Stockfish 16.10. plays great, but

[peter]

Another example
[fen]1q1k4/1n1p1b2/p1pPp1p1/r1P1PpPp/p1P2P1N/P1QK4/8/4B3 w - - 0 1[/fen]

SF15 solves instantly dev version has no idea

Analysis by Stockfish 161022:
...
1.Ke2 Rxc5 2.Bf2 Qa7 3.Qb2 a5 4.Nf3 Kc8 5.Nd2 Qa6 6.Ke1 h4 7.Qc3 h3 8.Qxh3 Rxc4 9.Nxc4 Qxc4 10.Qe3 c5 11.Qd2 Kb8 12.Be3 Ka7 13.Kf2 Kb6 14.Kg3 Qb3 15.Qc1 Kb5 16.Kh4 Be8 17.Bf2 Qd3 18.Qb2+ Kc6 19.Be3 Nd8 20.Qc1 Kd5 21.Qh1+ Kc4 22.Qc1+ Kb3 23.Qxc5 Nc6 24.Kg3 Qc3 25.Qxc3+ Kxc3
-+ (-7.01) Depth: 37/58 00:02:54 1009mN, tb=2003
1.Qb4 Kc8 2.Bf2 Qa7 3.Nf3 Rxc5 4.Ke2 a5 5.Qxa4 Qa6 6.Nd2 Rd5 7.Nf3 h4 8.Qc2 a4 9.Ke1 Ra5 10.Bxh4 c5 11.Kf1 Qc6 12.Qc3 Ra8 13.Bf2 Na5 14.Kg1 Nb3 15.Ne1 Kb7 16.Nf3 Qe4 17.Kg2 Kc8 18.Be3 Ra6 19.Kf2 Rb6 20.Kg3 Qc6 21.Bf2 Kb7
-+ (-7.22) Depth: 38/58 00:03:44 1282mN, tb=2192

Before patch is accepted some test suites must be run!?

Same like with first one, Jouni, here Qxa5 is chosen at once by 110922 at my hardware given above at more then one trial repeatedly.

BTW, I know this position or a very similar one for rather sure, but it isn't found by automatic database- search in my problem- storage, so there must be some displacement of any figure. Can you tell, if you changed an original one postition, used in some old test- suites too? I'd bet the well known one (discussed here in forum at least once too) was from an over the board- game, it's not a study, I just don't remember the opponents right now,
regards

[Jouni]

I took this from ACT4 test suite.

[peter]

I took this from ACT4 test suite.

Thanks, there I found it.
Yet pity that doesn't help me remembering the original one so far neither

[chessica]

The download adress from stockfish 16.x is where?

[Eduard]

Stockfish is not an engine to use for such positions. For Stockfish developers, only Elos count in the fish test. If a new version with a new code wins against the older one, then the newer one is better and replaces the old version. If you want more, you have to use a clone. I make my own.

[Eduard]

Such Positions are the reason why we need Clones!

1q1k4/1n1p1b2/p1pPp1p1/r1P1PpPp/p1P2P1N/P1QK4/8/4B3 w - - 0 1

My Engine with the pure code by Stockfish from 161022:

Analysis by Solista Standard:

1.Ke2 Rxc5 2.Bf2 Qa7 3.Nf3 Qb6 4.Qc1 Na5 5.Ne1 Nb3 6.Qc2 Qb7 7.Be3 Ra5 8.Nf3 Ke8 9.Qc3 c5 10.Kf2 Kf8 11.Qd3 Kg8 12.Bd2 Kg7 13.Be1 Rb5 14.cxb5 axb5 15.Qc3 Qe4 16.Qe3 Qc4 17.Kg3 Nc1 18.Ba5 b4 19.Nd2 Qc2 20.Bd8 Nd3
-+ (-6.89 --) Depth: 33/60 00:00:51 1048MN, tb=810
Black is clearly winning

My Solista Attack v2 Engine:

1q1k4/1n1p1b2/p1pPp1p1/r1P1PpPp/p1P2P1N/P1QK4/8/4B3 w - - 0 1

Analysis by Solista Attack v2:

1.Qxa5+ Nxa5 2.Bxa5+ Ke8 3.Bb6 Qb7 4.Kd2 Kf8 5.Kd3 Kg8 6.Kd4 Qa8 7.Kd3 Kg7 8.Kd2 Qb8 9.Ke2 Qh8 10.Kd2 Kg8 11.Ke3 Kh7 12.Kd3 Qe8 13.Kd2 Qb8 14.Kd3 Qb7 15.Kd4 Bg8 16.Ke3 Kh8 17.Ke2 Bf7 18.Kd3 Qc8 19.Kd2 Kg8 20.Ke3 Kg7 21.Ke2 Kh7 22.Ke3 Be8 23.Kd2 Qb8 24.Ke3 Kg7 25.Kd2 Kg8 26.Ke3 Qa8 27.Kd2 a5 28.Kd3 Qa6 29.Kd4 Qc8 30.Kd3 Qa8 31.Kd2
-/+ (-1.59) Depth: 55/60 00:00:46 906MN, tb=264
Black is clearly better

[peter]

I took this from ACT4 test suite.

