New test suite "Tactical Insanity"

[BBauer]

I tried.
Pos 56 has as bm Rc1.
IMHO Kd4, a3, Rb1 win as well.
So what? I expected something like that.
Kind regards
Bernhard

Thanks for the defect report.

I stumbeled over pos 135. Here Qh5+ should be found.
Clearly Qh5+ is impossible, but Qh5 gives a high score - for black! of course will black capture the white queen
and continue to win.
Did you ever took a look on any of your 140 positions?
Why you called this "test suit" "Tactical Insanity" is beyond me.
Anyway, kind regards
Bernhard

[Dann Corbit]

I tried.
Pos 56 has as bm Rc1.
IMHO Kd4, a3, Rb1 win as well.
So what? I expected something like that.
Kind regards
Bernhard

Thanks for the defect report.

I stumbeled over pos 135. Here Qh5+ should be found.
Clearly Qh5+ is impossible, but Qh5 gives a high score - for black! of course will black capture the white queen
and continue to win.
Did you ever took a look on any of your 140 positions?
Why you called this "test suit" "Tactical Insanity" is beyond me.
Anyway, kind regards
Bernhard

Thank you again for you excellent work in correcting this tactical position suite.
Believe it or not, I did examine the positions, but lacking your chess ability, I seem to have made at least two serious gaffes.
But with expert help from chess genius like you, I expect that my test set can be repaired.
Thank you again for your helpful assessments.

[Paloma]

Hi Dann,
about pos 135:

BBauer wrote:
Clearly Qh5+ is impossible, but Qh5 gives a high score - for black! of course will black capture the white queen
and continue to win.

But at your pv Qh5+ g6 Nxg6 hxg6 Qxg6+

is 2.Nxg6 also impossible, no wNight can reach field g6, so the FEN is incorrect!

I wonder wich engine outputs depth40 after 2066 sec. with ce 432 with this wrong FEN. (acd 40; acs 2066; bm Qh5+; c3 "Qh5+"; ce 432;)

So BBauer's impression >Did you ever took a look on any of your 140 positions?< seems to be right

[peter]

Thanks for the defect report.

I stumbeled over pos 135. Here Qh5+ should be found.
Clearly Qh5+ is impossible, but Qh5 gives a high score - for black! of course will black capture the white queen
and continue to win.
Did you ever took a look on any of your 140 positions?
Why you called this "test suit" "Tactical Insanity" is beyond me.
Anyway, kind regards
Bernhard

Thank you again for you excellent work in correcting this tactical position suite.
Believe it or not, I did examine the positions, but lacking your chess ability, I seem to have made at least two serious gaffes.
But with expert help from chess genius like you, I expect that my test set can be repaired.
Thank you again for your helpful assessments.

2 of 140 seems not too bad to me.
Albert Silver took out 100 of 300 from WAC, which is an old classic.
Should really start going on with looking at your new set closer too, Dann, who kows, what else could be found, but waiting for Bernhard's further investigations would spare others some work, wouldn't it? I always tend to spend too much time with single positions.

And yes, 135 is simply a wrong pv for a wrong fen, you mixed up two different positions probably.

[MikeGL]

The inspiration for this test is a mish-mash of things, but mostly these are positions where about half the engines I tried got the wrong answer after 30 seconds using 6 cores at 3.4 GHz.

Some are positions I have studied over time and found interesting.

All of them are positions which I did not have an ID for, so they are most likely not contained in other famous EPD test suites (but there are probably exceptions to that since I did not check carefully).

Thanks a lot for this collection.
Some positions are easy for SF9 with Multi-PV, but overall this is a good EPD test set IMO.


__________________________
Mike Libanan

[Dann Corbit]

I stumbeled over pos 135. Here Qh5+ should be found.
Clearly Qh5+ is impossible, but Qh5 gives a high score - for black! of course will black capture the white queen
and continue to win.
Did you ever took a look on any of your 140 positions?
Why you called this "test suit" "Tactical Insanity" is beyond me.
Anyway, kind regards
Bernhard

Thank you again for you excellent work in correcting this tactical position suite.
Believe it or not, I did examine the positions, but lacking your chess ability, I seem to have made at least two serious gaffes.
But with expert help from chess genius like you, I expect that my test set can be repaired.
Thank you again for your helpful assessments.

2 of 140 seems not too bad to me.
Albert Silver took out 100 of 300 from WAC, which is an old classic.
Should really start going on with looking at your new set closer too, Dann, who kows, what else could be found, but waiting for Bernhard's further investigations would spare others some work, wouldn't it? I always tend to spend too much time with single positions.

