trying to prove mate against 2 knights and 8 pawns

[Uri Blass]

After 1.e4 a6 2.Qf3 d6 3.Qc3
Slowchess say mate in 20 against black

Of course black has different options in the first moves but these are relatively strong moves by black that Wasp liked at big depth based on my analysis.

Other tries may be 1.e4 a6 2.Qf3 Nc6 3.Qg3 or 1.e4 Nc6
Note that the moves for white are based on a long analysis by Wasp4.50 and first moves that I checked for black are also based on long analysis of Wasp because I do not trust Stockfish that suggest 1.c4 that based on my intuition is not the fastest way to mate.

I believe that after some hundrends of lines it is possible to prove mate at move 3 by today engines in every line.

[fen]1n2k1n1/1pp1pppp/p2p4/8/4P3/2Q5/PPPP1PPP/RNB1KBNR b KQ - 1 3[/fen]

Slow64-avx2 [001]:
2/1 00:00 99 0 -30.32 Nb8-c6 Qc3xg7
3/3 00:00 475 0 -31.04 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7
4/4 00:00 1k 0 -31.08 Nb8-c6 Bf1xa6 Ke8-d7 Qc3-h3+ e7-e6 Ba6xb7
5/6 00:00 3k 0 -31.84 Nb8-c6 Bf1xa6 Nc6-d8 Ba6-d3 Ng8-f6 Qc3xc7
5/6 00:00 4k 0 -31.40 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Qg7xf7
6/6 00:00 4k 0 -31.40 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Qg7xf7
7/10 00:00 16k 1,638k -31.68 Ke8-d7 Qc3-h3+ e7-e6 Qh3xh7 Ng8-f6 Qh7xg7 Kd7-e7 Nb1-c3
7/10+ 00:00 22k 2,197k -31.67 Nb8-c6
7/10 00:00 25k 2,480k -31.56 Nb8-c6 Bf1xa6 Nc6-d8 Ba6-b5+ c7-c6 Qc3xg7 Ng8-f6 Bb5-d3
8/10 00:00 36k 3,610k -31.72 Nb8-c6 Bf1xa6 Nc6-d8 Ba6-d3 Nd8-e6 Bd3-c4 Ne6-d8 Qc3xg7 Ng8-f6
8/10+ 00:00 37k 3,653k -31.64 Ke8-d7
9/14 00:00 58k 5,800k -31.92 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Nd8xb7 Qc3-c6+ Ke8-d8 Qc6xb7 Ng8-f6 Nb1-c3 g7-g6 Qb7-b8+ Kd8-d7
9/14+ 00:00 58k 5,849k -31.91 Ke8-d7
10/14 00:00 101k 5,062k -32.04 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Nd8xb7 Qc3-c6+ Ke8-d8 Qc6xb7 Ng8-f6 Nb1-c3 h7-h6 Qb7-b8+ Kd8-d7
11/20 00:00 343k 5,718k -33.24 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Nd8xb7 Qc3-c6+ Ke8-d8 Qc6xb7 Ng8-f6 Nb1-c3 g7-g5 Qb7-b8+ Kd8-d7 Qb8-b5+ Kd7-c8 Qb5xg5 c7-c6 Qg5-f5+ Kc8-b7 Qf5-a5
11/20+ 00:00 345k 5,748k -33.23 Ke8-d7
11/20 00:00 475k 5,941k -33.16 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Nc3-d5 Nf6-e8 Qg7xf7 b7-b5 d2-d3 Kd7-d8 Bc1-g5 Kd8-d7 Nd5xe7
11/20+ 00:00 557k 6,194k -33.15 Ke8-d8
11/20 00:00 635k 5,776k -32.88 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 c7-c6 Qg7-f8+ Nf6-e8 Qf8xf7 Nb8-d7 Qf7xh7 Ne8-f6 Qh7-f7 Nd7-c5
12/20 00:00 719k 5,990k -33.00 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 c7-c6 Qg7-f8+ Nf6-e8 Qf8xf7 b7-b5 Qf7xh7 Nb8-d7 d2-d4 b5-b4 Nc3-a4 Ne8-f6 Qh7-h8+
13/20 00:00 1,319k 6,597k -33.44 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 c7-c6 Qg7xf7 b7-b5 d2-d3 Nb8-d7 Ng1-f3 c6-c5 Bc1-g5 h7-h5 Bg5xf6 Nd7xf6 e4-e5 d6xe5 Nf3xe5
13/20+ 00:00 1,338k 6,688k -33.43 Nb8-c6
14/22 00:00 2,082k 6,941k -33.96 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-d7 Nc3-d5 c7-c5 Nd5xf6 Nd7xf6 Qg7xf7 Nf6-d7 Qf7xh7 Nd7-f6 Qh7-f7 b7-b5 d2-d3 Kd8-d7
14/22+ 00:00 2,087k 6,955k -33.95 Nb8-c6
