Can your program see the draw?

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

Moderators: hgm, Harvey Williamson, bob

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
jwes
Posts: 778
Joined: Sat Jul 01, 2006 5:11 am

Can your program see the draw?

Post by jwes » Tue Mar 23, 2010 6:25 am

From 1234 Modern Endgame Studies. Finding the right moves is not that hard but does your program return a drawish score?
[D]8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - bm Kf7; id "MES.150";

P. Villanueva
Posts: 85
Joined: Sat May 17, 2008 8:57 pm
Location: Bilbao, Spain

Re: Can your program see the draw?

Post by P. Villanueva » Tue Mar 23, 2010 10:21 am

My KMT finds Kf7 but thinks that black wins.

FEN: 8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - 0 1

KMTChess dev:
1 00:00 12 12 -1,86 g7f6
2 00:00 54 54 -1,89 g7f6 a5c3
3 00:00 458 458 -2,43 g7f6 a5c7 f4f5 e6f5 f6f5
4 00:00 867 867 -2,43 g7f6 a5c7 f4f5 e6f5 f6f5
5 00:00 3.710 3.710 -2,88 g7f7 a5c3 g5g6 c3d4 g6g7
6 00:00 9.268 926.800 -1,78 g5g6!!
6 00:00 15.436 771.800 -1,37 g5g6 a5c3 g7f7 c3d4 g6g7 d4g7
7 00:00 19.992 666.400 -1,37 g5g6 a5c3 g7f7 c3d4 g6g7 d4g7 f7g7
8 00:00 33.098 661.960 -1,27 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 d7d6 g7f6
9 00:00 51.698 574.422 -1,34 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 c6c5 g7f6 c5c4
10 00:00 80.923 578.021 -1,32 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 d7d6 g7f6 d6d5 f6g5
11 00:00 128.357 534.820 -1,43 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 c6c5 g7f6 c5c4 f4f5 e6f5 f6f5
12 00:00 206.314 529.010 -1,43 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 c6c5 g7f6 c5c4 f4f5 e6f5 f6f5
13 00:00 333.797 538.382 -1,67 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 c6c5 g7f6 c5c4 f4f5 e6f5 f6f5 c4c3
14 00:01 544.994 539.598 -1,66 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 c6c5 g7f6 c5c4 f4f5 e6f5 f6f5 c4c3 f5e4
15 00:03 1.688.670 553.662 -2,03 g5g6 a5c3 g7f8 c3d4 g6g7 d4g7 f8g7 c6c5 g7f6 c5c4 f4f5 e6f5 f6f5 c4c3 f5e5 c3c2
16 00:04 2.541.012 544.113 -2,28 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1h8 f7f8 d5d4 g6g7 h8g7 f8g7 d4d3
17 00:06 3.500.079 561.810 -2,28 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 d6d7 g6g7 a1g7 f7g7 d5d4 g7f8 d4d3 f8f7
18 00:08 5.047.956 571.682 -2,28 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1h8 f7f8 d6d7 g6g7 h8g7 f8g7 d5d4 g7f6 d4d3 f6f7
19 00:12 7.492.306 576.774 -2,28 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1h8 f7f8 h8f6 f8f7 f6h8
20 00:38 20.799.615 544.492 -2,26 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1d4 f7f8 d6d7 f8f7 d4a1 f7f8 a1h8 g6g7 h8g7 f8g7 d5d4 g7h7
21 00:51 28.438.711 557.621 -2,28 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1h8 f7f8 d6d7 f8g8 h8f6 g8f7 f6a1 f7f8 a1b2 f8f7 e6e5 f4e5 b2e5 g6g7 e5g7 f7g7
22 02:24 76.319.468 529.996 -2,26 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1h8 f7f8 d6d7 f8g8 h8f6 g8f7 f6d4 f7f8 d4a1 g6g7 a1g7 f8g7 d5d4 g7h7
23 03:44 119.410.591 533.082 -2,55 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1h8 f7f8 h8b2 f8f7 b2a1 f7f8 a1c3 g6g7 c3g7 f8g7 d5d4 g7f6 d4d3 f4f5 d3d2
24 07:27 233.152.839 521.594 -2,54 g7f7 a5c3 d4d5 c6d5 g5g6 d7d6 f7f8 c3a1 f8f7 a1b2 f7f8 d6c7 f8f7 c7d7 f7f8 b2f6 f8f7 f6h8 f7f8 d5d4 g6g7 h8g7 f8g7 d4d3 g7g8

