? I didn't say "removed"... I said "washed out". Elo rating is a function of statistical performance. If you play in one rating pool (FIDE) and get a rating, then play in another rating pool, that FIDE rating is "washed out" over time... I've seen no evidence that SSDF ratings are anything other than the product of playing rated games between programs via SSDF testers, at least for the past 5-7 years. Which _still_ means SSDF ratings have _nothing_ to do with FIDE ratings. Because it exaggerates rating differences due to hardware and software changes.. > >http://home3.inet.tele.dk/hclykke/ > > >> >> >> >>> >>>It wouldn't be that accurate, but it might give you an idea of the range of >>>ability (of course, Elo ratings aren't that accurate to begin with). >>> >>>Has anybody tried to do something like this? >>> >>>KarinsDad :)
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.