15/26 00:00 4,359k 7,388k -34.13 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 c7-c6 Qg7-f8+ Nf6-e8 Qf8xf7 Nb8-d7 d2-d4 Ne8-f6 Bc1-g5 h7-h6 Bg5xh6 c6-c5 d4xc5 Nd7xc5 Bh6-g5 Nc5-d7 Bg5xf6 Nd7xf6 Ng1-f3
16/28 00:00 6,010k 7,154k -34.44 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 c7-c6 Qg7-f8+ Kd8-c7 Qf8xe7+ Nf6-d7 Qe7xf7 b7-b5 Qf7xh7 a6-a5 h2-h4 a5-a4 h4-h5 Nb8-a6 Nc3xb5+ c6xb5 Bf1xb5
16/28+ 00:00 6,232k 7,163k -34.43 Ke8-d7
16/28 00:00 6,445k 7,161k -34.08 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Nc3-d5 Nf6xd5 e4xd5 Nc6-d8 Qg7xh7 e7-e6 d5xe6+ Kd7xe6 h2-h4 d6-d5 h4-h5 Nd8-c6 Rh1-h3 Nc6-d4
17/30 00:02 17,355k 7,113k -35.16 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Nc3-d5 Nc6-d8 Nd5xf6+ e7xf6 Qg7xf6 Kd7-c8 d2-d4 b7-b5 d4-d5 Nd8-b7 Qf6xf7 Nb7-c5 Qf7xh7 Kc8-b7 h2-h4 b5-b4 h4-h5
17/30+ 00:02 17,374k 7,120k -35.15 Ke8-d8
17/30 00:02 17,692k 7,105k -34.88 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-d7 Nc3-d5 c7-c6 Nd5xf6 e7xf6 Qg7xf7 h7-h6 d2-d3 h6-h5 Qf7xh5 c6-c5 h2-h4 Kd8-c7 Qh5-f7 Kc7-c8 h4-h5
18/30 00:03 23,744k 6,882k -35.51 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 c7-c6 Qg7-f8+ Nf6-e8 Qf8xf7 h7-h6 d2-d4 c6-c5 d4xc5 Nb8-c6 Bc1xh6 Nc6-e5 Qf7-g8 Ne5-c6 c5xd6 Kd8-d7 d6xe7 Nc6xe7
18/30+ 00:03 24,615k 6,857k -35.50 Ke8-d7
19/34 00:10 67,774k 6,593k -37.20 Ke8-d8 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Nc3-d5 Nf6xd5 e4xd5 Nc6-a7 Qg7xf7 h7-h5 Qf7-f8+ Kd8-d7 Qf8-b8 Na7-b5 Qb8xb7 Nb5-d4 Bf1xa6 Kd7-e8 Qb7xc7 Ke8-f7 Qc7-c3 Nd4-f5 Qc3-h8 e7-e6 Qh8xh5+ Kf7-e7 d5xe6 Ke7xe6
19/34+ 00:10 68,553k 6,592k -37.19 Ke8-d7
19/34 00:10 69,784k 6,590k -36.08 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 b7-b5 Qg7xf7 h7-h6 Nc3-d5 Nf6xd5 e4xd5 Kd7-c8 Qf7xe7 h6-h5 d2-d3 Kc8-b7 Bc1-g5 h5-h4 Bg5xh4 Kb7-c8 Bh4-g5 Kc8-b7 h2-h4 a6-a5 h4-h5
19/34+ 00:10 69,977k 6,589k -36.07 Nb8-c6
20/34 00:12 82,610k 6,619k -36.32 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 b7-b5 Qg7xf7 h7-h6 Nc3-d5 Nf6xd5 e4xd5 Kd7-c8 Qf7xe7 h6-h5 d2-d3 Kc8-b7 Bc1-g5 h5-h4 Bg5xh4 Kb7-c8 Bh4-g5 Kc8-b7 h2-h4 b5-b4 h4-h5 a6-a5
20/35+ 00:13 85,984k 6,609k -36.31 Nb8-c6
21/38 00:30 184,595k 6,017k -37.20 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Nc3-d5 Nf6xd5 e4xd5 Nc6-d8 Qg7xh7 c7-c6 d5xc6+ Nd8xc6 Qh7xf7 Nc6-e5 Qf7-f5+ e7-e6 Qf5-h7+ Kd7-c6 Qh7-e7 Ne5-d7 Qe7xe6 b7-b5 h2-h4 Nd7-c5 Qe6-c8+ Kc6-b6 h4-h5 Nc5-b7 h5-h6
21/38+ 00:31 187,376k 6,017k -37.19 Ke8-d8
22/42 01:30 526,969k 5,818k -38.72 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Nc3-d5 Kd7-e8 Nd5xf6+ e7xf6 Qg7xf6 Nc6-b4 Bf1-c4 Ke8-d7 Ke1-d1 Kd7-c8 Qf6-h8+ Kc8-d7 Qh8xh7 b7-b5 Bc4xf7 Kd7-c6 a2-a3 Nb4xc2 Kd1xc2 Kc6-b6 h2-h4 a6-a5 h4-h5 a5-a4 d2-d4 c7-c6
22/42+ 01:30 527,584k 5,817k -38.71 Nb8-c6
22/42 01:48 642,378k 5,904k -38.52 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Ke8-d7 Qc3xg7 Ng8-f6 Bb7-d5 c7-c6 Bd5xf7 Kd7-c7 Nb1-c3 d6-d5 e4xd5 c6xd5 Nc3xd5+ Nf6xd5 Bf7xd5 Kc7-d6 Qg7xh7 e7-e6 d2-d3 e6xd5 Bc1-f4+ Kd6-e6 Qh7-h3+ Ke6-f7 Qh3-d7+ Kf7-g6 Qd7xd8