Stockfish also thinks that it's a win for black.
FEN: 8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - 0 1

Stockfish-16-ja:
1 00:00 22 105 -1,73 g5g6
2 00:00 39 185 -2,02 g5g6 Kd7e7
2 00:00 64 304 -1,89 Kg7f6 Ba5c3
3 00:00 160 761 -2,10 Kg7f6 Ba5c7 g5g6 Bc7xf4
4 00:00 345 1.635 -2,10 Kg7f6 Ba5c7 g5g6 Bc7xf4
5 00:00 834 3.897 -1,81 Kg7f6 Ba5b6 Kf6e5 Bb6c7+ Ke5e4 Kd7e7
6- 00:00 1.466 6.373 -2,38 Kg7f6 Ba5c3 Kf6e5 c6c5
7 00:00 3.782 15.890 -3,75 Kg7f6 Ba5c3 Kf6e5 c6c5 g5g6 Bc3xd4+ Ke5e4 Kd7e7 f4f5 e6xf5+ Ke4xf5
8 00:00 14.304 56.537 -3,67 Kg7f6 Ba5c3 Kf6e5 c6c5 Ke5e4 c5xd4 Ke4d3 Kd7e7 Kd3e4 Ke7f7 Ke4d3
9 00:00 18.744 71.269 -3,67 Kg7f6 Ba5c3 Kf6e5 c6c5 Ke5e4 c5xd4 Ke4d3 Kd7e7 Kd3e4 Ke7f7 Ke4d3
10- 00:00 22.557 82.324 -3,87 Kg7f6 Ba5c3 Kf6e5 c6c5 Ke5e4 c5xd4 Ke4d3 Kd7e7 Kd3e4 Ke7f7 Ke4d3 Kf7g7
11 00:00 34.883 119.872 -4,32 Kg7f6 Ba5c3 Kf6e5 c6c5 Ke5e4 c5xd4 Ke4d3 Kd7e7 Kd3e4 Ke7f7 Ke4d3 Kf7g7 Kd3e4 Kg7g6
12- 00:00 52.077 164.800 -5,21 Kg7f6 Ba5c3 g5g6 Bc3xd4+ Kf6g5 c6c5 f4f5 c5c4
12 00:00 90.910 257.535 -4,88 Kg7g6 Ba5b6 Kg6h5 Bb6xd4 Kh5g4 c6c5 Kg4f3 c5c4 Kf3e2 Kd7e7 Ke2d2 Ke7f7 Kd2c2 Kf7g6
13 00:00 161.719 387.815 -4,96 Kg7g6 Ba5b6 Kg6h5 Bb6xd4 Kh5g4 Kd7d6 f4f5 e6xf5+ Kg4xf5 c6c5 Kf5e4 Kd6e6 Ke4f4 Ke6d5 g5g6 c5c4 Kf4f3
14- 00:00 260.212 520.424 -5,69 Kg7g6 Ba5b6 Kg6h5 Bb6xd4 Kh5g4 Kd7d6 f4f5 e6xf5+ Kg4xf5 c6c5 Kf5e4 Kd6e6 Ke4f4 Ke6d5 g5g6 c5c4 Kf4f3 c4c3
15 00:00 598.082 800.645 -5,73 Kg7g6 Ba5b6 f4f5 e6xf5 Kg6xf5 Bb6xd4 Kf5e4 c6c5 Ke4d5 Kd7e7 Kd5e4 Ke7e6 Ke4f4 Ke6d5 Kf4f3 c5c4 g5g6 c4c3 Kf3e2 Kd5e4 Ke2d1
16 00:00 851.377 908.620 -6,14 Kg7g6 Ba5b6 f4f5 e6xf5 Kg6xf5 Bb6xd4 Kf5e4 c6c5 Ke4d5 Kd7e7 Kd5e4 Ke7f7 Ke4d3 Kf7g6 Kd3e4 Kg6xg5 Ke4d3 Kg5f5 Kd3c4 Kf5e4 Kc4b5 Ke4d5
