There are puzzles that end one move short of the answer, because the engine thinks there are "two" winning lines. But the second line is basically the same as the first one, only with a move repetition inserted.
lichess.org/training/GXa3c
In the above puzzle for example. The winning move is Bb1, but the engine sees two winning lines: Bb1, and Qg6+ Kh8 Qh6+ Kg8 Bb1. The winning idea is the exact same, but since the exact lines differ, the engine sees it as the end of the puzzle before Bb1!
I think the puzzle should go one move deeper and ask you to play Bb1. Otherwise you can win the puzzle without having found the winning idea. In a game the repetition could be useful with the clocktime maybe? But in a puzzle I think the repetition can be excluded and the winning line shoukd just be Bb1 straight away.
lichess.org/training/GXa3c
In the above puzzle for example. The winning move is Bb1, but the engine sees two winning lines: Bb1, and Qg6+ Kh8 Qh6+ Kg8 Bb1. The winning idea is the exact same, but since the exact lines differ, the engine sees it as the end of the puzzle before Bb1!
I think the puzzle should go one move deeper and ask you to play Bb1. Otherwise you can win the puzzle without having found the winning idea. In a game the repetition could be useful with the clocktime maybe? But in a puzzle I think the repetition can be excluded and the winning line shoukd just be Bb1 straight away.
