Author: Uri Blass
Date: 06:23:32 12/30/02
Go up one level in this thread
On December 30, 2002 at 08:27:40, John Lowe wrote: >On December 30, 2002 at 07:55:20, Sandro Necchi wrote: > >>On December 30, 2002 at 07:53:10, Sandro Necchi wrote: >> >>>On December 30, 2002 at 03:16:27, Uri Blass wrote: >>> >>>>On December 30, 2002 at 02:03:57, Sandro Necchi wrote: >>>> >>>>>On December 29, 2002 at 14:07:14, Lieven Clarisse wrote: >>>>> >>>>>>I disagree, testing without an opening book is a good test for chess engines! >>>>>> >>>>>>lieven. >>>>> >>>>>Hi, >>>>> >>>>>This is totally wrong. >>>>>Sorry, but the program has been developed considering the use of a massive book, >>>>>which is an important part of the program. So it is like to use the program >>>>>without legs... >>>>> >>>>>To me, the no use of the book or the use of a different book, it is like to test >>>>>a Ferrari F1 with a different engine or shape:-)) >>>>> >>>>>Ciao >>>>>Sandro >>>> >>>>Engines are used also for analysis and not only for engine-engine games from the >>>>opening book. >>> >>>OK, but why not use openings book until the end of the variations? >>> >>>Do you know that to develop the theory up to today level a huge amount of games >>>where needed? >>> >>>Chess is not perfect mathematics, so you cannot expect a program to find better >>>moves in the early stage of the game unless they can analyse until the endgame >>>and this would mean a huge more hardware power! >>> >>>> >>>>The question which engine is better from the opening position is not relevant >>>>for correspondence players who use chess engines to help them. >>> >>>If they rely on those moves instead of the theory ones they will not go too far! >>> >>>> >>>>The releavant question for them is which engine is better in the opening that >>>>they play. >>> >>>Why without the book. >>> >>>Why you do not ask GMs to play without theory? This is all nonsense! >>> >>>> >>>>Uri >>> >>>Sandro >> >>Uri, >> >>so how is Shredder 7? >> >>What about my statements? >> >>Any comment? >> >>Sandro > >Hi Sandro, > >To use a book or not..... > >Your comments about grandmasters don't hold water for me. > >The book is "crib-sheet" for the exam. The GM has the knowlege based on his >research of first class games. The GM understands the book! No I remember that I read in the past about a game when kasparov lost because he trusted his memory but unfortunately his memory betrayed him and he only remembered that a move is good but did not remember the reason that was written in his notebook. > >It's a matter of programming style - to have an exhaustive book or to have a >knowlege-based program. I might get more wins for my program if I use an >extensive crib-sheet but if I can get my program to "understand" the position on >the board - I've made a contribution to computer chess. Most programmers would >prefer the second (I hope?). > >Regards > >John Programmers can choose to have both or not to have both. I think that it is better if a program can find more correct moves in theory by calculation without the need to use books. Movei that is a free program is not a knowledge based program but most of my tests are tests without book when movei only changes the first move. I believe that it is better to use openings like 1.e4 a5 1.e4 a6...for tests and not the known theory. 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.