17 00:02 2.724.773 1.228.481 -6,50 Kg7g6 Ba5c3 Kg6h5 Bc3xd4 Kh5g4 Kd7d6 f4f5 e6xf5+ Kg4xf5 c6c5 Kf5e4 Kd6e6 g5g6 Ke6f6 g6g7 Kf6xg7 Ke4d3 Kg7f7 Kd3e2 c5c4 Ke2d2 Kf7e6
18- 00:03 4.738.721 1.341.273 -7,63 Kg7g6 Ba5c3 f4f5 Kd7e7 f5f6+ Ke7f8 Kg6h5 Bc3xd4 f6f7 Kf8xf7 Kh5h6 c6c5 g5g6+ Kf7g8 Kh6g5 Kg8g7
18 00:04 5.627.092 1.384.619 -6,50 Kg7f6 Kd7d6 g5g6 Ba5c3 Kf6g5 Bc3xd4 f4f5 e6xf5 Kg5xf5 Kd6d5 Kf5f4 c6c5 Kf4f3 Kd5e6 Kf3g4 c5c4 Kg4f3 Ke6f5 Kf3e2 Kf5xg6 Ke2d2 Kg6f7 Kd2c2 Kf7e6 Kc2d2
19- 00:04 6.242.943 1.398.822 -6,86 Kg7f6 Kd7d6 g5g6 Ba5c3 Kf6g5 Bc3xd4 f4f5 Kd6e7 Kg5f4 e6xf5 Kf4xf5 c6c5 Kf5g5 c5c4
20- 00:06 10.018.319 1.485.295 -8,32 Kg7f6 Kd7d6 g5g6 Ba5c3 Kf6g5 Bc3xd4 Kg5h5 c6c5 Kh5g4 c5c4 Kg4f3 Kd6d5 Kf3e2 Kd5e4 Ke2d2 Ke4xf4 Kd2c2 Kf4f5 Kc2d2 Kf5xg6
21 00:13 21.195.416 1.559.977 -13,01 Kg7f6 Kd7d6 g5g6 Ba5c3 Kf6g5 Bc3xd4 f4f5 Kd6e7 Kg5f4 e6e5+ Kf4g5 Bd4e3+ Kg5g4 Ke7f6 Kg4f3 Be3d4 Kf3e4 c6c5 Ke4d5 Kf6g7 Kd5c4 e5e4 Kc4b3 Kg7f6 g6g7 Kf6xg7 Kb3c4 Kg7f6 Kc4d5 e4e3 Kd5c4
21 00:16 26.704.149 1.588.114 -8,64 d4d5 c6xd5 Kg7f7 Ba5b6 g5g6 Bb6d4 Kf7f8 Kd7d6 Kf8f7 Bd4a1 Kf7f8 Kd6d7 Kf8f7 Ba1c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2f6 Kf8f7 Bf6c3 Kf7f8 Bc3f6
21 00:18 30.411.908 1.623.527 -8,44 Kg7f7 Ba5b6 d4d5 c6xd5 g5g6 Bb6d4 Kf7f8 Kd7d6 Kf8f7 Bd4c3 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2a1 Kf8f7 Kd6d7 Kf7f8 Ba1h8 Kf8g8 Bh8f6 Kg8f7 Bf6a1
22 00:19 31.973.956 1.638.849 -8,28 Kg7f7 Ba5b6 d4d5 c6xd5 g5g6 Bb6d4 Kf7f8 Kd7d6 Kf8f7 Bd4c3 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2c3 Kf8f7 Kd6d7
23 00:21 36.184.333 1.659.907 -8,28 Kg7f7 Ba5b6 d4d5 c6xd5 g5g6 Bb6d4 Kf7f8 Kd7d6 Kf8f7 Bd4c3 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2c3 Kf8f7 Bc3a1 Kf7f8 Ba1b2 Kf8f7 Kd6d7 Kf7f8 Bb2c3 Kf8f7 Kd7d6
