Mate Test Suite

Discussion of anything and everything relating to chess playing software and machines.

Moderator: Ras

Vinvin
Posts: 5334
Joined: Thu Mar 09, 2006 9:40 am
Full name: Vincent Lejeune

Re: Mate Test Suite

Post by Vinvin »

One more, mate in 6 :
[d]2r2rk1/2q2p1p/p1p1nPpQ/3p4/4p1B1/8/PPP3PP/R4R1K w - - 0 1
Rf5!!
BBauer wrote:
FWIW, I can get SF to solve the first position reasonably quickly for the cost of ~10 Elo, instantly for ~50.
The first position is solved by f4xe5 too. Nearly any engine will win this position. Same holds for other positions. So no need for elo loss.
Here my modification for your test set.

Code: Select all

8/2Nb4/pp6/4rp1p/1Pp1pPkP/PpPpR3/1B1P2N1/1K6 w - - bm Kc1 fxe5; #5;
8/8/2B1N3/3rp3/4k2K/7Q/2r3Pn/1b1N4 w - - bm Kh5 Kg5; #5;
qb6/3N2p1/r2p4/pP1k2p1/1pp3R1/6BB/2P1P3/2N1K3 w - - bm Bh4 bxa6; #5;
1K1N1b2/RPp1pr2/1kP5/2p5/P7/4B1P1/4p1b1/6n1 w - - bm Kc8 Bd2; #6;
n1N3br/2p1Bpkr/1pP2R1b/pP1p1PpR/Pp4P1/1P6/1K1P4/8 w - - bm Rh1 Nd6; #6;
1N1K1b1r/P3pPp1/4k1P1/rp1pB1RN/q4RP1/8/p2pB1p1/1b6 w - - bm Nd7 Rxa4; #6;
5R2/2P2pK1/2P2P2/1Pp1BP1P/b6p/1p1RPB2/1p2NPn1/6rk w - - bm Ra8 Bxb2 h6; #6;
3K4/1p1B4/bB1k4/rpR1p3/2ppppp1/8/RPPPPP2/r1n5 w - - bm b4 Rxa5; #7;
4k1r1/2pnp3/2B1N3/8/8/5R2/6P1/5K2 w - - bm Rf2 Ke1 Rf8+; #11;
n2Bqk2/5p1p/Q4KP1/p7/8/8/8/8 w - - bm Qc8; #13; 
8/5K2/3p3p/3p3P/pp1P4/rkb1R3/p1p3P1/n1B2B2 w - - bm Rd3; #16;
1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - - bm bxc6; #17;
8/6p1/p7/rp1K2p1/kb3pP1/2p2p1b/P1Np1P2/3N2R1 w - - bm Ke6; #20;
8/7p/6pP/5pP1/3BpP2/p1KpP3/pn1N4/k7 w - - bm Bh8 Be5; #21;
1B3N2/5p1B/2K2pn1/5krb/4p1p1/4P1P1/2P1PP2/8 w - - bm Ba7; #22;
1k3b1q/pP2p1p1/P1K1P1Pp/7P/2B5/8/8/8 w - - bm Bb5 Kd5; #27;
4rk2/2P2p2/p4P2/2p2b2/2p5/8/P7/2KR4 w - - bm Rd8; #29 ;
Kind regards
Bernhard
Top
zullil
Posts: 6442
Joined: Tue Jan 09, 2007 12:31 am
Location: PA USA
Full name: Louis Zulli

Re: Mate Test Suite

Post by zullil »

Vinvin wrote:One more, mate in 6 :
[d]2r2rk1/2q2p1p/p1p1nPpQ/3p4/4p1B1/8/PPP3PP/R4R1K w - - 0 1
Rf5!!
This one Stockfish can solve quickly, so it's much simpler (for SF) than most of the others in this thread:

Code: Select all

info depth 34 seldepth 50 multipv 1 score cp 235 lowerbound nodes 1160409450 nps 22175688 hashfull 464 tbhits 0 time 52328 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 244 lowerbound nodes 1162834815 nps 22168658 hashfull 464 tbhits 0 time 52454 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 258 lowerbound nodes 1163795864 nps 22167540 hashfull 464 tbhits 0 time 52500 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 279 lowerbound nodes 1164912012 nps 22167688 hashfull 465 tbhits 0 time 52550 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 310 lowerbound nodes 1166016029 nps 22169290 hashfull 466 tbhits 0 time 52596 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 357 lowerbound nodes 1167429786 nps 22168773 hashfull 466 tbhits 0 time 52661 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 427 lowerbound nodes 1168856482 nps 22169764 hashfull 466 tbhits 0 time 52723 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 532 lowerbound nodes 1172274263 nps 22171510 hashfull 466 tbhits 0 time 52873 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 690 lowerbound nodes 1174663842 nps 22174765 hashfull 466 tbhits 0 time 52973 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 926 lowerbound nodes 1179524729 nps 22180272 hashfull 467 tbhits 0 time 53179 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 1280 lowerbound nodes 1182834371 nps 22187435 hashfull 468 tbhits 0 time 53311 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 1810 lowerbound nodes 1187697293 nps 22199119 hashfull 470 tbhits 0 time 53502 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 2606 lowerbound nodes 1190085172 nps 22201424 hashfull 473 tbhits 0 time 53604 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 3800 lowerbound nodes 1190984105 nps 22203697 hashfull 473 tbhits 0 time 53639 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 5589 lowerbound nodes 1191440545 nps 22203927 hashfull 473 tbhits 0 time 53659 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 8274 lowerbound nodes 1191780863 nps 22204476 hashfull 473 tbhits 0 time 53673 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 seldepth 50 multipv 1 score cp 12301 lowerbound nodes 1193332985 nps 22212288 hashfull 473 tbhits 0 time 53724 pv f1f5
info depth 34 currmove f1f5 currmovenumber 1
info depth 34 currmove h6e3 currmovenumber 6
info depth 34 currmove h6h3 currmovenumber 17
info depth 34 currmove g4e2 currmovenumber 18
info depth 34 currmove g4d1 currmovenumber 19
info depth 34 currmove h6g5 currmovenumber 20
info depth 34 currmove f1f3 currmovenumber 22
info depth 34 seldepth 50 multipv 1 score mate 6 nodes 1194441861 nps 22216387 hashfull 474 tbhits 0 time 53764 pv f1f5 a6a5 f5h5 g6h5 g4f5 c7h2 h1h2 f8d8 h6h7 g8f8 h7h8
Top
F. Bluemers
Posts: 880
Joined: Thu Mar 09, 2006 11:21 pm
Location: Nederland

Re: Mate Test Suite

Post by F. Bluemers »

Dirty:

Code: Select all

12 230 70 952501 Rae1 d4 Rxe4 Rb8 b3 Rbe8 Rxd4 c5 Rd5 Rd8 Bxe6 fxe6
12 31989 115 2180527 Rf5 d4 Rh5 gxh5 Bf5 Qxh2+ Kxh2 Rfe8 Qxh7+ Kf8 Qh8#
12 31989 117 2223045 Rf5 d4 Rh5 gxh5 Bf5 Qxh2+ Kxh2 Rfe8 Qxh7+ Kf8 Qh8#
13 31989 121 2304836 Rf5 d4 Rh5 gxh5 Bf5 Qxh2+ Kxh2 Rfe8 Qxh7+ Kf8 Qh8#
Best
Fonzy
Top
User avatar
Eelco de Groot
Posts: 4724
Joined: Sun Mar 12, 2006 2:40 am
Full name:   Eelco de Groot

Re: Mate Test Suite

Post by Eelco de Groot »

This one Stockfish can solve quickly, so it's much simpler (for SF) than most of the others in this thread:
Okay, completely O.T. to the thread subject but 797 million nodes here, with 4 threads:

2r2rk1/2q2p1p/p1p1nPpQ/3p4/4p1B1/8/PPP3PP/R4R1K w - -

Engine: Rainbow Serpent Sf20150412_019 MOD MP (32 bit, 4 threads, 512 MB)
by Tord Romstad, Marco Costalba and Joona Kiiski

22/53 0:01 +1.01 1.h4 Rb8 2.h5 Rb7 3.hxg6 hxg6 4.a4 Qe5
5.Bxe6 fxe6 6.Qxg6+ Kh8 7.Ra3 e3
8.f7 Qg7 9.Qh5+ Qh7 10.Qxh7+ Kxh7
11.Rxe3 Rxb2 12.Rf4 Kg6 13.Rxe6+ Kg7
14.Rg4+ (3.877.509) 2887
.
.
.
37/58 3:09 +8.64++ 1.Rf5 (688.065.566) 3625
.
.
.
37/58 3:36 +88.59++ 1.Rf5 (780.981.749) 3606

37/58 3:41 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (797.179.815) 3606

38/58 3:41 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (797.783.777) 3606

39/58 3:41 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (798.492.922) 3606

40/58 3:42 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (800.601.365) 3605

41/58 3:42 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (801.684.331) 3605

42/58 3:42 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (803.253.270) 3605

43/58 3:43 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (804.623.548) 3605

44/58 3:43 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (806.525.652) 3605

45/58 3:44 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (808.852.067) 3606

46/58 3:45 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (813.410.194) 3604