And yes, 135 is simply a wrong pv for a wrong fen, you mixed up two different positions probably.

I have an automated process which ties together the output of the engines with the EPD position (for engines other than Stockfish, which I modified to write output directly to disk as it processes the data).

No doubt, it is analysis for another position. I made sure that the key moves were legal, but certainly should have paid closer attention to 135.

[BBauer]

Thank you again for you excellent work in correcting this tactical position suite.
Believe it or not, I did examine the positions, but lacking your chess ability, I seem to have made at least two serious gaffes.
But with expert help from chess genius like you, I expect that my test set can be repaired.
Thank you again for your helpful assessments.

2 of 140 seems not too bad to me.
Albert Silver took out 100 of 300 from WAC, which is an old classic.
Should really start going on with looking at your new set closer too, Dann, who kows, what else could be found, but waiting for Bernhard's further investigations would spare others some work, wouldn't it? I always tend to spend too much time with single positions.

And yes, 135 is simply a wrong pv for a wrong fen, you mixed up two different positions probably.

I have an automated process which ties together the output of the engines with the EPD position (for engines other than Stockfish, which I modified to write output directly to disk as it processes the data).

No doubt, it is analysis for another position. I made sure that the key moves were legal, but certainly should have paid closer attention to 135.

Never mind! Not everyone is a chess genius like me. Anyway, thank you for your kind words which encourage me to continue my critics.
Lets have a look at position 131.
[D] r5k1/R5p1/p4p1p/1p1b1P2/8/4B3/2B2K2/8 w - -
You give: r5k1/R5p1/p4p1p/1p1b1P2/8/4B3/2B2K2/8 w - - acd 64; acs 1200; bm Bb3; c3 "Bb3"; ce 306; pm Bb3; pv Bb3;
So the pv is: Bb3.
Isn`t this a little short?
A chess genius like me would consider the following variant:
1.Bb3 Bxb3 2.Rxa8+ Kh7 3.Rxa6 Bc2 4.Rb6 Bxf5 5.Rxb5 and my Stockfish gives a score of -2.93 which doesn´t change. So what? We have no winning move.
Is this an insane tactical position?
My anwer: Yes!
Kind regards
Bernhard

Ps: WAC ws mentioned. I would have removed not only 100 positions, but at least 270 positions.

[BBauer]

Thank you again for you excellent work in correcting this tactical position suite.
Believe it or not, I did examine the positions, but lacking your chess ability, I seem to have made at least two serious gaffes.
But with expert help from chess genius like you, I expect that my test set can be repaired.
Thank you again for your helpful assessments.

2 of 140 seems not too bad to me.
Albert Silver took out 100 of 300 from WAC, which is an old classic.
Should really start going on with looking at your new set closer too, Dann, who kows, what else could be found, but waiting for Bernhard's further investigations would spare others some work, wouldn't it? I always tend to spend too much time with single positions.

And yes, 135 is simply a wrong pv for a wrong fen, you mixed up two different positions probably.

I have an automated process which ties together the output of the engines with the EPD position (for engines other than Stockfish, which I modified to write output directly to disk as it processes the data).

No doubt, it is analysis for another position. I made sure that the key moves were legal, but certainly should have paid closer attention to 135.

Let´s have a look at pos 107.
[D]r1b1kbnr/pppp1Npp/8/8/8/5n2/PPPPBP1P/RNBQKR2 w Qkq - acd 38; acs 3600; bm Bxf3; c3 "Bxf3"; ce 1397; pm Bxf3;
Clearly Bxf3 is the best move. It is also the ONLY move. This tests your program in finding only moves.
Insane test position.
Keep up the good work, as they say.
Kind regards
Bernhard

[Dann Corbit]

2 of 140 seems not too bad to me.
Albert Silver took out 100 of 300 from WAC, which is an old classic.
Should really start going on with looking at your new set closer too, Dann, who kows, what else could be found, but waiting for Bernhard's further investigations would spare others some work, wouldn't it? I always tend to spend too much time with single positions.

And yes, 135 is simply a wrong pv for a wrong fen, you mixed up two different positions probably.

I have an automated process which ties together the output of the engines with the EPD position (for engines other than Stockfish, which I modified to write output directly to disk as it processes the data).

No doubt, it is analysis for another position. I made sure that the key moves were legal, but certainly should have paid closer attention to 135.

