Still no way to post a position diagram in this Forum ???!!!

[ernest]

Here is the analysis I have for that position:
[d]4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - bm Bc3+;
4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - acd 46; acs 5484; bm Bc3+; ce 401; pm Bc3+; pv Bc3+ Rxc3 Nc7 Qe2 e8Q+ Qxe8 Nxe8 Rxc4 Kxh3 Rc2 f4 Rxb2 Kg4 Rf2 Kg5 c4 Bxc4 Rxf4;

Hi Dann,

Actually, the moves Bc3+ Rxc3 are obvious, but then the only winning move indeed seems to be Nc7!!!
How did you find it ?
And is there an original link for that Neghina position ?

[zullil]

Here is the analysis I have for that position:
[d]4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - bm Bc3+;
4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - acd 46; acs 5484; bm Bc3+; ce 401; pm Bc3+; pv Bc3+ Rxc3 Nc7 Qe2 e8Q+ Qxe8 Nxe8 Rxc4 Kxh3 Rc2 f4 Rxb2 Kg4 Rf2 Kg5 c4 Bxc4 Rxf4;

Hi Dann,

Actually, the moves Bc3+ Rxc3 are obvious, but then the only winning move indeed seems to be Nc7!!!
How did you find it ?
And is there an original link for that Neghina position ?

His post indicates an engine search to depth 46, taking 5484 seconds. Given that depth, I doubt that the engine has recognized that all other knight moves allow a draw via Qxg3+. In particular, that Ng7 is also inadequate. Interesting position!

[d]4N2k/4PB1p/7P/2p5/2P5/2rq2Pp/1P3P1K/8 w - - 0 2
Cfish-dev with MultiPV = 4:

+6.65 2. Nc7 Qe2 3. e8=R+ Qxe8 4. Nxe8 Rxc4 5. f4 Rc2+ 6. Kxh3 Rxb2 7. Kg4 Rb4 8. Nd6 Rd4 9. Nb5 Rb4 10. Nc3 Rd4 11. Ne2 Rb4 12. Kg5 c4 13. f5 Rb5 14. Bh5 Rb3 15. f6 c3 16. f7 Rb5+ 17. Kh4 Rb8 18. Bg4 c2 19. Bf5 Rd8 20. Kg5 Rb8 21. g4 Rf8 22. Bxc2 Rxf7 23. Nf4 Re7 24. Bf5 Re1 25. Nd5 Rf1 26. Nc3 Ra1 27. Ne4 Ra5 28. Kh5 Ra1 29. Nc5 Rf1 30. Kg5 Rf2 31. Ne6 Kg8 32. Kf6 Kh8 33. Nf8 Rg2 34. Kg5 Re2 35. Nxh7 Re8 36. Nf6 Rf8 37. Nd7 (depth 54, 0:23:07)

0.00 2. Ng7 Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 Rh2+ 5. Kg4 Rh4+ 6. Kf5 Rf4+ 7. Ke5 Re4+ 8. Kd6 Re6+ 9. Kd5 Rd6+ 10. Kxc5 Rc6+ 11. Kd4 Rxc4+ 12. Ke3 Re4+ 13. Kd3 Re3+ 14. Kd4 Re4+ (depth 54, 0:23:07)

0.00 2. Nd6 Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 Rh2+ 5. Kg4 Rh4+ 6. Kf5 Rf4+ 7. Ke6 Rf6+ 8. Kd5 Rxd6+ 9. Kxc5 Rc6+ 10. Kd5 Rc5+ 11. Ke4 Rxc4+ 12. Kf5 Rf4+ 13. Ke6 Re4+ 14. Kf6 Rxe7 15. Kxe7 (depth 54, 0:23:07)

0.00 2. Nf6 Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 Rh2+ 5. Kg4 Rh4+ 6. Kf5 Rf4+ 7. Ke5 Re4+ 8. Kd6 Re6+ 9. Kd5 Rd6+ 10. Ke4 Rd4+ 11. Kf5 (depth 54, 0:23:07)

[peter]

Actually, the moves Bc3+ Rxc3 are obvious, but then the only winning move indeed seems to be Nc7!!!
How did you find it ?
And is there an original link for that Neghina position ?

