If this is your first visit, be sure to check out the FAQ by clicking the link above. You may have to register before you can post: click the register link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. 


Thread Tools  Display Modes 
#1




Engine puzzle position
I have stumbled upon a position where quite a few engines  I thought at
first it was just x64 types but Fritz 10 also sees the same phantom "Good move" which evaluates as a draw where none exists. The position is hopelessly lost but that isn't the point. Something is wrong. I found it analysing a lost game where there was a clear and bizarre discontinuity in the evaluation function that made no sense to me. Crafty x64 23.05 also sees the wrong right move but x32 20.14 is OK. r2r4/1p2p1kp/p5p1/3BNbPn/8/2N4P/PPP2P2/R3K2R b KQ  0 21 It is a close run thing between 22. ... Rac8, or 22 ... Nf4 but the phantom move which evaluates as a draw is 22. ... Bc8 If you actually play this move Bc8 then the evaluation becomes correct again with white to move. There is something odd about this position that breaks some engines. Any ideas what it might be? My guess is that it is some sort of draw by repetition bug common to several of the engines since the offending thread displays as just a single move with no continuation line : 354: Deep Fritz 14  M, Blitz 5m+5s4m+2s r2r4/1p2p1kp/p5p1/3BNbPn/8/2N4P/PPP2P2/R3K2R b KQ  0 1 Analysis by Deep Fritz 14 x64: 1. = (0.00): 21...Bc8 2. + (4.91): 21...Nf4 22.000 Nxh3 23.Rdf1 Rac8 24.Bxb7 Rc5 25.f4 Rd4 26.Re1 e6 27.Nd3 Rxd3 28.cxd3 Nxf4 29.Bxa6 Bxd3 30.Rh4 Ng2 31.Rxe6 Bxa6 32.Re7+ Kg8 33.Rhxh7 Rxg5 34.Reg7+ Kf8 35.Rf7+ Kg8 36.Rhg7+ Kh8 37.Ne4 3. + (4.92): 21...Bxc2 22.Bxb7 Rab8 23.Nc6 Rxb7 24.Nxd8 Rxb2 25.Ne6+ Kf7 26.Nc5 Nf4 27.a4 Bf5 28.Ra2 4. + (4.92): 21...Rac8 22.000 Nf4 23.Bf3 Rxd1+ 24.Rxd1 Rc5 25.Re1 Nxh3 26.Bxb7 Nxg5 27.Bxa6 h5 28.Bd3 Nf7 29.Nxf7 Kxf7 30.Kd2 Bxd3 31.Kxd3 Rf5 32.a4 h4 33.Ne4 Rd5+ 34.Ke3 Ra5 35.Rh1 Rxa4 36.Rxh4 5. + (5.18): 21...Rab8 22.000 Nf4 23.h4 Rbc8 24.Bf3 b5 25.Rxd8 Rxd8 26.Nc6 Re8 27.Rd1 Nh3 28.Rd2 6. + (5.34): 21...Ra7 22.000 Nf4 23.h4 Rc8 7. + (5.37): 21...e6 22.Bxb7 Rab8 23.Nc6 Rxb7 24.Nxd8 Rxb2 8. + (5.60): 21...Rf8 22.000 Ra7 23.Rd4 Bc8 24.Re1 b5 25.Nd3 Bxh3 26.Rde4 e6 27.Bxe6 Bg2 28.Rd4 Re7 29.Rd7 Rxd7 30.Bxd7 Ba8 31.Re6 Bb7 32.Rb6 Rb8 33.Nd5 (, 21.07.2017) And after playing the supposedly drawing move it is obviously junk: 354: Deep Fritz 14  M, Blitz 5m+5s4m+2s r1br4/1p2p1kp/p5p1/3BN1Pn/8/2N4P/PPP2P2/R3K2R w KQ  0 1 Analysis by Deep Fritz 14 x64: 1. + (5.69): 22.000 Ra7 23.h4 Nf4 