Author: Robert Hyatt
Date: 07:51:22 12/19/03
Go up one level in this thread
On December 19, 2003 at 05:26:59, Peter Berger wrote: >On December 19, 2003 at 00:16:03, Robert Hyatt wrote: > >>On December 18, 2003 at 18:32:21, Peter Berger wrote: >> >>>To call Crafty's >>>book wide is really wrong - I will try to reach you the last time with this >>>message, aware that I am _most_ probably not the first one to come up with this >>>anyway. To answer your second question: given 1. e4 e5, Crafy being white: what >>>is the probability that it will play 2. Bc4 against a computer? >> >> >>10% of the time: >> >> move played % score learn CAP sortv P% P >> Nf3! 152593 81 0.11 0.27 -655.36 14041.0 90 Y >> Bc4! 6913 3 0.07 0.00 -655.36 12046.3 10 Y >> >>that 3% you are quoting is the "percent of the time this was played >>in all the games in enormous.pgn, after 1. e4 e5. Bc4 was played 3% >>of the time. Crafty plays it about once out of every 10 games. >> >>However, once you are past that point, the book is _wide open_. So >>I don't understand your usage of the term "narrow". IE from here on >>in the game the program can choose from _any_ variation in the Bishop's >>opening that it wants... >> > >You are correct with the 10% number for 2. Bc4, I get the same number here. I >looked at the wrong column of the table. > >Why did I call Crafty's book narrow? One example: if 1. e4 e5 2. Nf3 Nc6 is >played Crafty as black will _always_ play only the following line against a >computer: 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 >O-O 8. c3 d6 9. h3 Bb7 - there are several similar lines , but I am sure you are >aware of this anyway. No I'm not aware of it, as if it were to lose that line, it should certainly change to something else because of the learning code... Also note that Crafty modifies the book width/etc stuff internally when playing on ICC, based on the opponent, the rating, the title, etc... > >Or are you? I just had a look at Crafty's most recent games on ICC: > >CSMath(C)-Crafty 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 d6 >glories(C)-Crafty 1. e4 e5 2. Nf3 Nc6 3. Bb5 a6 4. Ba4 d6 >StrongBad(C)-Crafty 1. d4 Nf6 2. c4 g6 > >These lines would not be possible with the bookc.bin provided for download. I don't know what you mean. The three files on the ftp machine are _exactly_ what I use here to play on ICC, no changes of any kind, except that the book.bin file on the ftp machine has no learning data, I cleared it as I explained how you could do the same in an earlier post. I clear my book learning fairly often in fact, otherwise the book gets _way_ narrow. >Yesterday I had a short conversation with another WB engine author who said that >testing against Crafty is mostly useless as it's only a booklearner test - so I >am not alone with this observation. I don't know what that means either. "A book learner test" _still_ requires a chess engine to play once the book runs out. > >>I don't believe we ever discussed how it should be set up. > >I won't start quoting ;). But it is true that it is a user experiment done by me >just as I see fit, although I set up Crafty as closely as possible to your >recommendations for an experiment like this. > >Peter
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.