Author: Vincent Diepeveen
Date: 04:17:54 09/01/01
Go up one level in this thread
On August 31, 2001 at 14:57:02, Uri Blass wrote: >On August 31, 2001 at 14:15:51, Roy Eassa wrote: > >>On August 31, 2001 at 13:08:19, Christophe Theron wrote: >> >>>On August 30, 2001 at 13:25:22, Chris Taylor wrote: >>> >>>> >>>> >>>>r4r1k/4bppb/2n1p2p/p1n1P3/1p1p1BNP/3P1NP1/qP2QPB1/2RR2K1 w - - 0 1 >>>> >>>>Analysis by Gambit Tiger 2.0: >>>> >>>>1.Nf6 Bf5 2.Ra1 Qb3 3.Ne1 Nxe5 4.Bxa8 Bxf6 5.Bxe5 Rxa8 6.Bxf6 gxf6 >>>> ³ (-0.30) Depth: 7 00:00:00 112kN >>>>1.Ne1 Rac8 2.Ra1 Qb3 3.Rdc1 a4 4.h5 >>>> = (-0.10) Depth: 7 00:00:00 115kN >>>>1.Rxc5 Bxc5 2.Rc1 a4 3.Rxc5 a3 4.Qd1 axb2 5.Nd2 Qa1 6.Nb1 >>>> = (0.16) Depth: 7 00:00:00 131kN >>>> ² (0.40) Depth: 11 00:00:04 1229kN >>>>1.Nf6 Rac8 2.Nxh7 Kxh7 3.Rc4 a4 4.Nxd4 Nxd4 5.Rxd4 b3 6.Re1 a3 7.bxa3 Qxa3 8.Qh5 >>>> ² (0.50) Depth: 11 00:00:16 4371kN >>>> = (0.24) Depth: 13 00:00:50 13234kN >>>>1.Rxc5 Bxc5 2.Rc1 a4 3.Rxc5 a3 4.Bc1 Rac8 5.Ng5 Ne7 6.Rxc8 Rxc8 7.Nxf7+ Kg8 >>>>8.Ngxh6+ gxh6 9.Nxh6+ >>>> ± (0.72) Depth: 13 00:01:11 19275kN >>>> = (0.12) Depth: 16 00:15:18 280852kN >>>>1.Nf6 Rac8 2.Nxh7 Kxh7 3.Nd2 Kg8 4.Ne4 Nb3 5.Nf6+ Bxf6 6.exf6 Nxc1 7.Rxc1 Na7 >>>>8.fxg7 Rxc1+ 9.Bxc1 >>>> ² (0.40) Depth: 16 00:27:04 495999kN >>>>1.Nxh6 gxh6 2.Bxh6 Qa4 3.Ng5 Bg6 4.h5 Bf5 5.g4 Bh7 6.Bxf8 >>>> ± (1.08) Depth: 16 00:52:57 980973kN >>>> ± (1.16) Depth: 19 06:05:50 7150366kN >>>> >>>>(Taylor, Notts 30.08.2001) >>>> >>>>Chris. >>> >>> >>> >>>This looks like an analysis by Junior, not by Tiger. >>> >>>Tiger does not skip from ply 16 to ply 19. >>> >> >>Interesting! What makes Junior skip from depth 7 to 11, then 11 to 13, then 13 >>to 16, then 16 to 19? I don't recall seeing this behavior in chess programs. > >Junior does not skip from depth 7 to 11 because 11-7>3 >Junior's depth is the number of half plies and not the number It is sometimes even 1/3 of the plies :) Of course Amir is not going to comment too soon here, so here my guess from openings position for a random line (nullmove not included): e4 normal move costs 2 ply e5 normal move costs 2 ply Nf3 normal move costs 2 ply Nf6 normal move costs 2 ply Nxe5 capture costs 1 ply Nxe4 capture costs 1 ply Nf3 move goes back costs 3 ply Nf6 move goes back costs 3 ply Ba6 bad move costs 3 ply bxa6 capture move costs 1 ply o-o normal move costs 2 ply Be7 normal move costs 2 ply Qe2 normal move costs 2 ply O-O normal move costs 2 ply Rd1 move costs 3 ply as it doesn't go to (half) open file Re8 move goes to open file costs 2 ply In total this line costs junior 9x2 + 3x1 + 4x3 ply = 18 + 3 + 12 = 33 units. Normal program this costs 16 ply. I have experimented for years with this in diep and all i could figure out was that it works great at blitz, but as soon as you get above that level it doesn't work for me. Best regards, Vincent >of plies and going always from depth x to x+1 >can make Junior too slow. > >depth x also does not mean x/2 plies of normal programs because Junior does >a lot of extensions and usually depth 18 of Junior >may be similiar to depth 14 or 15 of other programs. > >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.