Author: Magoo
Date: 16:25:16 04/29/03
Go up one level in this thread
Ok, i forgot to mention thats including the Quiscent "nodes"... i guess it was wrong of my, i should no include them into the node counter... The real nodes searched (in the alphabeta) are fewer... (by a big amount of course) On April 29, 2003 at 18:23:28, Uri Blass wrote: >On April 29, 2003 at 18:11:00, Magoo wrote: > >>(sorry about this repost, i hit space by misstake). >>This is a position where my engine plays black: >> >>bRbNbB..bK....bR >>bPbPbPbP..bPbPbP >>..........bN.... >>........bQ...... >>wP..wP.......... >>........wPwN.... >>..wP..wQ..wPwPwP >>wRwN....wKwB..wR >> >>nodes: 690975 <= LOL!! >> >>Notice that the White Knight is attacking the Black Queen. >> >>Move List: (in order listed, last iteration) >>Capture: Score: 1000003 move e5b2 <= LOSING >>Capture: Score: 1000003 move e5h2 <= LOSING >>Capture: Score: 1000003 move e5e3 <= LOSING >>Score: 14 move e5e6 <= OK >>Score: 14 move e8f8 >>Score: 13 move e5e4 >>Score: 1 move e8d8 >>Score: 0 move e5d6 >>. >>. >>. >> >>and after this, i get bestmove = e5c5, thats move # 18 of 43 possible moves. >>Maybe i should implement Killer Moves? Or maybe not always put the captures >>first, losing captures are not good moves.. so... > >If you get 500,000 nodes for 4 plies then >you first need to correct your bugs. > >Even tscp1.73 that use only mvv/lva and history get better result than you get. >Tscp1.73 does not use hash or killer moves for order of moves. > >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.