Author: Uri Blass
Date: 06:00:53 07/07/03
Go up one level in this thread
On July 07, 2003 at 05:30:38, Reinhard Scharnagl wrote: >Hi Uri, > >a good solution depends on whether: > >a) each bit of the eight has the same probability to be set or > they are decreasing or they are increasing, >b) the average number of the bits set, > >so please explain on that. > >Reinhard I calculate exact ditribution of the bits in the first 10,000,000 nodes from the opening position and here is the result: I have 20,357,657 cases of calling the switch command. The average number of bits set is 1.27405 in 2895517 cases bit 0 is set in 2910299 cases bit 1 is set in 2123957 cases bit 2 is set in 2733137 cases bit 3 is set in 5285584 cases bit 4 is set in 4632172 cases bit 5 is set in 4706124 cases bit 6 is set in 5121911 cases bit 7 is set When I think about it I am not surprised from the fact that bits 4,5,6,7 are set more often because they are about bishop direction and pawns only control bishop directions(it may suggest that I may start with them in the switch) I guess that I am going to go with Tim's suggestion and only start from the opposite direction because I do not think that trying to do more optimization is very important considering the fact that I call this switch only 2 times per node and movei is not a fast searcher. 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.