Never mind! Not everyone is a chess genius like me. Anyway, thank you for your kind words which encourage me to continue my critics.
Lets have a look at position 131.
[D] r5k1/R5p1/p4p1p/1p1b1P2/8/4B3/2B2K2/8 w - -
You give: r5k1/R5p1/p4p1p/1p1b1P2/8/4B3/2B2K2/8 w - - acd 64; acs 1200; bm Bb3; c3 "Bb3"; ce 306; pm Bb3; pv Bb3;
So the pv is: Bb3.
Isn`t this a little short?
A chess genius like me would consider the following variant:
1.Bb3 Bxb3 2.Rxa8+ Kh7 3.Rxa6 Bc2 4.Rb6 Bxf5 5.Rxb5 and my Stockfish gives a score of -2.93 which doesn´t change. So what? We have no winning move.
Is this an insane tactical position?
My anwer: Yes!
Kind regards
Bernhard

Ps: WAC ws mentioned. I would have removed not only 100 positions, but at least 270 positions.

I see from your pv that you have quite a sense of humor.
Thanks for the nice belly laugh.

[Dann Corbit]

Sting99:
FEN: r5k1/R5p1/p4p1p/1p1b1P2/8/4B3/2B2K2/8 w - - 0 1

Sting99:
2 00:00 310 1,135 +0.48 Rd7 Rc8
3 00:00 593 2,095 +0.68 Rd7 Rc8 Bd3
4 00:00 1,154 3,979 +0.52 Rd7 Rc8 Bd3 Bc6
5- 00:00 1,435 4,815 +0.36 Rd7 Rc8
5+ 00:00 1,593 5,257 +0.64 Rd7 Rc8 Bd3 Bc4 Be4
5 00:00 1,797 5,834 +0.80 Rd7 Rc8 Bd3 Bc4 Be4
6- 00:00 2,293 7,012 +0.60 Rd7 Rc8
6- 00:00 2,464 7,377 +0.44 Rd7 Rc8 Bd3 Bc4 Be4 a5
6 00:00 2,851 8,385 +0.40 Rd7 Rc8 Bd3 Bc4 Be4 a5
7- 00:00 4,606 13,048 +0.28 Rd7 Rc8
7 00:00 12,551 34,767 +0.20 Rc7 Re8 Bd3 Be4 Bxe4 Rxe4 Ke2
8 00:00 25,663 68,617 +0.16 Rd7 Rc8 Bd3 Bc4 Ke2 a5 Bxc4+ Rxc4 Kd3
9- 00:00 42,970 111,033 +0.04 Rd7 Rc8
9 00:00 44,887 115,688 +0.08 Rd7 Rc8 Bd3 Bc4 Ke2 b4 Bxc4+ Rxc4 Kd3 Rc3+ Ke4
10 00:00 102,993 245,221 0.00 Rd7 Rc8 Bd3 Bc4 Bxc4+ Rxc4 Rd8+ Kf7 Rd7+ Kg8
11 00:00 139,371 318,926 0.00 Rd7 Rc8 Bd3 Bc4 Bxc4+ Rxc4 Rd8+ Kf7 Rd7+ Kg8
12+ 00:00 166,804 373,163 +0.08 Rd7
12 00:00 245,259 514,169 0.00 Rd7 Rc8 Bd3 Bc4 Ke2 b4 Bd4 b3 Ra7 h5 Bxc4+ Rxc4 Kd3 Rc6
13+ 00:00 366,198 735,337 +0.08 Rd7
13 00:00 428,985 824,971 0.00 Rd7 Rc8 Bd3 Bc4 Ke2 b4 Ra7 Bxd3+ Kxd3 Rc3+ Ke4 b3 Ra8+ Kf7
14- 00:00 752,510 1,290,754 -0.08 Rd7 Bc4
14- 00:00 1,142,742 1,777,203 -0.16 Rd7 Bc4
14+ 00:00 1,495,490 2,077,069 +0.08 Rd7 Bc4 Bd4 a5 Be4 Re8 Bf3 a4 Ra7 Rd8 Ke3 Re8+ Kd2 Rd8 Kc3 a3 Bc6
14 00:00 1,527,946 2,076,013 0.00 Rd7 Bc4 Bd4 a5 Be4 Re8 Bf3 a4 Ra7 Rd8 Ke3 Re8+ Kd2 Rd8 Ke3
15 00:00 1,670,079 2,166,120 0.00 Rd7 Bc4 Bd4
16 00:00 1,880,491 2,301,702 0.00 Rd7 Bc4 Bd4 a5 Be4
