Can any chess engine or mate finder solve the following:

[Dann Corbit]

[d]8/kn4b1/P2B4/8/1Q6/6pP/1q4pP/5BK1 w - - acd 60; acs 285; bm Qxb2; c3 "Bc5+"; ce 224; dm 45;


what does 'c3 "Bc5+"'; on the above EPD code mean?
It seems 1.Bc5+ also wins so it's an alternative best move I think.

Bc5+ was proposed as a solution by at least one author

[zd3nik]

On deeper analysis, position 4 seems to be a forced mate in 68 with best defence from black. {Best defence for black simply meant avoiding any pawn move if possible}.

[Mate in 68]

[pgn]
[Event "Talkchess"]
[Site "Talkchess"]
[Date "2015.06.21"]
[Round "?"]
[White "2700"]
[Black "2700"]
[Result "1-0"]
[BlackElo "2700"]
[Time "02:40:14"]
[WhiteElo "2700"]
[TimeControl "600"]
[SetUp "1"]
[FEN "k1b5/1p1p1p1p/1PpPpP2/2B5/8/2p1p2P/2PbP3/3K4 w - - 0 1"]
[Termination "unterminated"]
[PlyCount "118"]
[WhiteType "human"]
[BlackType "human"]

1. Bd4 Kb8 2. Be5 Ka8 3. Bf4 Kb8 4. Bh6 Ka8 5. Bg7 Kb8 6. Bf8 Ka8 7. Be7
Kb8 8. Bd8 Ka8 9. Bc7 c5 10. Bd8 Kb8 11. Be7 Ka8 12. Bf8 Kb8 13. Bh6 Ka8
14. Bg7 Kb8 15. Bf8 Ka8 16. Be7 Kb8 17. Bd8 Ka8 18. Bc7 c4 19. Bd8 Kb8 20.
Be7 Ka8 21. Bf8 Kb8 22. Bh6 Ka8 23. Bg7 Kb8 24. Bf8 Ka8 25. Be7 Kb8 26. Bd8
Ka8 27. Bc7 e5 28. Bd8 Kb8 29. Be7 Ka8 30. Bf8 Kb8 31. Bh6 Ka8 32. Bg7 Kb8
33. Bf8 Ka8 34. Be7 Kb8 35. Bd8 Ka8 36. Bc7 e4 37. Bd8 Kb8 38. Be7 Ka8 39.
Bf8 Kb8 40. Bh6 Ka8 41. Bg7 Kb8 42. Bf8 Ka8 43. Be7 Kb8 44. Bd8 Ka8 45. Bc7
h5 46. Bd8 Kb8 47. Be7 Ka8 48. Bf8 Kb8 49. Bh6 Ka8 50. Bg7 Kb8 51. Bf8 Ka8
52. Be7 Kb8 53. Bd8 Ka8 54. Bc7 h4 55. Bd8 Kb8 56. Be7 Ka8 57. Bf8 Kb8 58.
Bh6 Ka8 59. Bg5 Be1 60. Kxe1 {Mate in 8}
[/pgn]

If white pushes his white pawn to h4 anytime, it will be a draw because the point of holding h3 pawn is to force black to play his h7 pawn into h4 and that will be taken by white's D.S.Bishop and a winning h-passer for white.


edit: modifying the position without both h-pawn is just draw

This sequence was also how I analyzed this position. Due to the high number of "in between" moves it's no wonder chess engines can't see this one. Add in null move pruning and there's no way!

I find this position to be much more interesting as an engine test position after move 36.

[d] k1b5/1pBp1p1p/1P1P1P2/8/2p1p3/2p1p2P/2PbP3/3K4 w - - 0 37

It's interesting as a test position at this point because Bd8 leads to a mate that is within the search horizon of most engine's. And it's also interesting because h4 is a blunder - one that heuristics could perhaps detect without needing to calculate all variations to full depth.

[MikeGL]

Sorry there is a mistake on my post above, I entered your above EPD into engines and it claims mate in 23, so from move 36 + 23 it is just 59 moves. Therefore the correct solution is just 59 moves.

So position 4 of Dann is [White to play Mate in 59]
Thanks for pointing that out.

Edit: On my old dual core machine, critter1.6(32bit) finds mate in 23 instantly on your EPD, but stockfish6(32bit) seems to have difficulty on this dual-core, searching for h4 only.

