Author: John Coffey
Date: 11:58:33 09/21/98
Go up one level in this thread
On September 21, 1998 at 14:22:05, Robert Hyatt wrote: > >No... you are ignoring the exponent, completely. When you do a null-move >search, you search two plies less deeply. If *this* causes a cutoff, >you just reduced the total nodes by a factor of *25*. If this happens >often enough, reducing the branching factor from 5 to 2.5 is not just easy, >it becomes trivial... > I will take your word for it, but part of this doesn't make sense to me: I am not ignoring the exponent at all, as I understand that reducing the number of nodes looked at gives exponential returns. But to go from a branching factor of 5 to 2.5 still requires reducing the search time for half of the branches looked at. Granted you may look at one branch 25 times less, but there are other branches that will be looked at more fully. If on the otherhand I search the first move that wins a piece, and all but one of my opponents responses regains material, then I could do a null move after all but one of my opponent's responses, thus saving close to 80 or 90%. Maybe this is the piece of the puzzle that I am missing? Thanks for the response. Best wishes, John Coffey
This page took 0.02 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.