Author: Uri Blass
Date: 08:30:48 01/11/04
Go up one level in this thread
On January 11, 2004 at 11:15:37, Mike Byrne wrote: >On January 11, 2004 at 10:28:29, Uri Blass wrote: > >>On January 11, 2004 at 09:08:55, Mike Byrne wrote: >> >>>On January 11, 2004 at 02:46:18, Uri Blass wrote: >>> >>>>On January 11, 2004 at 00:07:22, Mike Byrne wrote: >>>> >>>>>On January 10, 2004 at 20:23:32, Bob Durrett wrote: >>>>> >>>>>>On January 10, 2004 at 18:29:20, Mike Byrne wrote: >>>>>> >>>>>>>On January 10, 2004 at 17:47:25, Drexel,Michael wrote: >>>>>>> >>>>>>>>On January 10, 2004 at 17:22:53, Kurt Utzinger wrote: >>>>>>>> >>>>>>>>>On January 10, 2004 at 15:29:16, David Dahlem wrote: >>>>>>>>> >>>>>>>>>>White to move and draw. Which engines can find it? >>>>>>>>>> >>>>>>>>>>[D]3B4/1r2p3/r2p1p2/bkp1P1p1/1p1P1PPp/p1P1K2P/PPB5/8 w - - 0 1 >>>>>>>>>> >>>>>>>>>>1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. f5 1/2-1/2 >>>>>>>>>> >>>>>>>>>>Regards >>>>>>>>>>Dave >>>>>>>>> >>>>>>>>>Crafty_1908 WB2: (P4 1.8/32 MB hash) >>>>>>>>> 3 00:00 -4.04 1. c4+ Kxc4 2. Bd3+ Kd5 3. Bxa6 cxd4+ 4. Kd3 >>>>>>>>> 4 00:00 -4.01 1. c4+ Kxc4 2. Bd3+ Kd5 3. Bxa6 cxd4+ 4. Kd3 Ra7 >>>>>>>>> 4 00:00 -3.64 1. Ba4+!! >>>>>>>>> 4 00:00 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 >>>>>>>>> 5 00:00 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 >>>>>>>>> 6 00:00 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 7. Kf3 >>>>>>>>> 7 00:01 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 7. Kf3 Rc6 >>>>>>>>> 8 00:01 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 7. Kf3 Rc6 8. Ke4 >>>>>>>>> 9 00:01 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 7. Kf3 Rc6 8. Ke4 Rb6 >>>>>>>>>10 00:03 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 7. Kf3 Rc6 8. Ke4 Rb6 9. Kf3 >>>>>>>>>11 00:08 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 7. Kf3 Rc6 8. Ke4 Rb6 9. Kf3 Rc6 >>>>>>>>>12 00:11 0.00 1. Ba4+ Kxa4 2. b3+ Kb5 3. c4+ Kc6 4. d5+ Kd7 5. e6+ Kxd8 6. >>>>>>>>>f5 Rba7 7. Kf3 Rc6 8. Ke4 Rb6 9. Kf3 Rc6 >>>>>>>> >>>>>>>>Interesting. Crafty 19.06 can't find 1.Ba4+. >>>>>>>> >>>>>>>>Michael >>>>>>>> >>>>>>>>New game >>>>>>>>3B4/1r2p3/r2p1p2/bkp1P1p1/1p1P1PPp/p1P1K2P/PPB5/8 w - - 0 1 >>>>>>>> >>>>>>>>Analysis by Crafty 19.06: >>>>>>>> >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Bxa6 cxd4+ >>>>>>>> -+ (-4.45) Depth: 1/18 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Bxa6 cxd4+ >>>>>>>> -+ (-4.45) Depth: 1/18 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Bxa6 cxd4+ >>>>>>>> -+ (-4.45) Depth: 2/18 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Bxa6 cxd4+ >>>>>>>> -+ (-4.45) Depth: 2/18 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Bxa6 cxd4+ >>>>>>>> -+ (-4.45) Depth: 2/18 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Bxa6 cxd4+ 4.Kd3 >>>>>>>> -+ (-4.28) Depth: 3/19 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Bxa6 cxd4+ 4.Kd3 Ra7 >>>>>>>> -+ (-4.25) Depth: 4/19 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Be4+ Ke6 4.f5+ Kd7 5.Bxb7 axb2 6.Bxa6 b1Q 7.Bxa5 dxe5 >>>>>>>>8.dxc5 Qxa2 >>>>>>>> -+ (-5.36) Depth: 5/23 00:00:00 >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Be4+ Ke6 4.f5+ Kd7 5.Bxb7 axb2 6.Bxa6 b1Q 7.Bxa5 dxe5 >>>>>>>>8.dxc5 Qxa2 >>>>>>>> -+ (-5.36) Depth: 6/24 00:00:00 392kN >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Be4+ Ke6 4.f5+ Kd7 5.Bxb7 axb2 6.Bxa6 b1Q 7.Bxa5 dxe5 >>>>>>>>8.dxc5 Qxa2 >>>>>>>> -+ (-5.36) Depth: 7/26 00:00:01 806kN >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Be4+ Ke6 4.Bxb7 gxf4+ 5.Kxf4 dxe5+ 6.dxe5 Ra7 7.Bxa5 >>>>>>>>axb2 8.Bc8+ Kd5 9.Bb6 b1Q 10.Bxa7 Qxa2 >>>>>>>> -+ (-5.73) Depth: 8/27 00:00:04 2236kN >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Be4+ Ke6 4.f5+ Kd7 5.Bxb7 cxd4+ 6.Kxd4 dxe5+ 7.Kc5 Ra7 >>>>>>>>8.bxa3 Bxd8 9.Bc6+ Kc7 10.axb4 Rxa2 >>>>>>>> -+ (-6.00) Depth: 9/28 00:00:08 5405kN >>>>>>>>1.c4+ Kxc4 2.Bd3+ Kd5 3.Be4+ Ke6 4.f5+ Kf7 5.Bxb7 axb2 6.Bxa6 b1Q 7.Bxa5 Qxa2 >>>>>>>>8.e6+ Kf8 9.Bb6 Qxa6 10.dxc5 dxc5 11.Bxc5 >>>>>>>> -+ (-6.55) Depth: 10/30 00:00:31 21256kN >>>>>>> >>>>>>>It becomes a block pawn position, so when crafty is compiled with /DDETECTDRAW, >>>>>>>it will find it. >>>>>> >>>>>>Would somebody who is familiar with Crafty's /DDETECTDRAW code please tell >>>>>>everybody WHY is works [in plain English, please]? What is special about that >>>>>>code which almost every other chess programmer has failed to put into his/her >>>>>>engines? What does that code do [other than produce the correct move] which is >>>>>>unique? >>>>>> >>>>>>Bob D. >>>>> >>>>>A picture is worth a thousand words, play the best moves for the above position >>>>>and take a look at the ending position. >>>>> >>>>>[d]3k4/1r2p3/r2pPp2/b1pP1Pp1/1pP3Pp/pP2K2P/P7/8 b - - 0 6 >>>>> >>>>>the code simply identifies a locked position where the K is safely behind bars >>>>>(pawns) , i.e., the game must be drawn. >>>> >>>>The question is if it does not detect draw in wrong cases(for example if black >>>>has a queen instead of a bishop. >>>> >>>>Uri >>> >>> >>>New game >>>3k4/1r2p3/r2pPp2/q1pP1Pp1/1pP3Pp/pP2K2P/P7/8 b - - 0 1 >>> >>>Analysis by Crafty 19.08 SE: >>> >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Ke4 Rxh3 10.a5 Rxa5 >>> -+ (-23.39) Depth: 8 00:00:00 101kN >>>6...Qa4 >>> -+ (-23.79) Depth: 9 00:00:00 112kN >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Ke4 Rxh3 10.a5 Rh1 11.Kd3 Rxa5 >>> -+ (-23.89) Depth: 9 00:00:00 114kN >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Ke4 Rxh3 10.a5 Rh1 11.Kd3 Rxa5 >>> -+ (-23.89) Depth: 9 00:00:00 255kN >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Ke4 Rxh3 10.a5 Rh1 11.Kf3 Rxa5 >>> -+ (-23.99) Depth: 10 00:00:00 285kN >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Ke4 Rxh3 10.a5 Rh1 11.Kf3 Rxa5 >>> -+ (-23.99) Depth: 10 00:00:01 566kN >>>6...Qa4 >>> -+ (-24.39) Depth: 11 00:00:01 573kN >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Kf2 Rab6 10.Ke2 Rb2+ 11.Kf3 R6b3+ 12.Ke4 Re2# >>> -+ (-#7) Depth: 11 00:00:11 6025kN >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Kf2 Rab6 10.Ke2 Rb2+ 11.Kf3 R6b3+ 12.Ke4 Re2# >>> -+ (-#7) Depth: 11 00:00:13 7559kN >>>6...Qa4 7.bxa4 b3 8.axb3 Rxb3+ 9.Kf2 Rab6 10.Ke2 Rb2+ 11.Kf3 R6b3+ 12.Ke4 Re2# >>> -+ (-#7) Depth: 12 00:00:50 27457kN >>> >>>(Byrne, Secane, PA USA 11.01.2004) >>> >>>not in this case ..you put any type of piece (Queen, light square bishop or >>>knight) that can a force a win and it will see it >> >>Of course it is going to see it by search but the question was about static >>evaluation. >> >>You give only analysis at big depth and not at depth 1. >> >>Uri > >it needs 3 plies to in search tos ee that it bust through the locked pawns --- >that is how chess programs work ..search and evaluation ..not just one simple >static score at sd 1.. The static score is important because with wrong static score it may fail in a different position because of believeing in a wrong draw. It is obvious to me that it is going to cause serious mistakes when Crafty may prefer in the search losing position and not position that the opponent has 0.1 pawns advantage and when it is going to see that it is going to be too late and the only way to avoid it is going to be by giving the opponent bigger positional advantage. Giving 0.00 score for position that the side with the material advantage can win in parts of the cases is simply against intuition and I do not understand why people even test it instead of testing something more normal like having some average between the static score and 0 to do the score closer to draw in blocked position. 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.