Author: John Merlino
Date: 10:10:21 06/25/04
Go up one level in this thread
On June 25, 2004 at 12:23:07, Mihaly Szalai wrote: >On June 25, 2004 at 11:45:21, John Merlino wrote: > >>On June 25, 2004 at 11:25:17, José Antônio Fabiano Mendes wrote: >> >>> Emir Dizdarevic vs GM A. Miles, Biel 1985 >>> [D]r4rk1/pbpn2pp/1p2p3/5p2/2NP3q/1P1BP3/PBQ2PP1/3R1RK1 b >>> And here GM Miles played 15...Bf3!!, winning in a few moves. >>> Source ==> http://www.chessgames.com/perl/chessgame?gid=1097658 >> >>A little comparison. First the Chessmaster 9000 engine (The King v3.23) with >>default settings (64MB Hash) on a P4-2.4: >>Time Depth Score Positions Moves >>0:00 1/4 2.01 19119 1...Rf6 2.f3 Rh6 3.Ne5 Nxe5 4.dxe5 >>0:00 1/4 0.00 19892 1...Bxg2 2.Kxg2 Qg5+ 3.Kh3 Qg4+ >> 4.Kh2 Qh4+ 5.Kg1 Qg5+ 6.Kh1 Qh6+ >> 7.Kg1 Qg5+ >>0:00 1/5 0.00 25628 1...Bxg2 2.Kxg2 Qg5+ 3.Kh3 Qg4+ >> 4.Kh2 Qh4+ 5.Kg1 Qg5+ 6.Kh1 Qh6+ >> 7.Kg1 Qg5+ >>0:00 1/6 0.00 49655 1...Bxg2 2.Kxg2 Qg5+ 3.Kh1 Qh6+ >> 4.Kg1 Qg5+ 5.Kh1 >>0:01 1/7 -0.25 212719 1...Bxg2 2.f3 Qh1+ 3.Kf2 Qh3 4.Ke2 >> Bxf1+ 5.Rxf1 b5 6.Ne5 Nxe5 7.dxe5 >>0:02 1/8 -0.14 456402 1...Bxg2 2.f3 Qh1+ 3.Kf2 Bxf1 4.Bxf1 >> Rad8 5.Bg2 Qh4+ 6.Kg1 b5 7.Na5 c5 >>0:06 1/9 -0.10 1250892 1...Bxg2 2.f3 Qh1+ 3.Kf2 Bxf1 4.Rxf1 >> Qh6 5.Rg1 Rf7 6.Ba3 Qh3 7.Bf1 >>0:11 1/9 -1.74 2369544 1...Bf3 2.Bxf5 Rxf5 3.Qxf5 exf5 >> 4.gxf3 Qh3 5.Nd2 Nf6 6.Rc1 Nd5 >> 7.Rfd1 >>0:17 1/10 -2.32 3993044 1...Bf3 2.Nd2 Bxg2 3.f3 Bh3 4.Ne4 >> Bxf1 5.Rxf1 fxe4 6.Bxe4 Qg3+ 7.Kh1 >> Qh3+ 8.Kg1 Rac8 >>0:30 2/11 -2.67 6932181 1...Bf3 2.Nd2 Bxg2 3.f3 Rf6 4.Bxf5 >> Rxf5 5.Qxf5 exf5 6.Kxg2 Re8 7.e4 >> Re6 8.Rfe1 Rg6+ 9.Kf1 >>1:15 3/12 -2.97 18746024 1...Bf3 2.Nd2 Bxg2 3.f3 Rf6 4.Ne4 >> fxe4 5.Qxg2 exd3 6.Qh2 Qg5+ 7.Kf2 >> Rxf3+ 8.Kxf3 Rf8+ 9.Qf4 Qh5+ 10.Kf2 >> Rxf4+ 11.exf4 >>2:50 4/13 -3.02 45770011 1...Bf3 2.Nd2 Bxg2 3.f3 Bh3 4.Ne4 >> fxe4 5.Bxe4 Qg3+ 6.Kh1 Bxf1 7.Rxf1 >> Rf6 8.Qh2 Qxh2+ 9.Kxh2 Rh6+ 10.Kg1 >> Rd8 11.f4 Rh3 >> >>Now the Chessmaster 10th Edition engine (The King v3.33) on the same position: >>Time Depth Score Positions Moves >>0:00 1/4 2.61 6342 1...Rad8 2.Ba3 Rfe8 3.Nd2 >>0:00 1/4 0.00 14088 1...Bxg2 2.Kxg2 Qg4+ 3.Kh1 Qh3+ >> 4.Kg1 Qg4+ 5.Kh1 >>0:00 1/5 0.00 25939 