Thanks, there I found it.
Yet pity that doesn't help me remembering the original one so far neither

Haha, mixed it up with A.B.Petrosian-L.Hazai 1970(?) 45...


[fen]2k5/q1p5/3p4/pPpPp1pp/N1P1Pp2/P4PbP/KQ4P1/8 b - - 0 45[/fen]


Despite of the Queen- sac and the kind of fortress arising then, it's yet something rather different, not only as for the colour of the Queen to be sacrificed, isn't it?

[peter]

The download adress from stockfish 16.x is where?

Guess Stockfish of 16th10th (October) was meant in thread- title.
As a matter of fact, there were 5 of them, SF development versions, that were coming up latest sunday

[Uri Blass]

it has lost SF15 ability to solve fast some famous positions e.g.

[fen]4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 0 1[/fen]

1.Qa3 Rxe6 2.Qxa7 Qf7 3.Qb8+ Qe8
= (0.00) Depth: 57/27 00:02:10 702mN, tb=107023
1.Qa3 Rxe6 2.Qxa7 Qf7 3.Qb8+ Qe8
= (0.00) Depth: 58/27 00:02:17 741mN, tb=113257
1.Qa3 Rxe6 2.Qxa7 Qf7 3.Qb8+ Qe8 4.Qb7 Qe7 5.Qc8+ Qe8 6.Qxe8+ Rxe8 7.a4 Kf7 8.Be3 Rb8 9.Bd4 Rhc8 10.h4 Nc6 11.Bf2 Ke7 12.b4 Ne5 13.Bd4 Nc6
= (0.00) Depth: 59/27 00:02:43 878mN, tb=167175
1.Qa3 Rxe6 2.Qxa7 Qf7 3.Qb8+ Qe8
= (0.00) Depth: 60/27 00:03:48 1212mN, tb=290004

SF15 solves instantly in every try!

You cannot apply razoring to any depth with only an eval condition to limit it and then expect that to be a "nearly non-functional change".
https://github.com/official-stockfish/S ... cefc3ef1fa
Ask any of the senior programmers in this forum what then happens I am just joking about the certainty but likelihood > 60 % that is what causes this. I suspected this would happen when I saw that patch yesterday And if you go back to the commit just before that, unfortunately I still have no working compiler to make the change in the source myself, sorry but the Dev build from Abrok.eu will have to do:

Code: Select all

4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - -

Engine: SF 2022101609 Old Razoring (512 MB)
gemaakt door the Stockfish developers (see AUTHORS f

28/09  0:01   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (1.347.380) 884 

29/11  0:01   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (1.553.374) 891 

30/08  0:02   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Da7 (1.824.438) 890 

31/08  0:02   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Da7 (2.104.072) 888 

32/13  0:02   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 De7 5.Db8+ De8 (2.522.165) 888 

33/13  0:03   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (3.203.078) 878 

34/09  0:04   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (3.771.263) 873 

35/10  0:05   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 De7 5.Da8+ (4.596.170) 866 

36/10  0:06   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 De7 5.Da8+ (5.383.908) 858 

37/13  0:07   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 De7 5.Db8+ De8 (6.321.960) 852 

38/13  0:11   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (9.450.400) 829 

39/13  0:14   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 De7 5.Dc8+ De8 (11.749.920) 820 

40/13  0:18   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 De7 5.Dc8+ De8 (14.825.002) 816 

41/13  0:22   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (17.858.012) 810 

42/11  0:25   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 De7 5.Dc8+ De8 (20.336.586) 810 

43/11  0:27   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (22.599.380) 812 

44/09  0:32   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 Df7 (26.154.062) 810 

45/09  0:38   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 Df7 (30.539.928) 803 

46/09  0:46   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Db7 Df7 (36.851.247) 787 

47/09  0:57   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (45.375.875) 791 

48/07  1:08   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (53.414.922) 783 

49/07  1:20   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (63.196.444) 781 

50/07  1:36   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (74.947.725) 775 

