Variation on the Mate in 74

[Eelco de Groot]

See the post from Andrew Case http://talkchess.com/forum3/viewtopic.php?f=2&t=83218 for the original, this one appears a bit shorter but not easy for the Huntsman. Stockfish is probably fast on a big rig, I tried both Crystal 6 and Hutsman together. Crystal finds larger scores (like Stock I suppose) but not yet Mate until much farther down. But I only used 512 MB hash for both engines.

I think it is (about) Mate in 44 but could be shorter

[d]qn6/b1Kp2p1/p1pB4/p7/P1NkP3/2pP4/2B4p/8 w - -

[Eelco de Groot]

My result so far:


With Crystal I got this PV:

[pgn] [Event "?"] [Site "?"] [Date "2024.01.21"] [Round "?"] [White "?"] [Black "?"] [Result "*"] [SetUp "1"] [FEN "qn6/b1Kp2p1/p1pB4/p7/P1NkP3/2pP4/2B4p/8 w - -"] 1. Bxh2 Kc5 2. Bd6+ Kd4 3. Kc8 Bb6 4. Bh2 Kc5 5. Bd6+ Kd4 6. Bf4 Kc5 7. Be3+ Kb4 $8 8. Nxb6 Qa7 9. Nc4 Qa8 10. Kc7 d5 11. Nb6 Qa7+ * [/pgn]

and here Crystal gives

[d]1n6/q1K3p1/pNp5/p2p4/Pk2P3/2pPB3/2B5/8 w - - Engine: Crystal 6 PMT (512 MB)
gemaakt door the Stockfish developers (see AUTHORS f



34 44:17 +M32 12.Kd6 Ka3 13.Pc4+ dxc4 14.Lxa7 Kb2 15.Ld1 cxd3 16.Le3 d2 17.e5 Pd7 18.Kxd7 c5 19.e6 c2 20.Lxc2 Kxc2 21.Lxd2 Kxd2 22.e7 c4 23.e8D c3 24.Dd8 c2 25.Dxa5+ (3.283.282.834) 1235
34 44:17 -9.11 12.Kc8 d4 13.Lxd4 De7 14.Le3 De8+ 15.Kb7 Pd7 16.Kc7 Pc5 17.Lxc5+ Kxc5 18.Pd7+ Kd4 19.Pb8 Ke3 20.d4 Df7+ 21.Pd7 Da2 22.Ld1 Db1 23.e5 Dxd1 24.e6 c2 25.e7 (3.283.282.834) 1235
34 44:17 -17.55 12.Kd8 d4 13.Lf4 Dxb6+ 14.Kc8 c5 15.Ld1 Ka3 16.Lc1+ Ka2 17.Lg4 Pc6 18.Le6+ Kb1 19.Lf4 c2 20.Ld5 Kb2 21.Kd7 Pb4 22.Lg5 Pxd3 23.Ke7 c4 24.Kf7 c3 25.Kg8 (3.283.282.834) 1235

It is not certain 1.Bxh2 is shortest, if not it is an Avoid Move (well, sort of anyway for a mate problem). But other moves (Crystal starts with 1.Bf4 for instance) are probably only marginally better if at all. It is also easy to lose the M32, now Crystal thinks 12.Kd6 is M37 when it had to start again because I needed the pgn.

[peter]

See the post from Andrew Case http://talkchess.com/forum3/viewtopic.php?f=2&t=83218 for the original, this one appears a bit shorter but not easy for the Huntsman. Stockfish is probably fast on a big rig, I tried both Crystal 6 and Hutsman together. Crystal finds larger scores (like Stock I suppose) but not yet Mate until much farther down. But I only used 512 MB hash for both engines.