24 00:26 44.044.808 1.685.861 -8,28 Kg7f7 Ba5b6 d4d5 c6xd5 g5g6 Bb6d4 Kf7f8 Kd7d6 Kf8f7 Bd4c3 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2c3 Kf8f7 Bc3a1 Kf7f8 Ba1b2 Kf8f7 Kd6d7 Kf7f8 Bb2c3 Kf8f7 Kd7d6
25 00:29 49.918.591 1.697.738 -8,32 Kg7f7 Ba5b6 d4d5 c6xd5 g5g6 Bb6d4 Kf7f8 Kd7d6 Kf8f7 Bd4c3 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2c3 Kf8f7 Bc3a1 Kf7f8 Ba1b2 Kf8f7 Kd6d7 Kf7f8 Bb2c3 Kf8f7 Kd7d6
26+ 00:37 64.673.430 1.728.727 -8,12 Kg7f7 Ba5c3 d4d5 c6xd5 g5g6 Kd7d6 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Bb2d4 Kf7f8 Kd7d6 Kf8f7 Bd4c3
27 00:45 80.065.294 1.758.903 -8,44 Kg7f7 Ba5c3 d4d5 c6xd5 g5g6 Kd7d6 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Bb2d4 Kf7f8 Kd7d6 Kf8f7 Bd4h8 Kf7g8 Bh8a1 Kg8f7 Ba1b2 Kf7f8 Bb2c3 Kf8f7 Bc3a1 Kf7f8 Kd6d7 Kf8f7 Ba1h8 Kf7g8 Bh8f6 Kg8f7 Bf6a1 Kf7f8 Ba1b2
28 00:58 103.919.000 1.763.580 -8,44 Kg7f7 Ba5c3 d4d5 c6xd5 g5g6 Kd7d6 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2c3 Kf8f7 Bc3a1 Kf7f8 Ba1d4 Kf8f7 Kd6d7 Kf7f8 Bd4a1 Kf8f7 Ba1h8 Kf7g8 Bh8f6 Kg8f7 Bf6a1 Kf7f8 Ba1h8 Kf8g8
29 01:40 175.345.827 1.746.872 -8,28 Kg7f7 Ba5c3 d4d5 c6xd5 g5g6 Kd7d6 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4c3 Kf7f8 Bc3b2 Kf8f7 Kd7d6 Kf7f8 Bb2c3 Kf8f7 Bc3a1 Kf7f8 Ba1d4 Kf8f7 Kd6d7 Kf7f8 Bd4f6 Kf8f7 Bf6a1 Kf7f8 Kd7d6 Kf8f7 d5d4 g6g7
30 02:15 236.035.384 1.746.792 -8,44 Kg7f7 Ba5c3 d4d5 c6xd5 g5g6 Kd7d6 Kf7f8 Bc3f6 Kf8f7 Bf6d4 Kf7f8 Kd6d7 Kf8f7 Bd4b2 Kf7f8 Kd7d6 Kf8f7 Bb2a1 Kf7f8 Ba1d4 Kf8f7 Kd6d7 Kf7f8 Bd4c3 Kf8f7 Bc3a1 Kf7f8 Kd7d6 Kf8f7 Ba1b2 Kf7f8 Bb2c3 Kf8f7 Kd6d7 Kf7f8 Kd7d6