51/20  3:03   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Dxe8+ Txe8 5.a4 Kf7 6.Le3 Tb8 
                       7.Ld4 Thc8 8.b4 Pc6 9.Lf2 Pe5 10.Ld4 (135.366.677) 738 

52/65  4:50   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 
                       4.Dxe8+ Txe8 5.a4 Kf7 6.Le3 Tb8 
                       7.Ld4 Thc8 8.Kf2 Pc6 9.Le3 Ke7 
                       10.Ke1 Pe5 11.Ld4 Pc6 (209.897.190) 722 

53/37  5:01   0.00     1.Da3 Txe6 2.Dxa7 Df7 3.Da8+ De8 (217.812.753) 721 

It is not near non functional change but stockfish with 7 cores can solve the position after enough time(less than 10 minutes in first try)

FEN: 4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 0 1

Stockfish_22101610_x64_avx2:
NNUE evaluation using nn-ad9b42354671.nnue enabled
1/1 00:00 682 341k +0.64 Qd6-d5 Rc6xe6
2/2 00:00 1k 687k +0.44 Qd6-d5 Qe8xe6
3/3 00:00 2k 838k +0.44 Qd6-d5 Qe8xe6
4/3 00:00 2k 1,194k +0.56 Qd6-a3 Rc6xe6 Qa3xa7
5/4 00:00 3k 1,492k +1.25 Qd6-a3 Rc6xe6 Qa3xa7
6/5 00:00 4k 2,186k +2.09 Qd6-d5 Rc6xe6 Be2-b5
7/8 00:00 13k 4,421k +0.87 Qd6-a3 Rc6xe6 Qa3xa7 Re6-e7 Qa7xb6 Kg8-f7 Ra1-f1
8/11 00:00 17k 4,303k +0.74 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-b8+ Qe7-e8 Qb8-b7 Qe8-e7 Qb7-c8+ Qe7-e8
9/15 00:00 45k 7,534k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-b8+ Qe7-e8
10/11 00:00 66k 7,297k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-b8+ Qe7-e8 Qb8-a7
11/9 00:00 67k 7,477k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-b8+ Qe7-e8
12/10 00:00 74k 8,269k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-b8+ Qe7-e8 Qb8-b7 Qe8-e7 Qb7-b8+
13/16 00:00 96k 7,983k +0.31 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-f7 Qa7-b8+ Qf7-e8 Qb8-b7 Qe8-e7 Qb7xe7 Re6xe7
14/23 00:00 315k 7,322k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
15/11 00:00 366k 7,186k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
16/24 00:00 424k 7,179k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
17/11 00:00 450k 7,258k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
18/13 00:00 506k 7,235k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
19/24 00:00 550k 7,232k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
20/10 00:00 717k 7,172k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-a8+
21/9 00:00 819k 7,314k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
22/9 00:00 927k 7,415k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
23/33 00:00 1,756k 7,471k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
24/9 00:00 1,923k 7,513k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
25/12 00:00 2,161k 7,557k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8 Qc8-b7
26/11 00:00 2,386k 7,573k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
27/11 00:00 2,765k 7,597k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
28/9 00:00 3,118k 7,551k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
29/9 00:00 3,731k 7,553k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
30/9 00:00 4,008k 7,534k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
31/9 00:00 4,755k 7,488k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
32/16 00:00 5,298k 7,452k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8 Qc8-c7 Qe8-f7 Qc7-b8+ Qf7-e8 Qb8-c7
33/15 00:00 6,160k 7,440k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
34/15 00:01 7,610k 7,360k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
35/25 00:01 8,394k 7,324k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
36/25 00:01 10,959k 7,272k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
37/10 00:01 12,492k 7,258k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-a8+
38/9 00:01 14,002k 7,225k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
39/9 00:02 15,966k 7,179k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
40/9 00:02 17,747k 7,136k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
41/15 00:03 22,676k 7,031k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
42/15 00:04 33,955k 6,813k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8 Qc8-c7 Qe8-f7 Qc7-d8+ Qf7-e8
43/15 00:05 39,277k 6,740k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
