Author: Arturo Ochoa
Date: 14:37:56 08/26/02
Go up one level in this thread
On August 26, 2002 at 14:44:47, Uri Blass wrote: >On August 26, 2002 at 13:53:14, Dann Corbit wrote: > >>http://f11.parsimony.net/forum16635/messages/33526.htm > >It is better to play with the program against other opponents and not against >itself to get a good estimate and not to play more than 2 games in a >match(otherwise the problem can be aggresive learning and not lack of book). Hello: No, it is not better and it doesn't have any sense. As you declared in the long thread below, you believe (but it is not demonstrated) that a engine with book is not better than a engine without it. You also believe (but is it not demonstrated too) that a small book is enough for a chess engine. Small book until move ten is not enough to beat anything of high quality done by Kure or Noomen. His lines are very deep and you will see that the amateur engine doesnt have any chance against these "things". The Dann Corbits result is interesting and valuable because he proves that an Engine with Book can improve its level being a random book. More amazing, I am running a Match that I have called "Kramer Vs Kramer" where I will show how your suppositions are not real in any sense and how an engine is helped a lot with a Book. The most impressive of this Match that I am running is the Score of the Engine with Book against the same Engine without Book. Of course, a Match of 20 Games that cannot prove it at all, but the real positive thing is this engine is one of the strongest available free Winboard engine in the map and the book used is a combination of a public available plus other lines that I have added and the Engine shows another level in the Opening of course in the score. It is nonsense to say that chess programs can solve every position. If this were true, programs could beat humans in strategy. After a many tests that I have done privately, the current Engines cannot solve any strategical problem successfully. In this sense, all the engines including the Top are like childs in these kind of positions. The other supposition about "correct moves" without book is not true at all. There are many positions where the Programs doesnt know how to react. The big value (OK, I will put BIG AS BIG) of the work of Jeroen Noomen or Alexander Kure is they finds positions after the opening where the programs can react pretty acceptable. Sometimes, they fail and the top programs can have a lost game, but this is the biggest value of these Book Creators: to create opening books that are very solid (and not only killer). Maybe, you don't give a coin for this, but this is a fact: The Success of Fritz + the Kure Book or Tiger + the Jeroen Noomen Book. More, the Match run by Corbit showed another difference. I will wait until my Match is finished and all these unproved supositions are a complete nonsense. > >It may interesting to compare your book with changing the first move every game >so you can play a match of 40 games without losing the same game twice(the only >problem can be different order of moves). > >Uri No, it is not intersting because an engine without book will repeat the same moves if it doesnt have any kind of learning. The same apply for the engine plus a book. If you want to prove your supposition, please, post here concrete cases about what you are supposing, instead of, saying what it is interesting for you and not valuable for anybody. I liked a lot to propose this kind of ideas until I joined the first Computer Tournament with Diep and I had to come back to the reality. It is not lucky that Diep has improved a lot in openings during the last two Tournaments. The Official Computer Tournaments (I don't me the Internet Tournaments that are run through the Internet) gives another sense what you must do in this Matter. The process of learning about what it is convenient or not for a chess program is a matter of my private preparation that I am not going to reveal here. Of course, as an Active and Rated Player, I know by experience what it must be done. Interesting to test an engine with book and the same engine without book against another program? No, it is not interesting and it doesnt prove anything. Crafty Vs. Crafty + Book is interesting because it showed how the same engine can improve with the thing called Book that you seems to miss at all. The real thing that I called this kind of Matches: Kramer Vs Kramer in other to prove with facts (time plus dedication to run it) about the valuable thing called book. Regards, Arturo Ochoa M.
This page took 0.01 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.