Author: Simon Finn
Date: 05:54:12 10/22/01
Go up one level in this thread
On October 22, 2001 at 06:49:03, Andrew Williams wrote: >On October 21, 2001 at 18:22:26, Jon Dart wrote: > >>Recently Arasan had this position against Postmodernist: >> >>[D] 8/6P1/2B2n2/p4P2/P2k4/1pp5/1K6/8 w - - >> >>White played Kc1 and lost after ... c2 64. Kd2 Kc4 65. Bb5+ Kb4 66. Bd3 Kxa4 >>67. Bc4 Ka3 68. Kc1. >> >>I think White is dead after Kc1 and has to play Kxb3 here. But it seems to be >>extraordinarily difficult for a number of programs, including mine, to see that >>this is bad. > >PM was expecting Kxb3 and would have replied Kd3, with a score of +1.41 > >> >>Even at this point: >> >>[D] 8/6P1/2B2n2/p4P2/P1k5/1p6/2pK4/8 w - - >> >>Crafty keeps predicting a draw, with the pv Bb5+ Kb4 Bd3 Ng8 Kc1 Kc5 Bb5 Kd4 Kd2 >>Nf6 Bc6. Arasan's score goes - after about ply 12 but it doesn't see disaster. >>Hiarcs, however, seems to figure this out pretty fast. >> >>--Jon > >I've tried this position on my PIII-450 at work. The AMD 1200 at home is >approximately 4 times faster. PM's score is never positive for this second >position. The score is down at -2.07 after ply 14. The PV is: > >14= -207 115s 5927918 1.Bb5 Kb4 2.Bd3 Kxa4 3.Kc1 Kb4 4.Bxc2 bxc2 > 5.Kxc2 a4 6.Kb1 Ng8 ... which produces a drawn position. Easy for a human to see, but hard for a machine to evaluate correctly. Simon > > >In the game, after Bb5 PM got down to depth 16: > >16= 165 16s 4167280 65...Kb4 66.Bd3 Kxa4 67.Bc4 Ka3 68.Kc1 a4 > 69.g8=Q Nxg8 70.Bxg8 Ka2 71.f6 Ka1 > >In both cases, the PVs are incomplete. > >Regards > >Andrew
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.