47/58 3:47 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (818.084.432) 3603

48/58 3:48 +M6 1.Rf5 Qa7 2.Rh5 gxh5 3.Bf5 Qg1+
4.Rxg1 Rfe8 5.Qxh7+ Kf8 6.Qh8+ (822.593.598) 3602

best move: Rf1-f5 time: 3:50.101 min n/s: 3.602.746 nodes: 827.343.168

Apparently that was an unlucky run, next try:

33/54 0:29 +7.61++ 1.Rae1 (101.736.239) 3423

33/54 0:30 +7.67++ 1.Rf5 (104.736.633) 3416

33/54 0:30 +7.76++ 1.Rf5 (105.659.251) 3416

33/54 0:31 +7.90++ 1.Rf5 (106.223.683) 3416

33/54 0:31 +8.11++ 1.Rf5 (106.907.828) 3413
.
.
.
33/54 0:36 +88.06++ 1.Rf5 (122.984.512) 3411

33/54 0:36 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (124.110.568) 3412

34/54 0:36 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (124.515.644) 3413

35/54 0:36 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (124.921.299) 3413

36/54 0:36 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (125.411.570) 3414

37/54 0:36 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (125.936.026) 3415

38/54 0:37 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (126.532.682) 3416

39/54 0:37 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (127.412.160) 3408

40/54 0:37 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (128.465.078) 3410

41/54 0:38 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (129.773.305) 3411

42/54 0:38 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (131.150.124) 3414

43/54 0:39 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (133.780.150) 3411

44/54 0:39 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (135.974.756) 3415

45/54 0:40 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (138.529.389) 3417

46/54 0:41 +M6 1.Rf5 Rb8 2.Rh5 gxh5 3.Bf5 Qxh2+
4.Kxh2 Rfd8 5.Qxh7+ Kf8 6.Qh8+ (142.066.753) 3416

best move: Rf1-f5 time: 0:42.447 min n/s: 3.416.215 nodes: 143.518.241

Still much slower than Fonzy's engine :shock:
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
Top
Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Mate Test Suite

Post by Ferdy »

I run 3 engines at 30s/pos to see the common difficult position for them.
This could be interesting for Marek to solve too. This is not about counting how many bm are matched but about how the engine scored it.
There are positions that are not solved by bm but the engine showed big score already for other moves so this could be ignored for a moment.
Interesting positions are those that have low score and bm is not found at least at 30s with that settings. Longer engine thinking time is probably necessary here.
What engine can understand these positions given more time?
Pos 11, 12, and 13.

Code: Select all

Intel(R) Core(TM) i7-2600K CPU @ 3.40GHz
Physical Cores: 4
Hyper-Threading: ON
Hash: 128, Threads: 1, Time: 30s/pos
Test file: joseph_ellis_mates.epd
id name cheng4 0.38, TotalPos: 17, Correct: 7 (41.18%)

