A position some have claimed will never be solved ...

[MikeB]

A position some have said will never be solved by a chess program ...

[d]k2b3b/8/2N2b2/6b1/1b6/b7/3b2K1/b1b2B1n w - - 0 1

mover viewpoint fewer / Multi-PV setting = 1 / more
exclude: none best +tail
dep score nodes time (not shown: tbhits knps seldep)
26 +142.26! 22.3G 3:05.23 Ba6!
26 +113.15! 14.2G 2:02.10 Ba6!
26 +90.13! 12.4G 1:46.94 Ba6!
26 +71.74! 12.1G 1:44.39 Ba6!
26 +57.03! 12.0G 1:43.85 Ba6!
26 +45.29! 12.0G 1:43.68 Ba6!
26 +35.91! 12.0G 1:43.30 Ba6!
26 +28.42! 11.9G 1:42.94 Ba6!
26 +22.25! 11.9G 1:42.66 Ba6!
26 +17.50! 11.9G 1:42.27 Ba6!
26 +13.67! 11.7G 1:40.64 Ba6!
26 +10.66! 11.5G 1:39.11 Ba6!
26 +8.26! 11.3G 1:38.09 Ba6!
26 +6.37! 11.2G 1:37.23 Ba6!
26 +4.87! 11.1G 1:36.20 Ba6!
26 +3.68! 11.0G 1:35.06 Ba6!
26 +2.75! 10.9G 1:34.78 Ba6!
26 +2.03! 10.9G 1:34.53 Ba6!
26 +1.47! 10.9G 1:34.39 Ba6!
26 +1.03! 10.8G 1:33.34 Ba6!
26 +0.35! 10.7G 1:32.59 Ba6!
26 +0.23! 10.2G 1:29.15 Ba6!
26 +0.14! 10.0G 1:27.39 Ba6!
26 +0.08! 9.71G 1:24.72 Ba6!
26 +0.03! 9.43G 1:22.48 Ba6!
25 0.00 474.7M 0:05.40 Ba6 Bfb2 Kxh1 Bhe5 Kg2 Bda5 Kg1 Bb6+ Kh1
24 0.00 344.7M 0:03.85 Ba6 Bfb2 Kxh1 Bhe5 Kg2 Bda5 Kg1 Bb6+ Kh1
23 0.00 229.3M 0:02.50 Ba6 Bfb2 Kxh1 Bhe5 Kg2 Bda5 Kg1 Bb6+ Kh1
22 0.00 139.9M 0:01.48 Ba6 Bfb2 Kxh1 Bhe5 Kg2 Bda5 Kg1 Bb6+ Kh1
21 0.00 137.8M 0:01.45 Ba6 Bfb2 Kxh1 Bhe5 Kg2 Bda5 Kg1 Bb6+ Kh1
20 0.00 89.1M 0:00.90 Ba6 Bfb2 Kxh1 Bhe5 Kg2 Bda5 Kg1 Bb6+ Kh1
19 0.00 80.6M 0:00.80 Ba6 Bfb2 Kxh1 Bhe5 Kg2 Bda5 Kg1 Bb6+ Kh1
18 0.00 52.6M 0:00.49 Ba6 Bh6 Kxh1 Bde7 Kg2 Ba1b2 Kh1
17 0.00 51.2M 0:00.48 Ba6 Bh6 Kxh1 Bde7 Kg2 Ba1b2 Kh1
16 0.00 50.8M 0:00.47 Ba6 Bh6 Kxh1 Bde7 Kg2 Ba1b2 Kh1
15 0.00 50.6M 0:00.47 Ba6 Bh6 Kxh1 Bde7 Kg2 Ba1b2 Kh1
14 -6.95 42.9M 0:00.40 Kxh1 Kb7 Nxd8+ Kc7 Ne6+ Kc6 Nxg5 Kd5 Bg2+ Kd4 Bf1 Ke3 Ne4 Kf4 Nxf6 Be3 Nd5+ Kg3 Nxb4 Bhe5 Nd5
13 -8.68 25.3M 0:00.23 Kxh1 Kb7 Nxd8+ Kc7 Ne6+ Kd6 Nxg5 Be1 Ne4+ Kd5 Nxf6+ Ke5 Bg2 Be3 Ng4+ Kf4 Nxe3 Kxe3 Bh3
12 -8.98 16.7M 0:00.15 Kxh1 Kb7 Nxd8+ Kc7 Ne6+ Kd6 Nxg5 Ke5 Nf7+ Kf5 Bh3+ Kg6 Nxh8+ Kg7 Nf7 Kg8 Kg2 Bac3
11 -10.88 8.84M 0:00.08 Kxh1 B3b2 Bg2 Bd6 Nxd8+ Kb8 Nf7 Kc7 Nxg5 Bda3 Nf7 Be3 Nxh8 Bxh8
10 -13.79 3.46M 0:00.03 Kxh1 Bfe5 Bg2 Be1 Nxe5+ Kb8 Nc6+ Kc8 Nxd8 Kxd8 Kg1 Bed2 Be4
9 -14.25 1.72M 0:00.02 Kxh1 Bhg7 Bg2 Bd6 Nxd8+ Ka7 Nf7 Bde5 Nxg5 Bdxg5 Kg1
8 -14.44 762378 0:00.01 Kxh1 Bgf4 Nxd8 B6e5 Bg2+ Ka7 Nc6+ Kb6 Nxb4 Baxb4
7 -14.44 738487 0:00.01 Kxh1 Bgf4 Nxd8 B6e5 Bg2+ Ka7 Nc6+ Kb6 Nxb4
6 -15.78 222519 0:00.00 Kxh1 Bde7 Bg2 Bbd6 Nxe7+ Kb8 Nc6+ Kb7
5 -16.04 118924 0:00.00 Kxh1 Bhg7 Bg2 Bc5 Nxd8+ Ka7
4 -16.22 88183 0:00.00 Kxh1 Bfg7 Nxd8 Bac3
3 -17.36 28172 0:00.00 Kxh1 Bfd4 Nxb4
2 -17.43 11798 0:00.00 Kxh1 Bae5
1 NNUE evaluation using nn-62ef826d1a6d.nnue enabled
1 -18.84 1885 0:00.00 Nxb4
0 #

