"Poor Fish": very hard test suite

[Paloma]

I seems the 2nd ply in your Analysis ...Bf6 is not correct.

Deep Shredder 13 scores 1...Bxf4 much better (-0.38); (1...Bf6 -18.28)

Here Shredder's output after 1.f4 at depth 47/113

4K1k1/8/1p5p/1Pp3b1/5P2/1P6/P1B2P2/8 b - - 0 1
Analysis by Deep Shredder 13 x64 syz:

1. +/= (0.38): 1...Bxf4 2.Kd7 Bd2 3.Kc6 Ba5 4.Bf5 Kf7 5.f4 h5 6.Bh3 Kf6 7.Kc7 h4 8.Kc6 Kg6 9.Bc8 Kf6 10.a3 Ke7 ....
2. +- (18.28): 1...Bf6 2.a4 Bc3 3.f5 Kg7 4.Ke7 Bf6+ 5.Ke6 Bc3 6.Bd1 Kf8 7.Bh5 Be1 8.f3 Bf2 9.f6 Be3 10.Bg6 c4 ....

[MikeB]

I seems the 2nd ply in your Analysis ...Bf6 is not correct.

Deep Shredder 13 scores 1...Bxf4 much better (-0.38); (1...Bf6 -18.28)

Here Shredder's output after 1.f4 at depth 47/113

4K1k1/8/1p5p/1Pp3b1/5P2/1P6/P1B2P2/8 b - - 0 1
Analysis by Deep Shredder 13 x64 syz:

1. +/= (0.38): 1...Bxf4 2.Kd7 Bd2 3.Kc6 Ba5 4.Bf5 Kf7 5.f4 h5 6.Bh3 Kf6 7.Kc7 h4 8.Kc6 Kg6 9.Bc8 Kf6 10.a3 Ke7 ....
2. +- (18.28): 1...Bf6 2.a4 Bc3 3.f5 Kg7 4.Ke7 Bf6+ 5.Ke6 Bc3 6.Bd1 Kf8 7.Bh5 Be1 8.f3 Bf2 9.f6 Be3 10.Bg6 c4 ....

after Bxf4 , the game is lost very quickly for black

[d]4K1k1/8/1p5p/1Pp5/5b2/1P6/P1B2P2/8 w - - 0 2

Code: Select all

 info depth 37 seldepth 69 multipv 1 score cp 12725 nodes 1601148108 nps 41261386 hashfull 550 tbhits 1426563 time 38805 pv e8d7 f4d2 d7c6 d2a5 b3b4 c5b4 c2b3 g8g7 c6d5 g7f6 d5e4 h6h5 e4f4 f6g6 f4e5 h5h4 e5f4 g6f6 f2f3 h4h3 f4g3 f6e5 g3h3 e5f4 h3g2 f4e5 g2g3 e5f6 g3g4 f6e7 g4g5 e7d6 f3f4 d6c5 f4f5 c5b5 f5f6 b5c6 f6f7 b6b5 f7f8q a5c7 f8b4 c7h2 g5h4 h2g1 b4c3 c6b6 c3g3 g1c5 g3b8 b6a6 b8c8 a6b6 c8c5 b6b7 c5b5

Stockfish sees Bxf4 is actually worse Bf6 , with a score that shows a mate score is not too far away - but they both lose and f4 clearly wins..

one issue is that after Bd2 Kc6 Ba5 b4! - that's when Stockfish and Shredder see the position is lost, since after cxb4, Bb3 locks the bishop in at a5 and Bxb4 is no better really. below is shredder 13 analysis after Bxf4 with some back and forth, this is no better

Code: Select all

info depth 33 seldepth 80 time 174867 score cp -1678 nodes 434602688 nps 2485332 tbhits 945956 hashfull 343 pv g8f7 d7c6 h6h5 c6b6 h5h4 a2a4 h4h3 a4a5 f7e6 a5a6 f4b8 b6b7 b8f4 a6a7 h3h2 c2e4 e6e5 e4c6 e5d4 a7a8q d4c3 a8f8 f4d2 f8c5 c3b2 c5e5 d2c3 e5h2 c3d4 h2g3 b2c2 f2f4 d4c5 g3f3 c5f8 f4f5 c2b2 f5f6 f8c5 b3b4 c5d6 c6d5 d6f8 f3b3 b2c1 b3c3 c1b1 b7c7