I think it is (about) Mate in 44 but could be shorter

[d]qn6/b1Kp2p1/p1pB4/p7/P1NkP3/2pP4/2B4p/8 w - -

Don't have much hardware- time free at the moment neither, so just after a short Forward- Backward with Huntsman:

qn6/b1Kp2p1/p1pB4/p7/P1NkP3/2pP4/2B4p/8 w - - 0 1

Analysis by The Huntsman 1:

1.Lxh2 Kc5 2.Ld6+ Kd4 3.Kc8 g6 4.Lf4 Kc5 5.Le3+ Kb4 6.Lc1 Kc5 7.La3+ Kd4 8.Ld6 Lb6 9.Lf4 Kc5 10.Le3+ Kb4 11.Sxb6 Da7 12.Sc4 Da8 13.Kc7 d5 14.Sb6 Da7+ 15.Kd6 d4 16.Lxd4 Sd7 17.Sd5+ cxd5 18.Lxa7 dxe4 19.dxe4 Sf6 20.e5 Se8+ 21.Kd7 Sg7 22.Le3 Ka3 23.Lc1+ Kb4 24.e6 Sxe6 25.Kxe6 Kc5 26.Lxg6 c2 27.Le3+ Kb4 28.Lxc2 Kc3 29.Le4 Kb4 30.Kd5 Kc3 31.Lf5 Kb3 32.Ld2 Kxa4 33.Kc4 Ka3 34.Lc1+ Ka2 35.Kc3 a4 36.Kc2 a3 37.Le6+ Ka1 38.Lxa3 a5 39.Lb2#
Tiefe: 75/78 00:00:06 331MN

1.Bf4 I also tried but didn't find other ways than to return to Bxh2 one or more moves later on transposing to 1.Bxh2 again.
Yet there could be detours to be found stll with more hardware- time, regards

[Eelco de Groot]

Thanks Peter, yes you are probably right that it is only a transposition if you don't take immediately on h2. But I have not checked, with one thread it is too deep.

Crystal thinks it can shorten your variation but I'm not sure it is correct, I get two different variations.

[d]8/B2n4/p2K2p1/p7/Pk2P3/2p5/2B5/8 b - -
55/38 5:01 +M19 19...Pf8 20.Lc5+ Kc4 21.Le3 g5
22.Lxg5 Kd4 23.e5 Pd7 24.Kxd7 Kxe5
25.Ke7 Kd5 26.Lf4 Kc5 27.Ke6 Kd4
28.Kd6 Kc4 29.Le3 Kb4 30.Kd5 Ka3
31.Kd4 Kb2 32.Kd3 Ka2 (438.068.937) 1454

and one move earlier

[d]8/B2n4/p2K2p1/p2p4/Pk2P3/2pP4/2B5/8 b - - Engine: Crystal 6 PMT (512 MB)
gemaakt door the Stockfish developers (see AUTHORS f



61/48 22:23 +M19 18...dxe4 19.dxe4 Pf8 20.Lc5+ Kc4
21.Le3 Kb4 22.Ke7 Ka3 23.Lc1+ Ka2
24.Kxf8 g5 25.Lxg5 Kb2 26.e5 Kxc2
27.e6 Kb3 28.e7 Kxa4 29.e8D+ Kb3
30.Lf6 a4 31.Dc6 Kc2 (2.052.914.402) 1528

where Huntsman still gives

[d]8/B2n4/p2K2p1/p2p4/Pk2P3/2pP4/2B5/8 b - - Engine: The Huntsman 1 (512 MB)
gemaakt door the Stockfish developers (see AUTHORS f



59/49 20:07 +M24 18...dxe4 19.dxe4 Pf6 20.Ld4 Pg8
21.Lc5+ Kc4 22.Le3 Pf6 23.e5 Pe8+
24.Kd7 Kd5 25.Kxe8 Kxe5 26.Ke7 Kd5
27.Lf4 Kc4 28.Kd6 Kd4 29.Lc1 Kc4
30.Le3 g5 31.Lxg5 Kd4 (2.030.351.242) 1681

60/49 21:44 +M23 18...dxe4 19.dxe4 Pf6 20.Ld4 Pg8
21.Lc5+ Kc4 22.Le3 Pf6 23.e5 Pe8+
24.Kd7 Kd5 25.Kxe8 Kxe5 26.Ke7 Kd5
27.Lxg6 c2 28.Lxc2 Ke5 29.Lg5 Kd4
30.Kd6 Kc3 31.Le4 Kd4 (2.208.104.217) 1692

61/47 23:14 +M23 18...dxe4 19.dxe4 Pf6 20.Ld4 Pg8
21.Lc5+ Kc4 22.Le3 Pf6 23.e5 Pe8+
24.Kd7 Kd5 25.Kxe8 Kxe5 26.Ke7 Kd5
27.Lxg6 c2 28.Lxc2 Ke5 29.Lg5 Kd4
30.Kd6 Kc3 31.Le4 Kd4 (2.365.987.710) 1696

but Huntsman did not get to go backward from the end. M23 would make it M40 total, too long, but the other two can't be both right, but I stop here.