24.Bb3 Re8 25.Rd4 Nh3 26.f4 Nf2 27.Rg1 Bf5 28.Rg2 Nh3 29.Rgd2 b5 30.Nd5 Rf8 31.Nc6 Rd7 32.Ndxe7 2. + (5.59): 22.h4 Re8 23.000 Nf4 24.Rhe1 Rb8 25.Be4 Rf8 26.Nd7 3. + (5.34): 22.a4 Ra7 23.000 Rf8 24.Rhe1 Bxh3 25.Nd7 Re8 26.Nf6 Nxf6 27.gxf6+ Kxf6 4. + (5.20): 22.Nc4 Nf4 23.000 Rb8 24.Rhe1 e6 25.Bf3 Rxd1+ 26.Rxd1 Nxh3 27.Ne4 b5 28.Ne5 b4 29.Rd8 Rb5 30.Ng4 Bb7 31.Rd7+ Kh8 32.Ngf6 Bxe4 33.Rxh7# 5. + (5.17): 22.Nd3 Rf8 23.000 Ra7 24.Rhe1 b5 25.Be6 Bxe6 26.Rxe6 Nf4 27.Nxf4 Rxf4 28.Rde1 Kf8 29.Ne4 a5 30.Rb6 Rf5 31.Rc6 Rd7 32.Nc5 Rdd5 33.Rc8+ Kf7 6. + (5.09): 22.a3 Ra7 23.000 Rf8 24.Rhe1 b5 25.h4 Rc7 26.Bf3 Nf4 27.Kb1 Bf5 28.Be4 Bxe4 29.Rxe4 e6 30.Red4 Nd5 31.Ne4 7. + (5.03): 22.b3 Rf8 23.000 Ra7 24.Nd3 Nf4 25.Nxf4 Rxf4 26.f3 Rh4 27.Rde1 b5 28.Be6 Bb7 29.Ne4 Ra8 30.Kb2 Rd8 31.Rhg1 b4 8. + (4.96): 22.Rh2 Nf4 23.000 a5 24.Rd4 Nxh3 25.Rh4 Nxg5 26.f4 Nh3 27.Bg2 g5 28.R4xh3 Bxh3 29.Rxh3 gxf4 30.Bxb7 Rab8 31.Be4 h6 32.Nc6 (, 21.07.2017) Some engines can see as far as 000 from the ply before but the ones that fail still seem convinced it is a draw with no analysis given. I'm running under Deep fritz 14 user interface but most of the engines I have tried so far fail on this position. Crafty x32 a notable exception. Can anyone shed any light on what is going on? Is there anyone still reading these groups?  it's awfully quiet...  Regards, Martin Brown 
#2




Engine puzzle position
On 21/07/2017 17:14, Martin Brown wrote:
I have stumbled upon a position where quite a few engines  I thought at first it was just x64 types but Fritz 10 also sees the same phantom "Good move" which evaluates as a draw where none exists. The position is hopelessly lost but that isn't the point. Something is wrong. I found it analysing a lost game where there was a clear and bizarre discontinuity in the evaluation function that made no sense to me. Crafty x64 23.05 also sees the wrong right move but x32 20.14 is OK. r2r4/1p2p1kp/p5p1/3BNbPn/8/2N4P/PPP2P2/R3K2R b KQ  0 21 It is a close run thing between 22. ... Rac8, or 22 ... Nf4 but the phantom move which evaluates as a draw is 22. ... Bc8 If you actually play this move Bc8 then the evaluation becomes correct again with white to move. There is something odd about this position that breaks some engines. Any ideas what it might be? My guess is that it is some sort of draw by repetition bug common to several of the engines since the offending thread displays as just a single move with no continuation line : [...] You didn't post the game, so I have to ask: is there a position repetition after 21...Bc8? If so, a zero score would be acceptable. Cheers, Rainer 