But after Kf7, Bc3, d5, cxd5 g6

[d]8/3k1K2/4p1P1/3p4/5P2/2b5/8/8 b - - 0 3

Black can't advance its pawns any more. What a tricky position!

zullil
Posts: 5599
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Can your program see the draw?

Post by zullil » Tue Mar 23, 2010 10:54 am

jwes wrote:From 1234 Modern Endgame Studies. Finding the right moves is not that hard but does your program return a drawish score?
[D]8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - bm Kf7; id "MES.150";
A very interesting test position. Black's inability to make progress without blocking the bishop is the theme.

Is there any engine that evaluates this position as a draw?

[d]8/3k1K2/4p1P1/3p4/5P2/2b5/8/8 b - - 0 3

zullil
Posts: 5599
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Can your program see the draw?

Post by zullil » Tue Mar 23, 2010 11:27 am

jwes wrote:From 1234 Modern Endgame Studies. Finding the right moves is not that hard but does your program return a drawish score?
[D]8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - bm Kf7; id "MES.150";
Here's Stockfish-1.6.3, still "hoping" for success at depth 40:

Code: Select all

info depth 40 score cp -795 time 3552684 nodes 29749026644 nps 8373676 pv g7f7 a5c3 d4d5 c6d5 g5g6 c3a1 f7f8 a1b2 f8f7 d7d6 f7f8 b2c3 f8f7 c3h8 f7g8 h8f6 g8f7 f6d4 f7f8 d6d7 f8f7 d4c3 f7f8 c3a1 f8f7 d7d6 f7f8 d6c6 f8f7 a1h8 f7g8 h8c3 g8f7 c6d6 f7f8 c3b2 f8f7 d6d7 f7f8 b2c3 f8f7 c3h8 f7g8 h8f6 g8f7 f6d4 f7f8 d7d6 f8f7 d4c3 

MikeB
Posts: 3382
Joined: Thu Mar 09, 2006 5:34 am
Location: Pen Argyl, Pennsylvania

Re: Can your program see the draw?

Post by MikeB » Mon Feb 13, 2017 5:13 am

jwes wrote:From 1234 Modern Endgame Studies. Finding the right moves is not that hard but does your program return a drawish score?
[D]8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - bm Kf7; id "MES.150";
another end game study bites the dust...of course, who could fathom playing a kqqkqqb ending ...

a a close to mate score against white is now seen, this is with no back and forth..

Code: Select all

nfo depth 41 seldepth 101 multipv 1 score cp -12714 nodes 5250522807 nps 35636145 hashfull 950 tbhits 49645007 time 147337 pv g7f7 a5c3 d4d5 e6d5 g5g6 c3h8 f7g8 h8b2 g8f7 c6c5 f4f5 c5c4 f5f6 c4c3 g6g7 c3c2 g7g8q c2c1q g8e8 d7c7 e8e7 c7b6 e7b4 b6c6 b4a4 c6d6 a4b4 d6e5 b4e7 e5d4 e7b4 d4e3 b4b3 e3e4 b3g3 b2e5 g3h4 e4d3 f7e6 e5d4 h4h7 d3c4 f6f7 d4c5 h7h4 d5d4 e6f6 c4c3 h4g3 c1e3 g3c7 e3f2 f6g6 d4d3 c7a5 c5b4 a5a1 f2b2 a1a6 b2c2 a6f6 c3b3 f6e6 b3b2 e6e5 c2c3 e5b5 b2c2 b5f5 b4f8 f5f4 d3d2 f4a4 c2b2 a4b5 b2c1 b5f1 d2d1q f1f4 c3d2 f4c7 d2c2 g6g5
[pgn][Event "Computer Chess Game"]
[Site "Mac-Pro.local"]
[Date "2017.02.12"]
[Round "-"]
[White "Komodo 10.3 64-bit"]
[Black "McBrain 2017 v1.3 64 POP"]
[Result "0-1"]
[TimeControl "1/15"]
[FEN "8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - 0 1"]
[SetUp "1"]

