"Poor Fish": very hard test suite
Moderators: hgm, Rebel, chrisw
-
- Posts: 144
- Joined: Fri Sep 12, 2014 4:57 am
- Location: Randwick Australia
Re: Grandelius - Raznikov "Poor Fish": very hard t
Eelco : Neither Komodo10 or Stockfish 7 64 could find 50] Nd3 after 20 minutes of searching.Once I entered the move into their program they registered White with a plus score of (+2.00). Commercial computer chess programs still have a bit of improvement to go it seems; especially when
dealing with truly subtle moves.
Allan
dealing with truly subtle moves.
Allan
-
- Posts: 1476
- Joined: Mon Jan 28, 2013 2:51 pm
Re: "Poor Fish": very hard test suite
Hi Marco,
Try
[d]4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24
24. Qxe5 is the solution, from the game Gusev - Auerbach, Chelyabinsk, 1946, http://www.chessgames.com/perl/chessgame?gid=1472988.
Try
Code: Select all
4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24
24. Qxe5 is the solution, from the game Gusev - Auerbach, Chelyabinsk, 1946, http://www.chessgames.com/perl/chessgame?gid=1472988.
mcostalba wrote:Actually multi pv gives even more help to these already super strong engines, anyhow I have removed the position you found from the list and here is the final one (some of them are very difficult to find even with multi pv):tpoppins wrote:Or is this test supposed to be MultiPV=1-only?
The last 2 are special, the last one perhaps or is not correct or is super hard, the one before the last:Code: Select all
rn1q1rk1/2pbb3/pn2p3/1p1pPpp1/3P4/1PNBBN2/P1P1Q1PP/R4R1K w - - bm Nxg5; id "arasan19.16"; c0 "Dann Corbit, CCC 2015"; r5n1/p1p1q2k/4b2p/3pB3/3PP1pP/8/PPPQ2P1/5RK1 w - - bm Qf4; id "arasan19.23"; c0 "Shulman-Marciano, Ubeda 1997"; rn1qr1k1/1p2bppp/p3p3/3pP3/P2P1B2/2RB1Q1P/1P3PP1/R5K1 w - - bm Bxh7+; id "arasan19.42"; c0 "Polugaevsky-Torre, London 1984"; 1n3rk1/3rbppp/p2p4/4pP2/Ppq1P3/1N2B3/1PP3PP/R2Q1R1K w - - bm f6; id "arasan19.44"; c0 "Kuzmin-Groszpeter, Kusadasi 1990"; 8/5p2/3p2p1/1bk4p/p2pBNnP/P5P1/1P3P2/4K3 b - - bm d3; id "arasan19.115"; c0 "Bondarevsky-Petrosian, Soviet ch-19 Moscow 1951"; 8/4nk2/1p3p2/1r1p2pp/1P1R1N1P/6P1/3KPP2/8 w - - bm Nd3; id "arasan19.116"; c0 "Grandelius-Raznikov, EU Youth Chess Ch B18, Albena 2011"; 3r1r1k/pp5p/4b1pb/6q1/3P4/4p1BP/PP2Q1PK/3RRB2 b - - bm Qxg3+; id "arasan19.137"; c0 "Apicella-Lautier, Clichy 2001"; r3k3/1p4p1/1Bb1Bp1p/P1p1bP1P/2Pp2P1/3P4/5K2/4R3 w - - bm g5; id "arasan19.154"; c0 "Nakamura-Giri, FIDE Grand Prix, London 2012"; 1r1rb1k1/5ppp/4p3/1p1p3P/1q2P2Q/pN3P2/PPP4P/1K1R2R1 w - - bm Rxg7+; id "arasan19.155"; c0 "Vescovi-Gschwendtner, Groningen op 1994 (Aemis Test Suite #30)"; r1bq1rk1/pp2bppp/1n2p3/3pP3/8/2RBBN2/PP2QPPP/2R3K1 w - - bm Bxh7+; id "arasan19.157"; c0 "Trent-Deslandes, Cap d'Agde CCAS op 2008"; r1b2rk1/1pq1nppp/pbn1p3/8/3N4/3BBN2/PPP1QPPP/3R1RK1 w - - bm Bxh7+; id "arasan19.159"; c0 "Leu-Muck, DDR-ch H137 corr 1973"; 8/4bBpp/3p4/P6P/2PN2p1/3k1b2/P7/6K1 w - - bm h6; id "arasan19.188"; c0 "Timman, 1c Corus {m} 2008"; r1b3r1/5p1k/p1n2P1p/P1qpp1P1/1p1p4/3P2Q1/BPPB2P1/R4RK1 w - - bm Kf2; id "arasan19.179"; c0 "Efremov-Fischer, Danube/4 ICCF 2012"; 4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - - bm f4; id "arasan19.189"; c0 "Šachová skladba 2011-12, 1st Prize";
[D] r1b3r1/5p1k/p1n2P1p/P1qpp1P1/1p1p4/3P2Q1/BPPB2P1/R4RK1 w - -
Is the one I like most
-
- Posts: 2684
- Joined: Sat Jun 14, 2008 9:17 pm
Re: "Poor Fish": very hard test suite
I have modified SF to run DBT in automatic mode, so SF checks for the validity of the bm and in the above case (and also many others) the best move is not valid, namely it misses the trailing '+'. It should be Rxb3+ instead of Rxb3. So SF skipped it.Vinvin wrote:Why this position is not selected ?mcostalba wrote:I have just finished testing with Double Blind Test (DBT) the full suite of 160 hard positions provided by Vincent. I have used current Stockfish modified by me to run automatically the DBT test on a epd file without need of human supervisor, so it ran overnight.
...
1r6/4k3/r2p2p1/2pR1p1p/2P1pP1P/pPK1P1P1/P7/1B6 b - - bm Rxb3;
I have now fixed these invalid bm and I have tested them again and here is the result (to append to the above results already posted).