17 00:00 2,279,974 2,541,777 0.00 Rd7 Bc4 Bd4 a5 Be4 Re8 Bf3 a4 Ra7 Rd8 Ke3 Re8+ Kd2 Rd8 Ke3
18+ 00:01 4,905,108 4,094,414 +0.08 Rc7
18 00:01 5,350,569 4,239,753 0.00 Rc7 a5 Rc5 Bc4 Ba4 h5 Bc1 h4 Bxb5 Bxb5 Rxb5 a4 Ba3 h3 Rd5 Rc8 Rd2 Kh7 Bd6 g5 fxg6+ Kxg6 Kg3 Rc3+
19+ 00:01 7,505,858 5,010,586 +0.08 Rc7
19+ 00:01 9,149,663 5,341,309 +0.16 Bb3
19+ 00:01 9,181,753 5,341,333 +0.28 Bb3 Rxa7 Bxd5+ Rf7 Bc5 a5 Be6 a4 Bb4 h5 Kg2 h4 Bc5 b4 Bxb4 h3+ Kxh3 Kh7 Bxf7 a3 Bxa3 g5 fxg6+ Kg7 Be6 Kxg6 Kg2 Kg7 Bf5 Kf7
19+ 00:01 9,206,337 5,333,914 +0.46 Bb3 Rxa7 Bxd5+ Rf7 Bc5 a5 Be6 a4 Bb4 h5 Kg2 h4 Bc5 b4 Bxb4 h3+ Kxh3 Kh7 Bxf7 a3 Bxa3 g5 fxg6+ Kg7 Be6 Kxg6 Kg2 Kg7 Bf5 Kf7 Bb4
19+ 00:01 9,340,640 5,340,560 +1.14 Bb3 Rxa7 Bxd5+ Rf7 Bc5 a5 Be6 a4 Bb4 h5 Kg2 h4 Bc5 b4 Bxb4 h3+ Kxh3 Kh7 Bxf7 a3 Bxa3 g5 fxg6+ Kg7 Be6 Kxg6 Kg2 Kg7 Bf5 Kf7 Bb4
19+ 00:01 9,376,666 5,339,787 +1.75 Bb3 Rxa7 Bxd5+ Rf7 Bc5 a5 Be6 a4 Bb4 h5 Kg2 h4 Bc5 b4 Bxb4 h3+ Kxh3 Kh7 Bxf7 a3 Bxa3 g5 fxg6+ Kg7 Be6 Kxg6 Kg2 Kg7 Bf5 Kf7
19+ 00:01 9,690,433 5,344,971 +2.67 Bb3 Rxa7 Bxd5+ Rf7 Bc5 a5 Be6 a4 Bb4 h5 Kg2 h4 Bc5 b4 Bxb4 h3+ Kxh3 Kh7 Bxf7 a3 Bxa3 g5 fxg6+ Kg7 Be6 Kxg6 Kg2 Kg7 Bf5 Kf7 Bb4 Kg7
19 00:01 11,006,863 5,607,164 +2.86 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Bc5 Bc2 Ra8 Bxf5 Ke3 Kg6 Rxa6 h5 Ra7 Be6 Bf8 Bf7 Rb7 h4 Kf4 h3 Kg3 f5 Rxb5
20+ 00:02 11,869,678 5,712,068 +2.98 Bb3
20 00:02 12,022,201 5,719,410 +2.98 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Bc5 Bc2 Ra8 Bxf5 Ke3 Kg6 Rxa6 h5 Ra7 Be6 Bf8 Kf5 Rxg7 h4 Bd6 Bc4 Rh7 Kg4 Rb7
21- 00:02 12,707,316 5,831,719 +2.86 Bb3 Rd8
21 00:02 12,767,109 5,827,069 +2.86 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Bc5 Bc2 Ra8 Bxf5 Ke3 Be6 Rxa6 Bc4 Ra7 Kg8 Bd4 h5 Kf4 Bd5 Rd7 Bc4 Rc7 Bd5 Kg3
22+ 00:02 13,093,727 5,845,413 +2.98 Bb3
22 00:02 14,067,465 5,973,445 +2.98 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Bc5 Bc2 Ra8 Bxf5 Rxa6 Be4 Ra7 Kg8 Ke3 Bd5 Kf4 Bc4 Rc7 h5 Bd4 Bd5 Rd7 g5+ Kf5
23- 00:02 14,877,778 6,082,493 +2.86 Bb3 Rd8
23 00:02 15,221,954 6,100,983 +2.78 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Bc5 Bc2 Ra8 Bxf5 Rxa6 Bd3 Ra7 Kg8 Bd4 Bc4 Rb7 Kf8 Ke3 h5 Kf4
24+ 00:02 15,942,056 6,143,374 +2.90 Bb3
24 00:02 16,477,240 6,182,829 +2.94 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Ra8 Bc2 Bc5 Bxf5 Rxa6 Bd3 Ra7 Kg8 Bd4 Bc4 Rb7 Kf8 Kg3 h5 Rc7 Kg8 Kf3 Kf8