Mihai didn't answer to my mail till now, maybe he's on vacation, so I just hope it will be ok for him, if I give the .pgn, I got from him back in 2012. Pity I'm not quite sure, which lines I added by myself then, could be one or two at the utmost, so practically it's almost all his commenting and analysing.
I hope Mihai doesn't disagree with this and he'll give a link to the public version soon too.

[pgn] [Event "Mihai Rep"] [Site "?"] [Date "2012.08.22"] [Round "?"] [White "White to move and win"] [Black "?"] [Result "1-0"] [Annotator "Martan,Peter"] [SetUp "1"] [FEN "4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - 0 1"] [PlyCount "23"] [EventDate "2012.??.??"] 1. Bc3+ Rxc3 (1... Qxc3 2. bxc3 Rh1+ (2... Re1 3. Bh5 {winning for White}) ( 2... Rf1 3. Nc7 Rxf2+ 4. Kg1 {winning for White}) (2... Rg1 3. Be6 Rg2+ 4. Kxh3 {winning for White}) 3. Kxh1 h2 4. Bh5 Kg8 5. Nf6+ Kh8 6. e8=R#) 2. Nc7 (2. Nd6 {2. Nf6; 2.Ng7} Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 (4. Kg1 h2+ 5. Kh1 Rc1+ 6. Kxh2 Rh1+ {same as Kxh3 directly}) 4... Rh2+ 5. Kg4 Rh4+ {draw by perpetual check}) (2. bxc3 Qf1 {winning for Black}) (2. Nf6 Qxg3+) (2. Ng7 Qxg3+) 2... Qxg3+ (2... Qe2 3. e8=Q+ {winning for White}) (2... Qe4 3. e8=Q+ {winning for White}) (2... Qd7 3. e8=Q+ Qxe8 4. Bxe8 {winning for White} Rxc4 5. b3 Rc2 6. Kxh3 Rxf2 7. Na6 Rf6 8. Nxc5 Rxh6+ 9. Kg4 Kg7 10. Bb5 Kf6 11. Kf4 Ke7 12. Bc4 Rg6 13. Ne4 h6 14. Ke5 Rb6 15. g4 Rc6 16. g5 h5 17. Ng3 Rb6 18. Nxh5 {L7#19}) 3. fxg3 Rc2+ 4. Kxh3 (4. Kh1 Rc1+ 5. Kh2 Rc2+ {repeating}) (4. Kg1 h2+ 5. Kh1 Rc1+ 6. Kxh2 Rh1+ 7. Kg2 Rg1+ 8. Kf3 Rxg3+ 9. Ke4 Re3+ 10. Kd5 Rxe7 11. Be8 Rxc7 12. Bc6 Rf7 13. Kxc5 Rf5+ 14. Bd5 Rf4 15. Bb7 {also winning for White, but not very different from the mainline}) 4... Rh2+ 5. Kg4 (5. Kxh2 {Patt}) 5... Rh4+ 6. Kf5 (6. Kg5 Rg4+ 7. Kf6 Rf4+ 8. Ke5 Re4+ 9. Kd5 Re5+ 10. Kc6 Re6+ 11. Kb7 Rxe7 12. Be8 Kg8 13. Kb6 Kf8 14. Bc6 Re5 15. Nd5 Re2 16. Kxc5 Rxb2 17. Kd6) 6... Rf4+ (6... Rh5+ 7. Bxh5 Kg8 8. e8=R#) 7. Ke5 Re4+ 8. Kd5 (8. Kd6 Re6+ 9. Kd5 Rd6+ 10. Ke4 (10. Kxc5 Rc6+ {draw}) 10... Re6+ 11. Kf3 Rxe7 {draw}) 8... Re5+ (8... Rxe7 9. Be8 Rxc7 10. Bc6 {winning for White}) 9. Kc6 Re6+ (9... Rxe7 10. Be8 {winning for White}) 10. Kb7 (10. Kxc5 Rc6+ {draw}) 10... Rb6+ (10... Rxe7 11. Be8 {winning for White}) 11. Kc8 Rb8+ 12. Kd7 {#6} 1-0 [/pgn]