[Ajedrecista]

Hello Peter:

Don't have much hardware- time free at the moment neither, so just after a short Forward- Backward with Huntsman:

qn6/b1Kp2p1/p1pB4/p7/P1NkP3/2pP4/2B4p/8 w - - 0 1

Analysis by The Huntsman 1:

1.Lxh2 Kc5 2.Ld6+ Kd4 3.Kc8 g6 4.Lf4 Kc5 5.Le3+ Kb4 6.Lc1 Kc5 7.La3+ Kd4 8.Ld6 Lb6 9.Lf4 Kc5 10.Le3+ Kb4 11.Sxb6 Da7 12.Sc4 Da8 13.Kc7 d5 14.Sb6 Da7+ 15.Kd6 d4 16.Lxd4 Sd7 17.Sd5+ cxd5 18.Lxa7 dxe4 19.dxe4 Sf6 20.e5 Se8+ 21.Kd7 Sg7 22.Le3 Ka3 23.Lc1+ Kb4 24.e6 Sxe6 25.Kxe6 Kc5 26.Lxg6 c2 27.Le3+ Kb4 28.Lxc2 Kc3 29.Le4 Kb4 30.Kd5 Kc3 31.Lf5 Kb3 32.Ld2 Kxa4 33.Kc4 Ka3 34.Lc1+ Ka2 35.Kc3 a4 36.Kc2 a3 37.Le6+ Ka1 38.Lxa3 a5 39.Lb2#
Tiefe: 75/78 00:00:06 331MN

1.Bf4 I also tried but didn't find other ways than to return to Bxh2 one or more moves later on transposing to 1.Bxh2 again.
Yet there could be detours to be found stll with more hardware- time, regards

I took a brief look on your proposed line: after 25.- ..., Kc5, SF 16 with Multi-PV = 7 suggested that 26.- Be3+ can force a checkmate in 13 moves (#38 in total) instead of 26.- Bxg6 (checkmate in 14 moves, #39 in total). I tried that position with Chest 5.2, using the parameters indicated by Chest into a Special SearchMode:

Code: Select all

FEN: 8/8/p3K1p1/p1k5/P7/2p5/2B5/2B5 w - - 1 26

ChestUCI:
FEN: 8/8/p3K1p1/p1k5/P7/2p5/2B5/2B5 w - -   (4+5)
Stellungs-Analyse:  C0/R0/K6/P2/X6   W:3/19
Suche nach Spezial-Matt [C0/R0/K6/P2/X6] in 14 ...  (Hash=512MB)
  12	00:21	  16.917.960	1.028.072	+M12	Bc1e3+
Suche abgeschlossen ...  (Zeit=41.27s)
Matt in 12 gefunden !  (1 Lösung in 00:41)
 12/12	00:41	  42.921.505	1.039.941	+M12	Bc1e3+ Kc5c4 Ke6d6 g6g5 Be3xg5 Kc4d4 Bg5h6 Kd4c4 Bh6g7 Kc4b4 Kd6d5 Kb4a3 Bg7xc3 Ka3a2 Kd5c4 Ka2a3 Bc2b1 Ka3xa4 Bc3d2 Ka4a3 Bd2c1+ Ka3a4 Bb1c2+

Improving the shortest checkmate by an additional move (#37 in total). This is following the first fifty plies of your line, of course.

Chest line features a long no-brainer for black, I mean, a long sequence of one only legal move by black move after move, from 29.- ..., Kd4 until 36.- Ka4.

Starting from 23.- Bc1+, it looks like 23.- ..., Ka2 ties with 23.- ..., Kb4 in #14 from there.

