Puzzles Don't Choose Challenging Responses

The same seems to be true of the chess lessons.

The demonstration of mating with bishop and knight here https://www.chess.com/terms/checkmate-chess?ncc=1#king-bishop-and-knight 

starts from this (Black to play) position.

Black plays 1...Kf8 but 1...Kxh7 would have been much more challenging.

Edit: Don't bother looking - they fixed it at last.

I just wrote a blog that covers this question of why puzzles often don't choose "challenging" responses. See Understanding soundness and motivations in chess puzzles, problems, and studies.

The underlying reason is that puzzle responses need to provoke unique winning moves, for the puzzle to work at all. The more "challenging" responses will likely allow you to win in more than one way, making the puzzle unsound. I used a clear-cut example in the post, but the OP puzzle also illustrates this. So according to Stockfish evaluation (depth 38), the more testing 24.Nxd7 enables two winning replies, 24...Nxd7 (-3.81) and 24...Nxe2+ (-2.82).  

Thank you @Rocky64, I will take a look and I appreciate your explanation, it makes sense

Ah, interesting!

Just found an extraordinarily dumb puzzle... please delete

https://www.chess.com/puzzles/problem/1332361

I understand that it is supposed to be a 50 move rule thing, but to promote the pawn, you have to push it, which resets the 50 move rule!

Along with rerating all of the 2600 back rank puzzles LOL

lol

You're assuming a pawn promotion (or piece capture). The original position probably came from a lower rated player game that didn't know how to mate with king and queen and an automated system pulled the position where there was a mate in X or draw situation.

Those types of positions are interesting, and probably should have some kind of clue that they are Mate in X type puzzles though.

Yeah it's always nice to try to find the quickest and most concise way to win

Ah. 

50 move rule puzzles are so stupid, like if you couldnt find mate in the last 40 moves how would you now find the quickest mate possible

Staff are removing them now when they are reported. Though, I actually like them as it helps with calculation.

In most occasions they could be replaced with a "mate-in-n" instruction since that covers their effective use. Chess.com (for some reason) does not support the "mate-in-n" type and simulates it with 50-move constructions. But if you really want to train the calculation skill why not simply introduce it as a regular puzzle type? Thousands of solvers of composed chess problems actually do exactly that and thrive by it! While chess.com's 50-move formula only confuses everyone!

Most puzzles are pulled from games automatically. Part of the puzzle process us evaluation of what the solution is, so no information, such as mate in X is given.

Yes I know! But the 50-move trickery violates that principle as you need to read out the FEN-configuration which tells you you must complete the game in X moves ... or move a pawn or force a capture which is also dependent on the knowledge of X. Still ignoring the neighborhood complaints about your X-es being too flashy!

I understand that it can be confusing and I believe that's the primary reason they're being pulled when reported. However, when presented with any puzzle, it going to be to find the best move or line in the position and if the position is a mating line it will always be mate in the fewest moves possible