1. Kf7 {-2.22/37} Bc3 {+1.98/42 15} 2. d5 {-2.65/30 1.7} exd5 {+3.80/29 6}
3. g6 {-3.84/32 10} Bh8 {+4.56/30 6} 4. Kg8 {-3.89/34 5} Bb2 {+12.37/31 22}
5. Kf7 {-4.10/34 4} c5 {+12.55/29 8} 6. f5 {-4.61/34 9} c4 {+23.00/29 22}
7. f6 {-4.86/36 7} c3 {+23.71/26 3} 8. g7 {-5.70/37 15} c2 {+30.22/29 22}
9. g8=Q {-250.00/36 16} c1=Q {+126.88/23 2.2} 10. Qe8+ {-250.00/44 7} Kd6
{+114.70/19 0.4} 11. Qe7+ {-250.00/45 1:08} Kc6 {+54.67/1 0.1} 12. Qe6+
{-250.00/41 8} Kc5 {+126.88/19 0.4} 13. Qc8+ {-250.00/42 4} Kd4
{+125.59/22 0.5} 14. Qg4+ {-250.00/44 37} Kc3 {+126.88/22 1.1} 15. Qg3+
{-250.00/41 17} Kc4 {+126.92/22 1.4} 16. Qc7+ {-250.00/43 5} Kd3
{+127.20/24 0.5} 17. Qb7 {-250.00/38 4} d4 {+127.27/27 0.8} 18. Kg6
{-250.00/39 13} Qg1+ {+127.29/25 0.7} 19. Kf7 {-250.00/40 36} Qh2
{+127.32/25 0.7} 20. Qb3+ {-250.00/39 9} Ke4 {+127.33/30 1.1} 21. Qe6+
{-250.00/38 2.8} Qe5 {+127.36/31 1.0} 22. Qg4+ {-250.00/39 4} Ke3
{+127.36/32 0.9} 23. Qg1+ {-250.00/37 6} Kd2 {+127.37/33 1.0} 24. Qg2+
{-250.00/39 17} Qe2 {+127.39/32 1.0} 25. Qg5+ {-250.00/39 10} Kc2
{+127.40/34 1.3} 26. Kf8 {-250.00/36 30} Ba3+ {+127.40/31 1.1} 27. Kg8
{-250.00/37 6} d3 {+127.41/32 1.2} 28. f7 {-250.00/35 15} d2
{+127.42/34 2.3} 29. Qg6+ {-250.00/36 9} Kb2 {+127.43/34 1.3} 30. Qb6+
{-250.00/36 12} Ka2 {+127.44/36 1.8} 31. Qa5 {-250.00/36 9} Qg4+
{+127.44/37 3} 32. Kh7 {-250.00/37 12} d1=Q {+127.45/40 2.6} 33. f8=Q
{-250.00/38 26} Qd3+ {+127.46/41 2.6} 34. Kh8 {-250.00/41 6} Qdh3+
{+127.47/45 2.2} 35. Qh5 {-250.00/43 22} Qhxh5+ {+127.48/49 1.6} 36. Qh6
{-1000.01/5 0.1} Qxh6# {+1000.01/127 0.1}
{Black mates} 0-1[/pgn]

Gregory Owett
Posts: 193
Joined: Fri Mar 10, 2006 9:26 am
Location: France

Re: Can your program see the draw?

Post by Gregory Owett » Mon Feb 13, 2017 4:56 pm

"...But after Kf7, Bc3, d5, cxd5 g6 ..."
cxd5 ?? but exd5 !! wins :P

Milos
Posts: 3387
Joined: Wed Nov 25, 2009 12:47 am

Re: Can your program see the draw?

Post by Milos » Tue Feb 14, 2017 12:07 am

jwes wrote:From 1234 Modern Endgame Studies. Finding the right moves is not that hard but does your program return a drawish score?
[D]8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - bm Kf7; id "MES.150";
It looks pretty much as a mate to me, sorry but that study is totally busted.

