Crafty 23.1 50-Move-Rule Draw Bug

[hgm]

I guess the bottom line is that quality and playing strength are independent dimensions. There are many things you can leave out of an engine without a measurable effect on their Elo performance, such as underpromotion to Bishop, KBNK matng ability, Ability to see zugzwang in complex positions, etc. Yet a user experiences such short-cuts / short-comings as a serious black mark on the quality of the program. Because users are not only interested in games. They also use engines to solve problems, and evaluate other highly artificial positions.

[bob]

Hi Bob.
The *rare* argument is not a good one.
Most positions in chess are *rare*.
IMHO a program should handle alle possible positions correct.
Such a program will I call a chess program.
In the past we have seen a lot of crap like cannot handle
- insufficient material
- underpromotion
- studies
Recently Kramnik said about a game Carlsen-Kramnik that Carlsen began to produce studies and mate threats.
kind regards
Bernhard

I agree. And in real games I do not have crafty resign or offer draws, which makes this a non-issue. But I have normally been running with the "obvious cases" being dealt with (KN vs KN, etc).

Version 23.2 won't claim any of those as draws, however, although this will lengthen tournaments and games for some (on occasion).

[hgm]

If you worry about wasting time of Human users, the solution would be to offer a draw, through he offer draw command, or at least to accept a draw (therough the offer draw command) when the opponent offers you one (through the draw command).

If you worry about wasting tester's time in engine-engine games, don't: if they want to save time by adjudicating games that under strict FIDE rules are not draws (e.g. KBKN or KRKR), it is a GUI task to make this adjudication. So they can simply instruct the GUI to adudcate 'trivial draws' in addition to 'insufficien-material draws'.

[bob]

If you worry about wasting time of Human users, the solution would be to offer a draw, through he offer draw command, or at least to accept a draw (therough the offer draw command) when the opponent offers you one (through the draw command).

If you worry about wasting tester's time in engine-engine games, don't: if they want to save time by adjudicating games that under strict FIDE rules are not draws (e.g. KBKN or KRKR), it is a GUI task to make this adjudication. So they can simply instruct the GUI to adudcate 'trivial draws' in addition to 'insufficien-material draws'.

This is exactly what 23.2 does. It will offer draws where it used to claim them for KB vs KN and such.