maximal depth that stockfish changes its mind

[Uri Blass]

Following study is depth 99 with 3 cores

[d]qrb5/rk1p2K1/p2P4/Pp6/1N2n3/6p1/5nB1/6b1 w - - 0 1
Stockfish_18091012_x64_modern:
1/1 00:00 73 73k -36.33 1.Kg6
2/2 00:00 10k 1,945k -32.74 1.Nd3 Kc6 2.Nxf2
3/3 00:00 10k 1,663k -32.74 1.Nd3 Kc6 2.Nxf2
4/4 00:00 10k 1,709k -32.74 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6
5/6 00:00 10k 1,742k -32.74 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6
6/7 00:00 11k 1,787k -32.74 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6 4.Bxa8
7/8 00:00 11k 1,832k -32.74 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6 4.Bxa8 f1Q
8/9 00:00 12k 1,919k -32.69 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6 4.Bxa8 f1Q 5.Be4
9/11 00:00 12k 2,001k -32.69 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6 4.Bxa8 f1Q 5.Be4 Qa1+ 6.Kf7
10/15 00:00 16k 2,012k -32.69 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6 4.Bxa8 f1Q 5.Be4 Qa1+ 6.Kh6 Qxa5 7.Kg6
11/17 00:00 18k 1,982k -32.86 1.Nd3 Kc6 2.Nxf2 gxf2 3.Bxe4+ Kxd6 4.Bxa8 f1Q 5.Be4 Qa1+ 6.Kh6 Qxa5 7.Kg6 Bb7 8.Bb1 Bd4
12/20 00:00 37k 2,819k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
13/12 00:00 37k 2,847k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
14/12 00:00 37k 2,871k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
15/16 00:00 38k 2,895k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
16/16 00:00 38k 2,728k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
17/18 00:00 39k 2,790k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
18/20 00:00 40k 2,852k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
19/23 00:00 42k 2,809k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
20/14 00:00 44k 2,903k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
21/24 00:00 45k 2,821k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
22/25 00:00 48k 2,975k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
23/26 00:00 50k 2,952k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
24/22 00:00 54k 3,009k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
25/27 00:00 58k 3,034k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
26/23 00:00 61k 3,040k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
27/37 00:00 65k 3,103k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
28/31 00:00 70k 3,190k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
29/27 00:00 77k 3,098k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
30/27 00:00 83k 3,084k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
31/26 00:00 102k 2,910k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
32/25 00:00 107k 2,887k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
33/28 00:00 115k 2,872k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
34/36 00:00 200k 2,410k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
35/34 00:00 205k 2,381k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
36/32 00:00 209k 2,400k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
37/32 00:00 212k 2,414k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
38/33 00:00 219k 2,463k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
39/34 00:00 227k 2,438k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
40/33 00:00 243k 2,500k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
41/38 00:00 275k 2,621k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
42/30 00:00 304k 2,693k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
43/31 00:00 346k 2,790k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
44/38 00:00 400k 2,900k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
45/29 00:00 471k 3,037k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
46/34 00:00 524k 3,121k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
47/39 00:00 595k 3,197k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
48/22 00:00 688k 3,259k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
49/30 00:00 798k 3,340k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
50/28 00:00 875k 3,380k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
51/29 00:00 986k 3,447k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
52/24 00:00 1,043k 3,478k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
53/19 00:00 1,246k 3,570k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
54/34 00:00 1,414k 3,608k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
55/31 00:00 1,573k 3,658k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
56/29 00:00 1,759k 3,687k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
57/41 00:00 1,997k 3,739k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
58/37 00:00 2,171k 3,810k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
59/43 00:00 2,539k 3,912k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
60/40 00:00 2,960k 4,011k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
61/37 00:00 3,708k 4,143k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
62/43 00:00 3,956k 4,177k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
63/31 00:01 4,266k 4,207k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
64/38 00:01 4,619k 4,230k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
65/41 00:01 5,451k 4,319k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
66/46 00:01 6,042k 4,356k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
67/44 00:01 6,242k 4,341k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
68/40 00:01 6,799k 4,295k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
69/39 00:01 8,600k 4,377k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
70/43 00:02 9,572k 4,475k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
71/39 00:02 10,367k 4,543k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
72/48 00:02 11,620k 4,637k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
73/47 00:02 12,953k 4,715k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
74/43 00:03 14,659k 4,797k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
75/49 00:03 16,148k 4,805k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
76/40 00:03 18,741k 4,906k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
77/40 00:04 20,741k 4,926k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
78/40 00:04 21,735k 4,916k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
79/36 00:05 25,729k 4,862k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
80/36 00:06 29,520k 4,829k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
81/47 00:06 33,235k 4,816k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
82/37 00:07 37,539k 4,889k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
83/48 00:08 41,395k 4,953k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
84/41 00:09 45,367k 4,994k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
85/42 00:10 55,302k 5,096k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
86/41 00:12 66,269k 5,168k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
87/51 00:13 72,347k 5,201k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
88/64 00:16 89,014k 5,262k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
89/57 00:17 92,752k 5,262k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
90/43 00:19 105,307k 5,288k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
91/53 00:22 117,116k 5,297k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
92/47 00:23 123,149k 5,318k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
93/55 00:28 151,192k 5,367k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
94/43 00:34 186,294k 5,401k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
95/54 00:37 205,169k 5,410k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
96/52 00:41 224,696k 5,388k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
97/51 00:44 243,146k 5,404k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
98/39 00:53 286,358k 5,396k 0.00 1.Kg6 Bh2 2.Kf5 Bg1
99/49+ 01:00 323,633k 5,390k +M13 1.Kf7
99/49 01:24 464,433k 5,521k +M13 1.Kf7 Bh2 2.Kg6 Bg1 3.Kf5 Bh2 4.Ke5 Bg1 5.Kd4 Bh2 6.Ke3 Bg1 7.Ke2 Bh2 8.Kf1 Bg1 9.Kxg1 Nh3+ 10.Kf1 Nhg5 11.Ke2 Nf3 12.Bxf3 g2 13.Bxe4+
100/26 01:34 527,182k 5,593k +M13 1.Kf7 Bh2 2.Kg6 Bg1 3.Kf5 Bh2 4.Ke5 Bg1 5.Kd4 Bh2 6.Ke3 Bg1 7.Ke2 Bh2 8.Kf1 Bg1 9.Kxg1 Nh3+ 10.Kf1 Nhg5 11.Ke2 Nf3 12.Bxf3 g2 13.Bxe4+

