Author: Ed Panek
Date: 16:04:10 05/12/01
Go up one level in this thread
On May 12, 2001 at 15:56:42, Jonas Cohonas wrote: >On May 12, 2001 at 15:44:36, Aaron Tay wrote: > >>On May 12, 2001 at 15:26:13, Jonas Cohonas wrote: >> >>>On May 12, 2001 at 14:43:12, Chris Kantack wrote: >>> >>>>I guess I'm on the side that Kramnik (having access to Deep Fritz months ahead >>>>of time), will have little trouble winning the match. Indeed, he may be able >>>>to play the entire match with 99% of the moves already "practiced". >>> >>>I doubt that Kramnik has an 8 processor cpu at home so he cannot "prepare" >>>anything unless he own or have access to identical hardware. On an 8 processor >>>cpu it would be a whole different game as to just playing against a dual or >>>single cpu, the program running on 8 processors would look deeper and come up >>>with bettter moves! >> >>You make it seem >>like Kramnik is going to beat fritz at home, [after numerous takebacks] >>then memorise the moves and then repeat them ..LOL . > >Maby you should read my post again.. where in the above post do i claim or imply >that Kramnik is going to beat Fritz (at any hardware) with takebacks?? >If Kramnik finds a winning line against Fritz 7 on a single processor cpu, then >i am quite sure that it will not be a winning line on an 8 processor cpu because >of the simple fact that he would get into a different game on the 8 pr unit than >on the single cpu that was my point, and i think it is important to remember >that when talking about how "easy" it will be for Kramnik when he can prepare >month's in advance. Also Kramnik have no way of knowing if that copy he gets 3 >month's before is infact the same as the one he is going to face!! > >Regards >Jonas Jonas, I wish I could agree with you, but 8 processors vs 1 at standard time control will not allow the program to overcome any deficiencies Kramnik will find in it. Searching one more ply for example will not change a poor evaluation for a position. A program that doesnt understand how to play certain positions will not "learn" how to play them by looking one ply deeper. 8 processors will help Fritz tactically, but Kramnik will have already conceded tactics to the beast on his home PIII 800 or anything else for that matter. A nominal computer today is better than Kramnik at tactics. Speeding it up isnt going to help if Kramnik understands the positions to avoid. Once he gets the program home he will _LEARN_ how to draw and then beat it. For an analogy imagine a boxer and a wrestler are going to have a street fight. The boxer will do what he does best and so will the wrestler. I would put my money on the wrestler. The reason being is that in order for the boxer to win he must force the wrestler to box which is difficult. In order for the wrestler to win he must force the boxer to wrestle which is easier to do.here the boxer has to land a devastating blow to the boxer to win. If the boxer and the wrestle engage or go to the floor it is all over. Kramnik will do the same...play safe openings and proceed with caution not allowing the computer any short term attacks. THe position will look even or that Fritz is winning, but in reality as the endgame approaches Kramnik will have a plan to beat or draw the computer. Ed
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.