Author: Uri Blass
Date: 09:42:57 03/06/02
Go up one level in this thread
On March 06, 2002 at 11:01:26, John Merlino wrote: >On March 06, 2002 at 03:20:11, Uri Blass wrote: > >>On March 05, 2002 at 21:16:51, Steve wrote: >> >>>On March 05, 2002 at 00:10:46, John Merlino wrote: >>> >>>>On March 04, 2002 at 22:25:28, Sally Weltrop wrote: >>>> >>>>>On March 04, 2002 at 22:23:28, Sally Weltrop wrote: >>>>> >>>>>>[D]6k1/6pp/2qb4/8/2pQrNN1/1p3RPP/5P1K/r1B5 w - - 0 36 >>>>>> >>>>>>of course 36)Qxa1 and Black resigns next move! >>>>> >>>>> >>>>>this is taken from the 9th rd of Linares 2002 >>>>> >>>>>Shirov,A (2715) - Ivanchuk,V (2717) [C88] >>>>>Linares SuperGM Linares (9), 04.03.2002 >>>>> >>>>>1.e4 e5 2.Nf3 Nc6 3.Bb5 a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 0-0 8.a4 Bb7 9.d3 >>>>>Re8 10.Nbd2 Bf8 11.c3 Na5 12.Ba2 c5 13.d4 d6 14.b4 exd4 15.bxa5 dxc3 16.Nf1 Nxe4 >>>>>17.axb5 axb5 18.a6 Bc6 19.Rxe4 Bxe4 20.Bxf7+ Kxf7 21.Ng5+ Kg8 22.Nxe4 Qc8 >>>>>23.Nxc3 Rxa6 24.Rb1 b4 25.Nd5 Ra7 26.Rb3 Ra1 27.h3 Re5 28.Nf4 Qe8 29.Rf3 c4 >>>>>30.Kh2 b3 31.Qd2 d5 32.Ne3 Bd6 33.g3 Qc6 34.Ng4 d4 35.Qxd4 Re4 36.Qxa1 1-0 >>>> >>>>Two blunders in a row! (perhaps a victim of time pressure heading towards time >>>>control?) >>>> >>>>After 34.Ng4, Ivanchuk is clearly in the lead: >>>> >>>>Time Depth Score Positions Moves >>>>0:00 1/3 -1.86 1524 34...Rf5 35.Qc3 Ra2 36.Ne6 >>>>0:00 1/4 -2.47 6025 34...Rf5 35.Qc3 Ra2 36.Rd3 Rc2 >>>>0:00 2/5 -2.31 21581 34...Rf5 35.Bb2 Ra2 36.Ne3 Rxf4 >>>> 37.gxf4 >>>>0:00 2/6 -1.98 64702 34...Rf5 35.Qc3 Ra7 36.Bb2 Bxf4 >>>> 37.Rxf4 Rxf4 38.gxf4 >>>>0:02 3/7 -1.90 217927 34...Rf5 35.Qc3 Ra7 36.Ne6 Rxf3 >>>> 37.Qxf3 Bb4 38.Nd4 Qd6 >>>>0:08 3/8 -1.47 707718 34...Rf5 35.Qc3 Ra7 36.Ne6 Rxf3 >>>> 37.Qxf3 h5 38.Nd4 Qc5 >>>>0:40 4/9 -1.49 3345172 34...Rf5 35.Qc3 Ra7 36.Ne3 Be5 >>>> 37.Qb4 Rff7 38.Nexd5 Bxf4 39.Nxf4 >>>> Qxf3 40.Qb8+ Rf8 41.Qxa7 >>>>2:45 4/10 -1.51 14440411 34...Rf5 35.Qc3 Ra7 36.Ne6 Rxf3 >>>> 37.Qxf3 h5 38.Ne5 Bxe5 39.Qf8+ >>>> Kh7 40.Qf5+ g6 41.Qxe5 >>>> >>>>But when the game is over.... >>>> >>>>Time Depth Score Positions Moves >>>>0:00 1/3 3.38 3069 36...Re1 37.Nh5 Qb7 38.Ne3 >>>>0:00 1/4 3.50 10479 36...Re1 37.Nh5 Qb7 38.Qa5 Qxf3 >>>> 39.Qxe1 >>>>0:00 2/5 3.28 34108 36...Re1 37.Nh5 Qb7 38.Kg2 Bb4 >>>> 39.Qb2 >>>>0:01 2/6 5.73 116307 36...Re1 37.Qxg7+ Kxg7 38.Nh5+ >>>> Kh8 39.Bb2+ c3 40.Bxc3+ Qxc3 41.Rxc3 >>>> Re8 42.Rxb3 Re2 >>>>0:01 2/6 4.06 165811 36...Re7 37.Kg2 Re1 38.Nh5 Bf8 >>>> 39.Qc3 Qe6 40.Ne3 >>>>0:02 2/6 3.96 190722 36...Rxf4 37.Rxf4 Bxf4 38.Bxf4 >>>> c3 39.Ne5 Qc5 40.Qa8+ Qf8 >>>>0:04 3/7 4.00 355447 36...Rxf4 37.Rxf4 Bxf4 38.Bxf4 >>>> c3 39.Qd1 Qb7 40.Qd8+ Kf7 41.f3 >>>>0:07 3/8 4.87 709846 36...Rxf4 37.Rxf4 Bxf4 38.Bxf4 >>>> c3 39.Qa7 h5 40.Be5 Qg6 41.Qb8+ >>>> Kh7 42.Qxb3 hxg4 43.Qxc3 gxh3 44.Kxh3 >>>> Qh5+ 45.Kg2 >>>>0:32 3/8 4.76 2874367 36...Qb7 37.Bb2 h5 38.Ne3 Bxf4 >>>> 39.Rxf4 Rxf4 40.gxf4 Qe4 41.Bxg7 >>>> Qxf4+ 42.Kg1 >>>>0:54 3/8 4.71 4928413 36...Bxf4 37.Bxf4 Re6 38.Ne5 Qd5 >>>> 39.Kg2 Rb6 40.Qb2 >>>>1:41 4/9 5.06 9349513 36...Bxf4 37.Bxf4 Re6 38.Ne5 Qd5 >>>> 39.Kg2 Re7 40.Qc1 Rc7 41.Qb2 >>>> >>>>ouch.... I don't feel so bad now when I hang a pawn.... ;-) >>>> >>>>jm >>> >>>14.4 million positions and a search depth of 4 ply? What am I missing? >> >>a lot of extension and a different definition. >> >>I believe that the real depth is 9 >> >>The meaning of 4 is probably brute force depth >>with no null move pruning and no other pruning. >> >>The meaning of 9 is probably the real depth when chessmaster used only the null >>move algorithm to prune. >> >>The fact that usually in endgame I see x/x+1 numbers and not x/x+5 >>supports my opinion because I see no reason to prune less in the endgame >>(The opposite-endgame are the best candidate for more pruning rules >>unless you talk about null move pruning) >> >>Note that I did not write chessmaster so >>I do not know if my guess is correct. >> >>Uri > >The first number IS the brute force depth. All possibilities of plies of that >first number (in this case, 4) have been searched. The second number is the >MINIMUM extension depth. This is why Chessmaster can report a Mate in 15 (29 >plies) at depth 2/5, as it has done in this position (in less than one second, >no less ;-): It seems more correct to say that 4/9 means that chessmaster searches 9 plies with some pruning in the last 5 plies and not that it searches 4 plies with extensions. I have the feeling that chessmaster extensions include everything that is not pruned by null move or futility pruning or other pruning techniques that are used by the top chess programs. Otherwise chessmaster cannot search to depth 1 in the middlegame with practical chances to win top programs but it is exactly what happened when you give chessmaster big number for selectivity. Can you post a position when chessmaster misses tactics of 5 plies at depth 1/11 not because of zunzwang? Extensions cannot see everything about tactics because there is tactics of moves that seem to be quiet inspite of the fact that there is no zunzwang. 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.