[pgn]
[Event "Talkchess"]
[Site "Talkchess"]
[Date "2015.06.21"]
[Round "?"]
[White "2700"]
[Black "2700"]
[Result "1-0"]
[BlackElo "2400"]
[Time "02:40:14"]
[WhiteElo "2400"]
[TimeControl "600"]
[SetUp "1"]
[FEN "k1b5/1p1p1p1p/1PpPpP2/2B5/8/2p1p2P/2PbP3/3K4 w - - 0 1"]
[Termination "unterminated"]
[PlyCount "118"]
[WhiteType "human"]
[BlackType "human"]

1. Bd4 Kb8 2. Be5 Ka8 3. Bf4 Kb8 4. Bh6 Ka8 5. Bg7 Kb8 6. Bf8 Ka8 7. Be7
Kb8 8. Bd8 Ka8 9. Bc7 c5 10. Bd8 Kb8 11. Be7 Ka8 12. Bf8 Kb8 13. Bh6 Ka8
14. Bg7 Kb8 15. Bf8 Ka8 16. Be7 Kb8 17. Bd8 Ka8 18. Bc7 c4 19. Bd8 Kb8 20.
Be7 Ka8 21. Bf8 Kb8 22. Bh6 Ka8 23. Bg7 Kb8 24. Bf8 Ka8 25. Be7 Kb8 26. Bd8
Ka8 27. Bc7 e5 28. Bd8 Kb8 29. Be7 Ka8 30. Bf8 Kb8 31. Bh6 Ka8 32. Bg7 Kb8
33. Bf8 Ka8 34. Be7 Kb8 35. Bd8 Ka8 36. Bc7 e4 37. Bd8 Kb8 38. Be7 h6 39. Bf8 h5 40. Bh6 Ka8 41. Bg7 Kb8 42. Bf8 Ka8 43.
Be7 Kb8 44. Bd8 Ka8 45. Bc7 h4 46. Bd8 Kb8 47. Be7 Ka8 48. Bf8 Kb8 49. Bh6 Ka8 {{Mate in 10}}
[/pgn]

[MikeGL]

Embarassing flawed analysis.
Actually it's just Mate in 57, due to improvement of h3 up to h5 push of white, which reduces the distance to mate.

[Mate in 57]

[pgn]
[Event "Talkchess"]
[Site "Talkchess"]
[Date "2015.06.21"]
[Round "?"]
[White "2700"]
[Black "2700"]
[Result "1-0"]
[BlackElo "2400"]
[Time "02:40:14"]
[WhiteElo "2400"]
[TimeControl "600"]
[SetUp "1"]
[FEN "k1b5/1p1p1p1p/1PpPpP2/2B5/8/2p1p2P/2PbP3/3K4 w - - 0 1"]
[Termination "normal"]
[PlyCount "93"]
[WhiteType "human"]
[BlackType "human"]

1. Bd4 Kb8 2. Be5 Ka8 3. Bf4 Kb8 4. Bh6 Ka8 5. h4 Kb8 6. h5 Ka8 7. Bg7 Kb8
8. Bf8 Ka8 9. Be7 Kb8 10. Bd8 Ka8 11. Bc7 c5 12. Bd8 Kb8 13. Be7 Ka8 14.
Bf8 Kb8 15. Bh6 Ka8 16. Bg7 Kb8 17. Bf8 Ka8 18. Be7 Kb8 19. Bd8 Ka8 20. Bc7
c4 21. Bd8 Kb8 22. Be7 Ka8 23. Bf8 Kb8 24. Bh6 Ka8 25. Bg7 Kb8 26. Bf8 Ka8
27. Be7 Kb8 28. Bd8 Ka8 29. Bc7 e5 30. Bd8 Kb8 31. Be7 Ka8 32. Bf8 Kb8 33.
Bh6 Ka8 34. Bg7 Kb8 35. Bf8 Ka8 36. Be7 Kb8 37. Bd8 Ka8 38. Bc7 e4 39. Bd8
Kb8 40. Be7 Ka8 41. Bf8 Kb8 42. Bh6 Ka8 43. Bg7 Kb8 44. Bf8 Ka8 45. Be7 Kb8
46. Bd8 Ka8 47. Bc7 {Mate in 10} 1-0
[/pgn]

Sorry, the editing window (15 min) already lapsed, so I can't edit my previous post.

