Author: Sandro Necchi
Date: 00:02:04 09/15/02
Go up one level in this thread
On September 14, 2002 at 02:05:20, Sandro Necchi wrote: >On September 13, 2002 at 16:44:35, robert flesher wrote: > >>Hmmmmm.. well i have all the M-chess programs from versions 5-8. M-chess with >>its fighting spirit and romantic style makes it very fun to play. However!, >>against todays programs it can win a few games, BUT! mostly it gets outsearched >>by its opponents. My thoughts are that M-chess gained many of its victories from >>its excellent, one may even say COOKED! openings book. > >To judge my opening book as COOKED lines is quite superficial! >I already explain what the aim was and the kind of work I have done. If people >don't want to understand it is their problem, not mine. For those who don't know I explain what I did: I improved the book in every version at 360°. What does this means? I improved the book on everything I knew that should be improved adding also some long lines. This means improved variations which the program should play improving variations or novelties from Chess informant, from local chess club and or from computer games. I did not made specific improvements against specific computer programs to score better in the SSDF list even I admit sometimes those new lines/improved lines where effecting the score as played by the computer opponents. If this was effecting the list a lot (according to someone) it was because they had not learning features, which was not my fault. Today programs would probably loose only one time. Also I was including new ideas and testing possible reply also using some available programs, but not addressing those improvements on specific computer opponents. The reason why I was adding long lines was not to increase the strenght of the program by letting the book play most of the time (according to someone), but to let the owner get a move that the program would play in some minutes (at normal tournament time level) immediately saving time to the computer for the rest of the game a psicologically put in trouble human players by forcing them to speed up their moves and unabling them to use the computer time. So this was address to human players and not computers. Also do not forget the unique learning feature of M-Chess which was adding moves to the book up to 49 each side, so won games would have been added by the program if played often and with good score. This means that I have anticipated the computer, but this would have been done anyway! I mean that a line that a human would kill (up to middle-high level) would not have inserted even if scoring 100% or less against a computer or more programs. So definetively the COOKED sentence could come out ONLY by someone looking computer games only, while normal users that would play themself against the program would find the new book improved over the previous one. This was my real target and this is why I get upset when someone try to change things because of SUPERFICIAL judgement or because someone told this!! So this is the right story. Sandro > > >Style will not change the >>fact that it is tactically much weaker that say TIGER, FRITZ, HIARCS, >>CHESSMASTER etc.... But these are just my thoughts. Cheers~ > >I did make a better version changing the parameters. I did not gave this info to >anyone to respect Marty's work. >The problem is that in order to make improvements against today's programs there >is a need of improving the depth expecially in the endgame. This means that even >if version 8 can be improved changing the parameters, to make a real jump it is >needed tha Marty takes over and makes what needed, which I doubt for the time >being. > >Sandro
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.