Code: Select all
DBI 2
1r6/4k3/r2p2p1/2pR1p1p/2P1pP1P/pPK1P1P1/P7/1B6 b - - bm Rxb3+; c1 "diff=9";Brzozka-Bronstein; DBI 2
1rb4k/p5np/3p1rp1/1ppB4/2N2P2/1P2R1P1/P1P4P/4R1K1 w - - bm Re8+; c1 "diff=5";Bagirov-Grigorian; DBI 2
2b1r3/r2ppN2/8/1p1p1k2/pP1P4/2P3R1/PP3PP1/2K5 w - - bm Nd6+; c1 "diff=10"; DBI 2
5N2/6N1/q2k3p/8/6p1/4K1PP/B5P1/8 w - - bm Nf5+; DB 2
DBI 1
8/4K2p/6pr/5pk1/1pB2R2/p4PP1/8/8 w - - bm Rg4+; c1 "diff=7"; DBI 1
6b1/5p2/1p3Bp1/1P3pP1/5P2/pp1B2K1/pr4P1/k7 w - - bm Kh2; DBI 1
Code: Select all
+---+---+---+---+---+---+---+---+
| | r | | | | | | |
+---+---+---+---+---+---+---+---+
| | | | | k | | | |
+---+---+---+---+---+---+---+---+
| r | | | p | | | p | |
+---+---+---+---+---+---+---+---+
| | | p | R | | p | | p |
+---+---+---+---+---+---+---+---+
| | | P | | p | P | | P |
+---+---+---+---+---+---+---+---+
| p | P | K | | P | | P | |
+---+---+---+---+---+---+---+---+
| P | | | | | | | |
+---+---+---+---+---+---+---+---+
| | B | | | | | | |
+---+---+---+---+---+---+---+---+
Fen: 1r6/4k3/r2p2p1/2pR1p1p/2P1pP1P/pPK1P1P1/P7/1B6 b - - 0 1
Key: 2C8F898CDDC44011
Checkers:
Warm-up best move: ..Ra7, score: 554
Forcing best move: ..Rxb3+
After forcing best move, score: 101
-
- Posts: 2684
- Joined: Sat Jun 14, 2008 9:17 pm
Re: "Poor Fish": very hard test suite
Here we go:Gusev wrote:Hi Marco,
TryCode: Select all
4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24
Code: Select all
+---+---+---+---+---+---+---+---+
| | | | | q | | k | r |
+---+---+---+---+---+---+---+---+
| p | | | | | | | p |
+---+---+---+---+---+---+---+---+
| | p | r | Q | P | p | p | B |
+---+---+---+---+---+---+---+---+
| | | | | n | | | |
+---+---+---+---+---+---+---+---+
| | | | | P | | | |
+---+---+---+---+---+---+---+---+
| | | P | | | | | |
+---+---+---+---+---+---+---+---+
| P | P | | | B | | | P |
+---+---+---+---+---+---+---+---+
| R | | | | | | K | |
+---+---+---+---+---+---+---+---+
Fen: 4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24
Key: C153DD9E6280F0EE
Checkers:
Warm-up best move: Qa3, score: 0
Forcing best move: Qxe5
After forcing best move, score: -146
Can you confirm that Qxe5 is really better than Qa3, if this is teh case then this is a DBI 2 move!
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: "Poor Fish": very hard test suite
Sure, Qa3 is only draw, Qxe5 is winning.mcostalba wrote:Here we go:Gusev wrote:Hi Marco,
TryCode: Select all
4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24
So it seems it is a very hard move because score of bm is still lower.Code: Select all
+---+---+---+---+---+---+---+---+ | | | | | q | | k | r | +---+---+---+---+---+---+---+---+ | p | | | | | | | p | +---+---+---+---+---+---+---+---+ | | p | r | Q | P | p | p | B | +---+---+---+---+---+---+---+---+ | | | | | n | | | | +---+---+---+---+---+---+---+---+ | | | | | P | | | | +---+---+---+---+---+---+---+---+ | | | P | | | | | | +---+---+---+---+---+---+---+---+ | P | P | | | B | | | P | +---+---+---+---+---+---+---+---+ | R | | | | | | K | | +---+---+---+---+---+---+---+---+ Fen: 4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24 Key: C153DD9E6280F0EE Checkers: Warm-up best move: Qa3, score: 0 Forcing best move: Qxe5 After forcing best move, score: -146
Can you confirm that Qxe5 is really better than Qa3, if this is teh case then this is a DBI 2 move!
See the PGN is here : https://www.dropbox.com/s/4a879s438wv09 ... 2.pgn?dl=0
-
- Posts: 2684
- Joined: Sat Jun 14, 2008 9:17 pm
Re: "Poor Fish": very hard test suite
Thanks, but how can we verify that the black played in the best way? Perhaps he missed some good refutation?Vinvin wrote: Sure, Qa3 is only draw, Qxe5 is winning.
See the PGN is here : https://www.dropbox.com/s/4a879s438wv09 ... 2.pgn?dl=0
BTW I have finally finished with the 120secs DBT test of all your positions, fixing incorrect notation of checking moves and also fixing a issue when the position has black to move, in this latter case I missed scores are always relative to side to move, not from White point of view, so some weak positions wrongly slipped in.
