Author: Robert Hyatt
Date: 10:27:06 06/19/01
Go up one level in this thread
On June 19, 2001 at 08:54:12, Dieter Buerssner wrote: >On June 19, 2001 at 00:38:34, Robert Hyatt wrote: > >>On June 18, 2001 at 15:46:13, Bas Hamstra wrote: > >>>Not at all. Have you never tried Crafty without nullmove? See below my engine >>>output, rootposition. Where is the 6? I have never had 6, in my opinion 4,5 on >>>average is normal. >>> >> >>The "norm" is sqrt(W) where W is the number of moves at a position in the >>tree. In the middlegame, this is around 40 to 50. In the endgame, much >>less. The simple formula for total nodes searched will convince you that >>one more ply takes sqrt(W) more nodes. > >I think, this is only true without hashtables. Not only pruning techiques, also >hashtables should decrease the branching factor. I cannot remember any quatative >data of this effect, but it may be enough to explain the number Bas has given. > >Also, without hash-tables, without pruning but with search extensions, I think >even a higher branching factor has to be expected, because the "average search >depth" (however you define this ...) will be larger than the shown search depth. >Adding some plies depth will add these plies at least, but also some more nodes >due to the more extensions. > >Regards, >Dieter In the middlegame, I don't notice a lot of improvement with hashing over non-hashing, other than I depend on that for move ordering, rather than for scoring/cutoffs. In endgames it becomes huge of course. However, I could (at some point) run a few tests with no null-move, to see what my branching factor comes out at. If it is below 4 I will personally be greatly surprised. Just as I was surprised when I calculated this for Deep Blue from the 1997 log files...
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.