Author: Uri Blass
Date: 22:20:58 07/03/01
Go up one level in this thread
On July 04, 2001 at 00:44:00, Robert Hyatt wrote: >On July 03, 2001 at 14:52:23, Sune Larsson wrote: > >>On July 03, 2001 at 11:24:55, Robert Hyatt wrote: >> >>>On July 03, 2001 at 09:08:15, Uri Blass wrote: >>> >>>>On July 03, 2001 at 08:55:14, Robert Hyatt wrote: >>>> >>>>>On July 03, 2001 at 03:24:25, Adolfo Bormida wrote: >>>>> >>>>>>On July 03, 2001 at 02:10:32, Albert Silver wrote: >>>>>> >>>>>>>On July 03, 2001 at 01:08:54, Tony Asdourian wrote: >>>>>>> >>>>>>>>CT just beat a second GM, this one rated 2550 or so. Apparently made a really >>>>>>>>nice sac in the process. I've always been quite impressed with Fritz6, but >>>>>>>>looking at what CT 14.0 and GT2 are doing reminds me of the old Novag and RISC >>>>>>>>2500 dedicated computers, which for their time played with attacking flair but >>>>>>>>were also more than just tactical monsters, they seemed to "enjoy" playing >>>>>>>>aggresively. It is fun to watch this new program play so well. The online >>>>>>>>interview with Christopher Theron he gave recently is also the most revealing >>>>>>>>I've read from a top programmer in quite a while. >>>>>>> >>>>>>>It's game against Rodriguez was good, but this one was really pathetic. What the >>>>>>>heck was Ricardi thinking when he played the opening? He fell for a fairly basic >>>>>>>tactic on move 7 leaving him much worse and by move 13 he's down a pawn. After >>>>>>>that it was just a matter of time. Really not a lot to say about it other than >>>>>>>it was a point for Tiger. >>>>>>> >>>>>>> Albert >>>>>> >>>>>>Hi Albert >>>>>>I'm not agree! Ricardi was played the oppenning not well. Then in move 13th. He >>>>>>sac a pawn for some compensation according Ricardi's comments after the game 1) >>>>>>Can castling 2)Use the Open c file. If you can see the evaluation 12..., Ne7 is >>>>>>the second choice at ply 14 according ChessTiger. I think when you say "this one >>>>>>was really pathetic" you are using a very strong word and you no have respect >>>>>>for a 2550 GM! (who beat Karpov last year). Can you be sure than you can beat >>>>>>the "patetic Ricardi" after move 13? >>>>>> >>>>>>No offense, not missunderstand me. English is not my native language and only >>>>>>give my view. >>>>>> >>>>>>Regards, Adolfo >>>>> >>>>>Ricardi isn't a very good computer opponent. Against Crafty on ICC, he has >>>>>won 23, lost 129, drawn 16. >>>> >>>>It means almost 20% >>>> >>>>What is the time control of these games? >>>> >>>> Against scrappy he has lost 2 but that is all he >>>>>has played. Other GMs do much better than that. >>>> >>>>What is the score of Crafty and Scrappy against GM's at standard time control at >>>>the same time that Ricardi got the +23 -129 =16? >>>> >>>>Uri >>> >>> >>>Crafty has done well at longer time controls. But again, because most GM >>>players prefer to play "real chess" rather than "anti-computer chess" against >>>it. I am sure Ricardi could do better, _if_ he was willing to change his style >>>to avoid the strengths of the computer. See my response to Dan to see what I >>>mean... >> >> Ok, another aspect is that playing your "normal chess" or as you say, >> "real chess" aginst a top program, opens up real good options for training >> and gives you a tremendous challenge. I find those possibilities more >> interesting than repeated "anti-computer" strategies. Though the latter would >> fall into the category "computer weaknesses" and might serve as study material >> for the programmers to do list. >> >> Sune > > >I agree 100%. A GM knows he will never face a computer in a real rated FIDE >event. So preparation is a waste of time. But let computers play in any FIDE >event, and let them actually start winning _prizes_ and I'll bet there will be >a _whole_ lot of studying going on, real fast. :) It depends on the number of computers in Fide events. If only one computer is going to play then I believe that it can get easily not only the GM norm's but also rating of more than 2600 because it is a waste of time for GM's to prepare for a computer when more than 99% of their opponents in Fide events are humans. If more than 20 programs are going to play in tournaments then the situation may be different and humans are going to learn better to play against computers. The interesting question is what is the rating that computers can get in the second case. 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.