Author: Gordon Rattray
Date: 17:08:31 08/18/00
Does anyone else see the following with Junior 6a... [D] 8/5ppp/8/1pK5/kP6/8/5PPP/8 w Now, while using "infinite analysis", play through the following moves. Note: I've also tested this on a slower PC and noticed that in order to reproduce the problem I had to play through them slower (maybe 10 to 20 seconds per move roughly on PII 266Mhz). 1.g3 g6 2.f4 h5 3.h3 f6 4.g4 h4 5.g5 f5 I now see the following analysis... Analysis by Junior 6.0: 6.Kd4 Kxb4 -+ (-1.96) depth: 3 00:00:00 6.Kb6 Kxb4 -+ (-1.88) depth: 3 00:00:00 6.Kc6 Kxb4 -+ (-1.81) depth: 3 00:00:00 6.Kc6! -+ (-1.51) depth: 6 00:00:00 6.Kc6! Kxb4 7.gxf6 ² (0.32) depth: 6 00:00:00 6.Kc6! ² (0.62) depth: 9 00:00:00 6.Kc6! Kb3 7.Kxb5 Kc3 8.Kc5 +- (2.02) depth: 9 00:00:00 6.Kc6! +- (2.32) depth: 12 00:00:00 2kN 6.Kc6! Ka3 7.Kxb5 Kb3 8.Kc5 Kc3 9.b5 +- (2.98) depth: 12 00:00:00 3kN 6.Kc6! +- (3.28) depth: 15 00:00:00 6kN 6.Kc6 Ka3 7.Kxb5 Kb3 8.Kc5 Ka3 9.b5 +- (3.31) depth: 15 00:00:00 8kN 6.Kc6 Ka3 7.Kxb5 Kb3 8.Kc5 Ka4 9.Kc4 Ka3 10.Kb5 +- (3.27) depth: 18 00:00:00 16kN 6.Kc6! +- (3.57) depth: 21 00:00:00 34kN 6.Kc6! Kxb4 7.gxf6 Kc4 8.f7 b4 9.f8Q b3 10.Qg7 Kd3 11.Qxg6+ +- (7.53) depth: 21 00:00:00 292kN, tb=27 6.Kc6 Kb3 7.Kxb5 Kc3 8.Kc5 Kd3 9.b5 Ke4 10.b6 Kxf4 11.b7 Kg3 12.b8Q+ Kxh3 13.Qf4 Kg2 14.Qxh4 Kf3 +- (7.03) depth: 24 00:00:01 661kN, tb=38 6.Kc6 Kb3 7.Kxb5 Kc3 8.Kc5 Kd3 9.b5 Ke4 10.b6 Kxf4 11.b7 Kg3 12.b8Q+ Kxh3 13.Qf4 Kg2 14.Qxh4 Kg1 15.Qg3+ +- (7.05) depth: 27 00:00:02 1069kN, tb=215 6.Kc6! +- (7.35) depth: 30 00:00:03 1999kN, tb=1488 6.Kc6! +- (7.65) depth: 33 00:00:12 7721kN, tb=13504 Looking at around depth 6, the line 6.Kc6! Kxb4 7.gxf6 is illegal!! And Junior thinks that White is winning as the depth goes deeper (White is not winning)!!! Anyone else find such pecularities with this position? I don't see any problems if the final position is setup directly. But by playing through the moves, maybe a hash table problem is encountered?! Or is it the handling of "en passant" captures? Here's the specifics of my setup... Junior 6a PIII 933Mhz, 128Mb Hash, all 5 man Nalimov bases being used This just gets "better"!!!! ;-) While writing this message I was continuing my tests on the other PC and after the moves 1.g3 g6 2.f4 h5 3.h3 f6 4.g4 h4 I see this analysis... New position 8/8/5pp1/1pK5/kP3PPp/7P/8/8 w - - 0 1 Analysis by Junior 6.0: 5.f5 g5 6.Kc6 Kxb4 -+ (-1.69) depth: 3 00:00:00 5.g5 fxg5 +- (8.94) depth: 3 00:00:00 5.g5 fxg5 +- (8.94) depth: 6 00:00:00 5.g5 fxg5 +- (8.94) depth: 9 00:00:00 5.g5 fxg5 +- (8.94) depth: 12 00:00:00 6kN 5.g5 fxg5 +- (8.94) depth: 15 00:00:00 38kN 5.g5 fxg5 +- (8.94) depth: 18 00:00:00 190kN, tb=1 5.g5 hxg3 6.gxf6 g2 7.Kd6 g1Q 8.Ke7 Qe1+ 9.Kf7 Qxb4 10.Kxg6 Qxf4 11.Kh7 Qf5+ 12.Kg8 Qxf6 13.Kh7 -+ (-9.30) depth: 21 00:00:40 34230kN, tb=36179 5.Kd4 Kxb4 -+ (-3.45) depth: 21 00:00:40 34765kN, tb=36459 5.Kd4 Kxb4 6.Kd5 Kc3 7.Ke6 b4 8.Kxf6 b3 9.Kxg6 b2 10.Kh5 Kb4 11.Kxh4 b1Q 12.Kg5 Kc4 -+ (-5.37) depth: 23 00:00:45 37827kN, tb=39626 Look at depth 18!! hxg3??? What is going on here?? lol Thanks in advance for any clarifications or suggestions... Gordon
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.