44/15 00:06 44,674k 6,696k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
45/15 00:07 49,046k 6,677k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
46/35 00:08 56,692k 6,624k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
47/23 00:12 78,563k 6,330k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
48/28 00:14 91,921k 6,143k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
49/23 00:16 103,431k 6,151k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
50/29 00:20 123,853k 6,146k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8
51/62 00:54 318,032k 5,848k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8 Qc8xe8+ Re6xe8 a2-a4 Kg8-f7 Bh6-e3 Re8-b8 Be3-d4 Rh8-c8 Kg1-f2 Ne5-c6 Bd4-e3 Nc6-e5 Be3-d4
52/35 00:54 319,115k 5,850k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8
53/35 00:57 339,356k 5,855k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8 Qc8xe8+ Re6xe8 a2-a4 Kg8-f7 Bh6-e3 Re8-b8 Be3-d4 Rh8-c8 b2-b4 h7-h5 Ra1-f1 Kf7-e7 h2-h4 Rc8-c7 Kg1-g2 Rc7-d7 Rf1-a1 Ne5-c6 Bd4-e3 Rd7-c7 Kg2-f1 Nc6-e5 Be3-d4 Ne5-c6
54/35 00:58 341,772k 5,857k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
55/35 01:07 394,981k 5,841k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8 Qc8xe8+ Re6xe8 a2-a4 Kg8-f7 Bh6-e3 Re8-b8 Be3-d4 Rh8-c8 b2-b4 h7-h5 Ra1-f1 Kf7-e7 h2-h4 Rc8-c7 Kg1-g2 Rc7-d7 Rf1-a1 Ne5-c6 Bd4-e3 Rd7-c7 Kg2-f1 Nc6-e5 Be3-d4 Ne5-c6
56/62 02:05 679,477k 5,399k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8 Qc8xe8+ Re6xe8 a2-a4 Kg8-f7 Bh6-e3 Re8-b8 Be3-d4 Rh8-c8 b2-b4 h7-h5 h2-h4 Ne5-c6 Bd4-e3 Nc6-e5 Ra1-f1 Kf7-e7 Be3-d4 Rc8-c7 Kg1-g2 Rc7-d7 Rf1-a1 Ne5-c6 Bd4-e3 Rd7-c7 Kg2-f2 Nc6-e5 Be3-d4 Ne5-c6
57/39 02:06 680,552k 5,399k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
58/39 02:32 821,960k 5,382k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7 Qb7-c8+ Qe7-e8
59/39 02:37 851,202k 5,399k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
60/39 04:07 1,330,321k 5,383k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
61/39 04:43 1,519,611k 5,358k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
62/31 05:05 1,640,859k 5,370k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8 Qa8-b7 Qe8-e7
63/31 05:33 1,795,816k 5,389k 0.00 Qd6-a3 Rc6xe6 Qa3xa7 Qe8-e7 Qa7-a8+ Qe7-e8
64/36+ 07:12 2,353,264k 5,443k +0.05 Qd6xe5
64/36+ 07:13 2,360,455k 5,445k +0.10 Qd6xe5
64/75+ 08:46 2,941,274k 5,591k +6.33 Qd6xe5
64/81 09:27 3,230,172k 5,693k +6.80 Qd6xe5 f6xe5 Ra1-f1 Qe8-e7 Be2-d1 Rc6-c4 Bd1-b3 b6-b5 a2-a4 a7-a6 Bb3-a2 Qe7-g7 Bh6xg7 Kg8xg7 Rf1-f7+ Kg7-h6 Ba2xc4 b5xc4 Rf7-a7 Kh6-g5 Ra7xa6 Rh8-e8 Ra6-c6 Kg5-f4 a4-a5 Kf4xe4 Rc6xc4+ Ke4-d5 Rc4-a4 Re8xe6 a5-a6 Re6-e8 a6-a7 Re8-a8 Kg1-f2 g6-g5 Kf2-e3 h7-h5 Ra4-a5+ Kd5-c6 Ke3-e4 Kc6-b7 Ke4xe5 g5-g4 b2-b4 Ra8-e8+ Ke5-f4 Re8-f8+ Kf4-g5 Kb7-a8 Ra5-f5 Rf8-e8 Kg5-f4 Re8-e2 Rf5xh5 Re2-c2
65/86+ 10:29 3,725,036k 5,917k +7.03 Qd6xe5
65/86+ 12:20 4,513,360k 6,094k +7.62 Qd6xe5
65/86+ 55:05 23,858,842k 7,218k +10.53 Qd6xe5
65/103 1:33:52 40,744,024k 7,233k +11.38 Qd6xe5 f6xe5 Ra1-f1 Rc6-c8 Be2-d1 b6-b5 Bd1-b3 Rc8-c4 a2-a4 a7-a6 Bb3-a2 Qe8-e7 Kg1-g2 Qe7-g7 Ba2xc4 b5xc4 Bh6xg7 Kg8xg7 Rf1-f7+ Kg7-h6 e6-e7 Rh8-e8 Kg2-g3 g6-g5 Rf7-f6+ Kh6-g7 Rf6xa6 h7-h5 Ra6-a7 Kg7-f7 h2-h4 g5-g4 a4-a5 Kf7-e6 Ra7-a6+ Ke6xe7 Ra6-h6 Re8-a8 a5-a6 Ke7-f7 Rh6xh5 Ra8xa6 Rh5xe5 Ra6-b6 Re5-f5+ Kf7-g6 h4-h5+ Kg6-g7 Kg3xg4 Rb6xb2 Rf5-c5 Kg7-f7 Rc5xc4 Kf7-e6 Rc4-c5 Rb2-g2+ Kg4-f4 Rg2-f2+ Kf4-g3 Rf2-f1 h5-h6 Rf1-h1 Kg3-g4 Rh1-g1+ Kg4-h5 Rg1-e1 Kh5-g6 Re1-g1+ Rc5-g5 Rg1-f1 Rg5-f5 Rf1-h1 h6-h7 Rh1-h4 Rf5-g5
66/90 2:26:56 69,116,918k 7,839k +12.03 Qd6xe5 f6xe5 Ra1-f1 Rc6-c8 Be2-d1 b6-b5 Bd1-b3 Rc8-c4 a2-a4 a7-a6 Bb3-a2 Qe8-e7 Kg1-g2 Qe7-g7 Ba2xc4 b5xc4 Bh6xg7 Kg8xg7 Rf1-f7+ Kg7-h6 e6-e7 Rh8-e8 Kg2-g3 g6-g5 Rf7-f6+ Kh6-g7 Rf6xa6 Kg7-f7 Ra6-a7 h7-h5 a4-a5 Kf7-e6 Ra7-b7 Ke6-d6 Rb7-b6+ Kd6xe7 a5-a6 Ke7-f7 h2-h4 g5-g4 Rb6-h6 Re8-a8 Rh6xh5 Ra8xa6 Rh5xe5 Ra6-b6 Re5-f5+ Kf7-g6 h4-h5+ Kg6-g7 Rf5-c5 Rb6xb2 Kg3xg4 Kg7-f7 Rc5xc4 Kf7-e6 Rc4-c5 Rb2-g2+ Kg4-f4 Rg2-f2+ Kf4-g3 Rf2-f1 h5-h6 Rf1-h1 Kg3-g4 Ke6-f7 Rc5-c6 Kf7-e8 Kg4-f5 Ke8-d7 Rc6-a6 Kd7-c7 e4-e5 Rh1-h5+ Kf5-e4 Kc7-b7 Ra6-g6 Rh5-h1 Rg6-f6 Kb7-c8 Ke4-d5 Rh1-h5 Rf6-c6+ Kc8-b8