... and Confucius once said "永不說永不" ....
well that is probably not true, but he is on record for saying this "說自己可以做到的人和說自己不能做到的人……都是正確的" ...

it did find the mate-in-10 the first pass in just over 2 minutes ..a self-updated crystal which source can be found here
https://github.com/MichaelB7/Stockfish/tree/crystal

[Uri Blass]

Thanks.

My calculation was that in order to prove mate in 10 you probably need to calculate too many lines even if you know the idea.
I thought something like
if black has 30 moves in every ply then you need 30^9 nodes only to prove mate score by alphabeta algorithm
so I thought you need something different than alpha beta to solve the position but now when I think about it again I understand that hash tables can help significantly and I see that the bishops have only 32 squares and putting 9 bishops on 32 squares is only 32!/(9!*23!)=28048800 positions and even if you multiply it by some factor for king positions and for side to move it is something that chess engines can memorize.

Maybe we need something more complex to have something that is easy to prove but engines cannot prove by the alpha-beta algorithm.
.

[fridokar]

Even if the mate is not found in a short time, Fridolin 3.1 shows a positive score at least from the start.
https://sites.google.com/site/fridolinchess/

[peter]

A position some have said will never be solved by a chess program ...

[d]k2b3b/8/2N2b2/6b1/1b6/b7/3b2K1/b1b2B1n w - - 0 1

Sting has an alomost fully correct output- line in short time, yet doesn't make it for some longer to get the mating score into eval:

k2b3b/8/2N2b2/6b1/1b6/b7/3b2K1/b1b2B1n w - - 0 1

Analysis by Sting 24 (based on Stockfish 2.1.1) 64bit:

1.Kxh1
-+ (-30.62 ++) Depth: 7 00:00:04 5kN
...
1.Kxh1 Bd6 2.Bg2 Bb6 3.Bf3 Bhg7 4.Kg2
-+ (-31.55) Depth: 34 00:00:21 49862kN
1.Ba6
-+ (-31.47 ++) Depth: 35 00:00:36 134MN
1.Ba6
-+ (-31.39 ++) Depth: 35 00:00:38 152MN
1.Ba6 Bd6 2.Kxh1 Bb6 3.Kg2 Bac3 4.Bc8 Be1 5.Kf3 Bef2 6.Ke2 Bhg7 7.Kd3 Bdg3 8.Kc4 Bbc7 9.Kb5 Bad6 10.Ka6 Bcb4 11.Bb7#
-+ (-31.27 ++) Depth: 35 00:00:44 215MN
1.Ba6 Bd6 2.Kxh1 Bb6 3.Kg2 Bac3 4.Bc8 Be1 5.Kf3 Bef2 6.Ke2 Bhg7 7.Kd3 Bdg3 8.Kc4 Bbc7 9.Kb5 Bad6 10.Ka6 Bcb4 11.Bb7#
-+ (-31.09 ++) Depth: 35 00:01:12 584MN
1.Ba6 Bd6 2.Kxh1 Bb6 3.Kg2 Bac3 4.Bc8 Be1 5.Kf3 Bef2 6.Ke2 Bhg7 7.Kd3 Bdg3 8.Kc4 Bbc7 9.Kb5 Bad6 10.Ka6 Bcb4 11.Bb7#
-+ (-30.81 ++) Depth: 35 00:03:31 2634MN
Already 10 minutes now, neither a new iteration nor line.

24 threads of 12x3GHz Xeon X5670 CPU, 8G hash (maximum).

[Dann Corbit]

Regular (bleeding edge) stockfish has a terrible time with it:

Code: Select all

Searching: k2b3b/8/2N2b2/6b1/1b6/b7/3b2K1/b1b2B1n w - - 0 1
infinite: 0 time: 0 increment: 0 moves to go: 0

