Author: Brian Katz
Date: 11:07:07 10/18/02
I have mentioned in previous posts, that I have not been getting the same results from Deep-Fritz 7 as I have from Fritz 7 on ICC computer account games vs human opposition. I am using a 350 MHz Pent II 384 SDRAM Hash table set at 32 MB in game 5/12. I have been getting the best resutls with 32 MB even though this contradicts Steve Lopez'(of Chessbase) formula of 2x processor speed x seconds per move / by 1000 which would translate to approx 16 MB based on a 50 move game roughly 900 seconds of play. I only use 40 moves as the divisor since at least the First 10 moves are usually automatically played from the opening. Here is a position From an ICC computer account game at game 5/12 as Deep-Katzparov rated 2813 vs. a human opponent MK82 rated 2508 in which Fritz 7 played 24.Bc3...Since I have been noticing a problem with Deep Fritz 7's play on a single processor compared to Fritz 7, I decided to go over the game with Deep-Fritz 7 as the analysis engine to see if there is any real difference. Most of the game would have gone the same way, although I can't be exactly sure, since the analysis times may cause the program to eventually play a different move had I allowed Deep Fritz 7 to anlayse a little longer. Never the less here is a position where Deep Fritz 7 considers the move that Fritz 7 actually chose in the game, but would not have played itself. Deep-Fritz 7 selects 24.Rc2... rather than 24.Bc3... as played by Fritz 7. [D]6k1/r3pp1p/1p2b1p1/8/1B6/P5P1/1b2PP1P/2RR2K1 w Also here is the actual game score. Deep-Katzparov (2815) - MK82 (2508) [D77] ICC 5 12 Internet Chess Club, 18.10.2002 1.d4 Nf6 2.c4 g6 3.Nf3 Bg7 4.g3 0-0 5.Bg2 d5 6.0-0 dxc4 7.Na3 c5 8.dxc5 Qa5 9.Qd4 Nfd7 10.Nxc4 Qxc5 11.Qxc5 Nxc5 12.Be3 Nba6 13.Rac1 Be6 14.Ng5 Bg4 15.b4 Ne6 16.Bxb7 Nxb4 17.Bxa8 Rxa8 18.a3 Nd5 19.Nxe6 Bxe6 20.Bd2 Nb6 21.Nxb6 axb6 22.Bb4 Ra7 23.Rfd1 Bb2 24.Bc3 Bxc3 25.Rxc3 b5 26.Rd8+ Kg7 27.e4 Bc4 28.f4 e5 29.fxe5 Re7 30.a4 Rxe5 31.axb5 Bxb5 32.Rd5 Rxd5 33.exd5 Kf6 34.Re3 Bc4 35.d6 Be6 36.Kf2 h5 37.Ke2 g5 38.Kd3 Bf5+ Black resigns 1-0 All opinions and criticisms welcome as always. Best Brian Katz
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.