Author: Uri Blass
Date: 10:01:03 11/27/02
Go up one level in this thread
On November 27, 2002 at 11:51:02, Richard Pijl wrote: >On November 27, 2002 at 11:12:28, David Rasmussen wrote: > >>On November 27, 2002 at 10:49:00, Richard Pijl wrote: >> >>>On November 27, 2002 at 10:01:02, David Rasmussen wrote: >>> >>>>Some time ago, along the lines of these thoughts, I proposed "negative >>>>extensions". That is, if you can somehow classify a move as "probably not >>>>interesting", you can "extend" the depth by -1 or -0.75 or whatever seems >>>>reasonable. Exactly as you do with normal extensions. The good thing about this >>>>is that nothing gets pruned for good, everything will eventually get searched >>>>with iterative deepening, but you search what you think is interesting first. >>> >>>Isn't this just razoring? >>> >>>IIRC Beowulf is using this instead of futility pruning ... >>> >> >>Well, I would say that razoring is a special case of negative extensions. >>Razoring normally only happens near the frontier nodes. What I am talking about >>can be applied anywhere in the tree. And also, razoring specifically deals with >>the futility of getting the score up near alpha again. I am talking about all >>sorts of "futile" moves. I.e. a move that moves a rook off the only open file, >>is probably not good. That doesn't mean that we don't want to search this move >>to some specified depth, in time. But it means that we might put off searching >>it till later, by "extending" it negatively. >> > >Won't analyzing this kill your search speed? I mean, you will have to apply a >lot of patterns to really make a difference in decreasing the size of the tree, >but that will cost performance. And bad moves should have more beta cutoffs in >the subtree anyway, so the subtree is probably already quite small. I can only say that I disagree with you. I do not think that you need a lot of pruning rules to reduce the size of the tree significantly You need only good pruning rules. Uri
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.