 1   +0.00   00:00   12730  Kf3 
 2  -16.50   00:00   79582  Kxh1 Kb7 
 3  -16.43   00:00  115229  Kxh1 Bdf4 Nxd8 Bxd8 
 4  -16.48   00:00  132271  Kxh1 Bd6 Nxd8 Bxd8 
 5  -16.50   00:00  146821  Kxh1 Bb6 Nxb4 Baxb4 Bg2+ Kb8 
 6  -16.19   00:00  166493  Kxh1 Bb6 Nxb4 Baxb4 Bg2+ Kb8 
 7  -16.36   00:00  183105  Kxh1 Bb6 Nxb4 Baxb4 Bg2+ Kb8 Bd5 Bad4 Kg2 
 8  -16.15   00:00  195270  Kxh1 Bc5 Nxd8 Bxd8 Bg2+ Kb8 Be4 Bhc3 
 9  -16.29   00:00  220723  Kxh1 Bb6 Nxb4 Bdxb4 Bg2+ Kb8 Be4 Bcf4 Kg2 Bac3 
10  -16.15   00:00  241246  Kxh1 Bc7 Nxb4 Baxb4 Bg2+ Ka7 Be4 Bba3 Kg2 Bhg7 Kf1 
11  -16.28   00:00  357668  Kxh1 Bc7 Nxb4 Bdxb4 Bg2+ Ka7 Be4 Bfd4 Kg2 Bhg7 
12  -16.01   00:00  414907  Kxh1 Bf8 Nxd8 Bxd8 Bg2+ Ka7 Kg1 Bad4+ Kf1 Bfh6 Be4 
13  -15.80   00:00  793004  Kxh1 Bc5 Nxd8 Bxd8 Bg2+ Ka7 Be4 Bhf6 Kg2 Bh6 Kf1 Bcb2 Bd5 
14  -15.73   00:00   1211K  Kxh1 Bc5 Nxd8 Bxd8 Bg2+ Ka7 Be4 Bdb6 Kg2 Bab4 Kf1 Bcb2 Bd5 Bhd4 Be4 
15  -15.59   00:00   1778K  Kxh1 Bb6 Nxb4 Bdxb4 Bg2+ Ka7 Be4 Bgf4 Kg2 Ba1b2 Kf1 Bbe5 Bd3 Bcb2 Be4 
16  -15.38   00:00   2924K  Kxh1 Bc5 Nxd8 Bxd8 Bg2+ Ka7 Be4 Bc7 Kg2 Bhg7 Kf1 Bge7 Bd3 Bae5 Be4 Bcb2 Bd5 
17  -15.24   00:00   4792K  Kxh1 Bc7 Nxb4 Baxb4 Bd3 Ka7 Be4 Kb6 Kg1 Kc5 Kf1 Kd4 Bf5 Bfe5 Kg2 Kc3 Be4 
18  -15.17   00:00   7275K  Kxh1 Bb6 Nxb4 Baxb4 Bd3 Bac3 Kg2 Ka7 Be4 Bfe7 Kf1 Bgf4 Bd3 B4a5 Be4 B1b2 Bd5 
19  -15.02   00:00  11515K  Kxh1 Bc5 Nxd8 Bxd8 Kg2 Bc7 Bd3 Kb8 Be4 Bce3 Kf1 Bhg7 Kg2 Bac3 Kf1 Bab4 Bd5 Bge7 Bc4 
20  -14.97   00:00  14888K  Kxh1 Bc7 Nxb4 Baxb4 Bd3 Kb8 Kg1 Bge3+ Kf1 Bfd4 Be4 Ba7 Bd5 Bhf6 Be4 B1d4 Kg2 
21  -14.76   00:00  19751K  Kxh1 Bc5 Nxd8 Bxd8 Kg2 Bdc3 Bd3 Bde7 Be4+ Ka7 Bf5 Kb6 Bd3 Ka5 Be4 Ba1b2 Bd5 B1d2 Be4 
22  -14.60   00:00  33532K  Kxh1 Bb6 Nxb4 Bdxb4 Bg2+ Ka7 Be4 Bh4 Kg2 Bbd4 Bd3 Bbe1 Kf1 Bhg7 Kg2 Bac3 Kf1 Bcg5 Bf5 Kb6 Be4 
23  -14.27   00:02 100847K  Kxh1 Kb7 Nxd8+ Bxd8 Bd3 Kb6 Kg1 Bhe5 Kf1 Bb8 Bc2 B4d6 Bd3 Bdf6 Be4 Kc5 Bd3 Bfe5 Be4 Bad4 Bd3 Bcb2 
