Author: Sune Fischer
Date: 14:56:49 08/21/03
Go up one level in this thread
On August 21, 2003 at 17:40:11, Uri Blass wrote: >On August 21, 2003 at 17:27:05, Sune Fischer wrote: > >>On August 21, 2003 at 17:21:06, Uri Blass wrote: >> >>>>If you had a pruning rule that gained 50 elo in all games but made blunder moves >>>>every time there was a quadrople pawn on the board, would you do the pruning? >>> >>>I believe that if I have that pruning rule then I can easily change it not to >>>make the blunder without significant difference in strength. >> >>But for the sake of argument, suppose you couldn't. >>Which is more important to you? > >increasing strength. So I guess we fundamentally agree it is not an unbreakable principle if the reward is large enough. >Junior can also do verifying search with different bounds(-infinite,infinite) >when it get position when one side has only a bishop and in the rare case of >getting mate value it can change the assumption that bishop cannot mate for the >rest of the search. This is going to get complicated if the position is not yet at the root, but a few plies deep. >>In which case the mate is going to need 40 plies, so you'll miss it anyway. > >mate in 3 is 5 plies(5*4=20) and it is possible to get 20 plies in the endgame. >it can be even less than 5 plies if you do checks in the qsearch and in this >case it can be translated to 16 plies. I bet people will still be complaining that your program is unable to solve simple mates in 6 :) -S. >Uri
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.