Author: Robert Hyatt
Date: 11:30:22 01/02/02
Go up one level in this thread
On January 02, 2002 at 11:43:55, Benny Antonsson wrote: >On January 02, 2002 at 11:28:12, Robert Hyatt wrote: > >>On January 02, 2002 at 10:12:29, Benny Antonsson wrote: >> >>>This is an easy draw, but how do one make your program see it ? >>>Do you make special cases in the static evaluator or do you rely on EGTBs ? >>> >>>[D] 8/8/p4k2/1bKR1r2/8/8/8/8 w - - 0 60 >>> >>>White king reaches the a1 square after Rxf5+ Kxf5 and black cannot win. >> >> >>You need special knowledge in the endgame evaluation code. It is not >>difficult at all: >> >> nss depth time score variation (1) >> 1 0.00 0.00 1. Rxf5+ Kxf5 2. Kd6 >> 1-> 0.00 0.00 1. Rxf5+ Kxf5 2. Kd6 >> 2 0.00 0.00 1. Rxf5+ Kxf5 2. Kd5 Bf1 >> 2-> 0.00 0.00 1. Rxf5+ Kxf5 2. Kd5 Bf1 >> 3 0.00 0.00 1. Rxf5+ Kxf5 2. Kd5 Bf1 3. Kd6 >> 3-> 0.00 0.00 1. Rxf5+ Kxf5 2. Kd5 Bf1 3. Kd6 >> 4 0.00 0.00 1. Rxf5+ Kxf5 2. Kd5 Bf1 3. Kc5 Ke4 >> 4-> 0.00 0.00 1. Rxf5+ Kxf5 2. Kd5 Bf1 3. Kc5 Ke4 >> >> >>That is not an EGTB score, it is the result of the evaluation knowing >>that bishop + wrong rook pawn can't win if the opposing king can reach >>the queening square first. > >So... I have to implement a kind of "racing"-function ? ~ Which king reacher >(controls) the a1-square first. ? Basically, but it can be a fast static computation rather than some sort of tree search.
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.