Author: Robert Hyatt
Date: 05:00:13 10/13/98
Go up one level in this thread
On October 13, 1998 at 02:36:31, blass uri wrote: >I think that if the pawns are blocked it may be easy to built the tablebases >because the number of different position you need to evaluate for every case is >less than 300,000. >It is even more easy than KQvKR positions > >After you build the tablebases for blocked pawns you can build tablebases for >all the cases of almost blocked pawns(like blacka6, white a4) >and you can do a tablebases for all the cases when the pawns are in the same >file. > >These positions are in many cases not a win for the strong side. >There are cases when it is a draw If the weak side can take the pawn by the king >or a fortress posdition against a bishop >or even a win for the weak side if the knight cannot protect the pawn and stop >the another pawn from being a queen. > >Uri the only thing wrong with this is that these databases are very time- consuming to build. When you do a 5 piece file, you take the 5 pieces you are interested in, and put them on *every* possible square they can reach. And that produces some deep mates that make construction time-consuming...
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.