23/42 02:04 747,461k 5,981k -39.52 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Ke8-d7 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 Kd7-c8 Bd5xf7 Ne8-f6 Qh7-f5+ Kc8-b8 Bf7-c4 Nd8-b7 Bc4-a6 d6-d5 Ba6xb7 Kb8xb7 e4xd5 Nf6-e8 Qf5-e6 Ne8-d6 Qe6xe7 Nd6-c8 Qe7-e8 Nc8-b6 Qe8-c6+ Kb7-a6 Qc6xc7 Ka6-b5 Qc7-c6+ Kb5-a5 d5-d6
23/42+ 02:08 772,897k 5,992k -39.51 Ke8-d7
24/46 02:54 1,053,989k 6,038k -38.16 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Ke8-d7 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 c7-c6 Bd5xf7 Ne8-f6 Qh7-f5+ Kd7-c7 e4-e5 d6xe5 Qf5xe5+ Kc7-d7 Bf7-c4 Nf6-e8 Nb1-c3 Ne8-d6 Bc4-b3 Kd7-e8 h2-h4 Nd6-f7 Bb3xf7+ Nd8xf7 Qe5-g7 Nf7-d6 h4-h5 Ke8-d7 h5-h6 Nd6-f5 Qg7-g8 Kd7-c7
25/48 04:53 1,801,353k 6,131k -40.64 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Ke8-d7 Qc3xg7 Ng8-f6 Bb7-d5 Nf6xd5 e4xd5 e7-e6 d5xe6+ Kd7xe6 Qg7xh7 Nd8-c6 Nb1-c3 d6-d5 a2-a4 Nc6-a5 Qh7-g8 c7-c6 Qg8-e8+ Ke6-f6 Qe8-d8+ Kf6-g6 Qd8xa5 d5-d4 Nc3-e2 d4-d3 Ne2-f4+ Kg6-h6 Nf4xd3 f7-f6
25/48+ 04:54 1,803,481k 6,131k -40.63 Ke8-d7
26/52 12:23 4,297,765k 5,777k -42.60 Nb8-c6 Bf1xa6 Nc6-d8 Ba6xb7 Ke8-d7 Qc3xg7 Ng8-f6 Bb7-d5 Kd7-c8 Bd5xf7 Nf6-d7 Qg7xh7 Nd7-f6 Qh7-f5+ Kc8-b8 Bf7-c4 Nd8-b7 Bc4-a6 d6-d5 e4-e5 Nf6-e4 d2-d3 Ne4-c5 Bc1-e3 Nc5xa6 Qf5-f8+ Nb7-d8 Qf8xd8+ Kb8-b7 Qd8xe7 Na6-b8 Qe7-b4+ Kb7-c8 Qb4-f8+ Kc8-b7 e5-e6 Nb8-c6 Qf8-c5 d5-d4 Be3-g5 Kb7-c8 Qc5xc6
26/52+ 12:25 4,303,471k 5,774k -42.59 Ke8-d7
26/52+ 15:48 5,396,065k 5,690k -42.59 e7-e6
27/56 28:10 9,248,378k 5,470k -46.80 Nb8-c6 Bf1xa6 g7-g5 Ba6xb7 Nc6-e5 d2-d4 Ne5-g6 Qc3xc7 Ke8-f8 Bc1xg5 h7-h6 Bg5-e3 Ng6-h4 g2-g3 Nh4-g6 Qc7-c8+ Kf8-g7 a2-a4 e7-e6 a4-a5 Ng8-e7 Qc8-c7 h6-h5 a5-a6 f7-f6 Qc7xd6 Kg7-f7 Bb7-d5 e6xd5 e4xd5 Kf7-g7 a6-a7 Ne7-f5 Qd6-d7+ Ng6-e7 a7-a8Q h5-h4 g3xh4 Nf5xe3 Qd7xe7+ Kg7-g6 Qa8-e8+ Kg6-h6 f2xe3
27/56+ 28:15 9,270,215k 5,467k -46.79 e7-e6
27/56+ 30:04 9,802,868k 5,432k -46.79 Ke8-d7
27/56 33:55 10,947,452k 5,379k -46.19 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nf6xe4 Nc3xe4 Nb8-c6 Qg7xf7 Kd7-d8 Qf7xh7 Kd8-c8 h2-h4 b7-b5 h4-h5 Nc6-b4 Ne4xd6+ Kc8-d7 Qh7-f5+ e7-e6 Qf5-f7+ Kd7xd6 Qf7-f8+ Kd6-d5 Qf8xb4 c7-c5 Qb4-a5 Kd5-d6 Qa5xa6+ Kd6-e7 h5-h6 b5-b4 h6-h7 b4-b3 h7-h8Q b3xc2
27/56+ 36:01 11,640,560k 5,385k -46.18 g7-g6
28/56- 40:35 13,092,565k 5,375k -M77 Ke8-d7 Qc3xg7
28/56+ 40:37 13,103,751k 5,375k -95.00 Nb8-c6
28/56+ 43:10 13,895,554k 5,365k -95.00 g7-g6
28/56+ 43:30 13,995,700k 5,361k -95.00 d6-d5
28/56+ 44:49 14,382,161k 5,347k -M250 Ke8-d7
29/56 57:06 18,025,783k 5,260k -M25 Ke8-d7 Qc3xg7 Ng8-f6 Nb1-c3 Nb8-c6 Nc3-d5 Nf6xd5 e4xd5 Nc6-d8 Qg7xh7 Kd7-c8 h2-h4 c7-c6 d5xc6 d6-d5 Qh7-g8 Kc8-c7 h4-h5 f7-f6 c6xb7 Nd8xb7 Qg8xd5 e7-e6 Qd5xe6 Nb7-a5 Qe6xf6 Na5-c6 Bf1xa6 Nc6-a5 Qf6-e5+ Kc7-b6 Qe5-d6+ Kb6-a7 h5-h6 Na5-c4 Qd6-c7+ Ka7xa6 Qc7xc4+
29/56+ 57:26 18,128,219k 5,259k -M26 Nb8-c6
29/56 1:00:37 19,108,100k 5,253k -M28 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Kd7-c8 Bd5xf7 Nf6-d7 Bf7-e8 Nd8-c6 Be8xd7+ Kc8xd7 Qg7xh7 Nc6-b4 Qh7-f5+ Kd7-d8 Ke1-d1 Nb4-c6 Qf5-d5 Nc6-b4 Qd5-b5 c7-c5 c2-c3 Nb4-d3 Qb5xd3 c5-c4 Qd3xc4 Kd8-d7 Qc4-f7 Kd7-d8 Qf7-e6 Kd8-e8 Qe6-c8+ Ke8-f7 Ng1-f3 Kf7-g7 Qc8-e6 Kg7-f8 Nf3-g5 Kf8-g7 Qe6xe7+ Kg7-h6 Qe7-f6+