Code: Select all

FEN: 8/3k2K1/2p1p3/b5P1/3P1P2/8/8/8 w - - 0 1

Brainfish 191116 64 POPCNT:
 38/97	02:09	 4'088'729k	31'690k	-15.10	Kf7 Bc3 d5 exd5 g6 Bh8 Kg8 Bb2 Kf7 c5 f5 c4 f6 c3 g7 c2 g8Q c1Q Qe8+ Kd6 Qe6+ Kc5 Qc8+ Kd4 Qg4+ Kc3 Qg3+ Kc4 Ke6 Bd4 Qc7+ Bc5 Qd7 d4 Qa4+ Kc3 Kd5 Qg5+ Ke4 Qe3+ Kf5 d3 Qa1+ Kb3 Qb1+ Kc4 Qa2+ Kb5 Qb2+ Bb4 Qg2 Qc5+ Kg6 d2 Qe2+ Kb6 Kh6 Qd6 Qe3+ Kc6 Qe8+ Kc7 Qf7+ Kb6 Kg6 Qg3+ Kf5 Qh3+ Ke4 Qg4+ Ke5 d1Q
 39/99	03:11	 6'252'090k	32'622k	-22.22	Kf7 Bc3 d5 exd5 g6 Bh8 Kg8 Bb2 Kf7 c5 f5 c4 f6 c3 g7 c2 g8Q c1Q Qg2 Kc6 Qg4 Kd6 Qe6+ Kc5 Qc8+ Kd4 Qg4+ Kc3 Qg3+ Kc4 Qc7+ Kd3 Qb7 d4 Kg6 Ba3 f7 Qg1+ Kf6 Qf2+ Kg7 Qg3+ Kh7 Qh4+ Kg7 Qg5+ Kh7 Qf5+ Kg8 Qg6+ Kh8 Kd2 Qd5 Qd6 Qa5+ Qb4 Qf5 d3 Kh7 Kc3 Qe6 Qf4 Qa2 Qe4+ Kg7 Qe5+ Kh7 Qd6 Qa1+ Kd2 Qg7 Qh2+ Kg8 Kc2
 40/103	03:58	 7'901'726k	33'101k	-36.00	Kf7 Bc3 d5 exd5 g6 Bh8 Kg8 Bb2 Kf7 c5 f5 c4 f6 c3 g7 c2 g8Q c1Q Qg4+ Kd6 Qe6+ Kc5 Qc8+ Kd4 Qg4+ Kc3 Qg3+ Kc4 Qc7+ Kd3 Qb7 d4 Kg6 Ba3 f7 Qg1+ Kf6 Qf2+ Kg6 Qg3+ Kf5 Qh3+ Kg6 Qe6+ Kg7 Qe5+ Kg8 Qe7 Qb8 Qg5+ Kh8 Qh6+ Kg8 Qg6+ Kh8 Qd6 Qb5+ Kc3 Qf5 Qd8+ Kh7 d3 Qe5+ Qd4 Qe6 d2 Qc6+ Kb3 Qe6+ Kc2 Qc6+ Kb1
 41/103	05:00	 10'168'486k	33'825k	-74.18	Kf7 Bc3 d5 exd5 g6 Bh8 Kg8 Bb2 Kf7 c5 f5 c4 f6 c3 g7 c2 g8Q c1Q Qg4+ Kd6 Qb4+ Ke5 Qb8+ Ke4 Qg3 d4 Kg6 Ba3 Qg4+ Kd3 Qf5+ Kc3 Qf3+ Qe3 Qc6+ Kd2 Qg2+ Qe2 Qd5 Qd3+ Kg7 Qg3+ Kh7 Qd6 Qa5+ Qb4 Qg5+ Kc2 f7 d3 Qg2+ Qd2 Qf3 Qh2+ Kg8 Qc7 Qe4 Qc3 Kh7 Qc5 Qg2+ Kc3 Qe4 Qd6 Qe1+ d2 Qa1+ Kc4 Qa2+ Kb5 Qb1+ Bb4 Qf1+ Ka4 Qa1+ Kb3 Qb1+ Kc4 Qc2+ Kd4 Qa4 Ke3 Qd1 Qd3+ Kg7 Qd5 Qg1+ Kd3 Kg8 Kc2 Qg6+ Kc3
 42/103	07:30	 15'262'947k	33'899k	-128.13	Kf7 Bc3 d5 exd5 g6 Bh8 Kg8 Bb2 Kf7 c5 f5 c4 f6 c3 g7 c2 g8Q c1Q Qe8+ Kd6 Qe7+ Kc6 Qe8+ Kc5 Qf8+ Kc4 Qc8+ Kd4 Qg4+ Kc3 Qf3+ Kc4 Qe2+ Kb4 Qh2 d4 Qb8+ Kc4 Qc7+ Kd3 Qb7 Ba3 Qf3+ Qe3 Qg2 Kc3 Qd5 Kc2 Qg2+ Qd2 Qc6+ Qc3 Qg2+ Kb1 Qf1+ Kb2 Qe2+ Qc2 Qe5 Qc4+ Kg6 Qc3 Qb5+ Bb4 f7 Qc5 Qf1 Qc2+ Kg7 d3 Qf4 Qc3+ Kh7 Kb3 Qe4 d2 Qe6+ Ka3 Qd5 Kb2 Kg8 Qg3+ Kh7
 43/103	07:45	 15'757'864k	33'864k	-128.10	Kf7 Bc3 d5 exd5 g6 Bh8 Kg8 Bb2 Kf7 c5 f5 c4 f6 c3 g7 c2 g8Q c1Q Qe8+ Kd6 Qe7+ Kc6 Qe8+ Kc5 Qf8+ Kc4 Qc8+ Kd4 Qg4+ Kc3 Qf3+ Kc4 Qe2+ Kb4 Qh2 d4 Qb8+ Kc4 Qc7+ Kd3 Qb7 Ba3 Qf3+ Qe3 Qg2 Kc3 Qa2 Bb2 Qa5+ Kb3 Qb6+ Ka2 Qc6 Qc3 Qa8+ Ba3 Kg8 d3 f7 Qc5 Qg2+ Kb3 Qf3 Kc2 Kh7 Qc7 Qf5 Qc3 Qe4 Bf8 Qg6 Bb4 Kg8 Qc8+ Kh7 Bc3 Qg2+ d2 Qe4+ Kb2 Qf3