After all the fixes and after adding Gusev's position, here we are with the final DBT candidates (with updated references to the sources):
Code: Select all
DBI 2
1b6/2p5/1kB5/1PN5/8/3K4/8/q7 w - - bm Nb3; id "hard2016.1"; DBI 2
1r3rk1/5p1p/p2pb1p1/3Np1P1/2pRPR2/qP2Q1N1/P1PK3P/8 w - - bm Nf5; id "hard2016.9"; DBI 2
1r6/4k3/r2p2p1/2pR1p1p/2P1pP1P/pPK1P1P1/P7/1B6 b - - bm Rxb3+; id "hard2016.10"; Brzozka-Bronstein; DBI 2
1rb4k/p5np/3p1rp1/1ppB4/2N2P2/1P2R1P1/P1P4P/4R1K1 w - - bm Re8+; id "hard2016.14";Bagirov-Grigorian; DBI 2
2b1r3/r2ppN2/8/1p1p1k2/pP1P4/2P3R1/PP3PP1/2K5 w - - bm Nd6+; id "hard2016.19"; DBI 2
2r1r1k1/4bppp/1q4n1/3pP3/p1pP4/PnR2NBP/1P4P1/1B1QR2K w - - bm e6; id "hard2016.29"; DBI 2
3k4/2pq3p/pp5R/4P3/P6P/2PN4/1PN3K1/8 w - - bm Nd4; id "hard2016.39"; DBI 2
3k4/4pp2/n2n4/5BN1/3N4/b7/4K3/8 w - - bm Nxf7+; id "hard2016.40"; DBI 2
4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - - bm f4; id "arasan19.189"; c0 "Šachová skladba 2011-12, 1st Prize"; DBI 2
4k1br/1K1p1n1r/2p2pN1/P2p1N2/2P3pP/5B2/P2P4/8 w - - bm Kc8; id "hard2016.49"; DBI 2
4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - bm Qxe5; id "hard2016.52"; DBI 2
5N2/6N1/q2k3p/8/6p1/4K1PP/B5P1/8 w - - bm Nf5+; id "hard2016.55"; DB 2
5nr1/2Pp2pk/3Pp1p1/4P1P1/6P1/5K2/8/7n w - - bm c8=N; id "hard2016.58" DBI 2
8/1p2b3/6k1/5p2/p2NbP1p/P1B1P1p1/1P5P/6K1 w - - bm h3; id "hard2016.78"; c0 "fortress From Smirin-HIARCS"; DBI 2
8/3P3k/n2K3p/2p3n1/1b4N1/2p1p1P1/8/3B4 w - - bm Nf6+; id "hard2016.82"; c0 "study by Van Breukelen"; DBI 2
q7/b1p5/kp1p4/p2PN3/PPP5/1K6/8/5B2 w - - bm c5+; id "hard2016.103"; DBI 2
r5k1/p1pb2bp/3p3r/P1pPp1p1/2B1Pq2/1R2QPp1/1P4PP/5RBK b - - bm Rxh2+; id "hard2016.145"; DBI 2
r6r/4k3/p2bP2p/2n2PP1/2BQn2P/1pP1p3/1q1BK3/3R3R b - - bm Qc2; id "hard2016.149"; DBI 2
4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24 bm Qxe5; id "gusev"; c0 "Gusev - Auerbach, Chelyabinsk, 1946"; DBI 2
DBI 1
1rb2r1k/3p1pp1/p3p2p/1p2P3/4q3/1N2B1R1/PPP1N1P1/2KR1B2 w - - bm Nf4; id "hard2016.13"; DBI 1
2b1rk2/5p2/p1P5/2p2P2/2p5/7B/P7/2KR4 w - - bm f6; id "arasan19.41"; c0 "study by Smyslov"; DBI 1
2bq1b2/1p3k2/1r1p1p1p/pNpP1PpP/P1P3P1/6B1/1P5Q/2K1R3 w - - bm Re6; id "hard2016.22"; DBI 1
2q2r1k/p1p1b1pp/br1p1n2/2p1N1BP/Q1P1P3/8/PP1N1PP1/2KR3R w - - bm Ng6+; id "hard2016.27"; DBI 1
2r5/prkpR1p1/2p1ppK1/P1p1N1B1/P1P1P3/8/8/8 w - - bm Bf4; id "hard2016.36"; DBI 1
6R1/8/2pB3k/2P4p/5p1q/5P2/4P1K1/8 w - - bm Bf8+; id "hard2016.64"; DBI 1
6b1/5p2/1p3Bp1/1P3pP1/5P2/pp1B2K1/pr4P1/k7 w - - bm Kh2; id "hard2016.65"; DBI 1
6k1/p7/8/4B3/Pp3p1P/qP5P/P1Rb2K1/8 w - - bm Kf3; id "hard2016.68"; DBI 1
7q/b1p5/1p1Npkb1/pPP2ppP/P1P5/3B2P1/5P1R/K3R3 w - - bm hxg6; id "hard2016.74"; DBI 1
8/5p2/N1p3p1/3pP1P1/7P/2kP4/Pb6/7K w - - bm h5; id "hard2016.85"; DBI 1
8/bp4Pk/4KB2/R6p/2pNp1P1/p7/3p1Pnq/2r5 w - - bm Nc6; id "hard2016.94"; DBI 1
q1r5/k3ppp1/2R5/PP1KP2r/2p4Q/2p1p3/5pP1/3R4 w - - bm Qd4+; id "hard2016.101"; DBI 1
r1b2rk1/2q1bppp/p2pp3/2n3PQ/1p1BP3/1BN5/PPP2P1P/2KR2R1 w - - bm Bf6; id "hard2016.108"; DBI 1
r1bqkb1r/ppp3pp/n3pn2/1N1p1p2/3P1B2/4P3/PPP2PPP/R2QKBNR w KQkq - bm c4; id "hard2016.118"; DBI 1
r2qrbk1/1b3p2/3p2pB/1p1P1N2/1np1P3/5NRP/1P3PP1/1B1Qn1K1 w - - bm N3h4; id "hard2016.130"; DBI 1
8/4K2p/6pr/5pk1/1pB2R2/p4PP1/8/8 w - - bm Rg4+; id "hard2016.83"; DBI 1
Now I have started the final step, the last test they have to pass to be accepted in the list. This one it will be very hard and long and it serves 2 purposes, ensure that the position is really hard and verify that best move is really a best move. So the test will be to redo a DBT but this time with 1 hour time(!) per position on my more powerful QUAD set at 3 threads, followed by a forced move search also of 1 hour.
If after 1 hour normal search still don't find bm while instead forced move research rises definitely above baseline, so that bm is confirmed, then the test is considered passed...results in a few days
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: "Poor Fish": very hard test suite
Here are the results :Vinvin wrote:...mcostalba wrote:So this test set will:
Code: Select all
1) Contain only positions from human games or human studies (with reference) 2) Be sound, so that between best move(s) and the others there is an important difference 3) Be really difficult to solve for engines
3) I plan to rerun the latest set next week.