25 00:02 18,018,516 6,366,966 +2.98 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Ra8 Bc2 Bc5 Bxf5 Rxa6 Bd3 Ra7 Kg8 Bd4 Bc4 Rb7 Kf8 Kg3 h5 Ra7 Kg8 Rc7 Kf8
26 00:02 18,443,286 6,388,391 +2.98 Bb3 Rd8 Ra8
27- 00:03 21,103,275 6,661,387 +2.90 Bb3 Rd8
27 00:03 21,500,711 6,679,313 +2.90 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Ra8 Bc2 Bc5 Bxf5 Rxa6 Bd3 Ra7 Kg8 Bd4 Bc4 Rb7 Kf8 Kg3 h5 Bc5+ Kg8 Kf4 Be2 Bd4 Bd3 Ke3 Bc4
28- 00:03 29,396,758 7,483,899 +2.82 Bb3 Rd8
28 00:04 30,819,000 7,520,497 +2.82 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Ra8 Bc2 Bc5 Bxf5 Rxa6 h5 Ra7 Kg8 Ke3 Be6 Re7 Bc4 Kf4 Bd3 Rd7 Bc2 Kg3 Bb3 Rb7 Bc4 Kf4
29 00:07 73,807,998 9,688,631 +2.82 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Ra8 Bf7 Ra7 Bc4 Bc5 Kg8 Ra8+ Kh7 Kg3 a5 Rxa5 h5 Ra7 Kg8 Kf4 Bd5 Ra5 Bc4 Bd4 Kf7 Ra7+ Kg8 Rc7 Bd3 Rd7
30 00:10 111,702,240 10,384,144 +2.82 Bb3 Rd8 Ra8 Bxb3 Rxd8+ Kh7 Ra8 b4 Bc5 Bc4 Bxb4 a5 Bxa5 Bd3 Bb4 Bxf5 Ra7 Kg8 Kg3
31 00:14 153,574,146 10,734,946 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2 Kf4 b4 Rxa6 b3 Bd4 b2 Bxb2 h5 Ra7 Kg8 Bd4 Bd3 Rc7 Bb5 Bc3 Bd3 Rd7 Bb5 Rb7 Bf1 Bd4 Bd3 Bc5 Bc4 Bb4 Bd3 Rc7
32 00:18 208,500,423 11,075,131 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2 Kf4 b4 Rxa6 b3 Bd4 Bd3 Ra7 Kg8 Rb7 Bc2 Bc5 h5 Rb8+ Kh7 Rb6 Bd1 Bd4 h4 Rb8 h3 Kg3 Bc2 Rb7
33 00:19 212,098,550 11,028,418 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2
34 00:24 279,419,640 11,210,416 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2 Kf4 b4 Rxa6 b3 Bd4 Bd3 Ra7 Kg8 Rb7 Bc2 Bb2 Bd1 Rb8+ Kh7 Kg3 Bc2 Rb7 Kh8 Kf4 Kg8 Rb8+ Kh7 Rb5 Kg8 Bd4
35 00:27 301,984,859 11,163,537 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2 Kf4 b4 Rxa6 b3 Bd4 Bd3 Ra7 Kg8 Rb7 Bc2 Bb2 Bd1 Rb8+ Kh7 Kg3 Bc2 Rb7 Kh8 Kf4 Kg8 Rb8+ Kh7 Rb5 Kg8 Bd4
36 00:32 363,718,305 11,160,084 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2 Kf4 b4 Rxa6 b3 Bd4 Bd3 Ra7 Kg8 Rb7 Bc2 Bb2 Bd1 Rb8+ Kh7 Kg3 Bc2 Rb7 Kh8 Kf4 Kg8 Rb8+ Kh7 Rb5 Kg8 Bd4
37 00:49 568,521,291 11,449,888 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2 Kf4 b4 Rxa6 b3 Bd4 Bd3 Rb6 Bc2 Rb5 Kg8 Bb2 Kh7 Rb7 Kh8 Rb6 Kg8 Rb8+ Kh7 Rd8 h5 Rd7 Kg8
38 01:05 742,740,270 11,410,097 +2.82 Bb3 Bxb3 Rxa8+ Kh7 Kg3 Bc2 Kf4 b4 Rxa6 b3 Bd4 Bd3 Rb6 Bc2 Rb5 Kg8 Bb2 Kh7 Rb7 Kh8 Rb6 Kg8 Rb8+ Kh7 Rd8 Bb1