Author: Uri Blass
Date: 02:33:52 12/18/01
Go up one level in this thread
On December 18, 2001 at 05:07:12, Jouni Uski wrote: >On December 18, 2001 at 04:29:59, Uri Blass wrote: > >>On December 18, 2001 at 03:00:55, Steve Maughan wrote: >> >>>I received Shredder 6 yesterday. Having spent a couple of hours playing around >>>with it, my impressions are positive! I don't have Shredder 5.32 but do have >>>the original Shredder 5. The upgrade seems to be much more tactically nimble >>>than S5, aggressive and, like all top engines, accutely aware of king safety. >>> >>>As a test overnight I pitted it against Goliath 1.5 - with a twist. I used the >>>'Nunn Match' feature to force Shredder 6 to play my favourite opening gambits. >>>This isn't meant to be an indicator of playing strength, more an experiment to >>>see if Shredder 6 can cope when the heat is on. Goliath 1.5 makes a good >>>opponent for these matches as it has great tactical skills even though it's >>>positional understanding is not quite up to that of the top engines. >> >>I do not believe that goliath has great tactical skills. > >I believe. At least in finding tactical combinations in my testsuites it's equal >or faster than Gambit Tiger or Fritz. But of course it loses most games, because >of missing positional/endgame ability... > >Jouni I think that the main test for tactical ability is simply to avoid tacical errors Not finding a tactical combination is only one of the possible tactical errors that are done by programs. In the first nunn game I saw the following evauation by the programs: 14.Ba3 0.06/12 Rb8 -1.28/13 15.Rb5 1.52/12 Qc7 -1.38/12 16.Bb3 3.24/11 Ke8 2.26/12 I guess that 1 pawn difference is a positional difference but I do not believe that more than it is positional. I believe that shredder6 saw 15.Rb5 for tactical reasons when goliath did not see nothing even 1 ply later. I gave yace(only material) to analyze the position after 15.Rb5 for more time and here is the latest line: Shredder 6 - Goliath Light 1.5 1rbq1k1r/pp2n1pp/3p2n1/1R3p2/2BP4/B1P2Q2/P1P2PPP/4R1K1 b - - 0 1 Analysis by Yace 0.99.56: 15...Qc7 16.Bb3 Ke8 17.Rd5 b5 18.Bxd6 Qb6 19.Bc5 Qa5 20.Bxe7 Nxe7 21.Rxe7+ Kxe7 22.Qg3 Be6 23.Qxg7+ Bf7 24.Re5+ ± (0.90) Depth: 14 00:40:46 1268644kN (Blass, Tel-aviv 18.12.2001) Yace(only material) suggests at move 14 h5 with evaluation of 1.70 for black at depth 13 (it can mean advantage of 2 knights for 2 pawns and a bishop based on the 3.40 for knight,0.80 for pawn and 3.50 for bishop) Shredder 6 - Goliath Light 1.5 r1bq1k1r/pp2n1pp/3p2n1/5p2/2BP4/B1P2Q2/P1P2PPP/1R2R1K1 b - - 0 1 Analysis by Yace 0.99.56: 14...Rg8 15.Bxg8 Nxg8 µ (-1.00) Depth: 1 00:00:00 14...Qe8 15.Bxd6 -+ (-1.70) Depth: 1 00:00:00 14...Qd7 -+ (-2.50) Depth: 1 00:00:00 14...Qd7 15.Rf1 -+ (-2.50) Depth: 2 00:00:00 14...Qd7 15.Rf1 Ng8 -+ (-2.50) Depth: 3 00:00:00 14...Qd7 15.Rf1 Ng8 16.Rb5 -+ (-2.50) Depth: 4 00:00:00 14...Qd7 15.Rf1 Ng8 16.Rb5 Rb8 -+ (-2.50) Depth: 5 00:00:00 17kN 14...Qd7 15.Rf1 Qa4 16.Bb4 f4 17.Bd3 -+ (-2.50) Depth: 6 00:00:00 55kN 14...Qd7 15.Bd5 Nh4 16.Qh5 Nxd5 17.Bxd6+ Qe7 18.Bxe7+ Nxe7 -+ (-2.10) Depth: 7 00:00:00 174kN 14...Qd7 15.Bd5 Qc7 16.Bxb7 Nh4 17.Qh5 Bxb7 18.Qxh4 -+ (-1.80) Depth: 7 00:00:00 174kN 14...h5 15.Rxe7 Nxe7 16.Rxb7 Bxb7 17.Qxh5 Rxh5 18.Bxd6 -+ (-1.81) Depth: 7 00:00:01 174kN 14...h5 15.Rb5 Qc7 16.Bd5 Rb8 17.Be6 Bxe6 -+ (-2.50) Depth: 7 00:00:01 591kN 14...h5 15.Be6 Rb8 16.Rb5 Nd5 17.Bxc8 Nxc3 -+ (-2.10) Depth: 8 00:00:01 958kN 14...h5 15.Be6 Rb8 16.g3 Qc7 17.Bxc8 Rxc8 18.Rxb7 Qxc3 19.Qxc3 Rxc3 20.Rexe7 Rxa3 21.Rxg7 -+ (-1.80) Depth: 8 00:00:02 958kN 14...h5 15.Be6 Rb8 16.g3 Qc7 17.Bxc8 Rxc8 18.Rxb7 Qxc3 19.Qxc3 Rxc3 20.Rbxe7 -+ (-1.80) Depth: 9 00:00:04 2320kN 14...h5 15.Be6 Rb8 16.g3 Qc7 17.Bxc8 Rxc8 18.Rxb7 Qxc3 19.Qxc3 Rxc3 20.Rb8+ Kf7 -+ (-1.70) Depth: 10 00:00:09 5276kN 14...h5 15.Be6 Rb8 16.g3 Qc7 17.Bxc8 Rxc8 18.Rxb7 Qa5 19.Bxd6 Rxc3 20.Rb8+ Kf7 -+ (-1.70) Depth: 11 00:00:24 13810kN 14...h5 15.Be6 Rb8 16.g3 Bxe6 17.Rxe6 Kf7 18.Rxd6 Qc7 19.c4 Qxc4 20.Rxb7 Rxb7 -+ (-1.70) Depth: 12 00:01:04 36247kN 14...h5 15.Be6 Rb8 16.g3 Bxe6 17.Rxe6 Kf7 18.Rxd6 Qc7 19.c4 Qxc4 20.Rxb7 Rxb7 21.Qxb7 Qxa2 -+ (-1.70) Depth: 13 00:02:44 90705kN (Blass, Tel-aviv 18.12.2001) Uri
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.