The set with latest correction is here : http://home.scarlet.be/vincentlejeune/c ... 6beta6.epd
Code: Select all
Use Sec per move : 300 Sec per move
Analysis Lines : Minimum search depth=12
Activate abort analysis : True. Minimal analysis depth 24. Abort, if the correct move is held 3 additional ply
Solution 1st run 2nd run 3rd run
----------------------------------------------------------------------------------
1) Nc5-b3 Nc5-d7 Nc5-d7 Nc5-d7
2) Bb6-c7 Bb6xd8 Bb6xd8 Bb6xd8
3) Nf3-h2, g2-g4 Ra1-e1 Be3-d2 g2-g4 * 8 Sec
4) Rd6xf6 Qe5-d4 Qe5-d4 Qe5-d4
5) Bb1xh7 Bb1xh7 * 2 Sec Bb1xh7 * 3 Sec Bb1xh7 * 2 Sec
6) Rg1xg7 Rd1-d4 Rd1-d4 Rd1-d4
7) Nf3-d4 Nf3-d4 * 8 Sec Nf3-d4 * 30 Sec Nf3-d4 * 8 Sec
8) Bd3xh7 Rf1-h1 Rf1-h1 Rf1-h1
9) Ng3-f5 Ng3-f5 * 243 Sec Ng3-f5 * 266 Sec Nd5-f6
10) .. Rb8xb3, Kc3-c1 Ra6-a8 Ra6-a7 Ke7-e6
11) g2-g4 g2-g4 * 2 Sec g2-g4 * 4 Sec g2-g4 * 2 Sec
12) e4-e5 e4-e5 * 64 Sec e4-e5 * 68 Sec e4-e5 * 98 Sec
13) Ne2-f4 Be3-a7 Be3-a7 Be3-c5
14) Re3-e8 Nc4-d2 Nc4-d2 Nc4-d2
15) Be3xh6 Be3xh6 * 206 Sec Be3xh6 * 31 Sec Be3xh6 * 170 Sec
16) Rf1xf6 Rf1xf6 * 140 Sec Rf1xf6 * 37 Sec Bd4-a7
17) Rc8xc7 Kg7-f7 Kg7-f7 Kg7-f7
18) e3-e4 a2-a3 a2-a3 a2-a3
19) Nf7-d6 Nf7-h6 Nf7-h6 Nf7-h6
20) f5-f6 c6-c7 c6-c7 c6-c7
21) Rf2-f6 Rf2-f6 * 168 Sec Rf2-f6 * 66 Sec Rf2-f6 * 5 Sec
22) Re1-e6 Re1-e6 * 257 Sec b2-b3 Re1-e6 * 80 Sec
23) .. Rd6-f6 Rd6-d7 Rd6-f6 * 296 Sec Rd6-d7
24) Kh3-g2 Kh3-g2 * 14 Sec Kh3-g2 * 100 Sec Kh3-g2 * 3 Sec
25) .. h5-h4 h5-h4 * 0 Sec h5-h4 * 0 Sec h5-h4 * 0 Sec
26) c4-c5 Kd3-e3 Bf2-e1 Bf2-e3
27) Ne5-g6 Ne5-d3 Ne5-d3 Ne5-d3
28) Qc6-f3 Qc6-f3 * 133 Sec Qc6-f3 * 62 Sec Rc5-c1
29) e5-e6 Bb1-f5 Bb1-f5 Bb1-f5
30) Bd3xh7 Bd3xh7 * 211 Sec Bd3xh7 * 206 Sec Qd4xc5
31) Kh2-g3 Qf6-f4 Qf6-f4 Qf6-f4
32) Rf1-f5 Rf1-f5 * 52 Sec Rf1-f5 * 111 Sec Rf1-f5 * 117 Sec
33) Be3xc5 Be3xc5 * 10 Sec Be3xc5 * 43 Sec Be3xc5 * 12 Sec
34) Rd1-g1 Rd1-g1 * 19 Sec Rd1-g1 * 30 Sec Rd1-g1 * 29 Sec
35) Bd5xf7 Bd5xf7 * 27 Sec Bd5xf7 * 13 Sec Bd5xf7 * 24 Sec
36) Bg5-f4 Re7xd7 Bg5xf6 Bg5xf6
37) Bc2-a4 c3-c4 e5xf6 c3-c4
38) Qd8-a8 Qd8-f6 Qd8-f8 Qd8-f8
39) Nc2-d4 Kg2-f3 Kg2-f3 Kg2-f3
40) Ng5xf7 Bf5-b1 Bf5-c2 Bf5-c2
41) .. Bg7xd4 Bg7xd4 * 4 Sec Bg7xd4 * 33 Sec Bg7xd4 * 9 Sec
42) b2-b4 Bd4-g1 Bd4-g1 Bd4-g1
43) c2-c4 c2-c4 * 30 Sec c2-c4 * 23 Sec c2-c4 * 37 Sec
44) Rg3-g5 Rg3-h3 Rg3-h3 Rg3-h3
45) Qh3xh5 Qh3xh5 * 14 Sec Qh3xh5 * 13 Sec Qh3xh5 * 6 Sec
46) f3-f4 Ke8-d7 Ke8-d7 Ke8-d7
47) Re8xe5 Re8-c8 Re8-e7 Re8-c8
48) .. e6-e5, c3xc1 Nd7-f6 Nd7-f6 Qe7-d8
49) Kb7-c8 a5-a6 a5-a6 a5-a6
50) f5xe6 f5xe6 * 87 Sec f5xe6 * 128 Sec Qf3-c3
51) Kh7-h6 Kh7-h6 * 43 Sec Kh7-h6 * 58 Sec Kh7-h6 * 26 Sec
52) Qd6xe5 Qd6-a3 Qd6-a3 Qd6-a3
53) Ne4xf6 Ne4xf6 * 25 Sec Ne4xf6 * 21 Sec Ne4xf6 * 2 Sec
54) Qe2-h5 Qe2-h5 * 104 Sec Qe2-h5 * 16 Sec Qe2-h5 * 50 Sec
55) Ng7-f5 Ng7-f5 * 38 Sec Ng7-f5 * 54 Sec Ng7-f5 * 28 Sec
56) .. Rc2xf2 Rc2xf2 * 109 Sec Be7-c5 Rc2xf2 * 200 Sec
57) Rd3xb3 Rd3-e3 Rd3-e3 Rd3-e3
58) c7-c8N Kf3-g2 Kf3-g2 Kf3-g2
59) Qf5-e6 Rf7xc7 Rf7xc7 Rf7xc7
60) Rh5xg5 Rh5xg5 * 57 Sec Rh5xg5 * 19 Sec Rh5xg5 * 44 Sec
61) .. Qd7-h3 Qd7-h3 * 28 Sec Qd7-h3 * 25 Sec Qd7-h3 * 13 Sec
62) .. Re2xd2 Re2xg2 Re2xg2 Re2xg2