Pos  Correct  EngineBM  ScoreCP  D2Mate   epd
 1        0      f4e5    +1165    +1000   8/2Nb4/pp6/4rp1p/1Pp1pPkP/PpPpR3/1B1P2N1/1K6 w - - bm Kc1; dm 5; id "je 1"; c0 "b1c1;
 2        1      h4h5   +30025       +5   8/8/2B1N3/3rp3/4k2K/7Q/2r3Pn/1b1N4 w - - bm Kh5; dm 5; id "je 2"; c0 "h4h5";
 3        1      g3h4    +1532       +5   qb6/3N2p1/r2p4/pP1k2p1/1pp3R1/6BB/2P1P3/2N1K3 w - - bm Bh4; dm 5; id "je 3"; c0 "g3h4";
 4        1      b8c8   +32767       +6   1K1N1b2/RPp1pr2/1kP5/2p5/P7/4B1P1/4p1b1/6n1 w - - bm Kc8; dm 6; id "je 4"; c0 "b8c8";
 5        1      h5h1     +597    +1000   n1N3br/2p1Bpkr/1pP2R1b/pP1p1PpR/Pp4P1/1P6/1K1P4/8 w - - bm Rh1; dm 6; id "je 5"; c0 "h5h1";
 6        1      b8d7   +31470       +6   1N1K1b1r/P3pPp1/4k1P1/rp1pB1RN/q4RP1/8/p2pB1p1/1b6 w - - bm Nd7; dm 6; id "je 6"; c0 "b8d7";
 7        0      e5b2    +2337    +1000   5R2/2P2pK1/2P2P2/1Pp1BP1P/b6p/1p1RPB2/1p2NPn1/6rk w - - bm Ra8; dm 6; id "je 7"; c0 "f8a8";
 8        0      a2a5     +370    +1000   3K4/1p1B4/bB1k4/rpR1p3/2ppppp1/8/RPPPPP2/r1n5 w - - bm b4; dm 7; id "je 8"; c0 "b2b4";
 9        0      f3f5   +14169    +1000   4k1r1/2pnp3/2B1N3/8/8/5R2/6P1/5K2 w - - bm Rf2 Ke1; dm 11; id "je 9"; c0 "f3f2 f1e1";
10        0      a6d6     +254    +1000   n2Bqk2/5p1p/Q4KP1/p7/8/8/8/8 w - - bm Qc8; dm 13; id "je 10"; c0 "a6c8";
11        0      f1e2       +0    +1000   8/5K2/3p3p/3p3P/pp1P4/rkb1R3/p1p3P1/n1B2B2 w - - bm Rd3; dm 16; id "je 11"; c0 "e3d3";
12        0      f3g4     -790    +1000   1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - - bm bxc6; dm 17; id "je 12"; c0 "b5c6";
13        0      d5e4       +0    +1000   8/6p1/p7/rp1K2p1/kb3pP1/2p2p1b/P1Np1P2/3N2R1 w - - bm Ke6; dm 20; id "je 13"; c0 "d5e6";
14        1      d4h8     +980     +146   8/7p/6pP/5pP1/3BpP2/p1KpP3/pn1N4/k7 w - - bm Bh8; dm 21; id "je 14"; c0 "d4h8";
15        1      b8a7     +229    +1000   1B3N2/5p1B/2K2pn1/5krb/4p1p1/4P1P1/2P1PP2/8 w - - bm Ba7; dm 22; id "je 15"; c0 "b8a7";
16        0      c4d5       +0    +1000   1k3b1q/pP2p1p1/P1K1P1Pp/7P/2B5/8/8/8 w - - bm Bb5 Kd5; dm 27; id "je 16"; c0 "c4b5 c6d5";
17        0      d1d5       +0    +1000   4rk2/2P2p2/p4P2/2p2b2/2p5/8/P7/2KR4 w - - bm Rd8; dm 29; id "je 17"; c0 "d1d8";

Intel(R) Core(TM) i7-2600K CPU @ 3.40GHz
Physical Cores: 4
Hyper-Threading: ON
Hash: 128, Threads: 1, Time: 30s/pos
Test file: joseph_ellis_mates.epd
id name Sting SF 5 64bit, TotalPos: 17, Correct: 6 (35.29%)

Pos  Correct  EngineBM  ScoreCP  D2Mate   epd
 1        0      f4e5    +1940    +1000   8/2Nb4/pp6/4rp1p/1Pp1pPkP/PpPpR3/1B1P2N1/1K6 w - - bm Kc1; dm 5; id "je 1"; c0 "b1c1;
 2        0      h4g5    +9623       +8   8/8/2B1N3/3rp3/4k2K/7Q/2r3Pn/1b1N4 w - - bm Kh5; dm 5; id "je 2"; c0 "h4h5";
 3        0      b5a6    +1074    +1000   qb6/3N2p1/r2p4/pP1k2p1/1pp3R1/6BB/2P1P3/2N1K3 w - - bm Bh4; dm 5; id "je 3"; c0 "g3h4";
 4        1      b8c8    +2342       +6   1K1N1b2/RPp1pr2/1kP5/2p5/P7/4B1P1/4p1b1/6n1 w - - bm Kc8; dm 6; id "je 4"; c0 "b8c8";
 5        1      h5h1    +1191       +8   n1N3br/2p1Bpkr/1pP2R1b/pP1p1PpR/Pp4P1/1P6/1K1P4/8 w - - bm Rh1; dm 6; id "je 5"; c0 "h5h1";
 6        0      f4a4    +1474    +1000   1N1K1b1r/P3pPp1/4k1P1/rp1pB1RN/q4RP1/8/p2pB1p1/1b6 w - - bm Nd7; dm 6; id "je 6"; c0 "b8d7";
 7        0      e5b2    +3693    +1000   5R2/2P2pK1/2P2P2/1Pp1BP1P/b6p/1p1RPB2/1p2NPn1/6rk w - - bm Ra8; dm 6; id "je 7"; c0 "f8a8";
 8        0      a2a5     +579    +1000   3K4/1p1B4/bB1k4/rpR1p3/2ppppp1/8/RPPPPP2/r1n5 w - - bm b4; dm 7; id "je 8"; c0 "b2b4";
 9        1      f1e1   +10385      +18   4k1r1/2pnp3/2B1N3/8/8/5R2/6P1/5K2 w - - bm Rf2 Ke1; dm 11; id "je 9"; c0 "f3f2 f1e1";
10        1      a6c8   +10613      +15   n2Bqk2/5p1p/Q4KP1/p7/8/8/8/8 w - - bm Qc8; dm 13; id "je 10"; c0 "a6c8";
11        0      f7e6       +0    +1000   8/5K2/3p3p/3p3P/pp1P4/rkb1R3/p1p3P1/n1B2B2 w - - bm Rd3; dm 16; id "je 11"; c0 "e3d3";
12        0      f3g4    -1206    +1000   1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - - bm bxc6; dm 17; id "je 12"; c0 "b5c6";
13        0      d5e4     +185    +1000   8/6p1/p7/rp1K2p1/kb3pP1/2p2p1b/P1Np1P2/3N2R1 w - - bm Ke6; dm 20; id "je 13"; c0 "d5e6";
14        0      d4e5     +569    +1000   8/7p/6pP/5pP1/3BpP2/p1KpP3/pn1N4/k7 w - - bm Bh8; dm 21; id "je 14"; c0 "d4h8";
15        1      b8a7     +379    +1000   1B3N2/5p1B/2K2pn1/5krb/4p1p1/4P1P1/2P1PP2/8 w - - bm Ba7; dm 22; id "je 15"; c0 "b8a7";
16        1      c4b5    +2069    +1000   1k3b1q/pP2p1p1/P1K1P1Pp/7P/2B5/8/8/8 w - - bm Bb5 Kd5; dm 27; id "je 16"; c0 "c4b5 c6d5";
17        0      d1d5    +1063    +1000   4rk2/2P2p2/p4P2/2p2b2/2p5/8/P7/2KR4 w - - bm Rd8; dm 29; id "je 17"; c0 "d1d8";