30/56 1:02:21 19,645,453k 5,250k -M24 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Kd7-e8 Nb1-c3 c7-c6 Bd5-b3 h7-h6 Qg7xh6 Nf6-g4 Qh6-g7 Ng4-e5 Qg7-g8+ Ke8-d7 d2-d3 Kd7-c7 d3-d4 Ne5-g6 Bb3xf7 Ng6-h4 Qg8-e8 Nd8xf7 Qe8xe7+ Kc7-c8 Qe7xf7 Nh4xg2+ Ke1-f1 Ng2-h4 Qf7-e7 Nh4-g6 Qe7xd6 Ng6-e5 d4xe5 Kc8-b7 Qd6-d7+ Kb7-b6 Nc3-a4+ Kb6-a6 Qd7xc6+ Ka6-a7 Qc6-c7+ Ka7-a6 Qc7-b6+
30/56+ 1:02:48 19,785,693k 5,250k -M25 e7-e6
31/56 1:13:49 23,187,178k 5,234k -M21 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 e7-e6 Nb1-c3 c7-c6 Bd5-b3 Ne8-f6 Qh7-g7 Nf6-e8 Qg7-f8 Ne8-f6 Bb3xe6+ f7xe6 Qf8xf6 Nd8-b7 d2-d3 Nb7-c5 Bc1-g5 Kd7-c8 Qf6-f8+ Kc8-b7 Qf8xd6 Nc5-a6 Bg5-f4 Kb7-a7 Qd6xc6 e6-e5 Bf4xe5 Na6-c7 Qc6xc7+ Ka7-a6 Qc7-c6+ Ka6-a7 Nc3-b5+
31/56+ 1:13:53 23,203,793k 5,234k -M22 e7-e6
32/56 1:15:06 23,565,216k 5,229k -M21 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 h7-h6 Nb1-c3 Nf6-e8 Qg7xh6 c7-c6 Bd5-b3 Ne8-c7 Qh6-f8 Nc7-a8 Bb3xf7 Nd8xf7 Qf8xf7 Na8-b6 d2-d3 Nb6-c8 h2-h4 Kd7-c7 h4-h5 Kc7-d7 h5-h6 d6-d5 h6-h7 d5xe4 Nc3xe4 Nc8-d6 Ne4xd6 Kd7xd6 h7-h8Q Kd6-c7 Qf7xe7+ Kc7-b6 Qh8-b8+ Kb6-a6 Qe7-a7+
33/56 1:18:51 24,704,454k 5,221k -M20 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 e7-e6 Nb1-c3 c7-c6 Bd5-b3 Ne8-f6 Qh7-g7 Nf6-e8 Qg7-f8 Ne8-f6 Bb3xe6+ f7xe6 Qf8xf6 Nd8-b7 d2-d3 Nb7-c5 Bc1-g5 Kd7-c8 Qf6-d8+ Kc8-b7 Bg5-e7 e6-e5 Be7xd6 Nc5-d7 Qd8xd7+ Kb7-b6 Qd7-c7+ Kb6-a6 Qc7xc6+ Ka6-a7 Nc3-b5+
33/56+ 1:18:53 24,710,919k 5,221k -M21 e7-e6
33/56+ 1:20:53 25,305,958k 5,213k -M21 Ke8-d7
33/56+ 1:22:54 25,864,945k 5,200k -M21 g7-g6
34/56 1:23:50 26,121,312k 5,192k -M20 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 e7-e6 Nb1-c3 c7-c6 Bd5-b3 d6-d5 e4xd5 e6xd5 Qh7-f5+ Kd7-e7 Qf5-c8 Ne8-f6 Qc8-c7+ Nf6-d7 Bb3xd5 c6xd5 Nc3xd5+ Ke7-e6 Qc7xd8 f7-f6 Qd8-e7+ Ke6xd5 Qe7xd7+ Kd5-c5 d2-d3 Kc5-b6 Bc1-e3+ Kb6-a6 Qd7-c6+ Ka6-a5 Be3-d2+
35/56 1:24:02 26,180,202k 5,191k -M20 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 e7-e6 Nb1-c3 c7-c6 Bd5-b3 d6-d5 e4xd5 e6xd5 Qh7-f5+ Kd7-e7 Qf5-c8 Ne8-f6 Qc8-c7+ Nf6-d7 Bb3xd5 c6xd5 Nc3xd5+ Ke7-e6 Qc7xd8 f7-f6 Qd8-e7+ Ke6xd5 Qe7xd7+ Kd5-c5 d2-d3 Kc5-b6 Bc1-e3+ Kb6-a6 Qd7-c6+ Ka6-a5 Be3-d2+
36/56 1:24:08 26,207,505k 5,191k -M20 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 e7-e6 Nb1-c3 c7-c6 Bd5-b3 d6-d5 e4xd5 e6xd5 Qh7-f5+ Kd7-e7 Qf5-c8 Ne8-f6 Qc8-c7+ Nf6-d7 Bb3xd5 c6xd5 Nc3xd5+ Ke7-e6 Qc7xd8 f7-f6 Qd8-e7+ Ke6xd5 Qe7xd7+ Kd5-c5 d2-d3 Kc5-b6 Bc1-e3+ Kb6-a6 Qd7-c6+ Ka6-a5 Be3-d2+
37/56 1:24:28 26,294,452k 5,187k -M20 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 e7-e6 Nb1-c3 c7-c6 Bd5-b3 d6-d5 e4xd5 e6xd5 Qh7-f5+ Kd7-e7 Qf5-c8 Ne8-f6 Qc8-c7+ Nf6-d7 Bb3xd5 c6xd5 Nc3xd5+ Ke7-e6 Qc7xd8 f7-f6 Qd8-e7+ Ke6xd5 Qe7xd7+ Kd5-c5 d2-d3 Kc5-b6 Bc1-e3+ Kb6-a6 Qd7-c6+ Ka6-a5 Be3-d2+
38/56 1:24:52 26,397,172k 5,184k -M20 Nb8-c6 Bf1xa6 Ke8-d7 Ba6xb7 Nc6-d8 Qc3xg7 Ng8-f6 Bb7-d5 Nf6-e8 Qg7xh7 e7-e6 Nb1-c3 c7-c6 Bd5-b3 d6-d5 e4xd5 e6xd5 Qh7-f5+ Kd7-e7 Qf5-c8 Ne8-f6 Qc8-c7+ Nf6-d7 Bb3xd5 c6xd5 Nc3xd5+ Ke7-e6 Qc7xd8 f7-f6 Qd8-e7+ Ke6xd5 Qe7xd7+ Kd5-c5 d2-d3 Kc5-b6 Bc1-e3+ Kb6-a6 Qd7-c6+ Ka6-a5 Be3-d2+

