Author: Arturo Ochoa
Date: 10:58:21 02/16/05
Go up one level in this thread
On February 14, 2005 at 19:56:56, Uri Blass wrote: >On February 14, 2005 at 19:23:07, Thomas Mayer wrote: > >>Hi Arturo, >> >>> I did not tune the book against Movei.... I tune a book for a specific >>> engine. Anthony accepted my help and I did my best in 2 short months. Of >>> course, how could you understand that? >> >>we had that discussion already in several forums. Uri has his point of view >>(which is IMO totally wrong) and will not change that... senseless to discuss... >> >>You might remember CCT4 -> Since then I had never again such a fantastic tuned >>book for Quarks needs - it was the work of Leo Dijksman which he had done in >>several month with thousands of games and analysis. Of course it was Quark which >>was playing that tourney but it had a fantastic book that lead it to positions >>it can play. And that is what a good book cooker should do, he should help the >>engine to end up in positions after the opening where it knows what it should >>do. The more time he can spent on that the better it will do. >> >>Vincencts 700 Elo might be a bit to high, but 300 is for sure a good number - >>the difference between a random book and a well tuned book. Especially the >>weaker or more unbalanced your engine is, the more it is important to have a >>perfect book. > >It suggests that the engine is important because if the engine is not balanced >you may solve the problem by doing it balance. Every component of software is important including the book. > > Maybe for a weak engine even Vincents 700 is possible... > >I think that if the engine is too weak no book will help it unless you predict >perfectly the opoenent moves. > >I do not see how you can do tscp 700 elo better by a book and tscp is not the >weakest engine in the world. > >I guess that a book for weaker engines will be even less productive. How do you know that? What are you proofs? You only guess and how can you support this? > >If someone want to prove me wrong he is invited to try to improve tscp rating of >1699 in the WBEC site to 2400(slightly better than Amateur,Betsy,Leila) > >The rating should hold also against new opponents(that were still not released) >because when I talk about book the idea is not to have killer book that work >only against specific opponents with known books that is of course possible >after lot of work when you put win after win in the book. No, you are not so important so somebody wastes his valuable time in your absurd contrdictions. :) It is enough to point your nonsense here. > >Uri What you dont tunderstand is the Vincent's sense. Because your lack of credibility, he just put a extreme point. You think that a no-book engine can wint a Title.Prove that in Tournaments. You still think a tiny book is enough. How you can compete against the top engines. You will provide easy points of course. That is what I mean. Ok, my question is: how can I explain you in easy words if you cannot understand anything. It is just an "impossible mission" :)))
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.