[peter]

Hi Michael!

this one , the last one , is just super hard...
some back and forth using large hash tables with 12 cores , and a modified stockfish with mpv==2

[d]4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - - bm f4

Code: Select all

info depth 41 seldepth 76 multipv 1 score cp 2196 nodes 12639422838 nps 35086604 hashfull 999 tbhits 13800868 time 360235 pv f3f4 g5f6 a2a4 f6c3 f4f5 g8g7 e8e7 c3f6 e7e6 f6c3 c2d1 g7f8 f5f6 c3a5 d1h5 a5c3 h5g6 c3e1 e6d5 h6h5 g6h5 e1a5 d5c6 f8g8 c6d6 g8f8 d6d7 a5c3 d7c6 c3d4 h5e2 f8f7 e2c4 f7e8 c6b6 d4f6 b6b7 f6c3 b7a7 e8d8 b5b6 c3d4 b6b7 d8d7 b7b8q d4b2 b8g8 b2a1 g8d5 d7e7 d5f7 e7d8 f7f8 d8c7 f8e7 c7c6 e7e8 c6c7 e8b8 c7d7 a4a5 a1d4 b8b7 d7e8
 info depth 41 seldepth 76 multipv 2 score cp 239 nodes 12639422838 nps 35086604 hashfull 999 tbhits 13800868 time 360235 pv c2d1 g5h4 a2a4 h4g5 a4a5 b6a5 d1e2 g8g7 b5b6 g5f4 e8d7 g7f6 d7c6 f6e7 b6b7 h6h5 c6c5 f4c7 e2d3 e7d8 d3g6 h5h4 g6f5 c7e5 c5b6 e5d4 b6c6 d4e5 f5h3 e5b8 c6b6 d8e7 f3f4 e7d6 h3f5 d6d5 f5g4 d5d6 g4h3 d6e7 h3f5 e7f8 f5e6 b8f4 b6a7 f8e7 e6g4 e7f6 b7b8q f4b8 a7b8 f6g5

Here ist Eelco de Groot's SF- branch Rainbow Serpent (this one called Solstice Edition from 7.1.17) started with empty hash at 24 threads (but only 12 cores too) and 32G hash:

253: Salai,L - 10MV,C, 1st Prize, Sachova skladba 2011
4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - - 0 1

Analysis by Stockfish 070117 64:

1.Kd7 Bd2 2.Kc6 Ba5 3.a3 h5 4.b4 cxb4
+/- (1.51) Depth: 7/7 00:00:10 941kN
...
1.Kd7
+- (1.61 ++) Depth: 43/59 00:06:40 10945MN
1.f4
+- (1.78 ++) Depth: 43/59 00:09:33 16136MN
...
1.f4
+- (11.20 ++) Depth: 43/59 00:19:26 32882MN


I had two much shorter trials with less then 4 minutes Time to Solution (and Time to Eval as well) and a forth one about as long as the one copied but I couldn't wait long enough at none of them for a longer pv, the Serpent takes rather long to solve the fail high obviously.
The version I downloaded, Eelco had posted at Toga's Developers Forum then.

http://www.computerchess.info/tdbb/phpB ... e&start=10

http://www.computerchess.info/tdbb/phpB ... 30fc68b35e

The direct link was still working just now, when I tried it again.

The problem seems to be a one of nullmove pruning once again too, the key-move 4.b4! is to be found at once with engines the nullmove of which can be switched off or lowered, eg. komodo

[yanquis1972]

another position where (regular) houdini 5 excels...seems that some of the tactical mode code seeped into h5, as its frequently fastest of the big 3 on very difficult positions.


