Author: Uri Blass
Date: 06:26:01 12/30/02
Go up one level in this thread
On December 30, 2002 at 09:23:32, Uri Blass wrote: >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 I mean better for me. The point is that I am not interested in tuning movei for some specific line but to improve it and by changing the first move every game I may get cariety of positions. 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.