Author: José de Jesús García Ruvalcaba
Date: 11:07:33 02/07/99
Go up one level in this thread
On February 07, 1999 at 10:09:19, Micheal Cummings wrote: >Having started and experiment between CM6K and Rebel 10 which I have included >all of the CM6K opening to my knowledge into the rebel 10 opening book, I wonder >how Rebel or any other chess engine would perform using opening moves that are >not popular. > >I know that CM6K uses some weird and wonderful ones, which is why I am playing >12 games with it ans Rebel 10. >Since I have heard that Rebel 10 works well with its original opening book, I am >seeing how well it works without it. I would think that a truely strong chess >engine would be able to play more unusual openings to a capable level. > >I am wondering if anyone else has tried this. And I am more particular with the >less used openings, or even more popular ones which are used but a program does >not have in its opening book. >Since part of the chessmaster opening books for different personalities include >these kinds of openings which many other programs do not. I find it most >enjoyable to watch and see how thing develope. > >As in my first game I posted, in which Rebel 10 used a move which it does not >have in its opening book, which someone picked up on :) And thus went on to lose >the game. This by far does not show me anything, but I hope in coming games it >will. > >And as also pointed out by a fellow memberr when I played this game, gives me >the impression that Rebel 10 without its own original opening books leave its >engine lacking. As he stated this was not fair on Rebel 10 when I explained this >is what I was doing. So it has made me wonder more. > >I would like to hear from anyone knowing mroe on this matter. > >regards > >Micheal My personal opinion is that the opening book is an integral part of the program. If you want to test a program's *strength* you should use its own opening book. There is *nothing* wrong about testing an engine with a strange book (or with no book at all). These tests are interesting and have their value, but they do not reflect the program's overall strength. I do not care abaut "fairness". I do not think a match between computers can be unfair, as long as we do not forget the conditions and settings for both opponents. When you test in an unusual way and publish the results, be careful to state it. Omissions can be easily misunderstood. If nothing is said, default settings are presumed. Features of a good opening book: 1. It must be deep and wide, to avoid bad surprises. 2. It must avoid poor moves. 3. It should fit the style of the engine it is intended for. I guess CM6000's book does not fit the style of Rebel 10, that is not its purpose. José.
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.