Author: pavel
Date: 12:13:29 02/13/02
Go up one level in this thread
On February 13, 2002 at 15:07:36, Robert Hyatt wrote: >On February 13, 2002 at 15:01:25, pavel wrote: > >>Chess Tiger 14.0 - Shredder 6 >>[D]8/8/3r4/6p1/4k1B1/2K2R2/8/8 w - - 0 1 >> >>Analysis by Chess Tiger 14.0: >> >>69.Kc3-c4 Rd6-d2 70.Rf3-f8 Ke4-e3 71.Kc4-c5 Rd2-g2 72.Bg4-h5 g5-g4 73.Rf8-e8+ >>Ke3-f2 >> +- (2.34) Depth: 9 00:00:00 154kN, tb=54 >>69.Kc3-c4 Rd6-d2 70.Rf3-f8 Ke4-e3 71.Kc4-c5 Rd2-g2 72.Bg4-h5 g5-g4 73.Rf8-e8+ >>Ke3-f2 74.Kc5-d5 >> +- (2.34) Depth: 10 00:00:00 337kN, tb=111 >>69.Kc3-c4 Rd6-d1 70.Rf3-f8 Rd1-c1+ 71.Kc4-b5 Rc1-g1 72.Bg4-f5+ Ke4-d4 73.Rf8-g8 >>g5-g4 74.Rg8-d8+ Kd4-e5 75.Rd8-f8 >> +- (2.28) Depth: 11 00:00:03 1045kN, tb=432 >>69.Kc3-c4 Rd6-d1 70.Rf3-f8 Rd1-c1+ 71.Kc4-b3 Rc1-g1 72.Bg4-f5+ Ke4-d4 73.Kb3-c2 >>g5-g4 74.Kc2-d2 g4-g3 75.Kd2-e2 Rg1-g2+ 76.Ke2-f1 >> +- (2.18) Depth: 12 00:00:09 3250kN, tb=1522 >>69.Kc3-c4 Rd6-h6 70.Rf3-f7 Rh6-h4 71.Bg4-c8 g5-g4 72.Rf7-d7 Ke4-e3 73.Rd7-d3+ >>Ke3-f2 74.Rd3-d4 Rh4-h3 75.Rd4-d2+ Kf2-f1 76.Rd2-d7 >> +- (1.94) Depth: 13 00:00:35 12081kN, tb=4966 >>69.Rf3-g3 Ke4-f4 70.Rg3-g1 Rd6-a6 71.Kc3-d4 Ra6-a3 72.Bg4-e2 g5-g4 73.Rg1-f1+ >>Kf4-g5 74.Be2-d3 g4-g3 75.Kd4-e3 g3-g2 76.Rf1-f5+ Kg5-g4 >> +- (2.06) Depth: 13 00:00:45 15209kN, tb=7564 >> >>(Mahamood, MyTown 13.02.2002) >> >> >>Analysis by Shredder 6: >> >>69...Ke4-f4 70.Kc3-b4 >> ³ (-0.26) Depth: 1/2 00:00:00 >>69...Ke4-f4 70.Rg3-g2 >> ³ (-0.26) Depth: 2/4 00:00:00 >>69...Ke4-f4 70.Rg3-g1 Rd6-g6 >> ³ (-0.31) Depth: 3/6 00:00:00 >>69...Ke4-f4 70.Rg3-g1 Rd6-g6 71.Kc3-d3 >> = (-0.14) Depth: 4/8 00:00:00 >>69...Ke4-f4 70.Rg3-g1 Rd6-d8 71.Bg4-e6 Rd8-d6 72.Kc3-c2 >> ³ (-0.26) Depth: 5/10 00:00:00 >>69...Ke4-f4 70.Rg3-g1 Rd6-d8 71.Bg4-e6 Rd8-e8 72.Be6-d7 >> = (-0.12) Depth: 6/12 00:00:00 >>69...Ke4-f4 70.Rg3-g1 Rd6-a6 71.Kc3-d4 Ra6-a2 72.Kd4-d3 Ra2-a3+ 73.Kd3-e2 >> = (-0.06) Depth: 7/14 00:00:00 14kN, tb=292 >>69...Ke4-f4 70.Rg3-g1 Rd6-a6 71.Bg4-e2 Ra6-c6+ 72.Kc3-d4 Rc6-d6+ 73.Kd4-c5 >>Rd6-d2 74.Rg1-g2 >> = (-0.05) Depth: 8/16 00:00:00 30kN, tb=723 >>69...Ke4-f4 70.Rg3-g1 Rd6-a6 71.Bg4-e2 Ra6-c6+ 72.Kc3-d2 Rc6-d6+ 73.Be2-d3 g5-g4 >>74.Rg1xg4+ Kf4xg4 >> = (-0.01) Depth: 9/18 00:00:00 60kN, tb=1707 >>69...Ke4-f4 70.Rg3-g1 Rd6-a6 71.Bg4-e2 Ra6-c6+ 72.Kc3-d2 Rc6-d6+ 73.Be2-d3 g5-g4 >>74.Rg1xg4+ Kf4xg4 >> = (-0.01) Depth: 10/20 00:00:00 133kN, tb=4019 >>69...Ke4-f4 70.Rg3-g1 Rd6-a6 71.Bg4-e2 Ra6-c6+ 72.Kc3-d2 Rc6-d6+ 73.Be2-d3 g5-g4 >>74.Rg1xg4+ Kf4xg4 >> = (-0.01) Depth: 11/22 00:00:01 217kN, tb=6821 >>69...Ke4-f4 70.Rg3-g1 Rd6-g6 71.Kc3-d4 Rg6-d6+ 72.Kd4-c5 