Problem 38235

[tomohawk]

After 1.Bf6 Qf8 the computer gave 2.Qxf8+ as the only correct move, with 2.Qxg7+ getting the "try again" notice. Strangely, 2.Bxg7+, which transposes into the line after 2.Qxg7+ was marked as failure.

[richard]

It is not clear to me why toga gives any difference here, given enough time it does see them as equal.  The current generator I'm building new problems with would reject this position due to not having enough gap between best and second best. All the moves hover around the 2.75 ambiguity threshold and the bishop move appears to have been just below alternative threshold.  I need to generate a few more new problems (as some higher rated users are running out of suitable problems at the high end), but will in the near future do another verification run on the existing problems with the best to second best gap enforced and probably a lower "winning line" threshold.  I'm also planning a complete generator rewrite in the near future in order to get a bit more flexibily in the generator and allow me to explore a wider range of ideas in the generator with a newer cleaner underlying codebase (whilst still keeping the improvements in generation heuristics made over previous ancestors over the generator).

This is probably one of the worst problems I've seen in the current set :-)

Regards,
Richard.