Author: Robert Hyatt
Date: 06:56:37 04/21/01
Go up one level in this thread
On April 21, 2001 at 00:23:30, Uri Blass wrote: > >I dislike deciding that Deeper blue is better than all the PC's without seeing >games against other programs. I saw enough of its predecessor at 1/100th the speed and 1/10th the knowledge to know how well it played. We also had several reports of games played vs micros, by Hsu and Campbell. > >I read in one newspaper a claim that Kasparov played like an IM in the match(the >newspaper is a daily newspaper and not a newspaper only for chess). >It is not my opinion but people who believe it can doubt if Deeper blue was >better than other top programs of that time. I doubt you will find one impartial computer chess expert that would say that DB was no better than the top micro programs. That is simply an insane opinion, just as insane as thinking that a Volkswagen beetle will outrun a Mclaren F1. As far as Kasparov's play goes, I am reminded that when _any_ player runs into a very strong opponent, that player seems to play worse than normal. When in reality he is playing as strongly as always but the opponent makes him _look_ weaker. Kramnik handed Kasparov his head in a basket in their match. Yet I don't think Kasparov played horribly. Kramnik just played better. > >> >>DB never played Fritz. It was first built _after_ the 1995 WCCC event. >> >>In the present case, shredder was in the tournament with the other >>two potential competitors and it beat them both. not once either... > >Yes, but Crafty can get an hardware advantage and more than 8 processors and you >can also say that this crafty never played in tournament because the crafty that >played used clearly a slower hardware(Shredder also did not use 8 processors but >the difference for Crafty is clearly bigger). > >Uri That doesn't matter. I could run Crafty on the fastest hardware I can think of, and perhaps reach 30M nodes per second or so. And I would expect to win any match I would play against micro-based programs (not single games perhaps, but say 10-game or longer matches). Yet I would _not_ expect to win a match vs DB, although I would expect to win an occasional game here and there. Shredder ran on slower hardware than the other programs yet it still won both events. I see _nothing_ that suggests that his parallel version would somehow be weaker than the other two parallel programs. I don't see any indication that DF or DJ are any better than my program in terms of the parallel search efficiency. In fact, I'm not sure they are quite as good as mine (yet). So there is little to suggest that their parallel searches get more out of the hardware than SMK's parallel search. Or do we have to play a _new_ event each time Intel/AMD ups the clock speed a few mhz? Because one may get more out of that speed jump than the others? While I believe that this concept might be reasonable, I don't think the difference is enough to justify a monthly "I am the best" event...
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.