Rd6-a6 73.Kc5-d5 >> = (-0.01) Depth: 12/24 00:00:01 423kN, tb=12440 >>69...Ke4-f4 70.Rg3-g1 Rd6-g6 71.Kc3-d4 Rg6-d6+ 72.Kd4-c5 Rd6-a6 73.Bg4-d7 Ra6-g6 >>74.Rg1-f1+ Kf4-g3 75.Bd7-f5 Kg3-g2 76.Rf1-d1 >> = (0.00) Depth: 13/25 00:00:02 642kN, tb=18694 >>69...Ke4-f4 70.Rg3-g1 Rd6-g6 71.Kc3-d4 Rg6-d6+ 72.Kd4-c5 Rd6-a6 73.Bg4-d7 Ra6-g6 >>74.Rg1-g4+ Kf4-e5 >> = (0.00) Depth: 14/26 00:00:04 1086kN, tb=29673 >>69...Ke4-f4 70.Rg3-g1 Rd6-g6 71.Kc3-d4 Rg6-d6+ 72.Kd4-c5 Rd6-a6 73.Bg4-d7 Ra6-g6 >>74.Rg1-g4+ Kf4-e5 >> = (0.00) Depth: 15/27 00:00:06 1567kN, tb=40819 >>69...Ke4-f4 70.Rg3-g1 Rd6-g6 71.Kc3-d4 Rg6-d6+ 72.Kd4-c5 Rd6-a6 73.Bg4-d7 Ra6-g6 >>74.Rg1-g4+ Kf4-e5 >> = (0.00) Depth: 16/28 00:00:09 2284kN, tb=56293 >> >>(Mahamood, MyTown 13.02.2002) >> >> >> >>Why does chesstiger thinks it's better. >>A possible bug? >> >>pavs > > >It is missing a piece of knowledge that says "king, rook and bishop vs a >king and rook is a draw, so a king, rook and bishop vs a king, rook and >pawn can't possibly be won by the KRB side... > >IE: here is crafty's static evaluation of this position with no >search at all: > >note: scores are for the white side >material evaluation................. 3.60 >development......................... 0.00 >pawn evaluation..................... -0.24 >passed pawn evaluation.............. 0.16 >passed pawn race evaluation......... 0.00 >king safety evaluation.............. 0.00 >interactive piece evaluation........ -3.52 >total evaluation.................... 0.00 > > >Or after a search: > > time surplus 0.00 time limit 30.00 (3:30) > nss depth time score variation (1) > 1 0.00 0.00 1. Bh5 > 1-> 0.00 0.00 1. Bh5+ > 2 0.00 0.00 1. Bh5 Rd1 > 2-> 0.00 0.00 1. Bh5 Rd1 > 3 0.00 0.00 1. Bh5 Rd1 2. Kb4 > 3-> 0.00 0.00 1. Bh5 Rd1 2. Kb4 > 4 0.00 0.00 1. Bh5 Rd1 2. Kb4 Ra1 > 4-> 0.00 0.00 1. Bh5 Rd1 2. Kb4 Ra1 > 5 0.00 0.00 1. Bh5 Rd1 2. Kb4 Ra1 3. Bg4 > 5-> 0.01 0.00 1. Bh5 Rd1 2. Kb4 Ra1 3. Bg4 > 6 0.01 0.00 1. Bh5 Rd1 2. Kb4 Ra1 3. Bg4 Ke5 > 6-> 0.02 0.00 1. Bh5 Rd1 2. Kb4 Ra1 3. Bg4 Ke5 > 7 0.04 0.00 1. Bh5 Rd1 2. Kb4 Ra1 3. Bg4 Ke5 4. > Kc5 > 7-> 0.05 0.00 1. Bh5 Rd1 2. Kb4 Ra1 3. Bg4 Ke5 4. > Kc5 > >It is something the engine has to "know" as search won't be able to see >that eventually the pawn is lost, and then the resulting KRB vs KR is >_still_ a draw. Even without the knowledge, shouldn't TB kick in to help? note that the search might not be deep enough, also note that the game is currently running and CT is consistent with it's eval throughout the last 10+ moves. pavs
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.