Nordlandia wrote: ↑Fri Apr 02, 2021 7:14 pm
It looks like if white times it right, Ke8 and black is in Zugzwang.
So it is. Real problem is, before that works, white Knight at a8 has to be set free with cyclic zugzwang maneuvres of its own, because capturing Bishop f8 after single one free black Bishop is exchanged to soon against this one Knight, there are no more figures to move, if free black pawns at f- und h- line are blocked or taken. Capturing at f8 with white King then finally would end up stalemate. f8- Bishop has to be taken by the white Knight, not by the white King, when there are no more black figures left to move.
This one point having got, I just didn't see the way to set Knight at a8 free without exchanging it against a5- Bishop, shame on me.
After HekaStockfish with its Find Mate got stuck with mating score in Backward at move nr. 18, I changed to SF MOJO (Modification by Jörg Oster), a branch nullmove pruning can be switched off for.
With this one it worked back to 13th move:
N2b1b1k/4p1p1/4PpPp/B1K5/3Q4/8/3n1p1p/r4Nbr w - - 0 1
1. Qxa1 Nb3+ 2. Kc4 Nxa1 3. Kc3 Bxa5+ 4. Kb2 Kg8 5. Kxa1 Kh8 6. Kb2 Kg8 7. Kb3
Kh8 8. Kc4 Kg8 9. Kc5 Kh8 10. Kd5 Kg8 11. Kc6 Kh8 12. Kd7 Kg8 13. Ke8 h5
[d]N3Kbk1/4p1p1/4PpP1/b6p/8/8/5p1p/5Nbr w - - 0 1
Analysis by Stockfish-MOJO 281120:
14.Kd7 Kh8 15.Kc6 h4 16.Kd5 Bd8 17.Ke4 Kg8 18.Kf4 h3 19.Kg3 Ba5 20.Kxh3 Bd8 21.Kg3 Ba5 22.Kf4 Kh8 23.Ke4 Kg8 24.Kd4 Bd8 25.Kc5 Ba5 26.Kc6 Kh8 27.Kd7 Kg8 28.Ke8 f5 29.Kd7 Kh8 30.Kc6 Kg8 31.Kd5 Kh8 32.Ke5 Bc3+ 33.Kxf5 Bd4 34.Ke4 Ba7 35.Nc7 Kg8 36.Na6 Kh8 37.Kd5 Kg8 38.Kc6 Bd4 39.Nb8 Be5 40.Nd7 Bd6 41.Nb6 Be5 42.Kb7 Bd4 43.Nd7 Bc3 44.Kb6 Kh8 45.Nxf8 Kg8 46.Nh7 Bb4 47.Kc7 Bd6+ 48.Kd7 Bb4
+- (#50) Depth: 68/71 00:01:01 2343MN
About 2 to 5 moves to much DTM, I guess (at the end of the output- line 13 missing, so 61 all in all, I'm not sure about Hobacle's 59, maybe there are a few to be spared yet still) but yet rather quick Backward, best engine for keeping mate score in hash so far at my trials with this one position. Classical eval was used, 12x3GHz Xeon X5670 CPU (SSE4.1 popcnt) 32G hash.
Over night mating score got lost, here's the output afterwards at starting position:
]N2b1b1k/4p1p1/4PpPp/B1K5/3Q4/8/3n1p1p/r4Nbr w - - 0 1
Analysis by Stockfish-MOJO 281120:
1.Qxa1 Nb3+ 2.Kb4 Nxa1 3.Kc3 Bxa5+ 4.Kb2 Bd2 5.Kxa1 Bc3+ 6.Kb1 Bd2 7.Nb6 Be3 8.Nd7 Bd2 9.Ka1 Bc1 10.Nb6 Bg5 11.Kb2 Bd2 12.Ka2 Kg8 13.Kb3 Ba5 14.Nd7 f5 15.Kc2 Bb4 16.Kb2 Ba5 17.Nb8 Kh8 18.Kc2 Bc7 19.Nd7 Bd6 20.Kb1 Ba3 21.Ka1 Bc1 22.Ka2 Bg5 23.Kb3 Kg8 24.Nb8 Kh8 25.Kb2 Bd2 26.Kc2 Ba5 27.Nd7 Bb4 28.Kb3 Ba5 29.Ka4 Bc7 30.Ka3 Bg3 31.Ka2 Bf4 32.Kb2 Bg3 33.Kc3 Bf4 34.Nb6 Be3 35.Na8 Ba7 36.Kb4 Bd4 37.Kb3 Be5 38.Kc4 Bb8 39.Kb5 Ba7 40.Nc7
+- (137.72) Depth: 76/80 00:10:41 14601MN
Pity progress to mate isn't really to be seen there.
Anyhow a very fine study of cyclic zugzwang and one of the best april 1st. jokes I ever fell for completely regards
Peter.