[Eduard]

I took this from ACT4 test suite.

Thanks, there I found it.
Yet pity that doesn't help me remembering the original one so far neither

Haha, mixed it up with A.B.Petrosian-L.Hazai 1970(?) 45...


[fen]2k5/q1p5/3p4/pPpPp1pp/N1P1Pp2/P4PbP/KQ4P1/8 b - - 0 45[/fen]


Despite of the Queen- sac and the kind of fortress arising then, it's yet something rather different, not only as for the colour of the Queen to be sacrificed, isn't it?

Solver King ShashChess is the best here. I saw that this engine has a lot of code from Crystal in it. Since most test positions consist of 80% tactical, 15% Zugzwang and 5% blocking, it is no wonder that such an engine is good at it. In practical play, however, 99.99% of all positions I play with engines are postitionaly nature. There is a long way to go before such positions can occur, 99.9% of the time it doesn't happen at all. And then Shashchess (just my opinion) is weaker than Stockfish dev. Nevertheless: Compliments to the author who has developed a great solution engine here!

2k5/q1p5/3p4/pPpPp1pp/N1P1Pp2/P4PbP/KQ4P1/8 b - - 0 1

Analysis by ShashChess 24:

45...Qb6 46.Kb3 h4 47.Qd2 Qa7 48.Qe2 Kd7 49.Qb2 Kd8 50.Kc3 Qb6 51.Qb1 Kc8 52.Kd2 Kb7 53.Kd1 Kc8 54.Ke2 Kd8 55.Qd1 Kd7 56.Qb3 Kc8 57.Qb2 Kb7 58.Kd3 Kc8 59.Qc2 Kb8 60.Qd2 Kc8 61.Kc3 Kd7 62.Kb3 Kc8 63.Qd1 Kb8 64.Qg1 Kb7 65.Kc2 Kc8 66.Nxb6+ cxb6 67.Kb2 Kb7
+/= (0.31) Depth: 47/47 00:00:16 374MN, tb=1169
White is slightly better

Analysis by Solista Attack v2:

45...Qb6 46.h4
+- (2.52 --) Depth: 42/43 00:00:41 918MN, tb=51550