Author: Howard Exner
Date: 23:12:11 01/28/00
Go up one level in this thread
On January 29, 2000 at 01:10:51, Dann Corbit wrote: >On January 29, 2000 at 01:02:13, Howard Exner wrote: >>On January 28, 2000 at 19:29:36, Dann Corbit wrote: >>>On January 28, 2000 at 19:21:15, John Warfield wrote: >>> >>>>On January 28, 2000 at 16:49:13, Jeff Lischer wrote: >>>> >>>>>I calculate 2521 including all 23 games Rebel has played (Anand-2, GM >>>>>Challenge-15, Russek-4, and Israeli league-2). >>>> >>>> >>>>Looks like Rebel is GM Strength, no question. Ofcourse the rating would even be >>>>higher if only the k6-3 600 were used >>> >>>Why do you say that? It has not achieved the levels required to be a FIDE GM as >>>far as norms go. >> >>Forget the norms route, as we all now those will never be on the table for >>computers. >>Look at the rating and play over the games. Call it either a weak GM or a very >>strong IM ... it's rating fits into both. The statement, "looks like Rebel is GM >>strength" is not saying that Rebel has the gone through the hoops to get the >>coveted title, rather its rating falls clearly into the GM range. Chris Carson's >>summary of GM ratings covered the range and averages for a GM. >>> >>>Looks more like an IM to me, but a very good one. Still early. Wait until the >>>GM's start playing more anticomputer chess. >> >>How long are we going to have to wait for this? How do we know the opponents >>are not already anticomputer aware and that chess being the complex beast that >>it is, will not simply allow one opponent to dictate the course of a game at >>will. Also, how much in depth studying would it take for a GM or IM to unlock >>the secrets of antiGM play? Don't most GM's and IM's own a computer? Kasparov >>openly states that he uses them to test out new opening ideas. One would think >>lesser mortals would follow his example. > >Perhaps it is just a particular talent that some IM's have. But have you seen >Hawkeye shred computers on the internet? No, but his rating must greatly surpass his shredded computer foes. The only games that I know of being posted here from the net was the 4-0 trouncing of Ferret over IM Blatny at a time control somewhere around game/30. Computers have many weaknesses and I'll wager from Blatny's play on the internet that he is aware of them. How was this knowledge helpful to him? >If GM's used the same strategy, I >think they would do better. And look at that rebel rematch by the GM that lost >the first time and won the second. Looks like he did his homework to me. Was that due to anticomputer strategy or just good chess? Or was that the game where Rebel shut the door on his bishop with e5? If so no anticomputer stuff going on there, just self destruction on Rebel's part in this case. >I >have sent email to several of the participants. Believe it or not, I really >believe that some of them have *no idea* how to play against a computer. Some detailed thoughts as to why they make certain moves vs computers would be very interesting. I think anticomputer play is to a large degree knowing your opponents weaknesses and looking for opportunities to capitalize on them. There are two parts to this. "Knowing the weaknesses" of computers I believe is not too difficult a body of knowledge to acquire but the second part, "capitalizing on them" is where the challenge lies. Many of us here have sufficient knowledge of these weaknesses, programmers and players of various ratings alike. Are we going to automatically steer the game into our favor with that knowledge? Often by the unorthodox way in which computers play, the game moves out of these anticomputer positions into a wild free for all. I agree that many positions will arise in a game where the human will see exactly what themes are at play and many positions where computers will simply wander about in the dark. Computers do have many pluses in their corner also and it seems in Rebel's case the pros and cons of its play are averaging out at around 2520 fide. >>>I predict a plummet for all >>>computer programs in contests with people once they get the hang of it. >> >>Rebel's games vs humans are public knowledge know, plus opponents can go out and >>buy it, yet my feeling is it's rating will only move forward at a slow steady >>pace. > >Time will tell. No bets. ;-) > >>Dann, don't make me force you to write yet another poem! :) >We saw what happened the last time. And it wasn't a pretty sight. The "knight >fork" line still makes me cringe when I think of it. But that's the mental >image I had, so I had no choice but to write it. Maybe I'll request limerick mode.
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.