[peter]

After 1.e4 a6 2.Qf3 d6 3.Qc3
Slowchess say mate in 20 against black

Of course black has different options in the first moves but these are relatively strong moves by black that Wasp liked at big depth based on my analysis.

Other tries may be 1.e4 a6 2.Qf3 Nc6 3.Qg3 or 1.e4 Nc6
Note that the moves for white are based on a long analysis by Wasp4.50 and first moves that I checked for black are also based on long analysis of Wasp because I do not trust Stockfish that suggest 1.c4 that based on my intuition is not the fastest way to mate.

I believe that after some hundrends of lines it is possible to prove mate at move 3 by today engines in every line.

[fen]1n2k1n1/1pp1pppp/p2p4/8/4P3/2Q5/PPPP1PPP/RNB1KBNR b KQ - 1 3[/fen]

From the starting position it's #22 or less (didn't try very long at Backward of the latest SlowChess- line of yours that's 20 (+3) moves long, and so it could well be still not the shortest DTM.

[fen]1n2k1n1/1pp1pppp/p2p4/8/4P3/2Q5/PPPP1PPP/RNB1KBNR b KQ - 0 3[/fen]

3... Nc6 4. Bxa6 Kd7 5. Bxb7 Nd8 6. Qxg7 Nf6 7. Bd5 Ne8 8. Qxh7 e6 9. Nc3 c6 10. Bb3 d5 11. exd5 exd5 12. Nf3 Kc8 13. Ne5 Nd6 14. Qh8 Kc7 15. Qf8 N6b7 16. Qe7+ Kc8 17. Qd7+ Kb8 18. Bxd5 Na5 19. Nxc6+ Naxc6 20. Nb5 Ne6 21. Qxc6 Nc5 22. Qc7#

As for the fist moves of yours, I doubt them to be the very best ones for sure (3.Qc3 after 2.Qf3 looks like a detour to me compared to moving a Knight of White's before outposting the Queen so soon so far, how about 1.e4 2.Nf3 or 2.Nc3, see .pgn below), this one position

[pgn][Event "#19"]
[Site "?"]
[Date "2021.01.26"]
[Round "?"]
[White "2B27."]
[Black "Strangmüller, A.."]
[Result "1-0"]
[SetUp "1"]
[FEN "2b1kb2/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQ - 0 1"]
[PlyCount "37"]

1. e4 g6 2. Nc3 c6 3. Qe2 e6 4. d4 d6 5. Nf3 h6 6. Bf4 a6 7. O-O-O g5 8. Nxg5 hxg5 9. Bxg5 Be7 10. Qh5 Bxg5+ 11. Qxg5 Kd7 12. d5 b5 13. dxe6+ Kc7 14. exf7 Kb7 15. f8=Q a5 16. Qxd6 Ka8 17. Qxc6+ Bb7 18. Rd8+ Ka7 19. Nxb5# 1-0[/pgn]
is #19 in all lines, I have stored very many, that's just one example, so I can't believe 2 Knights should hold longer then 2 Bishops do, even if that could be true yet too.
These postions are some of the most difficult ones for engines to count best DTM because of the many near to each other moves along the lines, search- tree explodes from the very start. Specialised mate search programs like Gustav or Chest don't stand any chance neither, the are based on reduction of search trees, which doesn't work for such positions at all.

[peter]

After 1.e4 a6 2.Qf3 d6 3.Qc3
Slowchess say mate in 20 against black

Of course black has different options in the first moves but these are relatively strong moves by black that Wasp liked at big depth based on my analysis.

Other tries may be 1.e4 a6 2.Qf3 Nc6 3.Qg3 or 1.e4 Nc6
Note that the moves for white are based on a long analysis by Wasp4.50 and first moves that I checked for black are also based on long analysis of Wasp because I do not trust Stockfish that suggest 1.c4 that based on my intuition is not the fastest way to mate.

I believe that after some hundrends of lines it is possible to prove mate at move 3 by today engines in every line.

[fen]1n2k1n1/1pp1pppp/p2p4/8/4P3/2Q5/PPPP1PPP/RNB1KBNR b KQ - 1 3[/fen]

From the starting position it's #22 or less (didn't try very long at Backward of the latest SlowChess- line of yours that's 20 (+3) moves long, and so it could well be still not the shortest DTM.

[fen]1n2k1n1/1pp1pppp/p2p4/8/4P3/2Q5/PPPP1PPP/RNB1KBNR b KQ - 0 3[/fen]

3... Nc6 4. Bxa6 Kd7 5. Bxb7 Nd8 6. Qxg7 Nf6 7. Bd5 Ne8 8. Qxh7 e6 9. Nc3 c6 10. Bb3 d5 11. exd5 exd5 12. Nf3 Kc8 13. Ne5 Nd6 14. Qh8 Kc7 15. Qf8 N6b7 16. Qe7+ Kc8 17. Qd7+ Kb8 18. Bxd5 Na5 19. Nxc6+ Naxc6 20. Nb5 Ne6 21. Qxc6 Nc5 22. Qc7#

As for the fist moves of yours, I doubt them to be the very best ones for sure (3.Qc3 after 2.Qf3 looks like a detour to me compared to moving a Knight of White's before outposting the Queen so soon so far, how about 1.e4 2.Nf3 or 2.Nc3, see .pgn below), this one position

Edit-time over.
After 1.e4 a6 (which probably is not Black's best defense neither) 2.Qg4 I have #18 already, I'd lay my bets on #19 or less for shortest DTM with best moves of Black's, even it won't be possible to keep all moves in hash for shortest DTM- output probably.


[fen]1n2k1n1/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQ - 0 0[/fen]


1. e4 a6 2. Qg4 g6 3. Qf4 d6 4. Bc4 Nf6 5. e5 Nc6

Edit: After 1.e4 a6 (which probably is not Black's best defense neither) 2.Qg4 I have #20 already, I'd lay my bets on #19 or less for shortest DTM with best moves of Black's, even it won't be possible to keep all moves in hash for shortest DTM- output.

After

1. e4 a6 2. Qg4 g6 3. Qf4 d6 4. Bc4 Nf6 5. e5 Nc6:


4k3/1pp1pp1p/p1np1np1/4P3/2B2Q2/8/PPPP1PPP/RNB1K1NR w KQ - 0 1

Analysis by The Huntsman 1:

6.exf6 e5 7.Qh6 Nd4 8.Qg7 Kd7 9.Qxf7+ Kc6 10.Nc3 b5 11.Bd5+ Kb6 12.Qf8 Nxc2+ 13.Kf1 c6 14.Qb8+ Kc5 15.Qa7+ Kb4 16.a3+ Nxa3 17.bxa3+ Ka5 18.Qc7#
+- (#13) Depth: 50/26 00:00:39 1469MN


Edit- time will be over soon again, will return tomorrow,
Btw. fen- icon still doesn't work for me.

[Uri Blass]

After 1.e4 a6 2.Qf3 d6 3.Qc3
Slowchess say mate in 20 against black

Of course black has different options in the first moves but these are relatively strong moves by black that Wasp liked at big depth based on my analysis.