#3




Engine puzzle position
On 21/07/2017 21:23, Rainer wrote:
On 21/07/2017 17:14, Martin Brown wrote: I have stumbled upon a position where quite a few engines  I thought at first it was just x64 types but Fritz 10 also sees the same phantom "Good move" which evaluates as a draw where none exists. The position is hopelessly lost but that isn't the point. Something is wrong. I found it analysing a lost game where there was a clear and bizarre discontinuity in the evaluation function that made no sense to me. Crafty x64 23.05 also sees the wrong right move but x32 20.14 is OK. r2r4/1p2p1kp/p5p1/3BNbPn/8/2N4P/PPP2P2/R3K2R b KQ  0 21 It is a close run thing between 22. ... Rac8, or 22 ... Nf4 but the phantom move which evaluates as a draw is 22. ... Bc8 If you actually play this move Bc8 then the evaluation becomes correct again with white to move. There is something odd about this position that breaks some engines. Any ideas what it might be? My guess is that it is some sort of draw by repetition bug common to several of the engines since the offending thread displays as just a single move with no continuation line : [...] You didn't post the game, so I have to ask: is there a position repetition after 21...Bc8? If so, a zero score would be acceptable. The material balance is such that the winning side should not be looking for a draw and the losing side has no way to force perpetual check. The score for the move 22. ... Bc8 is totally anomolous (as is the fact that it does not have any continuation line in the annotation). It might be relevant that the strongest continuation for white is 000 and there may be a bug in the logic around queens side castling somewhere. Almost none of the engines see anything other than "best" is 21. ... Bc8 0.0 (and see no further than that) This is the result of applying blundercheck threshold 20cp 4s to the entire game (opening blunder in blitz) but played on in the hope of winning something back. [Event "Blitz 5m+5s4m+2s"] [Site "?"] [Date "2017.07.21"] [Round "?"] [White "Deep Fritz 14"] [Black "Martin"] [Result "10"] [ECO "B08"] [WhiteElo "1812"] [BlackElo "1815"] [PlyCount "75"] [TimeControl "240+2"] 1. d4 {0} d6 {4} 2. e4 {0} Nf6 {2} 3. Bd3 {0} g6 {2} 4. Nf3 {0} Bg7 {1} 5. h3 {0.23/20 0 last book move} Nbd7 {0.20/18 2} 6. Nc3 {0.20/21 0} OO {0.20/20 1} 7. e5 {0.20/20 0} dxe5 {0.47/21 6} ({Deep Fritz 14 x64:} 7... Ne8 8. Qe2 c5 9. Bg5 cxd4 10. exd6 Nxd6 11. Bxe7 Qb6 12. Bxf8 Nxf8 13. Ne4 Nxe4 14. Bxe4 Qb4+15. Nd2 Qxb2 16. OO