I think puzzle #UUtff should be removed.

From what I can tell, the puzzle has 2 winning moves. Moreover, the not-as-winning move is only unclear because ten moves down a line where white plays with computer like precision their pieces are well coordinated - which makes it a cringe puzzle at best, imo.
I might be coping because it ruined my run in storm, though. So I'm good with whatever your opinions on this are. Thank you for your time.
Puzzle link: lichess.org/training/UUtff

It seems perfectly simple to me , you win the queen you're up pawns xxx

Why should you be content with a knight when you can win the queen?
Even at depth 30 SF still gives +5.4 for the main move and only +1.4 for the fork.
Nothing wrong with the puzzle.

It's true that a puzzle with more than one "winning move" should not be in the system, and I think this case is borderline.

After 25...e5? 26.Qf5 Bxd4 27.Rb1 Qxa3 28.Rxb7 White does have some play and Black must be careful to consolidate and win with the two extra pawns.

How is this borderline? After the main line Black has 3 pawns more, plus a Q against 2 knights. That ́s way better than the 2nd line, where he only ends up with a 2 pawn advantage.

@Wasted_Youth said in #5:
> How is this borderline? After the main line Black has 3 pawns more, plus a Q against 2 knights. That ́s way better than the 2nd line, where he only ends up with a 2 pawn advantage.

Yes, but one requirement of the puzzles is that only one move wins at the start and at each stage in the follow-up.

The question is whether 25...e5 "wins" or not. The answer to that question, I suggest, is in the grey area.

There are often several moves which will hold the advantage; required however is always "the best move for side x", and there ́s absolutely no doubt that in this case that ́s 26...Bd6.

There was a thread a few days ago showing an example of a puzzle which really did have two equally good winning moves, both of them leading to mate in 39 moves:

lichess.org/forum/lichess-feedback/puzzle-hzyvj-is-wrong

It ́s interesting to put the second option there into the engine and watch the evaluation creep up with ever more depth of search. That doesn ́t happen here; even at depth 30 25...e5 remains greatly inferior to Bd6.

@Wasted_Youth said in #7:
> There are often several moves which will hold the advantage; required however is always "the best move for side x", and there ́s absolutely no doubt that in this case that ́s 26...Bd6.

Yes, the best move is required. Your phrase "hold the advantage" is key though. Holding the advantage is one thing, but clearly winning is quite another. And the latter, as I understand it, is what distinguishes what is allowed in puzzles and what is not.

>
> There was a thread a few days ago showing an example of a puzzle which really did have two equally good winning moves, both of them leading to mate in 39 moves:
>
> lichess.org/forum/lichess-feedback/puzzle-hzyvj-is-wrong
>
Yes, I remember that thread too. And the puzzle was removed. The problem here was that the search depth of the puzzle generator had been insufficient. If the puzzle generator had detected that the alternative move also won, it would not have selected the position for the puzzle base.

> It ́s interesting to put the second option there into the engine and watch the evaluation creep up with ever more depth of search. That doesn ́t happen here; even at depth 30 25...e5 remains greatly inferior to Bd6.
Yes. "Greatly inferior" is obvious. "No longer winning" is the question.