Author: Robert Hyatt
Date: 18:42:53 12/22/99
Go up one level in this thread
On December 22, 1999 at 19:43:03, John Warfield wrote: >On December 22, 1999 at 17:11:10, Robert Hyatt wrote: > >>On December 22, 1999 at 14:01:43, Graham Laight wrote: >> >>>On December 22, 1999 at 10:00:07, Robert Hyatt wrote: >>> >>>>>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. >>>> >>>> >>>>Baloney nowadays. No program would consistently play at near 2700 at > > > > > >>>>aegon. >>> >>>I respect your point of view as an experienced and practising computer chess >>>person. However, I feel compelled to ask: have you any evidence to support this >>>assertion? If not, why do you think it is so? >>> >>>At the end of the day, good chess is good chess. A machine that can beat more >>>computers is also likely to beat more humans. >>> >>>Graham >> >>I feel that way based on watching them play. > > > hmm Strange that a man of Science such as yourself would go on "feeling" and >no facts. Schredder 4 just got finish Drawing Karpov at 40/2, so they could very >well be near 2700, although I doubt it. By the way what is your rating? I would >like to know how much chess knowledge is behind these "feelings" > > I had a provisional rating of just over 2200. But that was provisional. I never played in the needed 24 games. I have played a FM on ICC (used to play him pretty regularly) and generally broke even with him. I'd guess he was really a 2000-2100 FIDE player as he was not playing much tournament chess when he was on ICC. I have a pretty good 'feeling' about the game. And a lot of my 'feelings' about how computers play come from both personal observation, plus lots of discussion with players _far_ stronger than I am, a couple of them being GMs and a couple more being active IM players... As far as going on 'feeling' and not 'fact' that is simply wrong. I haven't seen any great results by computers against GM players at 40/2. Drawing Karpov is not bad. But he is a long way from a 2700 player today... Once I see that programs do play evenly with GM players at 40/2, my estimate will change. But not until... > > > 2700 is an incredibly high >>rating, better than all but maybe 10 players. I don't believe a computer >>other than Deep Blue has a prayer of playing at that level, on any hardware >>forseeable for the next 5 years, assuming the PC platform. >> >>Rebel has almost hit 2500 against GM and IM players. to get to 2700 would >>require a herculean effort. IE it would have to win 3 games for every loss >>vs 2500 players, which seems impossible at present...
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.