63) e4-e5 b3-b4 e4-e5 * 283 Sec e4-e5 * 262 Sec
64) Bd6-f8 Rg8-f8 Kg2-f1 Rg8-f8
65) Kg3-h2 Kg3-f3 Kg3-f3 Kg3-f3
66) Be3xc5 Kb7-b6 Kb7-b6 Re5-e8
67) Bg2xe4 Bg2xe4 * 98 Sec Bg2xe4 * 75 Sec Bg2xe4 * 70 Sec
68) Kg2-f3 Rc2xd2 Rc2xd2 Rc2xd2
69) Nc3-d5 Nc3-d5 * 47 Sec Nc3-d5 * 89 Sec Nc3-d5 * 78 Sec
70) b3-b4 b3-b4 * 26 Sec b3-b4 * 53 Sec b3-b4 * 26 Sec
71) Qd3xg6 Qd3xg6 * 109 Sec h4-h5 h4-h5
72) .. Re3xf3 Rf6-h6 Rf6-h6 Rf6-h6
73) Kh4-h5 Qf2xf3 Qf2xf3 Qf2xf3
74) h5xg6 Nd6-b7 Nd6-b7 Nd6-b7
75) Kg1-h2 Kg1-h2 * 3 Sec Kg1-h2 * 9 Sec Kg1-h2 * 4 Sec
76) .. Rd7xd4, Kc3-c1 Rd7xd4 * 60 Sec Rd7xd4 * 215 Sec Rd7xd4 * 31 Sec
77) Bh3-f5 Bh3-f5 * 45 Sec Rc3-f3 Bh3-f5 * 234 Sec
78) h2-h3 Bc3-e1 Bc3-e1 Bc3-e1
79) Be2-h5 Nh3xf2 Nh3xf2 Nh3xf2
80) Bc4-e2 Bc4-f1 Bc4-f1 Bc4-f1
81) f7-f8Q f7-f8Q * 201 Sec f7-f8Q * 256 Sec f7-f8Q * 137 Sec
82) Ng4-f6 Ng4xe3 Ng4xe3 Ng4xe3
83) Rf4-g4 Rf4-g4 * 203 Sec Bc4-d5 Rf4-g4 * 112 Sec
84) Nf4-d3 h4xg5 h4xg5 h4xg5
85) h4-h5 e5-e6 Kh1-g2 e5-e6
86) c3-c4 Qd2-h6 c3-c4 * 143 Sec c3-c4 * 288 Sec
87) Na5-b3 Nd4xc2 Nd4xc2 Nd4xc2
88) Be3-c1 Be3-c1 * 0 Sec Be3-c1 * 0 Sec Be3-c1 * 130 Sec
89) .. g6-g5, d5-d4 g6-g5 * 29 Sec g6-g5 * 6 Sec d5-d4 * 9 Sec
90) .. Bf5-h3, Bc3-c1 a4-a3 a4-a3 a4-a3
91) Rf1-b1 Rf1-c1 Rf1-c1 Rf1-b1 * 1 Sec
92) g3-g4 g3-g4 * 106 Sec Bf2-d4 Bf2-d4
93) g4-g5 g4-g5 * 3 Sec g4-g5 * 14 Sec g4-g5 * 120 Sec
94) Nd4-c6 g7-g8B g7-g8Q g7-g8Q
95) Re5xf5 Re2-d2 Re2-d2 Re2-d2
96) f2-f4 Na8-c7 Na8-c7 Na8-c7
97) .. d3-d2 Qd4-e3 d3-d2 * 0 Sec d3-d2 * 1 Sec
98) Bh5-f3 Nc3-b5 Nc3-b5 Nc3-b5
99) Rh1-h6 Bf7-e8 Bf7-e8 Bf7-e8
100) Bd8-c7 g6xh7 g6xh7 g6xh7
101) Qh4-d4 Qh4xe7 Qh4xe7 Qh4xe7
102) .. Ne8-c7 Ne8-c7 * 95 Sec Ne8-c7 * 51 Sec Ne8-c7 * 51 Sec
103) c4-c5 c4-c5 * 219 Sec b4-b5 b4-b5
104) Bf1-e2 g2-g3 g2-g3 g2-g3
105) Bg5-f6 Bg5-f6 * 3 Sec Bg5-f6 * 31 Sec Bg5-f6 * 3 Sec
106) Be4-g6 Be4-g6 * 242 Sec Be4-d3 Be4-d3
107) Bd3xh7 Rf1-f3 Bd3xh7 * 209 Sec Bd3xh7 * 215 Sec
108) Bd4-f6 Qh5-h6 Qh5-h6 Qh5-h6
109) Bd3-g6 Nc3-b5 Nc3-b5 Rf1-f3
110) Re1-e5 c3-c4 c3-c4 c3-c4
111) Kg1-f2, ... Kg1-f2 * 16 Sec Kg1-f2 * 73 Sec Kg1-f2 * 41 Sec
112) Bd3xh7 Bd3xh7 * 13 Sec Nd2-f3 Bd3xh7 * 41 Sec
113) .. Nb6-c4, Bc3xc1 a7-a5 a7-a5 a7-a5
114) Qe1-h4 Qe1-h4 * 24 Sec Qe1-h4 * 46 Sec Qe1-h4 * 151 Sec
115) Ne4-g5 g2-g4 Kc1-b1 g2-g4
116) Bd3xh7 Bd3xh7 * 8 Sec Bd3xh7 * 32 Sec Bd3xh7 * 73 Sec
117) .. Bg7-h6, Nc3-c1 Qd8-e8 Qd8-e8 Qd8-e8
118) c2-c4 Ng1-f3 Ng1-f3 Ng1-f3
119) Rh1xh7 Rh1xh7 * 21 Sec Rh1xh7 * 8 Sec Rh1xh7 * 5 Sec
120) Qe2xe8 Bd4xf6 Bd4xf6 Bd4xf6
121) Nd5-f6 Nd5-f6 * 77 Sec Rf1-d1 h2-h3
122) f5-f6 Kg1-h1 Kg1-h1 Kg1-h1
123) Nd4-f5 Nd4-f5 * 10 Sec Nd4-f5 * 14 Sec Nd4-f5 * 69 Sec
124) Qf3-f6 Qf3-f6 * 178 Sec b3xa4 b3xa4
125) Qf3-f6 Qf3-f6 * 117 Sec Qf3-f6 * 170 Sec Qf3-f6 * 55 Sec
126) Qg5-h6 Qg5-h6 * 163 Sec Qg5-h6 * 284 Sec O-O-O
127) Nf3xg5 Nf3xg5 * 11 Sec Nf3xg5 * 23 Sec Nf3xg5 * 25 Sec
128) Nd4-c6 Nd4-c6 * 181 Sec Nd4-c6 * 118 Sec Nd4-c6 * 70 Sec
129) Nf3-h4 Qd1xe1 Qd1xe1 Qd1xe1
130) .. Rd8xd4 Rd8xd4 * 256 Sec Rd8xd4 * 143 Sec Rd8xd4 * 115 Sec
131) Bg7xh6 Bg7xh6 * 41 Sec Bg7xh6 * 26 Sec Bg7xh6 * 29 Sec
132) g2-g4 g2-g4 * 20 Sec g2-g4 * 80 Sec g2-g4 * 12 Sec
133) .. Kg8-g7 e4xd3 e4xd3 e4xd3