What's funnier is, it was included in EPD code of Dann at page 3 of this thread. "dm 56" (probably meant 'Distance to Mate 56 + bm Bd4 = 57) so Mate in 57.

Are there written rules for FEN code on those acronyms like 'bm', 'am', 'c3', 'dm' at the end of FEN code?

[MikeGL]

Nevermind my last sentence question above.
It is found at https://chessprogramming.wikispaces.com ... escription

Opcode mnemonics
acn analysis count nodes
acs analysis count seconds
am avoid move(s)
bm best move(s)
c0 comment (primary, also c1 though c9)
ce centipawn evaluation
dm direct mate fullmove count
draw_accept accept a draw offer
draw_claim claim a draw
draw_offer offer a draw
draw_reject reject a draw offer
eco Encyclopedia of Chess Openings opening code
fmvn fullmove number
hmvc halfmove clock
id position identification
nic _New In Chess_ opening code
noop no operation
pm predicted move
pv predicted variation
rc repetition count
resign game resignation
sm supplied move
tcgs telecommunication game selector
tcri telecommunication receiver identification
tcsi telecommunication sender identification
v0 variation name (primary, also v1 though v9)[/url]

[zullil]

I find this position to be much more interesting as an engine test position after move 36.

[d] k1b5/1pBp1p1p/1P1P1P2/8/2p1p3/2p1p2P/2PbP3/3K4 w - - 0 37

It's interesting as a test position at this point because Bd8 leads to a mate that is within the search horizon of most engines. And it's also interesting because h4 is a blunder - one that heuristics could perhaps detect without needing to calculate all variations to full depth.

Latest standard Stockfish cannot solve this. However, with null move reduction disabled:

Code: Select all

info depth 51 seldepth 50 multipv 1 score mate 23 nodes 2044894 nps 374934 hashfull 0 tbhits 0 time 5454 pv c7d8 a8b8 d8e7 h7h6 e7f8 h6h5 f8h6 b8a8 h6g7 a8b8 g7f8 b8a8 f8e7 a8b8 e7d8 b8a8 d8c7 h5h4 c7d8 a8b8 d8e7 b8a8 e7f8 a8b8 f8h6 b8a8 h6g5 a8b8 g5h4 b8a8 h4g3 a8b8 h3h4 b8a8 h4h5 a8b8 h5h6 b8a8 h6h7 a8b8 h7h8q d2c1 d1c1 b8a8 h8c8

[Paloma]

It's a mate in 56 according to ChestUCI V5.2

[pgn]
[Event "?"]
[Site "?"]
[Date "2015.06.22"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "k1b5/1p1p1p1p/1PpPpP2/2B5/8/2p1p2P/2PbP3/3K4 w - -"]

