Author: Mark Young
Date: 13:11:31 12/29/01
Go up one level in this thread
On December 29, 2001 at 14:07:22, ALI MIRAFZALI wrote: >There has been some speculation lately on the strenght difference between >Deep Blue and current top commercial programs.According to Mig from >www.kasparovchess.com programmers are divided on this topic some believing >that Deep Bluecan crush any commercial program and some believing that top >programs today have enough chess knowlegde to compensate the huge >hardware difference.As a step in this direction many people have analyzed >the kasparov Deep blue matches with varying speculative results.I think 6 >or 12 games is not enough for a scientific result .Here I present another >step in this direction.To my knowlegde this is one of the very few games >in which we have published analysis from IBM's Supercomputer Deep Blue >free to the general public.The game is Gulko(white ) vs Benjamin (Black) >My source is CHESS LIFE(a magazine published in the US)The issue number >is January 1997.This game is from the 1996 US championship.Please refer to >the above mentioned issue for the complete game score.The article was >written by Joel Benjamin.The following diagram is a variation from >Gulko vs Benjamin >[D]2b1k3/6p1/p1p3p1/P1n1Pp2/8/2N4N/1r1r2PP/R1KB3R w >In the article Benjamin claims that he thought that in the above diagram >black is winning but Deep Blue demonstrates otherwise .From the article: >29.Nf4 g5 30.Rb1! Rxb1+ 31.Kxd2 Rb2+ 32.Kc1 Rf2 33.Bh5+ Kd8 34.Rd1+ Kc7 >35.Nd3 Nxd3+ 36.Rxd3 Rxg2 37.Rg3 and White is winning -Deep blue.Here is >the analysis of CM8K(My favorite program)of the above diagram >29.Nf4 g5 30.Rb1 Rxb1+ 31.Kxd2 Rb2+ 32.Kc1 Rf2 33.Bh5+ Kd8 34.Nfe2 Rxg2 >35.Bf3 Rf2 .Which program analyzed better?? Please indicate the result >of your program.It would be interesting to know that if there is a >program that gets the same principal variation as Deep Blue in the above >diagram.Maybe even a better line?! Either the position is to simple to be a good example for Micros vs Deep Blue or micro's are now equal to Deep Blue. Today programs have not trouble seeing that White is better here and see it in less then 1 sec. New position 2b1k3/6p1/p1p3p1/P1n1Pp2/8/2N4N/1r1r2PP/R1KB3R w - - 0 1 Analysis by Chess Tiger 14.0: 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Kc2 g5 6.Rd6 +- (1.68) Depth: 8 00:00:00 113kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Rb1 g5 6.Rb6 +- (1.74) Depth: 9 00:00:00 159kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Kc2 g5 6.h4 Nd7 7.hxg5 Nxe5 +- (1.78) Depth: 10 00:00:01 291kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Rb1 g5 6.h4 gxh4 7.Rb6 Ne4+ 8.Kd3 +- (1.72) Depth: 11 00:00:02 663kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Rb1 g5 6.Rb6 Bd7 7.h4 gxh4 8.Kc2 +- (1.56) Depth: 12 00:00:05 1463kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Rb1 g5 6.Rb6 Bd7 7.h4 gxh4 8.Rb8+ Ke7 9.Rb6 +- (1.68) Depth: 13 00:00:20 5265kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Ne2 Ne6 6.h4 g6 7.Rf1 Nc5 8.Nxf4 Ne4+ 9.Kd3 +- (1.62) Depth: 14 00:00:32 8569kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Ne2 Ne6 6.h4 g6 7.Rf1 Bd7 8.Nxf4 Nxf4 9.Rxf4 +- (1.64) Depth: 15 00:01:20 21414kN 1.Nf4 g5 2.Rb1 Rxb1+ 3.Kxd2 Rxd1+ 4.Rxd1 gxf4 5.Ne2 Ne6 6.h4 g6 7.Rf1 c5 8.Nxf4 Nxf4 9.Rxf4 Be6 10.g3 +- (1.56) Depth: 16 00:03:56 64101kN (Young, MyTown 29.12.2001)
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.