Author: Brian Katz
Date: 03:29:39 12/27/03
Go up one level in this thread
On December 27, 2003 at 00:51:11, Mike Byrne wrote: > >[d]8/r7/8/5bk1/8/5B2/5RPP/6K1 b - - 0 1 >Crafty19.08 SE UCI 2 MB: > 7 00:00 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 > 8 00:00 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 > 9 00:00 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 >10 00:00 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 >11 00:00 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 >12 00:00 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 >13 00:01 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 >14 00:04 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 >15 00:06 0.01 1. ... Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.gxh3 I am assuming you are referring to this position without the use of Tablebases, since with Tablebases this position is a forced draw, as we all know. It is the Wrong colored Bishop for the Rook pawn to Promote as the Black King cannot be forced away from h8 and g8 and can always move back an forth from each of these squares or be stalemated. I found some interesting findings. First off this is with my slower computer Pent II 350 MHz 384 SDRAM, Hashtables 128 MB No Tablebases.... Fritz 8 finds 1...Ra1+ 2.Rf1 Rxf1+ 3.Kxf1 Bh3 4.Kf2 Bxg2 and rates it as =(0.12) and then after a period of time it lowers it to =(0.06) A draw for all practical purposes. With Yace Paderborn, Crafty 19.06, 19.01, Ruffian 1.0.0, Ruffian 1.0.1 the evaluation is almost immediately =(0.00) Crafty 18.11 has a little trouble with this position, playing similar moves, sacrificing the Bishop on h3 but not fully realizing that it should be =(0.00) with evaluations in the =(0.20's). Comet B27 and Comet B50 both have White winning with +-(3.36)and up +-(2.15) and up, respectively....Not finding the Bishop Sacrifice ...Bh3. Now with my AMD Athlon XP 2600+ 1 GIG DDR SDRAM Hashtables 768 MB....NO TABLEBASES Fritz 8 still begins with a =(0.12) evaluation and also drops to =(0.06) after a period of time. Isn't that interesting, the Great Fritz 8 on an Athlon 2600+ can find the true evaluation of the position when some weaker engines immediately find it on a Pent II 350 MHz. There is a difference here with Crafty 18.11, with a +-(0.29)The 350 MHz actually gives a better and lower evaluation with this engine. Yace Paderborn, Ruffian 1.0.0, Ruffian 1.0.1, Crafty 19.06, Crafty 19.03, Crafty 19.01, Crafty 18.12 all give =(0.00) Anaconda 1.0, after 40 seconds lowers the evaluation from the initial +(1.98) down to +(0.00) after 40 seconds. List 504 Oddly shows the evaluation as =(-0.15)and then changes it to =(0.15)after the move ...Bh3 is played. That is strange......How can this engine evaluate Black as being ahead ??? Shredder 6 has trouble with this position placing White ahead with +-(1.56), choosing not to exchange Rooks after the Check. However, if I force the sequence of moves upto the ...Bh3 Sacrifice, it evaluates the position as +-(4.85)and after Kf2 Bxg2 +-(5.52) or gxh3 rather than Kf2 with +-(4.76). Hiarcs 8 rates the positon as +-(1.16) but then drops the evaluation to =(0.00) after 30 seconds. I tried Fritz 5.32, Fritz 7 They also give white a plus rather than 0.00. EXchess 3.11 has lots of trouble with this position with evaluations in the +-(2.50's) Best I can find for now. Brian Happy Holidays
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.