Engine: Houdini 5.01 (4096 MB)
by Robert Houdart
Hash cleared
17/32 0:00 +1.20 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Be4 Kf7 6.f4 h4 7.Bg2 Be1 8.Kg5 Bxf2 9.a5 bxa5 (3.320.307) 29383 TB:2
18/35 0:00 +1.22 1.Kd7 Bd2 2.Ke6 Be1 3.a4 h5 4.Kf5 c4 5.a5 Bxa5 6.bxc4 Bb4 7.Kg5 Bc5 8.Kxh5 Kf7 9.Kg4 Bxf2 10.Kg5 Bc5 11.f4 Be7+ (4.037.187) 29468 TB:2
18/37 0:00 +1.23 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Bd3 Ke7 6.f4 h4 7.Kg4 Ke6 8.f5+ Kf6 9.Kxh4 Be1 10.Kg3 (6.201.860) 29532 TB:2
19/37 0:00 +1.25 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Bd3 Ke7 6.f4 h4 7.Kg4 Ke6 8.f5+ Kf6 9.Kxh4 Be1 10.Kg3 Kg5 11.Be4 Bd2 12.Kf3 Bc3 (6.722.020) 29482 TB:2
20/37 0:00 +1.22 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Be4 Kf7 6.Bd3 Ke7 7.f4 h4 8.Kg4 Ke6 9.f5+ Kf6 10.Kxh4 Be1 11.Kg3 Kg5 12.Be4 Bd2 (12.391.373) 30296 TB:2
21/42 0:00 +1.29 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Kd6 6.f4 Kd5 7.Kg5 Kd4 8.Bf1 Ke4 9.Bg2+ Kd4 10.a5 Bxa5 11.f5 Bb4 12.f6 c4 13.bxc4 (19.601.363) 30868 TB:4
22/42 0:00 +1.30 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Kf7 6.f4 h4 7.Bf1 Ke7 8.Kg5 Kd6 9.Bg2 Ke6 10.Bh3+ Ke7 11.Kxh4 Bxf4 12.Kg4 (21.819.136) 31037 TB:4
23/43 0:01 +1.34 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Kf8 5.f4 Ke7 6.Kg5 Kd6 7.Be4 h4 8.Bg2 Ke6 9.Bh3+ Kd5 10.Kg4 Ke4 11.Bf1 Kd5 12.Kxh4 Bxf4 13.Kg4 Bd2 (38.965.684) 31373 TB:5
24/44 0:02 +1.42 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Kd6 6.f4 h4 7.Bf1 Be1 8.Kg5 Bxf2 9.f5 c4 10.bxc4 Ke7 11.f6+ Kf7 12.Bg2 Be3+ 13.Kf5 Bd4 14.Bd5+ (67.115.008) 31658 TB:56
24/50 0:02 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Kf8 5.f4 Ke7 6.Kg5 Kd6 7.Bd3 Kd5 8.Bf1 h4 9.Kxh4 Bxf4 10.Kg4 Bd2 11.Kf5 Kd6 12.Bc4 Bc3 13.f4 Bd2 14.Bd3 (85.150.609) 31525 TB:107
25/51 0:03 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.Bc4 Ba5 12.f4 Bd2 13.Bf1 Ke7 14.Bd3 (94.837.478) 31392 TB:205
26/51 0:03 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (113.258.284) 31295 TB:349
27/51 0:04 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (126.445.787) 30908 TB:454
27/51 0:04 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (126.471.586) 30907 TB:454
28/51 0:04 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (137.229.787) 30858 TB:503
28/51 0:04 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (137.332.113) 30854 TB:504
29/51 0:06 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (185.806.041) 30515 TB:1.022
29/51 0:06 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (185.807.710) 30515 TB:1.022
30/51 0:06 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (191.079.158) 30518 TB:1.052
30/51 0:06 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (191.251.189) 30517 TB:1.066
30/51 0:07 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (216.125.586) 30388 TB:1.545
30/51 0:07 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (216.127.136) 30389 TB:1.545
31/54 0:08 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (272.065.653) 30249 TB:2.290
31/54 0:08 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd7 13.Bc6+ Kd6 14.Bf3 (272.066.968) 30246 TB:2.290
32/60 0:10 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Ke7 14.Kg5 (318.120.658) 30130 TB:3.071
32/60 0:10 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Ke7 14.Kg5 (321.414.618) 30123 TB:3.129
33/60 0:12 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Ke7 14.Kg5 (384.808.645) 29981 TB:4.399
33/60 0:13 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Ke7 14.Kg5 (401.923.787) 29938 TB:4.760
34/60 0:16 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Ke7 14.Kg5 (488.255.388) 29972 TB:5.847
34/60 0:16 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Ke7 14.Kg5 (488.257.250) 29972 TB:5.847
35/60 0:17 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Be3 14.Kg5 (527.670.918) 29847 TB:7.510
35/60 0:17 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Be3 14.Kg5 (527.672.450) 29847 TB:7.510
36/64 0:27 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Be3 14.Kg5 (808.127.347) 29285 TB:14.689
36/64 0:27 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kd6 13.Bc6 Be3 14.Kg5 (808.129.042) 29285 TB:14.689
36/64 0:31 +1.43 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kf7 13.Bc6 Ke7 14.Kg5 (916.446.588) 29231 TB:18.041
36/64 0:32 +1.43-- 1.Kd7 Bd2 2.Ke6 h5 3.Kf5 Kf7 4.a4 Ke7 5.Bd3 Be1 6.Kg5 h4 7.Kxh4 Bxf2+ 8.Kg5 Be3+ 9.Kg4 Bd2 10.Kf5 Kd6 11.f4 Ke7 12.Be4 Kf7 13.Bc6 Ke7 14.Kg5 (941.381.603) 29185 TB:18.815
36/64 0:34 +1.55++ 1.f4 (993.741.379) 29206 TB:20.621
36/64 0:36 +1.64++ 1.f4 (1.063.805.044) 29169 TB:23.312
36/64 0:38 +1.77++ 1.f4 (1.123.113.832) 28866 TB:27.183