134) Rf3-f6 Rf3-f6 * 83 Sec Rf3-f6 * 162 Sec Rf3-f6 * 61 Sec
135) Bd3xh7 Bd3xh7 * 38 Sec Bd3xh7 * 6 Sec Bd3xh7 * 129 Sec
136) Rd3-h3 Rd3-h3 * 195 Sec Rd3-h3 * 74 Sec Rd3-h3 * 143 Sec
137) Rd2-d1 Rd2-d1 * 0 Sec Rd2-d1 * 28 Sec Rd2-d1 * 160 Sec
138) Rd6xf6 Rd6xf6 * 37 Sec Rg1-g2 Rd6xf6 * 121 Sec
139) Qe2-h5 Bd5xc6 Bd5xc6 Qe2-h5 * 232 Sec
140) Bd3xh7 Bd4xc5 Bd3xh7 * 78 Sec Bd3xh7 * 231 Sec
141) Bg5-f6 Bg5-f6 * 6 Sec Bg5-f6 * 40 Sec Bg5-f6 * 5 Sec
142) .. c3xb2, c3xc1 c3xb2 * 73 Sec c3xb2 * 58 Sec c3xb2 * 35 Sec
143) .. e4-e3 e4-e3 * 184 Sec Ra8-f8 Ra8-f8
144) .. Rh6xh2 a7-a6 a7-a6 a7-a6
145) Qd2-f4 Qd2-f4 * 10 Sec Qd2-f4 * 35 Sec Qd2-f4 * 107 Sec
146) Nc3-d5 Qd2-e3 Qd2-e3 Qd2-e3
147) .. Bd7-h3 Bd7-h3 * 3 Sec Bd7-h3 * 68 Sec Bd7-h3 * 118 Sec
148) .. Qb2-c2, c3-c1 e3xd2 e3xd2 Qb2-c2 * 175 Sec
149) g2-g4 g2-g4 * 35 Sec g2-g4 * 155 Sec g2-g4 * 57 Sec
150) Bd3xh7 Ra1-c1 Ra1-c1 Ra1-c1
151) e5-e6 e5-e6 * 67 Sec e5-e6 * 70 Sec e5-e6 * 30 Sec
152) h3-h4 h3-h4 * 9 Sec h3-h4 * 0 Sec h3-h4 * 0 Sec
153) Na4xb6(?) Na4xb6 Na4xb6 Na4xb6
154) .. Rc5-b5(?) Rc5-b5 Rc5-b5 Rc5-b5
155) f5-f6 * 0 Sec Kh5-h6 * 0 Sec f5-f6 * 0 Sec
156) .. Kf6-e5 * 0 Sec Kf6-e5 * 0 Sec Kf6-e5 * 0 Sec
157) Bf5-g4 Bf5-g4 * 0 Sec Bf5-g4 * 0 Sec Bf5-g4 * 0 Sec
158) .. Ke8-f8 Ke8-f8 Ke8-f8
159) Kc6-d5, Bc4-d5 Kc6-d7 Kc6-d7 Bc4-a2
160) Bf4-e3, Bf4-d2, Bf4 Bf4-e3 * 3 Sec Bf4-e3 * 7 Sec Kh7-h6
1st run : 73 found
2nd run : 66 found
3rd run : 69 found
Found at least one time : 83 solutions
Found 3 times on 3 tries : 54 solutions
-
- Posts: 2684
- Joined: Sat Jun 14, 2008 9:17 pm
Re: "Poor Fish": very hard test suite
Thanks!Vinvin wrote:
Here are the results :Code: Select all
Use Sec per move : 300 Sec per move Analysis Lines : Minimum search depth=12 Activate abort analysis : True. Minimal analysis depth 24. Abort, if the correct move is held 3 additional ply
I have verified your results against DBT candidates and I am very satisfied to see that only a very small part of the DBT candidates was actually found in at least one of the 3 runs.
Here are the DBT candidates that did not survive your check:
Code: Select all
1r3rk1/5p1p/p2pb1p1/3Np1P1/2pRPR2/qP2Q1N1/P1PK3P/8 w - - bm Nf5; id "hard2016.9"; DBI 2
5N2/6N1/q2k3p/8/6p1/4K1PP/B5P1/8 w - - bm Nf5+; id "hard2016.55"; DB 2
q7/b1p5/kp1p4/p2PN3/PPP5/1K6/8/5B2 w - - bm c5+; id "hard2016.103"; DBI 2
r6r/4k3/p2bP2p/2n2PP1/2BQn2P/1pP1p3/1q1BK3/3R3R b - - bm Qc2; id "hard2016.148"; DBI 2
2bq1b2/1p3k2/1r1p1p1p/pNpP1PpP/P1P3P1/6B1/1P5Q/2K1R3 w - - bm Re6; id "hard2016.22"; DBI 1
8/4K2p/6pr/5pk1/1pB2R2/p4PP1/8/8 w - - bm Rg4+; id "hard2016.83"; DBI 1
To me it means that the DBT method is working great and in a very short amount of time and with slow hardware is able to detect a very hard position that would otherwise require much longer to be detected.
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: "Poor Fish": very hard test suite
Here's the analyze from SF-Matefinder after 1.Qxe5 fxe5 2.Rf1 :mcostalba wrote:Thanks, but how can we verify that the black played in the best way? Perhaps he missed some good refutation?Vinvin wrote: Sure, Qa3 is only draw, Qxe5 is winning.
