Author: Stuart Cracraft
Date: 13:55:24 07/18/04
Go up one level in this thread
On July 18, 2004 at 16:22:21, Roy Brunjes wrote: >In a recent game between Shredder 8 and Chessmaster 9000 on ICC the following >position arose: > >[D]r1b3k1/pp2n1b1/1qn1prpp/2pp4/5B1P/2PP1NP1/PPQ2PB1/RN3RK1 b - - 0 13 > >Chessmaster played 13. ... Rxf4!? > >Is this sound? I have analyzed it with the help of both Chessmaster and >Shredder 8 and it seems pretty good to me, but I do not have the CPU horsepower >and/or time necessary to test as many programs and all of the possible >variations as the group does here. > >The game was won in convincing fashion at 10m+10sec time control as follows: > >[Event "ICC Game"] >[Site ""] >[Date "2004.7.17"] >[Round ""] >[White "Shredder 8 running on Pentium IV 2.8 GHz"] >[Black "SKR_Jr_Mix running on Centrino 1.5 GHz"] >[TimeControl "300+5"] >[Result "0-1"] > >1.e4 e6 2.Qe2 c5 3.g3 Nc6 4.c3 g6 5.Bg2 Bg7 6.Nf3 Nge7 7.O-O d5 8.d3 O-O >9.e5 h6 10.h4 Qb6 11.Bf4 f6 12.exf6 Rxf6 13.Qc2 Rxf4 14.gxf4 Bd7 15.Qd2 >Rf8 16.Re1 Qc7 17.Na3 Rxf4 18.Rad1 Nf5 19.h5 Nce7 20.hxg6 Nxg6 21.Nc2 >Nfh4 22.Nxh4 Rxh4 23.Kf1 Nf4 24.Qe3 e5 25.Rd2 Nxg2 0-1 > > >Note1: I set CM9000 to a time control of 5m+5sec to allow for manual operation >of the program on ICC without losing on time. > >Note2: Yes, Shredder's opening line is weird -- it is not using the standard >book, so no flames about that please. > >Roy Personally I would not assay this sac and I've been known to throw away a lot of wood in my time in sacrificial lines. GA with a small transposition table, aspiration, verified null move (R=2), extremely simple evaluation, only a couple of extensions, a recent strong improvement in a test suite due to a bunch of hash fixes spurred by others comments on this board, not surprisingly also does not play Rxf4 (see below for an analysis). The reason I would not personally assay the move is that my black queen cannot easily come into the attack as rapidly as I would like after the sac. It will take me 2-3 moves to get the queen over and that gives white time to consolidate somewhat. NULLMV TT(524288) Alpha=-1023 Beta=977 Maxdepth=9999999 MaxTime=999 1. b6d8 0.00 21 49 b6d8 2. b6d8 0.01 -19 664 b6d8 c2d1 3. b6d8 0.02 -9 2724 b6d8 c2d1 c8d7 4. f6f8 0.15 -18 30688 f6f8 f1e1 b6d8 c2d1 5. f6f8 0.43 -8 88445 f6f8 f1e1 b6d8 c2d1 c8d7 6. f6f5 3.62 -18 700575 f6f5 f4c1 b6d8 c2d1 e6e5 b1d2 7. f6f5 8.44 -14 1655603 f6f5 g2h3 f5f8 b1d2 b6d8 c2d1 c8d7 8. e7f5 59.42 -11 11190314 e7f5 c2c1 f6f7 f1e1 b6d8 c1d1 9. f6f5 296.14 -5 50368361 f6f5 g2h3 f5f7 f1e1 e7f5 b1d2 b6d8 c2d1 c8d7 Still, it is better than the usual computerish move and I like to see this kind of thing even if it were to be proven unsound later but I would not personally play it. Stuart
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.