I don't. I apply extensions or reductions based solely on the static criteria I use, applied only at the node in question. There is a case to be made for other ways to do this, but to date I have not found something that is better. If you extend on getting out of check, you will reduce _some_ moves at the next ply. Wise or not? Hard to say. But every limit on this I have tried hurts overall play level in real testing. So far, anyway.metax wrote:I still have a dumb question...
If you extend a subtree because of check or anything else and you see that a reduction has been done before... do you extend two plies then? Probably not, but I just wanted to be sure...
Extensions/Reductions
Moderators: hgm, Rebel, chrisw
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: Extensions/Reductions
-
- Posts: 344
- Joined: Wed Sep 23, 2009 5:56 pm
- Location: Germany
Re: Extensions/Reductions
Doesn't this somehow neutralize the extensions and thus lead to decreased playing strength?
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: Extensions/Reductions
When you think carefully, even when you extend a specific move because (say) it checks the opponent, not _all_ moves below that node in the tree are relevant/important. It actually makes sense to still reduce those that are worthless, which lets you search the others even more deeply to make sure you are exposing all the tactical issues that are involved with the move.metax wrote:Doesn't this somehow neutralize the extensions and thus lead to decreased playing strength?