[peter]

And here another SF that can get the nullmove pruning switched off fully:

4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - - 0 1

Analysis by SF 171114IPx 64 POPCNTe:

1.Kd7 h5 2.Kc7 h4 3.Kxb6 Bd8+ 4.Kxc5 h3
+/- (0.82) Depth: 7/8 00:00:00 2kN
...
1.Kd7 Bd2 2.Ke6 Kg7 3.a4 Be1 4.f4 h5 5.Kf5 h4 6.Kg4 Kf6 7.Kxh4 Bxf2+ 8.Kh3 Be3 9.Kg4 Bd2 10.Kf3 Bc3 11.Ke4 Ke6 12.Bd1 Bd2 13.Be2 Kd6 14.Kf5 Kd7 15.Bf1 Ke7 16.Ke5 Kd7 17.Bg2 Bc3+ 18.Kd5 Ke7 19.Bh3 Bd2 20.Kc6 Ba5 21.Bd7 Kf7 22.Kd6 Kf6 23.Kc7 Kf7
+/= (0.59) Depth: 44/48 00:14:05 18227MN
1.f4
+/= (0.62 ++) Depth: 45/49 00:16:52 21919MN
...
1.f4 Bxf4 2.Kd7 Bd2 3.Kc6 Ba5 4.b4 cxb4 5.Bb3+ Kf8 6.Kd6 h5 7.Ke5 Ke7 8.Kf5 Kd6 9.f4 Kc5 10.Ke6 Kxb5 11.f5 h4 12.f6 h3 13.f7 h2 14.Bd5 b3 15.f8Q h1Q 16.Bxh1 bxa2 17.Qf1+ Kb4 18.Bd5 Kc3 19.Bxa2 Kd2 20.Bd5 Bc3 21.Be4 Ba5 22.Qd3+ Ke1 23.Qe3+ Kd1 24.Qf2 Bd2 25.Bb1
+- (4.60) Depth: 45/55 00:23:09 30367MN
1.f4
+- (4.63 ++) Depth: 46/55 00:23:29 30812MN
...
1.f4
+- (6.94 ++) Depth: 46/58 00:31:55 42429MN

This one gives a pv too, even if it takes a while,

[peter]

And here one more try with H5.01 started with empty hash again:

4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - - 0 1

Analysis by Houdini 5.01 Pro x64-popc:

1.Kd7 Bd2 2.a3 h5 3.Kc6 Ba5 4.b4 cxb4 5.axb4 Bxb4 6.Kxb6 Be1 7.Kc7 Bxf2 8.b6 Bg3+ 9.Kc8 h4
+/- (1.28) Depth: 12/24 00:00:00 287kN
...
1.Kd7 Bd2 2.Ke6 Kg7 3.a4 Ba5 4.f4 Bd2 5.f5 h5 6.f6+ Kf8 7.Be4 h4 8.Bg2 Bc3 9.f4 Ke8 10.Kf5 Kf7 11.Bd5+ Kf8 12.Kg5 h3 13.Kf5 h2 14.Kg4 Bxf6 15.Kg3 Bc3 16.Kxh2 Kg7 17.Kg3 Kf6 18.Kg4 Bd2 19.Bc6 Bc3 20.Kg3 Kf5 21.Kf3 Kf6 22.Ke4
+/- (1.32 --) Depth: 36/66 00:02:46 6046MN
1.f4
+/- (1.50 ++) Depth: 36/66 00:02:46 6046MN
...
1.f4 Bxf4 2.Kd7 Bd2 3.Kc6 Ba5 4.b4 cxb4 5.Bb3+ Kg7 6.Kd5 Kf6 7.Ke4 h5 8.Kf4 Kg6 9.Ke5 h4 10.Kf4 Kf6 11.Kg4 Ke5 12.f3 Kd4 13.Kxh4 Kc5 14.f4 Kxb5 15.Kh5 Kc5 16.Kg6 b5 17.f5 Bc7 18.f6 Bd6 19.Kh7 Kc6 20.f7 Bc5 21.Kg7 Bd4+ 22.Kg8 Bc5 23.f8Q Bxf8 24.Kxf8 Kb7 25.Ke7 Kc7 26.Kf6 Kd6 27.Kf5 Ke7 28.Ke5 Kd7 29.Kd5
+- (3.50) Depth: 36/70 00:08:46 19583MN
1.f4 Bxf4
+- (3.43 --) Depth: 37/70 00:08:53 19855MN
...
1.f4
+- (3.75 ++) Depth: 37/73 00:28:26 63247MN

[yanquis1972]

