New update of the ENGINES CRACKERS (3rd version)

[Vinvin]

Wrong test position in the set ?

The Engines Crackers 09

[d]2b1k3/8/6R1/2n5/8/B1r1N3/1pB5/6K1 w – – 0 1

The comments say "Mate in 66" but it's a forced draw by best play because of the "50 moves rule".
See study here : https://lichess.org/study/9YSDb3T0/A1FTu20A

[peter]

Wrong test position in the set ?

The Engines Crackers 09

[d]2b1k3/8/6R1/2n5/8/B1r1N3/1pB5/6K1 w – – 0 1

The comments say "Mate in 66" but it's a forced draw by best play because of the "50 moves rule".
See study here : https://lichess.org/study/9YSDb3T0/A1FTu20A

Both is right, Vincent, it's a cursed win, and even if 50 moves rule doesn't matter as for a study, yet engines can't see the win without being told so.
Here I have an old Cfish with UCI- option to extend the rule to more than 50 moves:

2b1k3/8/6R1/2n5/8/B1r1N3/1pB5/6K1 w - - 0 1

Analysis by Cfish 210718 64 POPCNT NUMA:

1. +- (3.14): 1.Bxb2 Rxe3 2.Rg8+ Kd7 3.Bf5+ Ne6 4.Rxc8 Kxc8 5.Kf2 Re4 6.Bxe4 Nc5 7.Bf5+ Kc7 8.Ke3 Kd6 9.Bg4 Kd5 10.Bf3+ Kc4 11.Bh5 Kd5 12.Bf7+ Ne6 13.Bf6 Kd6 14.Ke4 Nc5+ 15.Kf5 Nd7 16.Bd4 Kc6 17.Bb2 Kb6 18.Ke4 Kc7 19.Bc4 Nb6 20.Bb5 Kd6 21.Be2 Nd7 22.Bc4 Nc5+ 23.Kf5 Nd7 24.Bd4 Kc6 25.Bf7 Kd6 26.Bb3 Kc7 27.Be3 Kd6 28.Kf4 Ke7 29.Bc4 Kd6 30.Ke4 Nf6+ 31.Kf5 Ne8 32.Bd3 Ng7+ 33.Ke4

2. +/- (1.13): 1.Rg8+ Kf7 2.Bxb2 Kxg8 3.Bxc3 Ba6 4.Bb1 Kf7 5.Kg2 Bb7+ 6.Kh3 Ne4 7.Bb2 Ke6 8.Kh4 Ba8 9.Ba2+ Kd6 10.Nf5+ Kc6 11.Ba3 Nc3 12.Bc4 Nd5 13.Be2 Kd7 14.Nd4 Nc3 15.Bg4+ Kc7 16.Ne6+ Kb6 17.Kg5 Ka5 18.Kf5 Ka4 19.Bc5 Kb3 20.Ke5 Bg2 21.Nd4+ Kb2 22.Be6 Bf1 23.Bb4 Bd3 24.Nf5 Ba6 25.Kf4 Bb7 26.Ne3 Ba6 27.Ba5 Ne2+ 28.Ke5 Kc1 29.Bd5 Bd3 30.Ba2
regards

[Jouni]

Position 11 after 2 moves "takeback" and full hash:

Analysis by ShashChess 20.2

1.Rec1 f2+ 2.Kh1 Nxf1 3.Qxa3+ bxa3 4.b4 Nb6 5.Rc7+ Ka6 6.Ra7+ Rxa7 7.Rc6 Rh6+ 8.Rxh6 f6 9.Rxf6 Kb7 10.Rxb6+ Kc7 11.Rc6+ Kd7 12.Rc7+ Ke6 13.Rxe7+ Kd5 14.Rd7+ Kc4 15.Rd4+ Kb3 16.Rd3+ Ka4 17.Rxa3+ Kxb4 18.Rb3+ Kc4 19.Rb4+ Kd3 20.Rb3+
= (0.00) Depth: 37/54 00:00:05 39040kN, tb=45538

[Vinvin]

Both is right, Vincent, it's a cursed win, and even if 50 moves rule doesn't matter as for a study, yet engines can't see the win without being told so.
Here I have an old Cfish with UCI- option to extend the rule to more than 50 moves:

So, if I understand well : with the FIDE rules, this position is draw. And with other rules, it could be winning for white ?

[peter]

Both is right, Vincent, it's a cursed win, and even if 50 moves rule doesn't matter as for a study, yet engines can't see the win without being told so.
Here I have an old Cfish with UCI- option to extend the rule to more than 50 moves:

So, if I understand well : with the FIDE rules, this position is draw. And with other rules, it could be winning for white ?

It's a cursed win, yes, of course, mate's just not within 50 moves- boundary, yet Walter's comment about the DTM is correct. Fide rules say draw, but that doesn't matter as for the correctness of a composed study, you see?
It's agreement between composers and judges of studies as for their usual evaluation, of course you need not accept that as for your personal appraisal.
Tablebase- wins longer then 50 moves without pawn- move or material reduction are often so called cursed wins too, as you probably know regards

[Vinvin]

Both is right, Vincent, it's a cursed win, and even if 50 moves rule doesn't matter as for a study, yet engines can't see the win without being told so.
Here I have an old Cfish with UCI- option to extend the rule to more than 50 moves:

So, if I understand well : with the FIDE rules, this position is draw. And with other rules, it could be winning for white ?

It's a cursed win, yes, of course, mate's just not within 50 moves- boundary, yet Walter's comment about the DTM is correct. Fide rules say draw, but that doesn't matter as for the correctness of a composed study, you see?
It's agreement between composers and judges of studies as for their usual evaluation, of course you need not accept that as for your personal appraisal.
Tablebase- wins longer then 50 moves without pawn- move or material reduction are often so called cursed wins too, as you probably know regards

Thanks to make it very clear !
I didn't know this thing exists.

[Dann Corbit]

I think a chess engine that chooses a cursed win is broken, unless there is a UCI parameter that allows such things.
It can cause a possible win turn into a draw.

[Hai]

Wrong test position in the set ?

The Engines Crackers 09

[d]2b1k3/8/6R1/2n5/8/B1r1N3/1pB5/6K1 w – – 0 1

The comments say "Mate in 66" but it's a forced draw by best play because of the "50 moves rule".
See study here : https://lichess.org/study/9YSDb3T0/A1FTu20A

Top Chess Engines Testsuite 2024 v2
https://www.mediafire.com/file/cypaz2t0 ... 2.pgn/file
Number 38.
1-0
If a piece is captured in the meantime, which is the case, then there is no draw.