a higher standard for engines

[lech]

There is no standard for engines. Only strength is determined by ELO as result of never-ending computer matches. It can hide some defects, which don’t mean much for statistical comparisons. The defects are important for active users, though.
My text is a try to make a higher standard for engines.
http://free.of.pl/c/ccintorun/g/crazy%20queen.pdf
I hope it will find a reflection.

[BBauer]

You should give the FEN positions for white and for black.
And I agree, elo is not everything.
Programs coming with source code allow you to change the behavior.
For example you may change one single character in stockfish and it will behave much different.
I enjoyed your positions.
Kind regards
Bernhard

[lech]

Thanks Bernhard for a piece of advice. Now, all positions are saved in FEN for both White and Black options.
PDF-file: http://free.of.pl/c/ccintorun/g/crazy%20queen.pdf
Stockfish, mentioned by you, have a high IQ (by test) and I have respect for its authors.
Unfortunately, the high IQ is not necessary to achieve the top of ELO. But some need a reliable engine. I hope that creators of Stockfish (or others) can do it.

[Robert Flesher]

Am I missing something here? [d] 8/5pKN/5Pp1/5kP1/1B6/8/b7/7q b - - 0 1





1... Qxh7+ 2. Kxh7 Kxg5 3. Bd2+ Kxf6 4. Kh6 *

giving this position [d] 8/5p2/5kpK/8/8/8/b2B4/8 b - - 0 4


which is a draw!

[Michel]

It seems to me that black should not take the pawn on f6 but rather
play Kh5.

[lech]

of course 3...Kh5 wins.
The missing (engines) simple move 1...Qxh7 is hard to explain and accept for users.

[Robert Flesher]

It seems to me that black should not take the pawn on f6 but rather
play Kh5.

ah yes of course, my human greed just wanted another pawn!

[lech]

It is possible to arise a "crazy rook" too.

[d]7k/6rp/6pN/6P1/8/2B5/6r1/K7 b - - 0 1

1…Rxg5 wins
It is funny how engines try to shoot the white Bishop away.

[lech]

[d] 7R/6pb/6k1/Pr4Pp/7P/8/6K1/8 b - - 0 1

1...Kf7 wins
I hope my bug (wrong behavior) report was useful.