are you using tablebases peter? i'm able to reproduce immediately --

4096 MB Large Page Hash
10 threads used
Syzygy 6 men EGTB available - 510 tablebases found
Engine: Houdini 5.01 (4096 MB)
by Robert Houdart
18/34 0:00 +1.24 1.Kd7 Bd2 2.Ke6 Kg7 3.Kf5 h5 4.a4 Kf7 5.Bd3 Ke7 6.Be4 Kd6 7.f4 h4 8.Kg4 h3 9.Kxh3 Bxf4 (5.903.869) 30276 TB:3
18/36 0:00 +1.26 1.Kd7 Bd2 2.Ke6 Kg7 3.Kf5 h5 4.a4 Kf7 5.Bd3 Ke7 6.Be4 Kf7 7.f4 h4 8.Bg2 Be1 9.Kg5 c4 10.a5 cxb3 11.a6 b2 (6.946.728) 30335 TB:3
19/36 0:00 +1.16 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Bd3 Ke7 6.f4 h4 7.Kg4 Kf6 8.Kxh4 Bxf4 9.a5 bxa5 10.b6 Ke6 11.Kg4 Bb8 12.f4 Kd6 (11.966.413) 30371 TB:3
20/36 0:00 +1.18 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Bd3 Ke7 6.Be4 Be1 7.Kg5 Bxf2 8.Kxh5 Ke6 9.Bd3 Ke5 10.a5 bxa5 11.Bc4 (12.668.979) 30381 TB:3
21/38 0:00 +1.32 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 h4 8.Kg4 Kd5 9.Bf1 Ke4 10.Bh3 Kd4 11.f5 Ke5 12.Kxh4 (27.426.013) 31024 TB:7
22/38 0:00 +1.32 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 h4 8.Kg4 h3 9.Kxh3 Bxf4 10.Kg4 Bd2 11.Kf5 Kd5 12.Bc4+ Kd4 13.f4 Ke3 14.Kg4 (29.484.784) 31102 TB:7
23/42 0:01 +1.33 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 h4 8.Kg4 Kd5 9.Bf1 Ke6 10.Kg5 Kd5 11.Bg2+ Ke6 12.Bh3+ Ke7 13.Kxh4 Bxf4 14.Kg4 (34.427.881) 31156 TB:8
24/42 0:01 +1.34 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 h4 8.Kg4 Ke6 9.f5+ Kf6 10.Kxh4 Bg5+ 11.Kg3 Bd2 12.Kg4 Be1 13.Be4 Bd2 14.Bg2 (44.567.685) 31429 TB:17
24/44 0:01 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 h4 8.Kg4 Kd5 9.Bc4+ Ke4 10.Bf1 Kd5 11.Kxh4 Bxf4 12.Kg4 Bd2 13.f4 Ke6 14.Kg5 (53.593.973) 31507 TB:33
25/50 0:03 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Ke7 8.Kg5 h4 9.Kxh4 Bxf4 10.Kg4 Bd2 11.Kf5 Be1 12.f4 Bd2 13.Ke5 Be1 14.Be4 (109.129.227) 30637 TB:284
26/50 0:03 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kc7 8.Kg5 Kd6 9.Bc4 Kc7 10.Bf1 Kd6 11.Bg2 Ke7 12.Kxh5 Bxf4 13.Kg4 Bd2 14.Kf5 (113.441.261) 30610 TB:292
27/50 0:03 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kc7 8.Kg5 Kd6 9.Bc4 Kc7 10.Bf1 Kd6 11.Bg2 Ke7 12.Kxh5 Bxf4 13.Kg4 Bd2 14.Kf5 (116.570.202) 30603 TB:300
28/54 0:04 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kc7 8.Kg5 Kd6 9.Bc4 Kc7 10.Bf1 Kd6 11.Bg2 Ke7 12.Kxh5 Bxf4 13.Kg4 Bd2 14.Kf5 (130.631.678) 30592 TB:359
28/54 0:04 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kc7 8.Kg5 Kd6 9.Bc4 Kc7 10.Bf1 Kd6 11.Bg2 Ke7 12.Kxh5 Bxf4 13.Kg4 Bd2 14.Kf5 (130.633.350) 30593 TB:359
29/54 0:05 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kc7 8.Kg5 Kd6 9.Bc4 Kc7 10.Bf1 Kd6 11.Bg2 Ke7 12.Kxh5 Bxf4 13.Kg4 Bd2 14.Kf5 (162.051.330) 30512 TB:451
29/54 0:05 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kc7 8.Kg5 Kd6 9.Bc4 Kc7 10.Bf1 Kd6 11.Bg2 Ke7 12.Kxh5 Bxf4 13.Kg4 Bd2 14.Kf5 (162.053.115) 30512 TB:451
30/54 0:05 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kd5 8.Bf1 Kd4 9.Kg5 Kd5 10.Bg2+ Kd6 11.Kxh5 Bxf4 12.Kg4 Bd2 13.f4 Ke6 14.Kg5 (182.395.216) 30434 TB:638
30/54 0:05 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Kd5 8.Bf1 Kd4 9.Kg5 Kd5 10.Bg2+ Kd6 11.Kxh5 Bxf4 12.Kg4 Bd2 13.f4 Ke6 14.Kg5 (182.396.838) 30434 TB:638
30/54 0:06 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Ke7 8.Kg5 h4 9.Kxh4 Bxf4 10.Kg4 Bd2 11.Kf5 Be1 12.f4 Bd2 13.Bc4 Kd6 14.Be2 (205.716.244) 30471 TB:1.156
30/54 0:06 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 Kd6 7.f4 Ke7 8.Kg5 h4 9.Kxh4 Bxf4 10.Kg4 Bd2 11.Kf5 Be1 12.f4 Bd2 13.Bc4 Kd6 14.Be2 (205.872.977) 30472 TB:1.156
31/54 0:08 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Kd7 13.Bf1 Kd6 14.Be2 (246.990.027) 30594 TB:1.465
31/54 0:08 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Kd7 13.Bf1 Kd6 14.Be2 (252.732.972) 30556 TB:1.483
32/61 0:10 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Ke7 13.Bf1 Kd6 14.Be2 (323.815.556) 30416 TB:2.296
32/61 0:10 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Ke7 13.Bf1 Kd6 14.Be2 (323.874.121) 30416 TB:2.296
33/61 0:12 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Ke7 13.Bf1 Kd7 14.Be2 (392.763.585) 30482 TB:3.315
33/61 0:13 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Ke7 13.Bf1 Kd7 14.Be2 (402.642.694) 30466 TB:3.360
34/61 0:16 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Ke7 13.Bf1 Kd7 14.Be2 (497.421.755) 30428 TB:5.611
34/61 0:16 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd6 12.Kf5 Ke7 13.Bf1 Kd7 14.Be2 (497.825.041) 30429 TB:5.611
35/64 0:19 +1.40 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd7 12.Kf5 Kd6 13.Be2 Ke7 14.Kg5 (594.887.861) 30415 TB:7.273
35/72 0:30 +1.40-- 1.Kd7 Bd2 2.Ke6 Kg7 3.a4 h5 4.Kf5 Kf7 5.Be4 Ke7 6.Bd3 h4 7.Kg4 Be1 8.Kxh4 Bxf2+ 9.Kg5 Be1 10.f4 Bd2 11.Bc4 Kd7 12.Kf5 Kd6 13.Be2 Ke7 14.Kg5 (916.611.526) 30335 TB:15.653
35/72 0:34 +1.52++ 1.f4 (1.047.070.379) 30227 TB:21.589
35/72 0:38 +1.61++ 1.f4 (1.150.482.909) 30077 TB:26.434
35/72 0:41 +1.74++ 1.f4 (1.243.302.562) 30002 TB:34.654
35/72 0:43 +1.92++ 1.f4 (1.313.395.861) 29962 TB:41.067
35/72 0:48 +2.19++ 1.f4 (1.435.354.862) 29856 TB:47.824
35/72 0:56 +2.55++ 1.f4 (1.688.653.248) 29830 TB:62.559

