Author: John Merlino
Date: 09:18:59 03/19/02
Go up one level in this thread
On March 19, 2002 at 02:05:42, Jonas Cohonas wrote: >On March 18, 2002 at 20:10:34, John Merlino wrote: > >>On March 18, 2002 at 18:17:38, Art Basham wrote: >> >>>[D]r1bq1rk1/1p2bppp/p3p3/n3P3/4N3/1P1P1N2/PB4PP/R3QR1K w - - >>> >>>White played Nf6+...!! ...if gxf6, 2. Qg3+ etc, >>> >>>It took my old Mach 3 about 3 hrs. to find this one... >>> >>>from the Encl. of Middle Games >> >>A testing version of Chessmaster 9000 finds it in less than one second on my >>PIII-600: >> >>Time Depth Score Positions Moves >>0:00 1/3 0.50 1031 1.Qe3 Nc6 2.Rac1 Bd7 >>0:00 1/4 0.53 2656 1.Qe3 Nc6 2.Rac1 Bd7 3.a3 >>0:00 2/5 0.45 18089 1.Qe3 Nc6 2.Rac1 Bd7 3.a3 Qa5 >>0:00 2/5 0.66 33051 1.Nf6+ gxf6 2.exf6 Bxf6 3.Bxf6 >> Qxf6 4.Ne5 Qg5 5.Qxa5 Bd7 >>0:01 2/6 0.93 58616 1.Nf6+ gxf6 2.exf6 Bxf6 3.Bxf6 >> Qxf6 4.Qxa5 Bd7 5.Ne5 Qg5 6.Rae1 >>0:03 3/7 0.67 200096 1.Nf6+ gxf6 2.exf6 Bxf6 3.Bxf6 >> Qxf6 4.Qxa5 Bd7 5.Ne5 Qg5 6.Qc7 >>0:09 3/8 0.86 606043 1.Nf6+ gxf6 2.exf6 Bxf6 3.Bxf6 >> Qxf6 4.Qxa5 Bd7 5.Rae1 Bb5 6.Ne5 >> Qe7 7.Qc3 >>0:31 4/9 1.39 2235370 1.Nf6+ gxf6 2.exf6 Bxf6 3.Qg3+ >> Kh8 4.Ne5 Bd7 5.Rxf6 Qxf6 6.Qf2 >> Qg7 7.Ng6+ hxg6 8.Bxg7+ Kxg7 9.Qd4+ >> Kg8 10.Qxd7 >>1:36 4/10 1.53 7172020 1.Nf6+ gxf6 2.exf6 Bxf6 3.Qg3+ >> Kh8 4.Ne5 Bd7 5.Rxf6 Qxf6 6.Qf2 >> Qg7 7.Nxd7 f6 8.Bxf6 Rxf6 9.Nxf6 >> >>jm > >It looks like CM is going from strength to strenght, can you tell what will be >new in the CM9000 engine or is it too early? > >Regards >Jonas All I know is that Johan has found one significant endgame evaluation error, and has done quite a bit of research into speeding up the search (although I have no idea if this research has proved fruitful). He has also provided support for a new mode in the "Mentor Lines" window (which is actually going to be a separate window which will contain the equivalent of the "Chessmaster" checkbox data in CM8000's "Thinking Lines" window -- did that make sense?). The new "Mentor Lines" window will be able to show data in two different ways: 1) The way it does in CM8000, which a single PV updated over time, and 2) A "Best N Moves" list, in which the user can select anywhere from 2-N and the best N moves will be sorted by the best move at the top -- full PVs for each move included. Very useful for seeing alternate possibilities. jm
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.