SF10 excels in Arasan suite: 189/200

[Dann Corbit]

There are also 3 positions with a negative score, but I think it fairly likely that all three can hold the draw:

[d]2kr3r/pp4pp/4pp2/2pq4/P1Nn4/4Q3/KP2B1PP/2RR4 b - - acd 65; acs 9960; am Qxg2; bm Qc6; c0 "Eklund-Berthelsen, DSU Jubilee 100-B corr 2003"; c3 "Qc6"; c5 "dorszcz.0380"; ce -152; id "arasan20.141"; pm Qc6; pv Qc6 a5;

[d]1q2r1k1/3R1pb1/3R2p1/7p/p3N3/2P1BP1P/1P3PK1/8 b - - acd 62; acs 9960; bm Rxe4; c0 "Komodo 10.4-Deep Shredder 13, Andreas Strangmüller 2017"; c3 "Rxe4"; ce -105; id "arasan20.080"; pm Rxe4; pv Rxe4 Rd8+ Re8 Rxb8 Rxb8 Bc1 Ra8 f4 Kf8 Rd7 Ke8 Rb7 a3 bxa3 Bxc3 Rc7 Bd4 Rc4 Bb6 a4 Kd7 Kf3 Ba5 Be3 Kd6 Rd4+ Ke7 Re4+ Kd7 Ke2 Bc7 Rd4+ Ke7 Rc4 Ba5 Bd4 Ke6 Kd3 Kd5 Be3 Ke6 Rd4 f5 Ke2 Rb8 Rc4 Kd7 Rc5 Bc7 Rb5 Ra8 Rb4 Ba5 Rd4+ Ke6 Rc4 Kd5 Rc5+ Kd6 Rc1 Bc7 Rc4 Bd8 Rd4+ Kc6 Rb4 Ba5 Rb1 Bc7 Ra1 Kd5;

[d]k1b4r/1p3p2/pq2pNp1/5n1p/P3QP2/1P1R1BPP/2P5/1K6 b - - acd 54; acs 9960; am Nxg3; bm Rd8; c0 "Deep Fritz 13 Q6600-Arasan 17.2 Q6600, SSDF 2014"; c3 "Rd8"; ce -99; id "arasan20.050"; pm Rd8; pv Rd8 Rc3 Nxg3 Qe5 Nf5 Rc7 Qb4 Ka2 Ka7 Ne4 Qd4 Rxf7 Rd5 Qc7 Rd7 Rxd7 Qxd7 Qc3 Qc6 Kb2 Qxc3+ Kxc3 Kb6 Ng5 a5 Kc4 Kc7 c3 b6 b4 axb4 cxb4 Ne3+ Kd4 Nc2+ Ke5 Nxb4 Be4 Kd7 Nxe6 Ke7 Ng5 Bd7 Bxg6 Bxa4 Bxh5 b5 Be2 Nc6+ Kd5 b4 Kc5 b3 h4 Nd8 Bd3 Be8 Kd4 b2 Kc3 Ne6 Nxe6;

[Milton]

Jouni,
I would like to try to reproduce your result, but I do not have a lot of experience with using epd files. I think I have figured out how to process epd files with Arena but I have one question. Do you force your gui to use all 3 minutes for each test position or do you let it abort the search after X number of consecutive ply with the correction soulution? If the latter, how may consecutive ply do you require? Thanks!

Stockfish is really good in Arasan suite. With 180s limit and 4 cores i5 it only missed 11 positions ( 1 try)! They were 10,41,71,96,116,120,124,137,159,169 and 189. Houdini is not even closely so good. How many solves Lc0 Kai? Even this hard one were now solved


[d]r2q1r2/1b2bpkp/p3p1p1/2ppP1P1/7R/1PN1BQR1/1PP2P1P/4K3 w - - 0 1

Analysis by Stockfish 10 64 BMI2:

1.Bd2 Bc6 2.Rgh3 Rh8 3.Ne2 Bxg5 4.Rg4 Bxd2+ 5.Kxd2 c4 6.Rf4 Qc7 7.Qc3 Rac8 8.bxc4 dxc4 9.Rxc4 h5 10.Nd4 Qd7 11.Kc1 Be4 12.Re3 Rxc4 13.Qxc4 Bd5 14.Qc3 Rb8 15.f4 a5 16.Qxa5 Qb7 17.b3 Ra8 18.Qc3 Qa6 19.Re1 Rc8 20.Qe3
=/+ (-0.51) Depth: 33/54 00:01:11 433mN, tb=32
1.Rxh7+
=/+ (-0.41 ++) Depth: 34/45 00:01:27 530mN, tb=41
1.Rxh7+
=/+ (-0.32 ++) Depth: 34/45 00:01:27 532mN, tb=42
1.Rxh7+
= (-0.17 ++) Depth: 34/45 00:01:27 533mN, tb=42
1.Rxh7+
= (0.02 ++) Depth: 34/45 00:01:27 535mN, tb=42
1.Rxh7+
= (0.29 ++) Depth: 34/45 00:01:27 536mN, tb=42
1.Rxh7+ Kxh7 2.Rh3+ Kg8 3.Qf4 Bxg5 4.Qxg5 Qxg5 5.Bxg5 f6 6.Bxf6 Rxf6 7.exf6 Bc6 8.Ne2 d4 9.Rh4 Kf7 10.Nf4 Rc8 11.Nd3 Bb5 12.Rh7+ Kxf6 13.Kd2 a5 14.Ra7 c4 15.Nc1 Rb8 16.Rxa5 cxb3 17.Nxb3 Bc4 18.Nxd4 Rxb2 19.Ra7 Bd5
= (0.09) Depth: 34/47 00:01:29 542mN, tb=103
1.Rxh7+
= (0.19 ++) Depth: 35/49 00:01:29 546mN, tb=117
1.Rxh7+
= (0.28 ++) Depth: 35/49 00:01:30 551mN, tb=120
1.Rxh7+
+/= (0.43 ++) Depth: 35/55 00:01:32 563mN, tb=1247
1.Rxh7+ Kxh7 2.Rh3+ Kg8 3.Qf4 Bxg5 4.Qxg5 Qxg5 5.Bxg5 f6 6.Bxf6 Rxf6 7.exf6 Bc6 8.Ne2 d4 9.Rh4 Kf7 10.Rh7+ Kxf6 11.Rc7 Bf3 12.Rxc5 e5 13.Ra5 Bb7 14.Nc1 e4 15.Ra4 Ke5 16.Ne2 Rd8 17.Rb4 Ba8 18.Kd2 Rf8 19.Rxd4 Rxf2 20.Rd8 Bd5 21.Ke3 Rxh2 22.Re8+ Kf6
+/= (0.31) Depth: 35/55 00:01:34 574mN, tb=1568
1.Rxh7+
+/= (0.41 ++) Depth: 36/47 00:01:36 585mN, tb=1972
1.Rxh7+
+/= (0.50 ++) Depth: 36/55 00:01:37 597mN, tb=2286
1.Rxh7+
+/= (0.65 ++) Depth: 36/55 00:01:39 608mN, tb=2791
1.Rxh7+ Kxh7 2.Rh3+ Kg8 3.Qf4 Bxg5 4.Qxg5 Qxg5 5.Bxg5 f6 6.Bxf6 Rxf6 7.exf6 Bc6 8.Rh6 d4 9.Na4 Bxa4 10.bxa4 Kf7 11.Rh7+ Kxf6 12.Rc7 Rb8 13.b3 d3 14.Rxc5 dxc2 15.Rxc2 Rxb3 16.Rc6 a5 17.Ra6 Rh3 18.Rxa5 Rxh2 19.Ra8 Rh3 20.a5
+/= (0.65) Depth: 36/55 00:01:39 610mN, tb=2812
1.Qf6+
+/- (0.75 ++) Depth: 37/52 00:01:42 631mN, tb=3929
1.Qf6+
+/- (0.84 ++) Depth: 37/52 00:01:42 632mN, tb=3983
1.Qf6+
+/- (0.99 ++) Depth: 37/52 00:01:43 632mN, tb=3983
1.Qf6+
+/- (1.19 ++) Depth: 37/52 00:01:43 633mN, tb=3991
1.Qf6+
+/- (1.46 ++) Depth: 37/52 00:01:43 633mN, tb=4024
1.Qf6+
+- (1.83 ++) Depth: 37/52 00:01:43 634mN, tb=4071
1.Qf6+
+- (2.31 ++) Depth: 37/52 00:01:43 635mN, tb=4115
1.Qf6+
+- (2.93 ++) Depth: 37/52 00:01:43 636mN, tb=4259
1.Qf6+
+- (3.74 ++) Depth: 37/52 00:01:43 638mN, tb=4365
1.Qf6+
+- (4.76 ++) Depth: 37/52 00:01:44 639mN, tb=4538
1.Qf6+
+- (6.06 ++) Depth: 37/52 00:01:44 645mN, tb=4649
1.Qf6+ Bxf6 2.gxf6+ Qxf6 3.exf6+ Kxf6 4.Bg5+ Kg7 5.Bh6+ Kg8 6.Bxf8 Rxf8 7.Na4 Bc6 8.Nxc5 Ra8 9.Kd2 a5 10.Kc3 e5 11.f4 e4 12.Kd4 Rd8 13.Rh6 Kg7 14.Rgh3 Rb8 15.Rxh7+ Kf6 16.Rg3 Rb4+ 17.Ke3 d4+ 18.Ke2
+- (6.82) Depth: 37/52 00:01:46 654mN, tb=5057
1.Qf6+
+- (6.92 ++) Depth: 38/48 00:01:47 666mN, tb=5197
1.Qf6+
+- (7.01 ++) Depth: 38/49 00:01:49 681mN, tb=5320
1.Qf6+
+- (7.16 ++) Depth: 38/49 00:01:52 698mN, tb=5445

[Jouni]

I abort after 5 consecutive plies (default in Fritz GUI). I have no 10 hours to run suite.

[Milton]

I abort after 5 consecutive plies (default in Fritz GUI). I have no 10 hours to run suite.

Thanks for your response. I totally agree with you. I just wasn't sure if there was some generally accepted standard to use when analyzing epd files.

P.S.
I also have Fritz GUI but was too stupid to figure out how to use it to analyze epd files. It seems like there is an intermediate step required to convert the file to a database recognized by Fritz, whereas Arena can read the file directly.
Thanks again.

[Jouni]

You haven't used Fritz from version 1 like me? Create cbh-database and paste/append EPD. Then Process test set.