24  -14.19   00:02 123509K  Kxh1 Kb7 Nxd8+ Bxd8 Bd3 Kb6 Kg1 Kc5 Kf1 Bba5 Bg6 Bhg7 Bc2 B7h6 Kg2 B1c3 Kf1 Bab6 Kg2 Kd4 Bf5 B1b2 Kf1 
25  -14.06   00:04 216468K  Kxh1 Kb7 Nxd8+ Bxd8 Bb5 Kc7 Bd3 Bbe7 Kg1 Kd6 Bg6 Bhg7 Bd3 B7h6 Kf1 Bdf4 Kg2 Bhf8 Be4 Baf6 Bf5 Bc7 Bd3 
26  -13.80   00:08 419971K  Kxh1 Bc7 Nxb4 Baxb4 Kg2 Bb8 Bd3 Kb7 Kf1 Bh4 Be4+ Kc7 Bf5 Bfe7 Kg2 Bhc3 Be6 Bhe1 Bc4 Bef6 Bd5 Bab2 Be4 Bf2 
27  -13.51   00:14 783080K  Nxd8 Bxd8 Kxh1 Ka7 Bd3 Kb6 Kg1 Bag7 Be4 Bh4 Kf1 Bdh6 Bf5 Bd6 Be4 Bac5 Bf5 B1e3 Bc2 Bde5 Bf5 Bde7 Be4 B7d6 Bd5 
28  -13.43   00:18 953222K  Nxd8 Bxd8 Kxh1 Kb7 Kg2 Bae5 Be2 Kc6 Bd3 Bef4 Be4+ Kc5 Bg6 Kd4 Bf5 Kc3 Kf1 Bhe5 Ke2 Bge7 Bg6 Bec7 Be4 Bb6 Kf1 Kb2 Kg2 
29  -13.28   00:27   1459M  Nxd8 Bxd8 Kxh1 Ka7 Bh3 Kb6 Kg1 Bae5 Bf5 Bbe7 Bc8 Kc5 Kf1 Bhg7 Be6 Bgf8 Ke2 B2a5 Bf5 Bg3 Kf1 Bcf4 Bd3 Kd4 Bf5 Bac1 
30  -13.11   00:31   1691M  Nxd8 Bxd8 Kxh1 Ka7 Bh3 Kb6 Kg2 Kc5 Kf3 Bhf6 Ke2 Bae5 Be6 Kd4 Bf5 Bb6 Bg6 Bbc7 Kf1 Bcd8 Bc2 Ke3 Bg6 Ba1 Bf5 Bh8 Bd7 Bae5 Ba4 Bb8 Bc6 
31  -13.00   00:41   2187M  Nxd8 Bxd8 Kxh1 Kb7 Bd3 Kb6 Kg2 Kc5 Bg6 Bhe5 Kf2 Bh4+ Kf3 Kd4 Be8 Bb6 Bg6 Bh8 Ke2 Bbd8 Bc2 Bhg5 Bf5 Bbe7 Bd7 Bg7 Be8 Be5 Bg6 Bac5 Be8 
32  -12.90   00:56   3015M  Nxd8 Bxd8 Kxh1 Kb7 Bd3 Kb6 Kg2 Bdf4 Kf1 Kc5 Bf5 Kd4 Ke2 Bhe5 Bh7 Bfd2 Bf5 Bge7 Bg6 Kc3 Bf5 Bg7 Kd1 Beg5 Ke2 B5h6 Be4 Kd4 Bf5 Bh8 Bh7 Be5 Kd1 Kc3 Be4 
33  -12.69   02:18   7373M  Nxb4 Bae5 Kxh1 Baxb4 Kg2 Ka7 Bd3 Bbd6 Kf3 Kb6 Bf5 Kc5 Bd7 Kd4 Be6 B6b4 Kg4 Ke3 Kh3 Bh2 Kg2 Bfe7 Bd7 B2e5 Bf5 Kd4 Bc8 B7f6 Kf3 Bhg7 Kg4 Bdc7 
34  -12.61   03:04   9850M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Kc6 Kf3 Kd5 Ke2 Kd4 Bh3 Bfe7 Kf3 Bba3 Bd7 Be5 Ba4 Bf8 Bd7 Bfd6 Ba4 Bg3 Bb3 Bh2 Ba4 Bde5 Bd7 Bb6 Bc6 Bed6 Bd7 Kc5 Kg4 Bb8 
35  -12.61   03:06   9947M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Kc6 Kf3 Kd5 Ke2 Kd4 Bh3 Be5 Kf3 Bed6 Bd7 Bhe5 Bc6 Bba3 Bd7 Bb8 Ba4 Bb6 Bc2 Bbd6 Kg4 Kc5 Bf5 Bbc7 Be6 Bb8 Bd7 Bh2 Be8 Bh8 Kf5 
