Author: David Rasmussen
Date: 05:52:04 01/10/02
Go up one level in this thread
On January 09, 2002 at 23:23:51, Miguel A. Ballicora wrote: >On January 09, 2002 at 23:01:53, David Rasmussen wrote: > >>On January 09, 2002 at 22:52:10, Pham Minh Tri wrote: >> >>> >>>I agree that is a good idea. But how do you implement it in simple way? I am >>>afraid that that code is more expensive than Crafty code ;) >>> >> >>It depends on your data structures and your representation. I use bitboards, and > >I was going to answer and you just stole these _exact_ words by only few >minutes!! > Hey, that chip I implanted in your brain is paying off! >>I just added a recapture extension like this. It's pretty simple for me, as I >>just use my AttacksTo() function to see if there are other attackers to this >>square than the one that recaptured. AttacksTo() is just a check for each type >>of piece if such a piece attacks the given square. > >I guess this is similar to what I do. After there is a recapture, >I do makemove(), I go to search() and there I check if the side that just >moved attacks the square. >David, let me know what happen with this modification if improves anything >and it works for you. >As always what it works for somebody might not be good for another one. > It seemed to do some good. I think it solved a couple more wacs, but more importantly, it made the tree smaller. Anyway, intuitively, it seems much more to the point to do this when the recapture is most likely the only good move there is, which is not at all certain if there are more than one recapture. Thanks for this very simple idea :) /David
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.