See the PGN is here : https://www.dropbox.com/s/4a879s438wv09 ... 2.pgn?dl=0
(SF-Matefinder is always a good tool to check test position -> http://chess.ultimaiq.net/stockfish.html )
Code: Select all
FEN: 4q1kr/p6p/1pr1P1pB/4p3/4P3/2P5/PP2B2P/5RK1 b - - 1 2
Stockfish_160914_x64_modern_matefinder:
37/71 02:04 1 397 414k 11 194k +3,43 2. ... Rc7 3.Bd1 b5 4.Bb3 Rc4 5.a4 a6 6.Rf2 Qe7 7.Kg2 Qg7 8.Bxg7 Kxg7 9.Rf7+ Kh6 10.Bxc4 bxc4 11.e7 Re8 12.Kg3 g5 13.Rf6+ Kg7 14.Rxa6 Rxe7 15.Kg4 Rb7 16.Kxg5 h6+ 17.Rxh6 Rxb2 18.Rc6 Ra2 19.Rxc4 Rxh2 20.Rc7+ Kf8 21.a5 Ra2 22.Kf6 Ke8 23.Ke6 Kd8 24.Rc5 Ra3 25.Kxe5 Kd7 26.Kd5 Ra1
37/71 02:04 1 397 414k 11 194k +3,35 2. ... Qe7 3.Bd1 Rc4 4.Bb3 b5 5.Kg2 a6 6.a4 Qe8 7.Rf3 Qe7 8.Ba2 Qg7 9.Bxc4 bxc4 10.Bxg7 Kxg7 11.Rf7+ Kh6 12.e7 Re8 13.Kg3 g5 14.Rf6+ Kg7 15.Rxa6 Rxe7 16.Kg4 Rb7 17.Kxg5 Rxb2 18.Ra7+ Kf8 19.Rxh7 Rb3 20.a5 Ra3 21.Kf6 Ke8 22.Ke6 Kf8 23.Rh8+ Kg7 24.Ra8 Rxc3 25.Kxe5 Ra3 26.Kd4 c3 27.Kd3 c2+ 28.Kxc2 Ra4 29.a6 Rxe4 30.Ra7+ Kg6 31.Rd7 Ra4
---------------------------------------------------------------------------
38/72 03:25 2 291 678k 11 173k +3,97 2. ... Rc8 3.Bd1 b5 4.Bc2 Rc7 5.Bb3 Rc4 6.Rf2 a6 7.a4 Qe7 8.Kg2 Qe8 9.Rf3 Qe7 10.Ba2 Qg7 11.Bxc4 bxc4 12.Bxg7 Kxg7 13.Rf7+ Kh6 14.e7 Re8 15.Kg3 g5 16.Rf6+ Kg7 17.Rxa6 Rxe7 18.Kg4 Rb7 19.Kxg5 Kf7 20.Kf5 Ke7 21.Kxe5 h5 22.Rh6 Rxb2 23.Rxh5 Ra2 24.Rh7+ Kd8 25.Kd6 Rd2+ 26.Kc5 Ra2 27.Kb5 Rb2+ 28.Kxc4 Re2 29.Kd5 Rd2+ 30.Kc6 Rc2 31.a5 Rxc3+ 32.Kb5 Ra3 33.a6 Ra2 34.a7
38/72 03:25 2 291 678k 11 173k +3,64 2. ... Qe7 3.Bd1 Rc4 4.Bb3 b5 5.a4 a6 6.Kg2 Qe8 7.Rf3 Qe7 8.Ba2 Qg7 9.Bxc4 bxc4 10.Bxg7 Kxg7 11.Rf7+ Kh6 12.e7 Re8 13.Kg3 g5 14.Rf6+ Kg7 15.Rxa6 Rxe7 16.Kg4 Rb7 17.Kxg5 Kf7 18.Kf5 Rxb2 19.a5 Ke7 20.Rh6 Rb3 21.a6 Rxc3 22.Kxe5 Ra3 23.Rxh7+ Kd8 24.a7 Kc8 25.Kd4 c3 26.Kd3 c2+ 27.Kxc2 Ra5 28.Kb3
-
- Posts: 5228
- Joined: Thu Mar 09, 2006 9:40 am
- Full name: Vincent Lejeune
Re: "Poor Fish": very hard test suite
I added some themes to the positions :mcostalba wrote:Thanks, but how can we verify that the black played in the best way? Perhaps he missed some good refutation?Vinvin wrote: Sure, Qa3 is only draw, Qxe5 is winning.
See the PGN is here : https://www.dropbox.com/s/4a879s438wv09 ... 2.pgn?dl=0
BTW I have finally finished with the 120secs DBT test of all your positions, fixing incorrect notation of checking moves and also fixing a issue when the position has black to move, in this latter case I missed scores are always relative to side to move, not from White point of view, so some weak positions wrongly slipped in.