1...Bxg2 2.Kxg2 Qg4+ 3.Kh1 Qh3+ >> 4.Kg1 Qg4+ 5.Kh1 >>0:00 1/6 0.00 49581 1...Bxg2 2.Kxg2 Qg4+ 3.Kh1 Qh3+ >> 4.Kg1 Qg4+ 5.Kh1 >>0:00 2/7 -0.09 160625 1...Bxg2 2.f3 Qh1+ 3.Kf2 Bxf1 4.Rxf1 >> Qh4+ 5.Ke2 b5 6.Ne5 Nxe5 7.dxe5 >>0:02 2/8 -0.04 440265 1...Bxg2 2.f3 Bxf1 3.Bxf1 b5 4.Ne5 >> Qg5+ 5.Kf2 Nxe5 6.dxe5 Qh4+ 7.Kg1 >>0:06 2/9 -0.07 1448085 1...Bxg2 2.f3 Qh1+ 3.Kf2 Qh3 4.Rg1 >> Bxf3 5.Rdf1 Bd5 6.Ba3 Qh4+ 7.Ke2 >> Qh2+ 8.Kd1 >>0:10 2/9 -0.70 2129285 1...Bf3 2.Nd2 Bxg2 3.f3 Qg3 4.Ne4 >> fxe4 5.Qxg2 Qxg2+ 6.Kxg2 exd3 7.Rxd3 >> Rad8 8.d5 e5 >>0:23 3/10 -2.18 4976599 1...Bf3 2.Nd2 Bxg2 3.f3 Bh3 4.Ne4 >> Bxf1 5.Rxf1 fxe4 6.Bxe4 Qg3+ 7.Qg2 >> Qxg2+ 8.Kxg2 >>0:38 3/11 -2.25 9188867 1...Bf3 2.Nd2 Bxg2 3.f3 Bh3 4.Ne4 >> Bxf1 5.Rxf1 fxe4 6.Bxe4 Rac8 7.Qg2 >> Nf6 >>1:10 3/12 -2.24 17445021 1...Bf3 2.Nd2 Bxg2 3.f3 Bh3 4.Ne4 >> Bxf1 5.Rxf1 fxe4 6.Bxe4 Qg3+ 7.Qg2 >> Qxg2+ 8.Kxg2 Rad8 9.Rh1 Nf6 >>3:17 4/13 -2.94 49229898 1...Bf3 2.Nd2 Bxg2 3.f3 Rf6 4.Ne4 >> fxe4 5.Qxg2 exd3 6.Qh2 Qg5+ 7.Kf2 >> Rxf3+ 8.Kxf3 Rf8+ 9.Qf4 Qh5+ 10.Kg3 >> Rxf4 11.Kxf4 >> >>First of all, note that both of these were done in the CM9 GUI, so I have no way >>of knowing if the CMX engine would perform differently inside the CMX GUI. >> >>The times and evals are pretty close, no way to determine anything significant >>from them. But does anybody notice something rather interesting? :-) >> >>jm > >The main line is different. Or something else? > >Btw John, have you noticed that the CMX.obk is exactly the same as the >CM9000.obk? > >Mihaly Something else is different. Take a look at the X/Y depths.... Also, I had not noticed that the opening book was the same. This does not surprise me, and I don't necessarily think it is a BAD thing. Just about all that could have been done with it (apart from actually hiring an opening book expert to create a new one or edit the CM9 version) was to update the PGN file that was used to create the book with games from the last two years. This would probably have made very little difference to the book -- a little larger, and a few changes in probabilities here and there, but otherwise pretty insignificant.... 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.