Author: Sarah Bird
Date: 01:10:46 07/03/99
Go up one level in this thread
On July 03, 1999 at 00:28:59, Melvin S. Schwartz wrote: > >On July 02, 1999 at 23:42:15, James T. Walker wrote: > >>On July 02, 1999 at 22:55:21, Steve Lopez wrote: >> >>>On July 02, 1999 at 20:45:09, Melvin S. Schwartz wrote: >>> >>>> >>>>Have you got it now? :-)))) >>> >>>That's the beauty of the Internet, man -- it allows one to make an ass of >>>oneself worldwide in a matter of seconds. :-D >>> >>>-- Steve Lopez >> >>Hello Steve, >>I also have Hiarcs 7.32. I'm very happy with the program including the opening >>book. However, if you play the moves Mel suggest, here is the problem: It's >>true that exf6 e.p. is in the book. So next you play Nd6+ and Hiarcs starts >>thinking since it's out of book and plays Bxd6. After playing Bxd6 though, you >>can now click the red backup arrow below the board and guess what shows up in >>the book moves window, "9...Bxd6 no" in red! So it appears it's in the book as >>a move not to make and following this guidline it does not make this "book move" >>but when thinking on it's own it selects this move. > >Hello Jim > >Glad to see someone finally found out I was correct about that. I guess Mr. >Lopez didn't make the effort to verify what I was talking about since he was so >anxious to say I was wrong. We argued about that so much and I got so frustrated >I had to visit my psychiatrist three times to get my head screwed back on. :) Oh >well, that's life. By the way, that Bxd6 thing seems like a bug in the program. >First of all, if Hiarcs didn't play Bxd6, it wouldn't be able to castle and its >position would not be very good. The thing is if instead of playing f5 it should >have played Qc7 which is in the tree as the favorite move with I believe 80% and >f5 something like 18%. So it was a poor choice to play such a move which by the >way is in the book. Qc7 instead of f5 would without question be the proper move. >Oh, and by the time I get around to eliminating all the bad lines, I think there >will be Hiarcs 8.32 and Fritz 6.32 to consider. :-) Since this all started about Hiarcs opening book being not to "GM Standard". It still comes down to a final position after the moves: 1. e4......c5 2. Nf3.....e6 3. d4......cxd 4. Nxd.....Nf6 5. Nc3.....Nc6 6. Nxc6....bxc 7. e5......Nd5 8. Ne4.....f5 9. Nd6+....Bxd6 10.exd.....0-0 11.Be2.....Qb6 12.c4......Nf6 13.0-0 At this point my hiarcs evaluate white as up 0.27 after 5 minutes. Fritz 5.32 sees it as 0.41 after the same time frame. I just don't see this as a faulty book line. As for the pawn on d6 which you like so much to me it looks like more of a weakness than a strength. I also don't understand how, if you are playing this game with Hiarcs you get a timeout for analysis of it's book moves. Since if you were playing a correct game against the computer, 1) you would not see any of it's book moves, 2) you would not see any of it's thinking and\or + - =. I say this since you state it as some miracle that after 13 moves you can possibly be up 0.4 of a pawn. I put it to you that this is far from unusual to get out the opening up as white. To be up by more than 0.5 over Fritz now there's another story. Try this one: 1. e4....g6 2. d4....Bg7 3. Nc3...c5 4. dxc...Qa5 note: Hiarcs book has dxc no but not Fritz 5. Bd3... note: Hiarcs book only lists Bd2, Bd3 is from Fritz However at this point Hiarcs jumps back into book and lists 5........Qxc5 with a variation to move 10 at which point it sees it as relativly equal. Fritz lists this same 5.......Qxc5 with the same variation to move 10 at which point it sees the same relatively equal position. Yet: 5........Bxc3+ 6. bxc...Qxc3+ 7. Bd2...Qxc5 now a pawn up !! After 3 minutes thought Fritz sees itself as losing by 0.62 choosing Nf3 Hiarcs sees it as equal and chooses Be3. The whole point of this little story being that: 1. The evauation of a position is much more than just point score. 2. Computers all evaluate positions differently. 3. Computers score positions differently. The final point being this is the earliest I can get over 0.5 with black against Fritz lol. Thanks Sarah. > >Regards, >Mel >>By the way, my Hiarcs did play some 1 b4, 1b3, 1 g3 moves when new, but after >>losing most of them it hasn't played any of those opening move in the last 150 >>games. So my advice to Mel is to beat it when it plays bad lines and it will >>learn it's lesson and stop playing them. :-) >>Jim Walker
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.