[peter]

are you using tablebases peter?

I wasn't in this case, thinking it wouldn't help because it didn't as for SF MateFinder. Seems I was wrong as for Houdini.
Yet it needs 30" Time to Eval (and Solution, just considering it's not the only one winning move here, 1.Bd1 is a Major Dual) only with empty hash without tbs but with option Tactical.
Can reproduce your times with full Syzygys too, about 30" indeed,

[Eelco de Groot]

are you using tablebases peter?

I wasn't in this case, thinking it wouldn't help because it didn't as for SF MateFinder. Seems I was wrong as for Houdini.
Yet it needs 30" Time to Eval (and Solution, just considering it's not the only one winning move here, 1.Bd1 is a Major Dual) only with empty hash without tbs but with option Tactical.
Can reproduce your times with full Syzygys too, about 30" indeed,

Well, it's pretty good from Houdini, even if it needs either Syzygy or Tactical settings to be this fast, it is an accomplishment indeed and especially if older Houdini's could not do it. If they could, maybe we could try one of Richard Vida's decompiled Houdinippolito's Or one of Richard's own programs.

1. Bd1 indeed seems to be a dual Peter, so I am not the first to discover it... A pity for the prizewinning surprise first move f4, it is not uniqely winning. Rainbow Serpent is at the moment a bit slower than the public one that you tried, but should pick up more depth (better branching) AND it finds Bd1{!} But it does a lot of nullmoving (very high R reduction if nullmove is allowed) even compared to the Solstice one.


4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - -

Engine: Rainbow Serpent 20170105_038 HT (i7 6700, 3 hyperthreads, 512 MB)
by T. Romstad, M. Costalba, J. Kiiski, G. Linscott

27/41 0:02 +1.77 1.Kd7 Bd2 2.f4 h5 3.Bf5 Ba5 4.Kc7 Kf7
5.Bh3 h4 6.a3 Kf6 7.Bg4 Ke7 8.Bf5 Kf8
9.b4 cxb4 10.axb4 Bxb4 11.Kxb6 Be1
12.f3 Bd2 13.Kc6 Bxf4 14.b6 (38.408.622) 13653

28/41 0:03 +1.72 1.Kd7 Bd2 2.f4 h5 3.Kc6 Ba5 4.a3 Kf7
5.Bf5 Kf6 6.Bc8 h4 7.Bg4 Ke7 8.Bf5 Kf6
9.Bh3 Kf7 10.Kc7 Kf6 11.Bg4 Ke7
12.Bf5 Kf6 13.Bh3 Ke7 14.b4 (41.629.370) 13666