[Spliffjiffer]

hi Uri,
bad study as it seems after 15 sec SF-dev search:
SF says Kh7 is a dual !?

[Uri Blass]

hi Uri,
bad study as it seems after 15 sec SF-dev search:
SF says Kh7 is a dual !?

I agree bad study.

Here is the source
https://www.chess.com/forum/view/endgames/test-position

[Spliffjiffer]

but its easy to correct this to a real study:
we put white king to h7 and the only move with mate in 13 is:
[D]qrb5/rk1p3K/p2P4/Pp6/1N2n3/6p1/5nB1/6b1 w - - 0 1
1.Kg7!! (only move) mate 13

[zullil]

but its easy to correct this to a real study:
we put white king to h7 and the only move with mate in 13 is:
[D]qrb5/rk1p3K/p2P4/Pp6/1N2n3/6p1/5nB1/6b1 w - - 0 1
1.Kg7!! (only move) mate 13

Current SF-dev finds the mate at depth 98 (one thread deterministic search)

info depth 98 seldepth 56 multipv 1 score mate 13 nodes 347238338 nps 3095643 hashfull 904 tbhits 0 time 112170 pv h7g7 g1h2 g7g6 h2g1 g6f5 g1h2 f5e5 h2g1 e5d4 g1h2 d4e3 h2g1 e3e2 g1h2 e2f1 h2g1 f1g1 f2h3 g1f1 h3g5 f1e2 g5f3 g2f3 g3g2 f3e4

[Spliffjiffer]

to change "nothing" AND make the problem a "study" (or bettter sayed: a mate-problem) this position seems most adequate to my eyes (makes the move most surprising to my eyes):
[D]qrb5/rk1p1K2/p2P4/Pp6/1N2n3/6p1/5nB1/6b1 w - - 0 1
again...Kg7!! (only move), mate 13
respect to the original author, he/she must have seen the dual in the original position...its misterious but anyway: nice kingwalk