Ne6 17. Rab1 Qxa2 18. Bxb7 Bxb7 19. Rxb7 Qxc2 20. Qf3 Rf821. Rd1 a5 {0.20/20}) 8. dxe5 {0.38/21 0} Nh5 {1.02/18 8} ({Deep Fritz 14 x64:} 8... Ne8 9. Bc4 {0.38/21}) 9. g4 {0.89/18 0} Nf4 {3.61/20 7} ({Deep Fritz 14 x64:} 9... Nxe5 10. gxh5 Nxd3+ 11. Qxd3 Bf5 12. Qe2 Qd6 13. OO Bxh3 14. Re1 Bxc3 15. bxc3 Qd5 16. h6 Rfd8 17. Rb1 b6 18. c4 Qf5 19. Qe5 Qxe5 20. Nxe5 Bf5 21. Nc6 {0.89/18}) 10. Bxf4 {3.50/19 0} f6 {4.26/19 23} ({Deep Fritz 14 x64:} 10... Nc5 11. Be2 Ne6 12. Be3 Bd7 13. Qd2 Bc6 14. OOO Qxd2+ 15. Bxd2 Rad8 16. Be3 b6 17. Rxd8 Rxd8 18. Rd1 Rxd1+ 19. Bxd1 Bxf3 20. Bxf3 Bxe5 21. Kd2 Bf4 22. Nb5 Bxe3+ 23. fxe3 a6 {3.50/19}) 11. exf6 {4.16/18 0} Nxf6 {5.42/22 2} ({Deep Fritz 14 x64:} 11... exf6 12. Bc4+ Kh8 13. OO f5 14. g5 Qe7 15. Bb3 Nc5 16. Nd5 Qe4 17. Re1 {4.16/18}) 12. Bc4+ {5.42/22 0} Kh8 {5.42/21 2} 13. Qxd8 {5.42/21 0} Rxd8 {5.29/20 3 Deep Fritz 14 doubles} 14. Bxc7 {5.27/19 0} Rd7 {6.15/20 6} 15. Be5 {5.94/19 0} a6 {6.29/19 10} 16. g5 {5.44/19 0} Nh5 {5.49/20 4} 17. Be6 {5.20/22 0} Rd8 {5.20/21 5} 18. Bd5 {5.12/19 0} Bxe5 {5.65/21 34} 19. Nxe5{5.47/19 0} Kg7 {5.70/19 19} 20. Bb3 {4.74/19 0} Bf5 {4.74/18 16} 21. Bd5 {0.00/28 0} ({Deep Fritz 14 x64:} 21. Ne2 Rd6 {4.74/18}) 21... Rab8 {5.18/19 6} ({Deep Fritz 14 x64:} 21... Bc8 {0.00/28}) 22. a4 {3.15/19 0} ({Deep Fritz 14 x64:} 22. OOO Nf4 23. h4 Rbc8 24. Bf3 Rd6 25. Rhe1 Rxd1+ 26. Rxd1 Rc7 27. Rd4 Nh3 28. Be4 Nxf2 29. Bxf5 gxf5 30. Rf4 Rxc3 31. bxc3 {5.18/19}) 22... e6 {4.24/22 8} ({Deep Fritz 14 x64:} 22... Bxc2 23. Bf3 Nf4 24. Ne2 Nd3+ 25. Nxd3 Bxd3 26. Rc1 b5 27. a5 Bc4 28. Rc3 Rbc8 29. b3 Be6 30. Re3 Rd6 31. Nf4 Rc1+ 32. Ke2 Rc2+ 33. Kf1 {3.15/19}) 23. Be4 {4.24/21 0} Nf4 {4.24/19 21} 24. Rc1 {3.96/18 0} Rd4 {4.59/22 13} 25. Bxf5 {4.50/21 0} exf5 {4.50/21 3} 26. Rd1 {4.48/19 0} Rb4 {5.68/19 13} ({Deep Fritz 14 x64:} 26... Rdd8 27. Nd3 Ne6 28. h4 Nd4 29.Kd2 b5 30. axb5 axb5 31. Ne2 Ra8 32. Nxd4 Rxd4 33. Ra1 {4.48/19}) 27. Rd7+ {5.68/19 0} Kg8 {5.59/17 7} 28. Nd5 {4.10/18 0} ({Deep Fritz 14 x64:} 28. Kd2 Rxb2 29. Re1 Nxh3 30. Nd3 Nxg5 31. Nxb2 Nf3+ 32. Ke2 Nxe1 33. Kxe1 Re8+ 34. Kf1 b6 35. Nd5 Re6 36. Nd3 b5 37. Nc5 {5.59/17}) 28... Re4+ {5.45/21 9} ({Deep Fritz 14 x64:} 28... Nxd5 29. Rxd5 Rxb2 30. Nd7 Re8+ 31. Kd2 Rc8 32. Nf6+ Kh8 33. Rc1 {4.10/18}) 29. Kf1 {5.28/18 0} Rxe5 {6.33/20 8} 30. Nxf4 {6.14/19 0} Re4 {7.97/20 10} ({Deep Fritz 14 x64:} 30... b5 31. Nd3 Re4 32. axb5 axb5 33. Kg2 Rbe8 34. Ra1 R8e7 35. Ra8+ Kg7 36. Rxe7+ Rxe7 37. Rb8 Re2 38. Rb7+ Kg8 39. Nb4 Re5 40. Kf3 Rc5 41. c3 Re5 42. Nc6 Re6 43. Nd4 {6.14/19}) 31. Nd5 {7.77/20 0} Rxa4 {13.94/20 3} ({Deep Fritz 14 x64:} 31... Kf8 32. Kg2 {7.77/20}) 32. Nf6+ {13.53/20 0} Kf8 {12.20/18 5} 33. Rxh7 {8.61/18 0} ({Deep Fritz 14 x64:} 33. Nxh7+ Kg8 34. Nf6+ Kf8 35. Kg2 Rc8 36. Re1 Re4 37. Nxe4 fxe4 38. Rxb7 Rc4 39. Rd1 Rc8 40. Rd6 e3 41. fxe3 a5 42. Rf6+ Kg8 43. Rxg6+ Kf8 44. Rf6+ Kg8 45. c3 Rc5 46. h4 Rd5 47. Ra6 Rd2+ 48. Kf3 {12.20/18}) 33... Rd8 {12.54/21 13} ({Deep Fritz 14 x64:} 33... Ra1+ 34. Kg2 Rxh1 35. Rh8+ Kf7 36. Rxb8 Ra1 37. Rxb7+ Ke6 38. Rb6+ Kf7 39. c4 a5 40. c5 Rc1 41. c6 a4 42. Nd5 Ke6 43. Nf4+ Kd6 44. Nxg6 Kc5 45. Rb8 Kd6 46. h4 Rxc6 47. h5 Rc4 48. Rd8+ Ke6 49. Re8+ {8.61/18}) 34. Kg2 {11.99/19 0} Rd2 {17.14/23 10} ({Deep Fritz 14 x64:} 34... Rd6 35. Re1 Re4 36.Nxe4 fxe4 37. Rxb7 Rd4 38. Rb6 Kg7 39. Rxa6 Rd5 40. Ra7+ Kf8 41. Rxe4 Rxg5+ 42.Rg4 Rb5 43. Rc4 Rb8 44. Rcc7 Kg8 45. b3 Rd8 46. h4 Kf8 47. Ra6 Ke8 48. Rxg6 Kf8 49. h5 Rd2 50. Rgg7 {11.99/19}) 35. Re1 {17.05/21 0} Re4 {15.24/23 7} 36. Nxe4 {14.68/22 0} fxe4 {14.28/21 2} 37. Rxb7 {12.59/18 0} e3 {21.93/23 6} ({Deep Fritz 14 x64:} 37... Rd4 38. Rb6 Kf7 39. Rf6+ Kg7 40. Rxa6 Rd5 41. Rxe4 Rb5 42. Ra7+ Kf8 43. b3 Rxg5+ 44. Rg4 Rd5 45. Rxg6 Rd8 46. c4 Rb8 47. Rf6+ Kg8 48. Rd6 Kf8 49. c5 Rc8 50. c6 Ke8 51. Rg6 {12.59/18}) 38. Rxe3 {19.94/20 0} 10 Just in case it is relevant Deep Fritz is running on a 4 core i7 3770K but the other failing chess engines are mostly single core ones.  Regards, Martin Brown 
#4




Engine puzzle position
On 22/07/2017 10:16, Martin Brown wrote:
On 21/07/2017 21:23, Rainer wrote: On 21/07/2017 17:14, Martin Brown wrote: I have stumbled upon a position where quite a few engines  I thought at first it was just x64 types but Fritz 10 also sees the same phantom "Good move" which evaluates as a draw where none exists. The position is hopelessly lost but that isn't the point. Something is wrong. I found it analysing a lost game where there was a clear and bizarre discontinuity in the evaluation function that made no sense to me. Crafty x64 23.05 also sees the wrong right move but x32 20.14 is OK. r2r4/1p2p1kp/p5p1/3BNbPn/8/2N4P/PPP2P2/R3K2R b KQ  0 21 It is a close run thing between 22. ... Rac8, or 22 ... Nf4 but the phantom move which evaluates as a draw is 22. ... Bc8 If you actually play this move Bc8 then the evaluation becomes correct again with white to move. There is something odd about this position that breaks some engines. Any ideas what it might be? My guess is that it is some sort of draw by repetition bug common to several of the engines since the offending thread displays as just a single move with