[ernest]

Cfish-dev with MultiPV = 4:

+6.65 2. Nc7 Qe2 3. e8=R+ Qxe8 4. Nxe8 Rxc4 5. f4 Rc2+ 6. Kxh3 Rxb2 7. Kg4 Rb4 8. Nd6 Rd4 9. Nb5 Rb4 10. Nc3 Rd4 11. Ne2 Rb4 12. Kg5 c4 13. f5 Rb5 14. Bh5 Rb3 15. f6 c3 16. f7 Rb5+ 17. Kh4 Rb8 18. Bg4 c2 19. Bf5 Rd8 20. Kg5 Rb8 21. g4 Rf8 22. Bxc2 Rxf7 23. Nf4 Re7 24. Bf5 Re1 25. Nd5 Rf1 26. Nc3 Ra1 27. Ne4 Ra5 28. Kh5 Ra1 29. Nc5 Rf1 30. Kg5 Rf2 31. Ne6 Kg8 32. Kf6 Kh8 33. Nf8 Rg2 34. Kg5 Re2 35. Nxh7 Re8 36. Nf6 Rf8 37. Nd7 (depth 54, 0:23:07)

0.00 2. Ng7 Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 Rh2+ 5. Kg4 Rh4+ 6. Kf5 Rf4+ 7. Ke5 Re4+ 8. Kd6 Re6+ 9. Kd5 Rd6+ 10. Kxc5 Rc6+ 11. Kd4 Rxc4+ 12. Ke3 Re4+ 13. Kd3 Re3+ 14. Kd4 Re4+ (depth 54, 0:23:07)

0.00 2. Nd6 Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 Rh2+ 5. Kg4 Rh4+ 6. Kf5 Rf4+ 7. Ke6 Rf6+ 8. Kd5 Rxd6+ 9. Kxc5 Rc6+ 10. Kd5 Rc5+ 11. Ke4 Rxc4+ 12. Kf5 Rf4+ 13. Ke6 Re4+ 14. Kf6 Rxe7 15. Kxe7 (depth 54, 0:23:07)

0.00 2. Nf6 Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 Rh2+ 5. Kg4 Rh4+ 6. Kf5 Rf4+ 7. Ke5 Re4+ 8. Kd6 Re6+ 9. Kd5 Rd6+ 10. Ke4 Rd4+ 11. Kf5 (depth 54, 0:23:07)

Very enlightning, Zullil !
So Nc7 is the only Knight move that does not allow the stalemate chase...

[ernest]

Actually, the moves Bc3+ Rxc3 are obvious, but then the only winning move indeed seems to be Nc7!!!
How did you find it ?
And is there an original link for that Neghina position ?

Mihai didn't answer to my mail till now, maybe he's on vacation, so I just hope it will be ok for him, if I give the .pgn, I got from him back in 2012. Pity I'm not quite sure, which lines I added by myself then, could be one or two at the utmost, so practically it's almost all his commenting and analysing.
I hope Mihai doesn't disagree with this and he'll give a link to the public version soon too.