After all the fixes and after adding Gusev's position, here we are with the final DBT candidates (with updated references to the sources):
Code: Select all
DBI 2
Deep tactic, Nd7 is losing -> 1b6/2p5/1kB5/1PN5/8/3K4/8/q7 w - - bm Nb3; id "hard2016.1"; DBI 2
Deep tactic in a king attack -> 1r3rk1/5p1p/p2pb1p1/3Np1P1/2pRPR2/qP2Q1N1/P1PK3P/8 w - - bm Nf5; id "hard2016.9"; DBI 2
Break a fortress to enter in winngin EG (pawn promotion) -> 1r6/4k3/r2p2p1/2pR1p1p/2P1pP1P/pPK1P1P1/P7/1B6 b - - bm Rxb3+; id "hard2016.10"; Brzozka-Bronstein; DBI 2
Give an exchange to attack the king -> 1rb4k/p5np/3p1rp1/1ppB4/2N2P2/1P2R1P1/P1P4P/4R1K1 w - - bm Re8+; id "hard2016.14";Bagirov-Grigorian; DBI 2
Fortress (no square to enter) -> 2b1r3/r2ppN2/8/1p1p1k2/pP1P4/2P3R1/PP3PP1/2K5 w - - bm Nd6+; id "hard2016.19"; DBI 2
Tactic, king attack -> 2r1r1k1/4bppp/1q4n1/3pP3/p1pP4/PnR2NBP/1P4P1/1B1QR2K w - - bm e6; id "hard2016.29"; DBI 2
Trap the queen -> 3k4/2pq3p/pp5R/4P3/P6P/2PN4/1PN3K1/8 w - - bm Nd4; id "hard2016.39"; DBI 2
King locked, no progress possible -> 3k4/4pp2/n2n4/5BN1/3N4/b7/4K3/8 w - - bm Nxf7+; id "hard2016.40"; DBI 2
deflection to distract the bishop -> 4K1k1/8/1p5p/1Pp3b1/8/1P3P2/P1B2P2/8 w - - bm f4; id "arasan19.189"; c0 "Šachová skladba 2011-12, 1st Prize"; DBI 2
Long mate threat and underpromotion -> 4k1br/1K1p1n1r/2p2pN1/P2p1N2/2P3pP/5B2/P2P4/8 w - - bm Kc8; id "hard2016.49"; DBI 2
Rook trapped + continuous mate threat -> 4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - bm Qxe5; id "hard2016.52"; DBI 2
Fortress + stalemate -> 5N2/6N1/q2k3p/8/6p1/4K1PP/B5P1/8 w - - bm Nf5+; id "hard2016.55"; DB 2
Locked position + lack of good squares -> 5nr1/2Pp2pk/3Pp1p1/4P1P1/6P1/5K2/8/7n w - - bm c8=N; id "hard2016.58" DBI 2
Fortress -> 8/1p2b3/6k1/5p2/p2NbP1p/P1B1P1p1/1P5P/6K1 w - - bm h3; id "hard2016.78"; c0 "fortress From Smirin-HIARCS"; DBI 2
Long mate threat -> 8/3P3k/n2K3p/2p3n1/1b4N1/2p1p1P1/8/3B4 w - - bm Nf6+; id "hard2016.82"; c0 "study by Van Breukelen"; DBI 2
Locked position + lack of good squares -> q7/b1p5/kp1p4/p2PN3/PPP5/1K6/8/5B2 w - - bm c5+; id "hard2016.103"; DBI 2
continuous pin +mate threat -> r5k1/p1pb2bp/3p3r/P1pPp1p1/2B1Pq2/1R2QPp1/1P4PP/5RBK b - - bm Rxh2+; id "hard2016.145"; DBI 2
Tactic -> r6r/4k3/p2bP2p/2n2PP1/2BQn2P/1pP1p3/1q1BK3/3R3R b - - bm Qc2; id "hard2016.149"; DBI 2 -> a game by MVL, IIRC ???
Position already above ?? -> 4q1kr/p6p/1prQPppB/4n3/4P3/2P5/PP2B2P/R5K1 w - - 1 24 bm Qxe5; id "gusev"; c0 "Gusev - Auerbach, Chelyabinsk, 1946"; DBI 2
DBI 1
Tactic, attack the king and the queen -> 1rb2r1k/3p1pp1/p3p2p/1p2P3/4q3/1N2B1R1/PPP1N1P1/2KR1B2 w - - bm Nf4; id "hard2016.13"; DBI 1
Lack of good moves, blocked -> 2b1rk2/5p2/p1P5/2p2P2/2p5/7B/P7/2KR4 w - - bm f6; id "arasan19.41"; c0 "study by Smyslov"; DBI 1
Sack an exchange to attack -> 2bq1b2/1p3k2/1r1p1p1p/pNpP1PpP/P1P3P1/6B1/1P5Q/2K1R3 w - - bm Re6; id "hard2016.22"; DBI 1
King attack, demolition of the castle -> 2q2r1k/p1p1b1pp/br1p1n2/2p1N1BP/Q1P1P3/8/PP1N1PP1/2KR3R w - - bm Ng6+; id "hard2016.27"; DBI 1
Zugzwang + lack of space + king attack -> 2r5/prkpR1p1/2p1ppK1/P1p1N1B1/P1P1P3/8/8/8 w - - bm Bf4; id "hard2016.36"; DBI 1
King attack +continuous check discovery threat -> 6R1/8/2pB3k/2P4p/5p1q/5P2/4P1K1/8 w - - bm Bf8+; id "hard2016.64"; DBI 1
Long king walk to take on b6 + black lacks of good moves -> 6b1/5p2/1p3Bp1/1P3pP1/5P2/pp1B2K1/pr4P1/k7 w - - bm Kh2; id "hard2016.65"; DBI 1
Trapped queen + zugzwang -> 6k1/p7/8/4B3/Pp3p1P/qP5P/P1Rb2K1/8 w - - bm Kf3; id "hard2016.68"; DBI 1
Trapped pieces + zugzwang -> 7q/b1p5/1p1Npkb1/pPP2ppP/P1P5/3B2P1/5P1R/K3R3 w - - bm hxg6; id "hard2016.74"; DBI 1
Pawns promotion + Novotny sacrifce -> 8/5p2/N1p3p1/3pP1P1/7P/2kP4/Pb6/7K w - - bm h5; id "hard2016.85"; DBI 1
Mate network, king attack -> 8/bp4Pk/4KB2/R6p/2pNp1P1/p7/3p1Pnq/2r5 w - - bm Nc6; id "hard2016.94"; DBI 1
Mate attack -> q1r5/k3ppp1/2R5/PP1KP2r/2p4Q/2p1p3/5pP1/3R4 w - - bm Qd4+; id "hard2016.101"; DBI 1
King attack -> r1b2rk1/2q1bppp/p2pp3/2n3PQ/1p1BP3/1BN5/PPP2P1P/2KR2R1 w - - bm Bf6; id "hard2016.108"; DBI 1
Control the center in the opening ? -> r1bqkb1r/ppp3pp/n3pn2/1N1p1p2/3P1B2/4P3/PPP2PPP/R2QKBNR w KQkq - bm c4; id "hard2016.118"; DBI 1
Tactic : attack king and queen -> r2qrbk1/1b3p2/3p2pB/1p1P1N2/1np1P3/5NRP/1P3PP1/1B1Qn1K1 w - - bm N3h4; id "hard2016.130"; DBI 1
Zugzwang, long mate threat -> 8/4K2p/6pr/5pk1/1pB2R2/p4PP1/8/8 w - - bm Rg4+; id "hard2016.83"; DBI 1