Intel(R) Core(TM) i7-2600K CPU @ 3.40GHz
Physical Cores: 4
Hyper-Threading: ON
Hash: 128, Threads: 1, Time: 30s/pos
Test file: joseph_ellis_mates.epd
id name Stockfish 6 64 POPCNT, TotalPos: 17, Correct: 2 (11.76%)

Pos  Correct  EngineBM  ScoreCP  D2Mate   epd
 1        0      f4e5    +1614    +1000   8/2Nb4/pp6/4rp1p/1Pp1pPkP/PpPpR3/1B1P2N1/1K6 w - - bm Kc1; dm 5; id "je 1"; c0 "b1c1;
 2        1      h4h5    +5540       +8   8/8/2B1N3/3rp3/4k2K/7Q/2r3Pn/1b1N4 w - - bm Kh5; dm 5; id "je 2"; c0 "h4h5";
 3        0      b5a6    +1147    +1000   qb6/3N2p1/r2p4/pP1k2p1/1pp3R1/6BB/2P1P3/2N1K3 w - - bm Bh4; dm 5; id "je 3"; c0 "g3h4";
 4        0      e3d2    +1569    +1000   1K1N1b2/RPp1pr2/1kP5/2p5/P7/4B1P1/4p1b1/6n1 w - - bm Kc8; dm 6; id "je 4"; c0 "b8c8";
 5        0      c8d6     +620    +1000   n1N3br/2p1Bpkr/1pP2R1b/pP1p1PpR/Pp4P1/1P6/1K1P4/8 w - - bm Rh1; dm 6; id "je 5"; c0 "h5h1";
 6        0      f4a4    +1623    +1000   1N1K1b1r/P3pPp1/4k1P1/rp1pB1RN/q4RP1/8/p2pB1p1/1b6 w - - bm Nd7; dm 6; id "je 6"; c0 "b8d7";
 7        0      e5b2    +2760    +1000   5R2/2P2pK1/2P2P2/1Pp1BP1P/b6p/1p1RPB2/1p2NPn1/6rk w - - bm Ra8; dm 6; id "je 7"; c0 "f8a8";
 8        0      a2a5     +606    +1000   3K4/1p1B4/bB1k4/rpR1p3/2ppppp1/8/RPPPPP2/r1n5 w - - bm b4; dm 7; id "je 8"; c0 "b2b4";
 9        0      f3d3     +829    +1000   4k1r1/2pnp3/2B1N3/8/8/5R2/6P1/5K2 w - - bm Rf2 Ke1; dm 11; id "je 9"; c0 "f3f2 f1e1";
10        0      a6d6       +6    +1000   n2Bqk2/5p1p/Q4KP1/p7/8/8/8/8 w - - bm Qc8; dm 13; id "je 10"; c0 "a6c8";
11        0      f7e6       +0    +1000   8/5K2/3p3p/3p3P/pp1P4/rkb1R3/p1p3P1/n1B2B2 w - - bm Rd3; dm 16; id "je 11"; c0 "e3d3";
12        0      f3d1    -1145    +1000   1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - - bm bxc6; dm 17; id "je 12"; c0 "b5c6";
13        0      d5d4       +0    +1000   8/6p1/p7/rp1K2p1/kb3pP1/2p2p1b/P1Np1P2/3N2R1 w - - bm Ke6; dm 20; id "je 13"; c0 "d5e6";
14        0      d4e5     +644    +1000   8/7p/6pP/5pP1/3BpP2/p1KpP3/pn1N4/k7 w - - bm Bh8; dm 21; id "je 14"; c0 "d4h8";
15        1      b8a7     +421    +1000   1B3N2/5p1B/2K2pn1/5krb/4p1p1/4P1P1/2P1PP2/8 w - - bm Ba7; dm 22; id "je 15"; c0 "b8a7";
16        0      c4a2       +0    +1000   1k3b1q/pP2p1p1/P1K1P1Pp/7P/2B5/8/8/8 w - - bm Bb5 Kd5; dm 27; id "je 16"; c0 "c4b5 c6d5";
17        0      d1d5       +6    +1000   4rk2/2P2p2/p4P2/2p2b2/2p5/8/P7/2KR4 w - - bm Rd8; dm 29; id "je 17"; c0 "d1d8";
Pos 11:
[d]8/5K2/3p3p/3p3P/pp1P4/rkb1R3/p1p3P1/n1B2B2 w - - bm Rd3; dm 16; id "je 11"; c0 "e3d3";
Pos 12:
[d]1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - - bm bxc6; dm 17; id "je 12"; c0 "b5c6";
Pos 13:
[d]8/6p1/p7/rp1K2p1/kb3pP1/2p2p1b/P1Np1P2/3N2R1 w - - bm Ke6; dm 20; id "je 13"; c0 "d5e6";
Top
BBauer
Posts: 658
Joined: Wed Mar 08, 2006 8:58 pm