[pgn] [Event "Mihai Rep"] [Site "?"] [Date "2012.08.22"] [Round "?"] [White "White to move and win"] [Black "?"] [Result "1-0"] [Annotator "Martan,Peter"] [SetUp "1"] [FEN "4N2k/4PB1p/7P/B1p5/2P5/3q2Pp/1P3P1K/2r5 w - - 0 1"] [PlyCount "23"] [EventDate "2012.??.??"] 1. Bc3+ Rxc3 (1... Qxc3 2. bxc3 Rh1+ (2... Re1 3. Bh5 {winning for White}) ( 2... Rf1 3. Nc7 Rxf2+ 4. Kg1 {winning for White}) (2... Rg1 3. Be6 Rg2+ 4. Kxh3 {winning for White}) 3. Kxh1 h2 4. Bh5 Kg8 5. Nf6+ Kh8 6. e8=R#) 2. Nc7 (2. Nd6 {2. Nf6; 2.Ng7} Qxg3+ 3. fxg3 Rc2+ 4. Kxh3 (4. Kg1 h2+ 5. Kh1 Rc1+ 6. Kxh2 Rh1+ {same as Kxh3 directly}) 4... Rh2+ 5. Kg4 Rh4+ {draw by perpetual check}) (2. bxc3 Qf1 {winning for Black}) (2. Nf6 Qxg3+) (2. Ng7 Qxg3+) 2... Qxg3+ (2... Qe2 3. e8=Q+ {winning for White}) (2... Qe4 3. e8=Q+ {winning for White}) (2... Qd7 3. e8=Q+ Qxe8 4. Bxe8 {winning for White} Rxc4 5. b3 Rc2 6. Kxh3 Rxf2 7. Na6 Rf6 8. Nxc5 Rxh6+ 9. Kg4 Kg7 10. Bb5 Kf6 11. Kf4 Ke7 12. Bc4 Rg6 13. Ne4 h6 14. Ke5 Rb6 15. g4 Rc6 16. g5 h5 17. Ng3 Rb6 18. Nxh5 {L7#19}) 3. fxg3 Rc2+ 4. Kxh3 (4. Kh1 Rc1+ 5. Kh2 Rc2+ {repeating}) (4. Kg1 h2+ 5. Kh1 Rc1+ 6. Kxh2 Rh1+ 7. Kg2 Rg1+ 8. Kf3 Rxg3+ 9. Ke4 Re3+ 10. Kd5 Rxe7 11. Be8 Rxc7 12. Bc6 Rf7 13. Kxc5 Rf5+ 14. Bd5 Rf4 15. Bb7 {also winning for White, but not very different from the mainline}) 4... Rh2+ 5. Kg4 (5. Kxh2 {Patt}) 5... Rh4+ 6. Kf5 (6. Kg5 Rg4+ 7. Kf6 Rf4+ 8. Ke5 Re4+ 9. Kd5 Re5+ 10. Kc6 Re6+ 11. Kb7 Rxe7 12. Be8 Kg8 13. Kb6 Kf8 14. Bc6 Re5 15. Nd5 Re2 16. Kxc5 Rxb2 17. Kd6) 6... Rf4+ (6... Rh5+ 7. Bxh5 Kg8 8. e8=R#) 7. Ke5 Re4+ 8. Kd5 (8. Kd6 Re6+ 9. Kd5 Rd6+ 10. Ke4 (10. Kxc5 Rc6+ {draw}) 10... Re6+ 11. Kf3 Rxe7 {draw}) 8... Re5+ (8... Rxe7 9. Be8 Rxc7 10. Bc6 {winning for White}) 9. Kc6 Re6+ (9... Rxe7 10. Be8 {winning for White}) 10. Kb7 (10. Kxc5 Rc6+ {draw}) 10... Rb6+ (10... Rxe7 11. Be8 {winning for White}) 11. Kc8 Rb8+ 12. Kd7 {#6} 1-0 [/pgn]

Thanks a lot, Peter, great PGN, very complete ! and that was in 2012...
I wonder what engines today would find the solution (with high eval) in less than 1 hour (say on a quad-core).
Obviously those engines must be good at pursuing perpetual chess !

[zullil]

Thanks a lot, Peter, great PGN, very complete ! and that was in 2012...
I wonder what engines today would find the solution (with high eval) in less than 1 hour (say on a quad-core).
Obviously those engines must be good at pursuing perpetual chess !

[d]7k/4PBNp/7P/2p5/2P5/2rq2Pp/1P3P1K/8 b - - 1 2
How long for engines to find Qxg3+?