Author: Sune Fischer
Date: 12:59:12 07/11/02
Go up one level in this thread
On July 11, 2002 at 15:36:52, Alejandro Dubrovsky wrote: >On July 11, 2002 at 15:10:10, Sune Fischer wrote: > >>I fear I may have castle and promotion bugs. >> >>My new version produces correct perft numbers on the initial position up to ply >>6, but I would like to test for castle, ep and promotions also, so it would be a >>big help if I could get confirmation on some numbers: > >i'm testing as well, so my number are neither confirmations nor corrections. >great help :) It is :) >>// castle test: >>r3k2r/3q4/2n1b3/7n/1bB5/2N2N2/1B2Q3/R3K2R w KQkq - 0 1 >>perft 4 = 5660814 ? >>perft 5 = 269371355 ? >i get 5257077 and 246266720 Hm, my old version produces: perft 4 = 5664262 (don't know about 5 plies - too long a wait:). >> >>// ep test: >>rnbq1bnr/1pppkp1p/4p3/2P1P3/p5p1/8/PP1PKPPP/RNBQ1BNR w - - 0 1 >>perft 4 = 295376 ? >>perft 5 = 8296614 ? > >same Great, I was also pretty sure it was correct, it would have been unlikely to have the same bug in two different move generators. >>// promotion test: >>rn1q1bnr/1bP1kp1P/1p2p3/p7/8/8/PP1pKPpP/RNBQ1BNR w - - 0 1 >>perft 4 = 1874042 ? >>perft 5 = 58274872 ? > >what's the correct way to count in promotions? is it meant to be a separate >move for promote to queen, rook, bishop and knight, or is that just counted as >one. counting each one separately, i get 2010006 and 67867493 Yeah, that's the way I count them. My previous version agrees with your numbers, so I guess the bug is in my new version. Oh well, 1 of 3 is a good start :) -S.
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.