Re: Mate Test Suite

Post by BBauer »

This is a nice position too.
Thank you.
Bernhard
Top
Dann Corbit
Posts: 12870
Joined: Wed Mar 08, 2006 8:57 pm
Location: Redmond, WA USA

Re: Mate Test Suite

Post by Dann Corbit »

Ferdy wrote: I am curious were you able to verify the distance to mate of these positions?
Almost all of them are found in the ChestUCI database, which has distance to mate.
Top
lech
Posts: 1175
Joined: Sun Feb 14, 2010 10:02 pm

Re: Mate Test Suite

Post by lech »

Ferdy wrote:I run 3 engines at 30s/pos to see the common difficult position for them.
This could be interesting for Marek to solve too. This is not about counting how many bm are matched but about how the engine scored it.
...
I found many next bugs in the code of Stockfish 2.1.1 parent for my Sting project.
They are not that important in play, but they can crash my work for such difficult positions in analysis. The last my tool ("lazy King") will be added in the next version, and after it I will try to do Sting much more stronger, maybe the best. :D
Maybe, I can't be friendly, but let me be useful.
Top
Ferdy
Posts: 4856
Joined: Sun Aug 10, 2008 3:15 pm
Location: Philippines

Re: Mate Test Suite

Post by Ferdy »

Gives up queen in order to win. Takes time but there is hope, found bm after 51 minutes. Now the king should walk towards the rook carefully to get positive score. I am not concerned on time but for the engine to find the correct moves.
Pos 12:
[d]1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - - bm bxc6; dm 17; id "je 12"; c0 "b5c6";

Code: Select all

Deuterium v2015.1.35.142:

18/35	00:14	   9,171,901	615,977	-7.57	1.Qd1 Qf8 2.Bc4 Qd8 3.Qd2 cxb5 4.Bd5 Re5 5.h6 Rxf5 6.c6 dxc6 7.Bf3 Nb6 8.Qc2 Qg8+ 9.Ka1 Qh7 10.Qxc6
 19/33	00:19	  11,978,909	610,639	-8.31	1.Qd1 Qf8 2.Bc4 Qd8 3.Qd2 cxb5 4.Bd5 Re5 5.h6 Rxf5 6.h7 f3 7.Bb7 Rxc5 8.Bxf3 Qh8 9.Qd3 Nb6 10.Bb7 f5
 20/27	00:27	  16,661,542	607,287	-8.61	1.Qd1
 20/45	01:29	  56,484,455	629,395	-9.66	1.Qd1 Qf8 2.bxc6 Qg8+ 3.Kb1 g2 4.Bxg2 Qxg2 5.cxd7 Qe4+ 6.Qc2 Qb4+ 7.Qb2 Rb3 8.Qxb3 Qxb3+ 9.Kc1 Qc4+ 10.Kd2 Qa2+ 11.Kc3 Qa5+ 12.Kd3 Qxa6+ 13.Kd2 Qa5+ 14.Ke2 Qb5+ 15.Kf3 Qxd7 16.Kxf4 Nxd6 17.cxd6 Qxd6+ 18.Kg4
 21/36+	01:30	  57,096,775	630,082	-9.36	1.Qd1
 21/44	01:57	  74,086,005	632,268	-9.53	1.Qd1 Qf8 2.Bc4 Nxd6 3.cxd6 Qd8 4.Qd2 Qb6 5.h6 Qc5 6.Bb3 Qxb5 7.Qd1 Re2+ 8.Kb1 Qxf5+ 9.Bc2 Qb5+ 10.Kc1 Qxa6 11.Qd4 Qb6 12.Qxb6+ axb6 13.Bf5 g2 14.Kd1 g1Q+ 15.Kxe2
 22/39+	02:18	  88,120,439	637,869	-9.23	1.Qg4
 22/49	06:09	 256,284,153	693,880	-10.96	1.Qg4 Qf8 2.Bc4 Qd8 3.b6 Rc3 4.Bb3 Rxc5 5.Qxf4 Qxb6 6.Qxg3 Qxa6+ 7.Kb1 Rxf5 8.Bc2 Rxh5 9.Qc3 Rh1+ 10.Kb2 Nxd6 11.Qb4+ Nb5 12.Ba4 Kb7 13.Qf4
 23/35+	06:12	 258,645,750	693,972	-10.66	1.Qg4
 23/49	08:43	 371,577,192	709,998	-10.80	1.Qd1 Qf8 2.Bc4 Nxd6 3.cxd6 Qd8 4.Qd2 Qb6 5.bxc6 Qxc6 6.Qb4+ Kc8 7.Bf1 Qc2+ 8.Ka1 Rb3 9.Qc4+ Qxc4 10.Bxc4 g2 11.Bxb3 g1Q+ 12.Kb2 Qh2+ 13.Kc3 Qxh5 14.Kd4 Qxf5 15.Bd5 Qe5+ 16.Kc4 Qxd6 17.Bb7+ Kc7 18.Kc3
 24/34	09:45	 414,693,778	708,235	-11.10	1.Qd1
 24/49	11:41	 499,948,283	712,306	-11.13	1.Qd1 Qf8 2.Bc4 Nxd6 3.cxd6 Qd8 4.Qd2 Qb6 5.bxc6 Qxc6 6.Qb4+ Kc8 7.Qb7+ Qxb7 8.axb7+ Kxb7 9.Bd5+ Kb6 10.h6 g2 11.Bxg2 Re2+ 12.Kb3 Rxg2 13.h7 Rh2 14.Kc4 a5 15.h8Q Rxh8 16.Kd3 a4 17.Kc4 Kc6 18.Kd4
 25/45+	12:11	 520,086,867	711,334	-10.83	1.Qd1
 25/38+	12:21	 527,009,745	711,123	-9.83	1.Qd1
 25/53	16:37	 740,389,182	741,933	-10.45	1.Qd1 Qf8 2.Bc4 Nxd6 3.cxd6 Qd8 4.Qd2 Qb6 5.h6 Qc5 6.Bb3 Qxb5 7.Bc2 Qxa6+ 8.Kb1 Re2 9.Qb4+ Qb5 10.Qxb5+ cxb5 11.h7 Rh2 12.Be4 Rxh7 13.Kc2 Rh2+ 14.Kd3 g2 15.Bxg2 Rxg2 16.Ke4 Rf2 17.Kd4 a5
 26/42	17:18	 769,269,980	740,960	-10.75	1.Qd1
 26/55	29:21	1,359,636,883	772,014	-13.34	1.Qd1 Qf8 2.Bc4 Nxd6 3.cxd6 Qd8 4.Qd2 Qb6 5.h6 Qc5 6.Bb3 Qxb5 7.Bd1 Qxa6+ 8.Kb1 Qd3+ 9.Qxd3 Rxd3 10.Bc2 g2 11.Bxd3 g1Q+ 12.Kb2 Qh2+ 13.Kc3 f3 14.h7 Qxh7 15.Kd4 f2 16.Ba6 Qxf5 17.Ke3 Qc5+ 18.Kf3 Qa3+ 19.Kxf2 Qxa6
 27/48+	30:39	1,417,178,957	770,586	-13.04	1.Qd1
 27/36+	30:42	1,419,934,176	770,619	-12.04	1.Qd1
 27/56	50:59	2,250,537,421	735,495	 0.00	1.bxc6 Rxf3 2.c7+ Ka8 3.Bg2 Qxh5 4.Ka1 Qg4 5.Kb2 Qh5 6.Kc1 Qg4 7.Kd1 Qh5 8.Kc1
 28/38	51:05	2,254,144,203	735,400	 0.00	1.bxc6 Rxf3 2.c7+ Ka8 3.Bg2 Qxh5 4.Ka1 Qg4 5.Kb2 Qh5 6.Kc1 Qg4 7.Kd1 Qh5 8.Kc1
 29/39	51:22	2,266,142,070	735,241	 0.00	1.bxc6 Rxf3 2.c7+ Ka8 3.Bg2 Qxh5 4.Ka1 Qg4 5.Kb2 Qh5 6.Kc1 Qg4 7.Kd1 Qh5 8.Kc1
 30/41	51:42	2,280,946,510	735,122	 0.00	1.bxc6 Rxf3 2.c7+ Ka8 3.Bg2 Qxh5 4.Ka1 Qg4 5.Kb2 Qh5 6.Kc1 Qg4 7.Kd1 Qh5 8.Kc1
 31/42	52:20	2,306,846,025	734,587	 0.00	1.bxc6 Rxf3 2.c7+ Ka8 3.Bg2 Qxh5 4.Ka1 Qg4 5.Kb2 Qh5 6.Ka1
[...] in progress
Top
User avatar
Eelco de Groot
Posts: 4724
Joined: Sun Mar 12, 2006 2:40 am
Full name:   Eelco de Groot

