Author: Mike S.
Date: 12:36:23 07/09/01
Go up one level in this thread
On July 09, 2001 at 11:14:35, Kurt Widmann wrote: >How deep do both have to search before dropping 9... Qc6?? for Bxe6! instead? The following results are from P3/700 MHz, 32 MB Hash (or 29 MB with Crafty; see below). I use Fritz 6 Light. It switches away from 9...Qc6? to 9...Bg7 first, at a depth of 13/41 plies, and finds 9...Bx6 much later at the same depth: New game - Fritz 6 Light rnb1qbkr/ppp1pN1p/1n1pP1p1/8/8/5Q2/PPPP1PPP/RNB1K2R b KQ - 0 1 Analysis by Fritz 6 Light: 9...Qxf7-- 10.Qxf7# +- (3.19) Depth: 1/3 00:00:00 +- (#1) Depth: 1/3 00:00:00 9...Bxe6! 10.Nxh8 Kxh8 11.Qxb7 +- (1.81) Depth: 1/3 00:00:00 ³ (-0.31) Depth: 1/4 00:00:00 9...Bg7! 10.Nxh8 Kxh8 ³ (-0.38) Depth: 1/4 00:00:00 ³ (-0.47) Depth: 1/4 00:00:00 9...Qc6-- 10.Nh6+ Kg7 11.Qf7+ Kxh6 12.d3+ Kh5 13.Rg1 Qd5 14.g4+ ³ (-0.50) Depth: 1/7 00:00:00 ± (1.37) Depth: 13/41 00:01:13 39133kN 9...Bg7! 10.Nxh8 Kxh8 11.0-0 Nc6 12.Re1 Nd4 13.Qd1 Qc6 14.Nc3 ± (1.34) Depth: 13/41 00:01:13 39535kN -+ (-1.63) Depth: 13/41 00:06:54 213769kN 9...Nc6! -+ (-1.66) Depth: 13/41 00:08:04 248795kN 9...Bxe6! -+ (-1.69) Depth: 13/41 00:08:28 259952kN -+ (-1.69) Depth: 13/41 00:08:28 259952kN ---------- The next analysis is from the CB native version of Crafty 18.08. Unfortunately, it switches away from 9...Bxe6 at ply 9/18 to 9...Qc6?, but comes back to Bxe6 at ply 13/39: New game - Crafty 18.08 rnb1qbkr/ppp1pN1p/1n1pP1p1/8/8/5Q2/PPPP1PPP/RNB1K2R b KQ - 0 1 Analysis by Crafty 18.08: 9...Bxe6 10.Nxh8 Kxh8 11.Qxb7 c6 12.0-0 -+ (-1.78) Depth: 6/13 00:00:00 9...Nc6 10.Nxh8 Kxh8 11.Qe3 Nd5 12.Qe4 Ncb4 -+ (-2.01) Depth: 7/18 00:00:00 -+ (-2.09) Depth: 7/18 00:00:00 220kN 9...Bxe6 10.Nxh8 Bd5 11.Qg4 Kxh8 12.Nc3 Bc6 13.0-0 Kg8 -+ (-2.14) Depth: 7/18 00:00:00 253kN -+ (-2.62) Depth: 9/18 00:00:01 597kN 9...Qc6 10.Nh6+ Kg7 11.Qf7+ Kxh6 12.d3+ Kh5 13.g4+ Kh4 14.Bg5+ Kxg5 15.h4+ Kxg4 16.Rg1+ Kh3 17.Qf4 Qg2 18.Rxg2 Kxg2 19.Qg3+ Kh1 20.Qh3+ Kg1 21.Qg4+ Kh1 22.Qh3+ -+ (-3.02) Depth: 9/18 00:00:09 3823kN = (0.00) Depth: 13/39 00:08:27 216650kN 9...Bxe6 10.Nxh8 Kxh8 11.Nc3 Nc6 12.d3 Na4 13.0-0 Nxc3 14.bxc3 Qf7 15.Qxf7 Bxf7 16.Rb1 Rb8 ³ (-0.40) Depth: 13/39 00:08:34 219729kN -+ (-3.01) Depth: 13/39 00:09:26 239473kN 9...Qc6? 10.Nh6+! does this trick... difficult to expect for some programs, that white will sac another piece. It seems, that I could still score against some top programs with this (if their book allows that line until 5.Bxf7+ at least). Usually, I test with this position from the line after 6...Kg6 7.Qf3. Here, I saw that most newer programs find 7...Qe8 easily - so I thought, it's defused (at least for my puroses; I need a quick mate line... ± mean I still lose against the computer :o) Noordijk - Landau [B02] Rotterdam, 1927 [Gelenczei 1958] [D]rnbq1b1r/ppp1p1pp/1n1p2k1/4P1N1/8/5Q2/PPPP1PPP/RNB1K2R b KQ - 0 7 8...Qe8 (Landau took the Knight...) Regards, M.Scheidl
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.