------------

I see that Eelco is also working. He gets a #38 while I get a #37, but it is true that he got it from an earlier deviation (19.- ..., Nf8) than me (26.- Be3+). There might be room for improvement.

Regards from Spain.

Ajedrecista.

[peter]

I took a brief look on your proposed line: after 25.- ..., Kc5, SF 16 with Multi-PV = 7 suggested that 26.- Be3+ can force a checkmate in 13 moves (#38 in total) instead of 26.- Bxg6 (checkmate in 14 moves, #39 in total). I tried that position with Chest 5.2, using the parameters indicated by Chest into a Special SearchMode:

Code: Select all

FEN: 8/8/p3K1p1/p1k5/P7/2p5/2B5/2B5 w - - 1 26

ChestUCI:
FEN: 8/8/p3K1p1/p1k5/P7/2p5/2B5/2B5 w - -   (4+5)
Stellungs-Analyse:  C0/R0/K6/P2/X6   W:3/19
Suche nach Spezial-Matt [C0/R0/K6/P2/X6] in 14 ...  (Hash=512MB)
  12	00:21	  16.917.960	1.028.072	+M12	Bc1e3+
Suche abgeschlossen ...  (Zeit=41.27s)
Matt in 12 gefunden !  (1 Lösung in 00:41)
 12/12	00:41	  42.921.505	1.039.941	+M12	Bc1e3+ Kc5c4 Ke6d6 g6g5 Be3xg5 Kc4d4 Bg5h6 Kd4c4 Bh6g7 Kc4b4 Kd6d5 Kb4a3 Bg7xc3 Ka3a2 Kd5c4 Ka2a3 Bc2b1 Ka3xa4 Bc3d2 Ka4a3 Bd2c1+ Ka3a4 Bb1c2+

Improving the shortest checkmate by an additional move (#37 in total). This is following the first fifty plies of your line, of course.

Chest line features a long no-brainer for black, I mean, a long sequence of one only legal move by black move after move, from 29.- ..., Kd4 until 36.- Ka4.

Starting from 23.- Bc1+, it looks like 23.- ..., Ka2 ties with 23.- ..., Kb4 in #14 from there.

------------

I see that Eelco is also working. He gets a #38 while I get a #37, but it is true that he got it from an earlier deviation (19.- ..., Nf8) than me (26.- Be3+). There might be room for improvement.

Backward to 20. of Huntsman's line I'm at 36 all in all so far:

4n3/B7/p2K2p1/p3P3/Pk6/2p5/2B5/8 w - - 0 1

Analysis by The Huntsman 1:

21.Kd7 Sg7 22.Le3 g5 23.Lxg5 Ka3 24.Lh6 Kb2 25.Lxg7 Kxc2 26.e6 Kd3 27.Lh6 Ke4 28.e7 Kf3 29.e8D c2 30.De3+ Kg2 31.De2+ Kg3 32.Le3 Kh3 33.Df3+ Kh2 34.Lf2 c1D 35.Dg3+ Kh1 36.Dh3#
Tiefe: 80/32 00:00:41 3321MN

But at further Backward Huntsman deviates with 17...Ka3 (from old 17...cxd5 there), which he thinks would cost one move again:

8/q2n4/p1pK2p1/p2N4/P2BP3/k1pP4/2B5/8 w - - 0 1

Analysis by The Huntsman 1:

18.Lxa7 Sf8 19.Se3 Kb2 20.Ld4 g5 21.Ld1 g4 22.Lxg4 Kb3 23.Ld1+ Kb2 24.Kxc6 Kb1 25.Lxc3 Kc1 26.Kd6 Sg6 27.e5 Kb1 28.e6 Ka2 29.e7 Sxe7 30.Kxe7 Kb1 31.d4 Ka2 32.d5 Kb1 33.d6 Kc1 34.d7 Kb1 35.d8D Ka2 36.Dd2+ Ka3 37.Db2#
Tiefe: 68/40 00:00:24 1499MN

Whole variant would look like that then:

[pgn][Event "Mate in 37?"] [Site "?"] [Date "2024.01.20"] [Round "?"] [White "Eelco"] [Black "After, Popov"] [Result "1-0"] [SetUp "1"] [FEN "qn6/b1Kp2p1/p1pB4/p7/P1NkP3/2pP4/2B4p/8 w - - 0 1"] [PlyCount "73"] [EventDate "2024.??.??"] 1. Bxh2 Kc5 2. Bd6+ Kd4 3. Kc8 g6 4. Bf4 Kc5 5. Be3+ Kb4 6. Bc1 Kc5 7. Ba3+ Kd4 8. Bd6 Bb6 9. Bf4 Kc5 10. Be3+ Kb4 11. Nxb6 Qa7 12. Nc4 Qa8 13. Kc7 d5 14. Nb6 d4 15. Bxd4 Qa7+ 16. Kd6 Nd7 17. Nd5+ Ka3 18. Bxa7 Nf8 19. Ne3 Kb2 20. Bd4 g5 21. Bd1 Ng6 22. Kxc6 Nf4 23. Kc5 g4 24. Bxg4 Kb3 25. Bd1+ Ka2 26. Kc4 Kb2 27. e5 Kc1 28. Kxc3 Ne6 29. Bg4 Nc7 30. Nc4 Nb5+ 31. axb5 axb5 32. Be3+ Kb1 33. Na3+ Ka1 34. Be6 b4+ 35. Kc2 b3+ 36. Bxb3 a4 37. Bd4# 1-0 [/pgn]

Gustav can approve that quickly from 23. of Huntsman- line with 3 flights but further backward it gets stuck as for time to depth needed. Will stop that here again, regards

[peter]

Gustav can approve that quickly from 23. of Huntsman- line with 3 flights but further backward it gets stuck as for time to depth needed

Edit time over:
That was meant for old Huntsman- line, not for the new one above, e.g. at 23. there black King had already 5 flights.

[Spliffjiffer]

8/q2n4/p1pK2p1/p2N4/P2BP3/k1pP4/2B5/8 w - - 0 1

Analysis by The Huntsman 1:

18.Lxa7 Sf8 19.Se3 Kb2 20.Ld4 g5 21.Ld1 g4 22.Lxg4 Kb3 23.Ld1+ Kb2 24.Kxc6 Kb1 25.Lxc3 Kc1 26.Kd6 Sg6 27.e5 Kb1 28.e6 Ka2 29.e7 Sxe7 30.Kxe7 Kb1 31.d4 Ka2 32.d5 Kb1 33.d6 Kc1 34.d7 Kb1 35.d8D Ka2 36.Dd2+ Ka3 37.Db2#
Tiefe: 68/40 00:00:24 1499MN

Whole variant would look like that then:

[pgn][Event "Mate in 37?"] [Site "?"] [Date "2024.01.20"] [Round "?"] [White "Eelco"] [Black "After, Popov"] [Result "1-0"] [SetUp "1"] [FEN "qn6/b1Kp2p1/p1pB4/p7/P1NkP3/2pP4/2B4p/8 w - - 0 1"] [PlyCount "73"] [EventDate "2024.??.??"] 1. Bxh2 Kc5 2. Bd6+ Kd4 3. Kc8 g6 4. Bf4 Kc5 5. Be3+ Kb4 6. Bc1 Kc5 7. Ba3+ Kd4 8. Bd6 Bb6 9. Bf4 Kc5 10. Be3+ Kb4 11. Nxb6 Qa7 12. Nc4 Qa8 13. Kc7 d5 14. Nb6 d4 15. Bxd4 Qa7+ 16. Kd6 Nd7 17. Nd5+ Ka3 18. Bxa7 Nf8 19. Ne3 Kb2 20. Bd4 g5 21. Bd1 Ng6 22. Kxc6 Nf4 23. Kc5 g4 24. Bxg4 Kb3 25. Bd1+ Ka2 26. Kc4 Kb2 27. e5 Kc1 28. Kxc3 Ne6 29. Bg4 Nc7 30. Nc4 Nb5+ 31. axb5 axb5 32. Be3+ Kb1 33. Na3+ Ka1 34. Be6 b4+ 35. Kc2 b3+ 36. Bxb3 a4 37. Bd4# 1-0 [/pgn]

both moves, namely 20.d4 and 20.e5 seem to mate in 17 !?

[Spliffjiffer]

actually they even mate in 16