Other tries may be 1.e4 a6 2.Qf3 Nc6 3.Qg3 or 1.e4 Nc6
Note that the moves for white are based on a long analysis by Wasp4.50 and first moves that I checked for black are also based on long analysis of Wasp because I do not trust Stockfish that suggest 1.c4 that based on my intuition is not the fastest way to mate.

I believe that after some hundrends of lines it is possible to prove mate at move 3 by today engines in every line.

[fen]1n2k1n1/1pp1pppp/p2p4/8/4P3/2Q5/PPPP1PPP/RNB1KBNR b KQ - 1 3[/fen]

From the starting position it's #22 or less (didn't try very long at Backward of the latest SlowChess- line of yours that's 20 (+3) moves long, and so it could well be still not the shortest DTM.

[fen]1n2k1n1/1pp1pppp/p2p4/8/4P3/2Q5/PPPP1PPP/RNB1KBNR b KQ - 0 3[/fen]

3... Nc6 4. Bxa6 Kd7 5. Bxb7 Nd8 6. Qxg7 Nf6 7. Bd5 Ne8 8. Qxh7 e6 9. Nc3 c6 10. Bb3 d5 11. exd5 exd5 12. Nf3 Kc8 13. Ne5 Nd6 14. Qh8 Kc7 15. Qf8 N6b7 16. Qe7+ Kc8 17. Qd7+ Kb8 18. Bxd5 Na5 19. Nxc6+ Naxc6 20. Nb5 Ne6 21. Qxc6 Nc5 22. Qc7#

As for the fist moves of yours, I doubt them to be the very best ones for sure (3.Qc3 after 2.Qf3 looks like a detour to me compared to moving a Knight of White's before outposting the Queen so soon so far, how about 1.e4 2.Nf3 or 2.Nc3, see .pgn below), this one position

Edit-time over.
After 1.e4 a6 (which probably is not Black's best defense neither) 2.Qg4 I have #18 already, I'd lay my bets on #19 or less for shortest DTM with best moves of Black's, even it won't be possible to keep all moves in hash for shortest DTM- output probably.


[fen]1n2k1n1/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQ - 0 0[/fen]


1. e4 a6 2. Qg4 g6 3. Qf4 d6 4. Bc4 Nf6 5. e5 Nc6

Edit: After 1.e4 a6 (which probably is not Black's best defense neither) 2.Qg4 I have #20 already, I'd lay my bets on #19 or less for shortest DTM with best moves of Black's, even it won't be possible to keep all moves in hash for shortest DTM- output.

After

1. e4 a6 2. Qg4 g6 3. Qf4 d6 4. Bc4 Nf6 5. e5 Nc6:


4k3/1pp1pp1p/p1np1np1/4P3/2B2Q2/8/PPPP1PPP/RNB1K1NR w KQ - 0 1

Analysis by The Huntsman 1:

6.exf6 e5 7.Qh6 Nd4 8.Qg7 Kd7 9.Qxf7+ Kc6 10.Nc3 b5 11.Bd5+ Kb6 12.Qf8 Nxc2+ 13.Kf1 c6 14.Qb8+ Kc5 15.Qa7+ Kb4 16.a3+ Nxa3 17.bxa3+ Ka5 18.Qc7#
+- (#13) Depth: 50/26 00:00:39 1469MN


Edit- time will be over soon again, will return tomorrow,
Btw. fen- icon still doesn't work for me.

6.exf6 is mate in at most 13 moves when after 6...e5
7.Qh6 mate in 11 Nd4 8.Nc3 is mate in 10(better than 8.Qg7)

I am surprised by the move 5...Nc6 that does not seem good to me and 5...dxe5 seem more natural to me when
I could not see mate in 13 for white.
natural is 6.Qxe5 and here you need to show mate in 12 for white after Kd8 or Kd7 or Nc6.

Here is the analysis with 7 options that I get from slowchess

FEN: 1n2k3/1pp1pp1p/p4np1/4Q3/2B5/8/PPPP1PPP/RNB1K1NR b KQ - 0 6

Slow64-avx2 [001]:

36/52 1:11:40 21,332,696k 4,960k -M12 g6-g5 Qe5xc7 Nb8-c6 Qc7-c8+ Nc6-d8 Ng1-f3 e7-e6 Nf3-e5 Nf6-g8 b2-b3 b7-b5 Qc8-d7+ Ke8-f8 Qd7xd8+ Kf8-g7 Qd8xg5+ Kg7-f8 Bc1-a3+ b5-b4 Ba3xb4+ Ng8-e7 Ne5-d7+ Kf8-e8 Qg5xe7+
36/52 1:11:40 21,332,696k 4,960k -M12 h7-h6 Qe5xc7 Nb8-c6 Qc7-c8+ Nc6-d8 Ng1-f3 Nf6-g4 h2-h3 Ng4-f6 Nf3-e5 e7-e6 b2-b3 Ke8-f8 Qc8xd8+ Nf6-e8 Ne5-d7+ Kf8-g7 Bc1-b2+ Kg7-h7 Qd8xe8 e6-e5 Qe8xf7+ Kh7-h8 Qf7-g8+
36/52 1:11:40 21,332,696k 4,960k -M12 Ke8-f8 d2-d4 g6-g5 Bc1xg5 Nb8-c6 Bg5-h6+ Kf8-e8 Qe5xc7 Nc6-d8 Ng1-f3 Nf6-g4 Bh6-g7 b7-b5 Nb1-c3 f7-f6 Nc3-d5 Nd8-c6 Qc7-c8+ Ke8-f7 Nd5-f4+ Kf7xg7 Qc8-g8+ Kg7-h6 Qg8-f8+
36/52 1:11:40 21,332,696k 4,960k -M12 Nf6-d7 Qe5xc7 Nb8-c6 Ng1-f3 Nc6-d8 Nf3-g5 Nd7-f8 Qc7-c8 e7-e6 b2-b3 Nf8-d7 Bc1-a3 b7-b5 Qc8xa6 f7-f6 Qa6-d6 b5-b4 Ba3xb4 Nd8-c6 Qd6xc6 f6xg5 Bc4xe6 Ke8-d8 Qc6xd7+
36/52 1:11:40 21,332,696k 4,960k -M13 Nb8-c6 Qe5xc7 Ke8-f8 Ng1-f3 Nf6-e8 Qc7-c8 h7-h6 d2-d3 g6-g5 h2-h4 e7-e6 h4xg5 Nc6-e7 Qc8-d7 Ne7-g8 g5-g6 Ne8-d6 Qd7xd6+ Kf8-g7 g6xf7 Ng8-e7 Qd6xe7 Kg7-h7 f7-f8Q+ Kh7-g6 Qf8-f7+
36/52 1:11:40 21,332,696k 4,960k -M18 Ke8-d8 Bc4xf7 Nb8-d7 Qe5-e3 c7-c6 Bf7-e6 h7-h6 Qe3xh6 g6-g5 Be6xd7 Kd8xd7 Qh6xg5 Kd7-d6 h2-h4 a6-a5 h4-h5 a5-a4 h5-h6 Nf6-h7 Qg5-g6+ Kd6-d7 Qg6xh7 c6-c5 Qh7-g6 e7-e6 h6-h7 Kd7-c6 Qg6xe6+ Kc6-b5 Nb1-c3+ Kb5-a5 h7-h8Q a4-a3 b2xa3 c5-c4 Qh8-a8+
36/52 1:11:40 21,332,696k 4,960k -M19 Ke8-d7 Bc4xf7 Kd7-d8 Qe5-d4+ Nb8-d7 Qd4-e3 c7-c6 Bf7-e6 h7-h6 Qe3xh6 g6-g5 Be6xd7 Kd8xd7 Qh6xg5 Kd7-d6 h2-h4 a6-a5 h4-h5 a5-a4 h5-h6 Nf6-h7 Qg5-g6+ Kd6-d7 Qg6xh7 c6-c5 Qh7-g6 e7-e6 h6-h7 Kd7-c6 Qg6xe6+ Kc6-b5 Nb1-c3+ Kb5-a5 h7-h8Q a4-a3 b2xa3 c5-c4 Qh8-a8+