36  -12.61   03:06   9984M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Kc6 Kf3 Kd5 Ke2 Kd4 Bh3 Be5 Bc8 Bba3 Kf3 Bed6 Bd7 Bhe5 Kg4 Bh2 Kf3 Bac5 Be6 Bde5 Ke2 B5a3 Kf3 Bhg3 Bd7 Bb8 Bh3 Bgd6 Bd7 B6c7 Bc8 Be5 Bd7 Kc5 Be6 Bh8 
37  -12.07   10:30  39588M  Kxh1 Bd6 Nxd8 Bxd8 Kg2 Ka7 Kf3 Bhe5 Ke4 Kb6 Kf5 Bh2 Bg2 Kc5 Ba8 Bb8 Be4 Bb6 Bc2 Kc4 Be4 Kb4 Bd3 Bhe5 Ke6 Bg1 Kf5 Bbd6 Be4 Bb6 Bd5 Kc5 Ba8 Bg3 Ke6 Bh2 Be4 Ba7 
38  -11.75   18:01  69203M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bb6 Bb1 B5d6 Ke8 Bhe5 Kd7 Kc5 Ba2 Bbd8 
39  -11.75   18:01  69216M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bb6 Bb1 B5d6 Ke8 Bhe5 Kd7 Kc5 Ba2 Bbd8 Bb1 Bg3 Be4 Kb6 
40  -11.75   18:02  69313M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bb6 Bb1 B5d6 Ke8 Bhe5 Kd7 Kc5 Ba2 Bbd8 Bb3 Kb6 Be6 Bh8 Ke8 Bh2 Bf5 
41  -11.75   18:53  72654M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bb6 Bb1 B5d6 Ke8 Bbd8 Bc2 Kc5 Kd7 Kb6 Bf5 Bh2 Ke8 Ba7 Kf7 Kc5 Bg4 Bab8 
42  -11.75   22:33  87054M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bb6 Bb1 B5d6 Ke8 Bbd8 Bc2 Kc5 Be4 Bhe5 Bd3 Bh4 Bf5 Beg3 Bd3 Bde5 Ba6 Bh8 
43  -11.75   22:34  87087M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Kg6 Kc5 Kf7 Bg3 Be4 Kb6 Ke6 Bgd6 Bf3 Kc5 Kf7 Bb6 Ke8 Bh2 Bd1 Bed6 Bh5 
44  -11.75   22:35  87161M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Kg6 Kc5 Kf7 Bg3 Be4 Kb6 Ke6 Bgd6 Bf3 Kc5 Kf7 Bb6 Ke8 Bh2 Bd1 Bed6 Bh5 Bbd8 
45  -11.75   22:36  87236M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Kg6 Kc5 Kf7 Bg3 Be4 Kb6 Ke6 Bgd6 Bf3 Kc5 Kf7 Bb6 Ke8 Bh2 Bg4 Bbd6 Bh5 B2e5 
46  -11.75   22:42  87641M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Kg6 Kc5 Kf7 Bg3 Be4 Kb6 Ke6 Bgd6 Bf3 Kc5 Kf7 Bb6 Ke8 Bhe5 Bg2 Bbd8 Kf7 Bd8c7 Ke8 