jdart
Posts: 3816
Joined: Fri Mar 10, 2006 4:23 am
Location: http://www.arasanchess.org

Re: Can your program see the draw?

Post by jdart » Tue Feb 14, 2017 12:37 am

Here is another one I have been looking at recently:

[D] R7/7k/P1r5/7P/6K1/8/8/8 w - - 0 66

This is a tablebase draw, unless White puts his own Rook en prise. But Stockfish w/o 6-man bases gives it a large + score (Arasan, too).

--Jon

corres
Posts: 1558
Joined: Wed Nov 18, 2015 10:41 am
Location: hungary

Re: Can your program see the draw?

Post by corres » Tue Feb 14, 2017 8:08 am

I think in the view point of programming there is no any obstacle to "teach" engines for understanding of drawn positions. Nevertheless, these plus codes decrease the speed of engine and enhance the possibility of a drawn. So the outcome should be a weaker engine.

User avatar
Nordlandia
Posts: 2402
Joined: Fri Sep 25, 2015 7:38 pm
Location: Sortland, Norway

Re: Can your program see the draw?

Post by Nordlandia » Tue Feb 14, 2017 9:55 am

Fortress !?

[d]8/5pk1/4p1p1/2r4p/5P2/1Q3KP1/7P/8 b - - 0 0

[d]8/5pk1/1Q2p1p1/3r3p/4KP2/6PP/8/8 w - - 0 0

Post Reply