[Uri Blass]

FEN: 1n2k3/1pp1pp1p/p4np1/4Q3/2B5/8/PPPP1PPP/RNB1K1NR b KQ - 0 6

38/54 1:41:13 29,102,667k 4,792k -M13 Nb8-c6 Qe5xc7 Ke8-f8 Ng1-f3 Nf6-e8 Qc7-c8 h7-h6 d2-d3 g6-g5 h2-h4 e7-e6 h4xg5 Nc6-e7 Qc8-d7 Ne7-g8 g5-g6 Ne8-d6 Qd7xd6+ Kf8-g7 g6xf7 Ng8-e7 Qd6xe7 Kg7-h7 f7-f8Q+ Kh7-g6 Qf8-f7+
38/54 1:41:13 29,102,667k 4,792k -M17 Ke8-d8 Bc4xf7 Nb8-d7 Qe5-e3 c7-c6 Bf7-e6 h7-h6 Qe3xh6 Kd8-c7 Ng1-f3 a6-a5 Be6xd7 Nf6xd7 Qh6xg6 Nd7-f6 Qg6-f7 Nf6-d5 Nb1-c3 Nd5-b6 Qf7xe7+ Kc7-b8 d2-d3 Nb6-c8 Qe7-d8 b7-b5 Bc1-f4+ Kb8-b7 Qd8-c7+ Kb7-a6 Qc7xc8+ Ka6-b6 Bf4-e3+ c6-c5 Be3xc5+
38/54 1:41:13 29,102,667k 4,792k -M18 Ke8-d7 Bc4xf7 Kd7-d8 Qe5-d4+ Nb8-d7 Qd4-e3 c7-c6 Bf7-e6 h7-h6 Qe3xh6 Kd8-c7 Ng1-f3 a6-a5 Be6xd7 Nf6xd7 Qh6xg6 Nd7-f6 Qg6-f7 Nf6-d5 Nb1-c3 Nd5-b6 Qf7xe7+ Kc7-b8 d2-d3 Nb6-c8 Qe7-d8 b7-b5 Bc1-f4+ Kb8-b7 Qd8-c7+ Kb7-a6 Qc7xc8+ Ka6-b6 Bf4-e3+ c6-c5 Be3xc5+

stopped the analysis
I doubt if there is a mate in at most 12 moves after Kd7 or Kd8 and moves that are not 6.Qxe5 do not seem logical for me.

[peter]

I doubt if there is a mate in at most 12 moves after Kd7 or Kd8 and moves that are not 6.Qxe5 do not seem logical for me.

You're right about Qxe5, that's for rahter sure longer then after Kc6 (?!).
Letting run Huntsman over night after 1.e4 a6 2. d4 having seen some #20 and #19- lines soon but with suboptimal moves from Black yet too, this one oftenly most accurate engine as for shortest lines (SlowChess is fast but tends to stop pondering with small detours instead of searching for the very shortest ways) doesn't make it on its own with 30 threads of 16x3.4GHz and 32G hash at positions of 3 moves (at least not the ones I tried so far) away from starting one, even with prefilled hash.
Maybe I'll give SlowChess or HekaStockfish a try, but at the moment the whole thing's getting too hardware- time- consuming (and or time of my own for interactive Forward- Backward) right now.
If I'll have more free ressources again, I'll return to it but won't give more lines here without getting at least to #20, which I still think to be the utmost necessary. Of course the two Knights could give longer resistance indeed than the 2 Bishops do, yet I' still doubt that.

Edit: in meantime after

1. e4 a6 2. d4 Nf6 3. e5 Ng8 4. Nc3 c5 5. Qh5 cxd4

and some more Forward- Backward from and to there I have this:

1n2k1n1/1p1ppppp/p7/4P2Q/3p4/2N5/PPP2PPP/R1B1KBNR w KQ - 0 1

Analysis by The Huntsman 1:

6.Bc4 e6 7.Ne4 Kd8 8.Qxf7 Ne7 9.Qf8+ Kc7 10.Qxe7 Nc6 11.Qd6+ Kb6 12.Qxd7 g5 13.Nd6 Nd8 14.Nc8+ Kc5 15.Qd6+ Kxc4 16.b3+ Kc3 17.Qc5#
+- (#12) Depth: 65/24 00:01:10 2718MN

Not proving anything neither of course, 5...cxd4 could be suboptimal again and still to many moves to get there, just to give one more output line.