47  -11.75   23:10  89482M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Kg6 Kc5 Kf7 Bg3 Be4 Kb6 Ke6 Bgd6 Bf3 Kc5 Kf7 Bb6 Ke8 Bhe5 Bg2 Bbd8 Kf7 Bd8c7 Be4 Bb6 Bf5 
48  -11.75   24:22  94228M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Kg6 Kc5 Kf7 Kb6 Ke6 Bg3 Be4 Bgd6 Kf7 Bh2 Ke6 Kc5 Kf7 Bb6 Bf5 Bbd6 Bc2 B2e5 Ke6 Bbc7 Bb3 
49  -11.75   27:08 105212M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Kg6 Kc5 Kf7 Kb6 Ke6 Bg3 Be4 Bgd6 Kd7 Bhe5 Ke6 Bh2 Kf7 Bde5 Ke8 Kc5 Kf7 Bbc7 Bc2 Bh8 Bb3 Bhd6 Ke6 Bb8 
50  -11.75   27:10 105325M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bhg7 Kf3 Kc6 Kg4 Kc5 Kf5 Bae5 Be2 Kd4 Bg4 Bb8 Kg6 Kc5 Bf3 Bdc7 Kf7 Bh8 Be4 Bfe5 Bh7 Kd4 Kg6 Bbe7 Kf7 Bcd8 Ke6 Kc5 Kd7 B7d6 Ke6 Bhg7 Bg8 B8e7 Kd7 Bh8 Be6 Bf8 Kc8 Kb6 Kd7 Bbc7 Bg4 Bcd8 Ke6 Kc5 Kf5 
51  -11.75   29:21 113974M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bc7 Kf3 Kc6 Ke4 Bbd6 Kd3 Kc5 Kc2 Kb4 Kd3 Bcd8 Ke4 Kc5 Kf5 B2a5 Bd3 Bh4 Bc2 Bab2 Kg6 Bb8 Ba4 Kb6 Bc2 Bad2 Bf5 B2e5 Kf7 Bfe7 Be6 Bbd6 Bb3 Kc5 Be6 Bhg5 Bb3 B7f6 Ke6 Bb8 Kf5 Kd4 Be6 Bbd6 Kg6 B6c7 Bg4 Kc5 Bh3 Bcd6 
52  -11.75   29:27 114360M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bc7 Kf3 Kc6 Ke4 Bbd6 Kd3 Kc5 Kc2 Kb4 Kd3 Bcd8 Ke4 Kc5 Kf5 B2a5 Bd3 Bh4 Bc2 Bab2 Kg6 Bb8 Ba4 Kb6 Bc2 Bad2 Bf5 B2e5 Kf7 Bfe7 Be6 Bbd6 Bb3 Kc5 Be6 Bhg5 Bb3 Kd4 Be6 Bb8 Bc8 B7d6 Be6 Kc3 Bf5 B6e7 Kg8 Kd4 Bc8 B7d6 Kf7 
53  -11.75   29:48 115785M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bc7 Kf3 Kc6 Ke4 Bbd6 Kd3 Kc5 Kc2 Kb4 Kd3 Bcd8 Ke4 Kc5 Kf5 B2a5 Bd3 Bh4 Bc2 Bab2 Kg6 Bb8 Ba4 Kb6 Bc2 Bad2 Bf5 B2e5 Kf7 Bfe7 Be6 Bbd6 Bb3 Kc5 Be6 Bhg5 Bb3 Kd4 Be6 Bb8 Bc8 B7d6 Be6 Kc3 Bf5 Bh2 Be6 Kb4 Bf5 B6e7 
54  -11.75   30:19 117850M  Nxb4 Baxb4 Kxh1 Kb7 Kg2 Bc7 Kf3 Kc6 Ke4 Bbd6 Kd3 Kc5 Kc2 Kb4 Kd3 Bcd8 Ke4 Kc5 Kf5 B2a5 Bd3 Bh4 Bc2 Bab2 Kg6 Bb8 Ba4 Kb6 Bc2 Bad2 Bf5 B2e5 Kf7 Bfe7 Be6 Bbd6 Bb3 Kc5 Be6 Bhg5 Bb3 Kd4 Be6 Bb8 Bc8 B7d6 Be6 Kc3 Bd7 B6e7 Kg8 Kd4 Be6 Beb4 Bf5 