1. Bd4 h6 ( {Or} 1... Kb8 2. Be5 h6 3. Bf4 h5 4. Bg5 Ka8 5. Bh6
Kb8 6. Bf8 Ka8 7. Be7 Kb8 8. Bd8 Ka8 9. Bc7 e5 10. Bd8 Kb8
11. Be7 Ka8 12. Bf8 Kb8 13. Bg7 Ka8 14. Bh6 Kb8 15. Bg5 Ka8
16. Bh4 Kb8 17. Bg3 Ka8 18. Bxe5 Kb8 19. Bf4 Ka8 20. Bh6
Kb8 21. Bf8 Ka8 22. Be7 Kb8 23. Bd8 Ka8 24. Bc7 c5 25. Bd8
Kb8 26. Be7 Ka8 27. Bf8 Kb8 28. Bg7 Ka8 29. Bh6 Kb8 30. Bf8
Ka8 31. Be7 Kb8 32. Bd8 Ka8 33. Bc7 c4 34. Bd8 Kb8 35. Be7
Ka8 36. Bf8 Kb8 37. Bg7 Ka8 38. Bh6 Kb8 39. Bf8 Bc1
40. Kxc1 Ka8 41. Be7 Kb8 42. Bd8 Ka8 43. Bc7 h4 44. Bd8 Kb8
45. Be7 Ka8 46. Bf8 Kb8 47. Bh6 Ka8 48. Bg5 Kb8 49. Bxh4
Ka8 50. Bg3 Kb8 51. h4 Ka8 52. h5 Kb8 53. h6 Ka8 54. h7 Kb8
55. h8=Q Ka8 56. Qxc8#) 2. Be5 h5 3. Bf4 Kb8 4. Bg5 Ka8
5. Bh6 Kb8 6. Bf8 Ka8 7. Be7 Kb8 8. Bd8 Ka8 9. Bc7 e5
10. Bd8 Kb8 11. Be7 Ka8 12. Bf8 Kb8 13. Bg7 Ka8 14. Bh6 Kb8
15. Bg5 Ka8 16. Bh4 Kb8 17. Bg3 Ka8 18. Bxe5 Kb8 19. Bf4
Ka8 20. Bh6 Kb8 21. Bf8 Ka8 22. Be7 Kb8 23. Bd8 Ka8 24. Bc7
c5 25. Bd8 Kb8 26. Be7 Ka8 27. Bf8 Kb8 28. Bg7 Ka8 29. Bh6
Kb8 30. Bf8 Ka8 31. Be7 Kb8 32. Bd8 Ka8 33. Bc7 c4 34. Bd8
Kb8 35. Be7 Ka8 36. Bf8 Kb8 37. Bg7 Ka8 38. Bh6 Kb8 39. Bf8
Bc1 40. Kxc1 Ka8 41. Be7 Kb8 42. Bd8 Ka8 43. Bc7 h4 44. Bd8
Kb8 45. Be7 Ka8 46. Bf8 Kb8 47. Bh6 Ka8 48. Bg5 Kb8
49. Bxh4 Ka8 50. Bg3 Kb8 51. h4 Ka8 52. h5 Kb8 53. h6 Ka8
54. h7 Kb8 55. h8=Q Ka8 56. Qxc8#
[/pgn]

After 1. Bd4 Kb8 2. Be5 is 2... h6 the only move, 2... Ka8 leads to earlier mate
[pgn]
[Event "?"]
[Site "?"]
[Date "2015.06.23"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[SetUp "1"]
[FEN "k1b5/1p1p1p1p/1PpPpP2/2B5/8/2p1p2P/2PbP3/3K4 w - -"]

1. Bd4 Kb8 2. Be5 Ka8 3. Bf4 h5 4. Bh6
Kb8 5. Bf8 Ka8 6. Be7 Kb8 7. Bd8 Ka8 8. Bc7 e5 9. Bd8 Kb8
10. Be7 Ka8 11. Bf8 Kb8 12. Bg7 Ka8 13. Bh6 Kb8 14. Bg5 Ka8
15. Bh4 Kb8 16. Bg3 Ka8 17. Bxe5 Kb8 18. Bf4 Ka8 19. Bh6
Kb8 20. Bf8 Ka8 21. Be7 Kb8 22. Bd8 Ka8 23. Bc7 c5 24. Bd8
Kb8 25. Be7 Ka8 26. Bf8 Kb8 27. Bg7 Ka8 28. Bh6 Kb8 29. Bf8
Ka8 30. Be7 Kb8 31. Bd8 Ka8 32. Bc7 c4 33. Bd8 Kb8 34. Be7
Ka8 35. Bf8 Kb8 36. Bg7 Ka8 37. Bh6 Kb8 38. Bf8 Bc1
39. Kxc1 Ka8 40. Be7 Kb8 41. Bd8 Ka8 42. Bc7 h4 43. Bd8 Kb8
44. Be7 Ka8 45. Bf8 Kb8 46. Bh6 Ka8 47. Bg5 Kb8 48. Bxh4
Ka8 49. Bg3 Kb8 50. h4 Ka8 51. h5 Kb8 52. h6 Ka8 53. h7 Kb8
54. h8=Q Ka8 55. Qxc8#[/pgn]

[MikeGL]

Thanks for the clarification.
Just curious on how many minutes did ChestUCI V5.2 solved this mate in 56.
Would be interesting to have the settings/flags for ChestUCI you used.

[Paloma]

Solution time (on 2.8GHz): 2:59

C22/R0/K2/P5/X7

Position is on the latest ChestUCI Database @ http://fhub.jimdo.com/

[MikeGL]

thanks for the settings and URL link