Re: Mate Test Suite

Post by Eelco de Groot »

Ferdy wrote: What engine can understand these positions given more time?
Pos 11, 12, and 13.

Pos 11:
[d]8/5K2/3p3p/3p3P/pp1P4/rkb1R3/p1p3P1/n1B2B2 w - - bm Rd3; dm 16; id "je 11"; c0 "e3d3";
Pos 12:
[d]1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - - bm bxc6; dm 17; id "je 12"; c0 "b5c6";
Pos 13:
[d]8/6p1/p7/rp1K2p1/kb3pP1/2p2p1b/P1Np1P2/3N2R1 w - - bm Ke6; dm 20; id "je 13"; c0 "d5e6";
Pos 11 seems too hard but becomes easy if you give the first move Rd3. So that is probably only solvable without nullmove in the critical line. Pos 12 is doable for Stockfish with 4 threads. 1 thread would just take longer I think. But with 4 in 6 minutes. Could be a lucky run.

[d]1kn5/p2p4/P1pP1p1q/1PP2P1P/5p2/4rQp1/K7/5B2 w - -

Engine: Rainbow Serpent Sf20150412_021 MOD MP (Q6700, 4 threads, 512 MB)
by Tord Romstad, Marco Costalba and Joona Kiiski

19/31 0:01 -8.27 1.Qd1 Qf8 2.Bc4 Qd8 3.Qd2 cxb5 4.Bd5 Qe8
5.Be6 Rxe6 6.fxe6 Qxe6+ 7.Kb2 f3
8.c6 Qxd6 9.Qxd6+ Nxd6 10.h6 Kc7
11.h7 (4.221.739) 2860
.
.
.
26/49 0:27 -12.51-- 1.Qd1 Qf8 (98.668.911) 3561

26/49 0:29 -12.40++ 1.bxc6 (106.906.542) 3578
.
.
.
43/72 5:17 +109.58++1.bxc6 (1.177.498.595) 3707

43/72 5:59 +M17 1.bxc6 Rxf3 2.c7+ Ka8 3.Bg2 Qxh5
4.Ka1 Qg4 5.Kb2 Qh5 6.Kc1 Qg4 7.Kd2 Qh5
8.Kd1 Qg4 9.Ke2 Qh5 10.Bxf3+ Qxf3+
11.Kxf3 g2 12.Kxg2 f3+ 13.Kf1 f2
14.c6 (1.354.028.210) 3762

44/72 6:18 +M17 1.bxc6 Rxf3 2.c7+ Ka8 3.Bg2 Qxh5
4.Ka1 Qg4 5.Kb2 Qh5 6.Kc1 Qg4 7.Kd2 Qh5
8.Kd1 Qg4 9.Ke2 Qh5 10.Bxf3+ Qxf3+
11.Kxf3 g2 12.Kxg2 f3+ 13.Kf1 f2
14.c6 (1.431.118.079) 3778

best move: b5xc6 time: 6:48.239 min n/s: 3.778.449 nodes: 1.549.617.150
Debugging is twice as hard as writing the code in the first
place. Therefore, if you write the code as cleverly as possible, you
are, by definition, not smart enough to debug it.
-- Brian W. Kernighan
Top

Return to “Computer Chess Club: General Topics”