Author: José Carlos
Date: 11:08:16 03/23/00
Go up one level in this thread
On March 23, 2000 at 14:02:44, Tom Kerrigan wrote: >On March 23, 2000 at 13:57:51, John Coffey wrote: > >>The best example of this is a mate in 5 vs. a mate in 3. But more difficult is >>a slight improvement to the position in 5 moves vs. a slight improvement to the >>position in 3 moves. It would seem better to choose the shorter solution. >> >>John Coffey > >The problem with this rationale is that a move might look great at move 3 and >then lose the game 1 move later. > >So you really want to pick the move that looks great after 5 moves, instead of >the one that looks great after 3 moves. > >If the computer does not pick your "good enough" move after searching 5 moves >ahead, then it's possible that your move wasn't so great to begin with. > >-Tom The problem, if I understood right, is when you get a score of 5 plies depth (due to extensions, for example) in a 3 ply search, and then you search another move that returns the same eval in just 3 plies. You'll keep the first PV move cause you didn't improve but, would it be better to take the shorter way to that "advantage". If we speak of a mate in x, the answer is yes, but I'm not sure in other cases (except if you are going to be mated, in wich case you'd prefer the longest variation). José C.
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.