no continuation line : [...] You didn't post the game, so I have to ask: is there a position repetition after 21...Bc8? If so, a zero score would be acceptable. The material balance is such that the winning side should not be looking for a draw and the losing side has no way to force perpetual check. The score for the move 22. ... Bc8 is totally anomolous (as is the fact that it does not have any continuation line in the annotation). It might be relevant that the strongest continuation for white is 000 and there may be a bug in the logic around queens side castling somewhere. Almost none of the engines see anything other than "best" is 21. ... Bc8 0.0 (and see no further than that) This is the result of applying blundercheck threshold 20cp 4s to the entire game (opening blunder in blitz) but played on in the hope of winning something back. [...] [game score weeded] [Event "Blitz 5m+5s4m+2s"] [Site "?"] [Date "2017.07.21"] [Round "?"] [White "Deep Fritz 14"] [Black "Martin"] [Result "10"] [ECO "B08"] [WhiteElo "1812"] [BlackElo "1815"] [PlyCount "75"] [EventDate "2017.??.??"] [EventType "blitz"] 1. d4 d6 2. e4 Nf6 3. Bd3 g6 4. Nf3 Bg7 5. h3 Nbd7 6. Nc3 OO 7. e5 dxe5 8. dxe5 Nh5 9. g4 Nf4 10. Bxf4 f6 11. exf6 Nxf6 12. Bc4+ Kh8 13. Qxd8 Rxd8 14. Bxc7 Rd7 15. Be5 a6 16. g5 Nh5 17. Be6 Rd8 18. Bd5 Bxe5 19. Nxe5 Kg7 {1st occurrence} 20. Bb3 Bf5 21. Bd5 Rab8 (21... Bc8 {position repetition}) 22. a4 e6 23. Be4 Nf4 24. Rc1 Rd4 25. Bxf5 exf5 26. Rd1 Rb4 27. Rd7+ Kg8 28. Nd5 Re4+ 29. Kf1 Rxe5 30. Nxf4 Re4 31. Nd5 Rxa4 32. Nf6+ Kf8 33. Rxh7 Rd8 34. Kg2 Rd2 35. Re1 Re4 36. Nxe4 fxe4 37. Rxb7 e3 38. Rxe3 10 It's what I suspected: 21...Bc8 repeats the position, hence the draw score. UCI engines are always fed with the full game history, so they can detect the repetition. NB: there is a decadesold controversy whether it's ok to return a draw score after the first position repetition, or if this should be done only after the second repetition. One can argue either way, I'll let that aside. Cheeers, Rainer 
Thread Tools  
Display Modes  


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
Solving Chess!?  [email protected]  rec.games.chess.computer (Computer Chess)  7  January 20th 10 09:21 PM 
Fritz, engine off by default  Paul Stivers[_2_]  rec.games.chess.computer (Computer Chess)  7  February 19th 08 08:54 PM 
Fritz, engine off by default  Paul Stivers[_2_]  rec.games.chess.misc (Chess General)  0  February 18th 08 10:20 PM 
The cheating IBM  Chess One  rec.games.chess.computer (Computer Chess)  100  July 22nd 06 02:52 PM 