[fridokar]

After Nxb4 Baxb4 Kxh1 it's clearly a draw. Which engine then shows correctly an eval of "0.0" ?
Fridolin does it!

[MikeGL]

I think Lc0 might be able to get this mate easily.
Already seen similar endgames, where Lc0 trapped SF king while having material disadvantage but still won.

[Alayan]

Multiple bishops of the same color is a feature that doesn't happen in any sensible game of chess. Altering the rules to forbid promoting into a bishop of the same color as one already on the board would leave a game that is indistinguishable from chess when playing seriously from sensible positions.

A consequence is that many engines are terrible at handling positions with multiple bishops of the same color.

[OneTrickPony]

I think Lc0 might be able to get this mate easily.
Already seen similar endgames, where Lc0 trapped SF king while having material disadvantage but still won

It sucks at it. It thinks black is winning (95%) although it does show 1.Ba6 as a top choice

8 [+4.72] 1.Ba6 Bfd4 2.Kxh1 B4e3 3.Bc8 Bdb6 4.Be6 Kb7 5.Bd5 Kc7 6.Be4 Kd6 7.Kg2

(around 2M nodes so quite shallow still)

[Jouni]

[d]k2b3b/8/2N2b2/6b1/1b6/b7/3b2K1/b1b2B1n w - - 0 1
Which engine shows mate in 10? None I tried so far .Still difficult. I don't use Chest .