.
.
.
41/62 14:05 +1.70 1.Kd7 Bd2 2.Ke6 Kg7 3.Bd3 h5 4.Kf5 h4
5.Kg4 Kf6 6.Kxh4 Ke5 7.Kg3 Bc1 8.Kg4 Bh6
9.a4 Bd2 10.Bc2 Be1 11.f4+ Kf6
12.Kf3 Bc3 13.Ke4 Bb4 14.Kd5 (10.549.922.095) 12472

42/62 20:00 +1.70 1.Kd7 Bd2 2.Ke6 Kg7 3.Bd3 Be1 4.f4 h5
5.a4 Bd2 6.Kf5 Be1 7.Kg5 Bd2 8.Be4 Ba5
9.Kxh5 Kf6 10.Kh4 Bb4 11.Kg3 Be1
12.Kg4 Bd2 13.Bg2 Be1 14.Kf3 (15.015.557.525) 12511

43/62 27:02 +1.70 1.Kd7 Bd2 2.Ke6 Kg7 3.Bd3 h5 4.Kf5 h4
5.Kg4 Kf6 6.Kxh4 Bf4 7.Bc2 Bd2 8.Kg4 Ke5
9.Kg3 Bh6 10.Bd3 Bg5 11.Kg4 Bf4
12.Kh5 Kf6 13.a4 Bd2 14.f4 (20.357.240.083) 12547

44/62 44:56 +1.79++ 1.Bd1 (33.812.722.728) 12538

44/62 48:34 +1.87++ 1.Bd1 (36.636.416.123) 12570

44/62 50:55 +1.99++ 1.Bd1 (38.506.627.159) 12600

44/62 53:06 +2.16++ 1.Bd1 (40.151.573.823) 12600

44/62 55:41 +2.39++ 1.Bd1 (42.148.373.598) 12615

44/62 58:07 +2.69++ 1.Bd1 (44.000.075.864) 12615

44/62 61:22 +3.10++ 1.Bd1 (46.376.904.118) 12593

44/85 71:39 +3.62++ 1.Bd1 (54.062.196.842) 12572

44/85 85:18 +4.29++ 1.Bd1 (64.004.504.630) 12504

44/85 94:30 +5.15++ 1.f4 (70.401.477.563) 12415

44/85 94:31 +6.25++ 1.f4 (70.414.793.211) 12415

44/85 94:43 +7.63++ 1.f4 (70.561.212.000) 12414

44/85 103:48 +9.37++ 1.f4 (77.010.026.735) 12364

44/85 117:15 +11.58++ 1.f4 (86.771.596.072) 12333

44/85 150:12 +14.35++ 1.f4 (111.590.876.187) 12381

44/85 235:33 +17.83++ 1.f4 (177.951.768.021) 12590

44/85 351:35 +17.83 1.f4 Bxf4 2.Kd7 Bd2 3.Kc6 Ba5 4.b4 cxb4
5.Bb3+ Kg7 6.Kd5 Kf6 7.Ke4 Kg5 8.Ke5 h5
9.f4+ Kg6 10.f5+ Kg7 11.f6+ Kf8
12.Kf4 h4 13.Kg4 h3 14.Kxh3 (268.993.606.678) 12751

45/85 391:20 +17.91++ 1.f4 (300.284.600.323) 12788

45/85 412:16 +17.99++ 1.f4 (316.783.130.801) 12806

45/85 428:13 +18.11++ 1.f4 (329.370.628.785) 12819

45/85 448:53 +18.28++ 1.f4 (345.524.000.330) 12828

45/85 456:03 +18.51++ 1.f4 (351.167.515.655) 12833

45/85 460:41 +18.81++ 1.f4 (354.830.239.646) 12836

45/85 467:45 +19.22++ 1.f4 (360.352.448.837) 12839

45/85 474:53 +19.74++ 1.f4 (365.965.904.498) 12843

45/85 488:39 +20.41++ 1.f4 (377.033.229.511) 12859

45/85 504:13 +21.27++ 1.f4 {This is in Ferrari red } (389.381.078.376) 12870

[peter]

1. Bd1 indeed seems to be a dual Peter, so I am not the first to discover it... A pity for the prizewinning surprise first move f4, it is not uniqely winning.

Yes, this was discussed already when Marco brought it up here. 1.Bd1 can transpose if Black answers 1...Bd2 with 2.f4 and has other winning lines of its own after different answers from Black at first move,