Author: Alvaro Rodriguez
Date: 06:56:20 08/14/00
Go up one level in this thread
On August 14, 2000 at 08:52:18, Robert Hyatt wrote: >On August 14, 2000 at 05:45:28, Alvaro Rodriguez wrote: > >>On August 13, 2000 at 20:46:52, Robert Hyatt wrote: >> >>>On August 13, 2000 at 18:27:00, Alvaro Rodriguez wrote: >>> >>>>On August 13, 2000 at 18:16:30, Robert Hyatt wrote: >>>> >>>>>On August 13, 2000 at 15:21:17, Alvaro Rodriguez wrote: >>>>> >>>>>>On August 13, 2000 at 15:09:02, Robert Hyatt wrote: >>>>>> >>>>>>>On August 13, 2000 at 13:01:27, Jon Dart wrote: >>>>>>> >>>>>>>>>This is an interesting move IMHO as the bishop doesn't get lost at all , it >>>>>>>>>simply can't participate in the game anymore ; in fact it lasts until the 46th >>>>>>>>>move when white finally wins the bishop . It seems to me without special code >it is impossible to avoid this ( for example Fritz 6a would play like that >>>>>>>>> too) . >>>>>>>> >>>>>>>>Crafty has had special-case code to avoid this kind of move for some >>>>>>>>time. >>>>>>>> >>>>>>>>--Jon >>>>>>> >>>>>>> >>>>>>>Even more interesting, because it understands that problem, it will leave >>>>>>>the pawn 'hanging' because it thinks no one is silly enough to take it. If >>>>>>>a program doesn't have that bit of knowledge, it almost looks like Crafty >>>>>>>tries to 'tempt it'. :) >>>>>>> >>>>>>>I see that against more than one program on ICC, pretty regularly... >>>>>> >>>>>>Is it hard to implement that in a program? Because maybe other programmers >>>>>>should think about it, because like you said, no one is silly enough to take it >>>>>>:-) >>>>>> >>>>>>When you see those games in ICC, does crafty win them ? >>>>>> >>>>>>Regards, >>>>>>Alvaro >>>>> >>>>>Most of the time it wins. On occasion there is some long and complicated way to >>>>>safety get the bishop out. Which means Crafty has simply gambited a pawn away >>>>>for nothing. But these cases are very rare. A bishop at a2, with a white pawn >>>>>at b3 that is supported, is in great trouble. >>>> >>>>Big trouble for sure.. Is it hard to implement that in a program? If it is easy, >>>>I don´t understand why programmers don´t use that in their programs. Some does, >>>>but others don´t.. Maybe they think that it´s a good move. I don´t know.. >>>> >>>>Regards, >>>>Alvaro >>> >>> >>>I think that it is easy to lull yourself to sleep with the argument "this is so >>>rare it isn't worth handling.." There are many such cases. If you do that for >>>all of them, suddenly it is not rare, it is commonplace to see one of those >>>holes show up in every game. I think that plugging such holes, when they show >>>up, is a critical thing to fix. If you keep doing this, sooner or later you >>>don't have any holes left to plug... >> >>Robert, do you have some games from ICC where Crafty´s opponent played Bxa2 and >>Crafty responded with b3 ? Maybe it´s hard to get them.. >> >>Regards, >>Alvaro > >This is probably an answer you won't like, but the file "crafty.pgn" on my ftp >site has tens of thousands of ICC games. You will find plenty of games in there >with Bxa2, Bxa7, Bxh2 and Bxh7 moves... It even has full PGN from the >servers so you can see who played the move... > >Bob Hmmm, i guess that means it`s pretty difficult to look for those games..:-) Regards, Alvaro
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.