[peter]

1. e4 a6 2. d4 Nf6 3. e5 Ng8 4. Nc3 c5 5. Qh5 cxd4

and some more Forward- Backward from and to there I have this:

1n2k1n1/1p1ppppp/p7/4P2Q/3p4/2N5/PPP2PPP/R1B1KBNR w KQ - 0 1

Analysis by The Huntsman 1:

6.Bc4 e6 7.Ne4 Kd8 8.Qxf7 Ne7 9.Qf8+ Kc7 10.Qxe7 Nc6 11.Qd6+ Kb6 12.Qxd7 g5 13.Nd6 Nd8 14.Nc8+ Kc5 15.Qd6+ Kxc4 16.b3+ Kc3 17.Qc5#
+- (#12) Depth: 65/24 00:01:10 2718MN

Not proving anything neither of course, 5...cxd4 could be suboptimal again and still to many moves to get there, just to give one more output line.

What I fear most of all with such positions of course has happened already again, can't stop.

Taking Huntsman's lines to let SlowChess have Backard of them, helps this one engine much too.
At the moment I'm at the line above with two more plies backwards:

1. e4 a6 2. d4 Nf6 3. e5 Ng8 4. Nc3 c5

getting this output:

SlowChess Blitz 2.9 avx2 (32768 MB hash, 31 threads)
FEN: 1n2k1n1/1p1ppppp/p7/2p1P3/3P4/2N5/PPP2PPP/R1BQKBNR w HA - 0 5

[35] M14 00:01.8 31.69M Qh5 Nc6 Qxh7 Kf8 Nf3 cxd4 Nd5 e6 Nb6 d3 Bxd3 Nce7 Bg5 f6 exf6 Nxf6 Bxf6 gxf6 Bg6 Nxg6 Nxd7+ Ke8 Nxf6+ Kf8 Ng5 a5 Qf7#

So it's already "only" the first 4 moves to be doubted, 4...c5, 3...Ng8 and 2..Nf6, but I had already some as short lines after 3...Nd5 and ...Ne4 too, and I still don't think 1...a6 Black's best move neither.
Hope I'l make it to get off of PC now soon for real.

[peter]

1. e4 a6 2. d4 Nf6 3. e5 Ng8 4. Nc3 c5

getting this output:

SlowChess Blitz 2.9 avx2 (32768 MB hash, 31 threads)
FEN: 1n2k1n1/1p1ppppp/p7/2p1P3/3P4/2N5/PPP2PPP/R1BQKBNR w HA - 0 5

[35] M14 00:01.8 31.69M Qh5 Nc6 Qxh7 Kf8 Nf3 cxd4 Nd5 e6 Nb6 d3 Bxd3 Nce7 Bg5 f6 exf6 Nxf6 Bxf6 gxf6 Bg6 Nxg6 Nxd7+ Ke8 Nxf6+ Kf8 Ng5 a5 Qf7#

So it's already "only" the first 4 moves to be doubted, 4...c5,

4...Nc6 instead

1. e4 a6 2. d4 Nf6 3. e5 Ng8 4. Nc3 Nc6:

SlowChess Blitz 2.9 avx2 (32768 MB hash, 31 threads)
FEN: 4k1n1/1ppppppp/p1n5/4P3/3P4/2N5/PPP2PPP/R1BQKBNR w HA - 3 5

[34] M14 00:38.5 681.63M Nd5 Nb4 Nxb4 g6 Qf3 c6 Nd3 f5 Nc5 e6 Qb3 b5 Bc4 h6 Bxe6 Ne7 Bxd7+ Kd8 Qe6 Kc7 Qxe7 Kb6 Bc8 f4 b4 g5 Qb7#

So we stll had to find any better moves for Black in the first 4, of the 4th having already tried out two of the not really bad looking candidates. #18 or less might be correct at all, but we are probably at the point again, where further engine output wil become more and more unreliable.

[peter]

1. e4 a6 2. d4 Nf6 3. e5 Ng8 4. Nc3 Nc6:

SlowChess Blitz 2.9 avx2 (32768 MB hash, 31 threads)
FEN: 4k1n1/1ppppppp/p1n5/4P3/3P4/2N5/PPP2PPP/R1BQKBNR w HA - 3 5

[34] M14 00:38.5 681.63M Nd5 Nb4 Nxb4 g6 Qf3 c6 Nd3 f5 Nc5 e6 Qb3 b5 Bc4 h6 Bxe6 Ne7 Bxd7+ Kd8 Qe6 Kc7 Qxe7 Kb6 Bc8 f4 b4 g5 Qb7#

Instead of 3...Ng8 3...Nd5:

First output of SlowChess' after Forward- Backward of some new lines and go inifinite then there, engine cpomes to given depth and output and stops pondering then there with this one:

FEN: 1n2k3/1ppppppp/p7/3nP3/3P4/8/PPP2PPP/RNBQKBNR w HA - 1 4

[39] M16 10:44.7 11692.11M Qh5 c6 Qxh7 Kd8 c4 Nb4 e6 Nc2+ Kd1 Nxd4 exf7 Ne6 Qe4 a5 Qxe6 dxe6 f8=Q+ Kd7 Qxb8 e5 Bg5 Ke6 Qe8 c5 Bd3 e4 Bxe4 Ke5 Nf3+ Kxe4 Qg6#

Going into this line and trying 7.Qxg7 instead of c4 makes DTM shorter for one, so #18 all in all yet still.

[peter]

Going into this line and trying 7.Qxg7 instead of c4 makes DTM shorter for one, so #18 all in all yet still.

Even better then 5.Qxh7 5.Bc4:
[pgn][Event "?"]
[Site "?"]
[Date "2022.11.10"]
[Round "?"]
[White "Blass, U."]
[Black "2N, 27."]
[Result "1-0"]
[SetUp "1"]
[FEN "1n2k1n1/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQ - 0 0"]
[PlyCount "35"]

1. e4 a6 2. d4 Nf6 3. e5 Nd5 4. Qh5 c6 5. Bc4 h6 6. Nf3 Kd8 7. Bxd5 cxd5 8. Qxf7 Nc6 9. Nc3 e6 10. Qf8+ Kc7 11. Qd6+ Kd8 12. Nxd5 Ke8 13. Nc7+ Kf7 14. Qxd7+ Kg6 15. Nxe6 Ne7 16. Qxe7 b6 17. g4 Kh7 18. Qxg7# 1-0[/pgn]

Pity not to be kept in hash with correct DTM-outpu further then till 3rd move of Black's.