Author: Christopher R. Dorr
Date: 10:05:56 04/12/00
Go up one level in this thread
Of course, in short matches, anything can happen. Kaspy's negative results in short matches against Genius are no more indicative of his current prowess against computers than would be likely private scores of 10-0 against computers in training. I've never seen these published, but I'd bet heavily that he has had these kinds of runs in private. I still think that today's programs aren't much better than FIDE 2500 at slow time controls. Kasparov against 6 FIDE 2500's? Whitewash! I'd venture on 5-1 or 6-0. He has shown that he can destroy several human GMs (of 2500+ strength) in clock simuls before; there is no reason to believe that he couldn't do it again. And he would have more time than one would think. If he could knock out the first 8-10 moves almost instantly (which is likely), and at least 25% of all remaining moves are instant moves (forced recaptures or obvious moves), then he would be left with almost 2 hours for 20 or so moves. Given that many games would be thematically similar, I don't see time as a major limiting factor. I've played on both sides of simuls before, both against GMs (as the weaker player) and USCF 1800s and below (as the stronger player). While some of the GMs I've played in tournament games have taken the full 2 hours to beat me, the ones I've played in simuls have spent perhaps a total of 20 minutes on my games, and beaten me equally badly. I usually will take my full time to beat a USCF 1800 in a tournament game. But in simuls, I have a seriously positive record against them, taking a mere fraction of that time.This is the advantage that *kaspy* has; he knows when to spend time, and when not to. I think it would be a very interesting match, but I'd still put my $$$ on Kaspy. Chris On April 12, 2000 at 09:07:37, blass uri wrote: >On April 12, 2000 at 08:59:21, Christopher R. Dorr wrote: > >>I disagree, >> >>Kasparov has taken on teams of several GM's at once before, and won handily. I >>believe it was the German National team two or three years ago (Somebody please >>correct me if I'm wrong). If Today's programs are GMs at 40/2 (which I believe >>to be the case, but which has not been conclusively proven yet), then they are >>fairly average GMs, i.e. 2500-2550 FIDE. I would bet that Kasparov could defeat >>32 FIDE 2500 GM's in a simul. Probably not with a lopsided score, but still >>probably something in the neighborhood of 18-14. I'm about USCF 2200, and >>believe that I could defeat a team of 1850s/1900s with a similar score, and that >>is a similar rating differential. Additionally, I think it would be even easier >>for Kaspy against the comps, as even the best micros still have identifiable and >>exploitable weaknesses, and Kaspy has a great deal of experience against comps. >>Against today's comps, I'd venture a score of perhaps 20-12 in favor of Kaspy. >> >>Chris > >I disagree > >Kasparov had problems against Genius3 in 25 minutes/game and it was with old >hardware(he lost 1.5:.5 against p90 and won 1.5:.5 against p120). > >kasparov probably can get only 50% if he plays against 6 computers in a >simultan(3 hours/game) because he has only 30 minutes against every computer and >I assume that kasparov is at the same level as computers in 30 minutes/game. > >2 hours/40 moves is something close to 3 hours/game so I expect 50% against 6 >computers and less than 50% against 32 computers. > >Uri
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.