Underpromotion in Puzzles?

Is this just convention? I frequently see this site's puzzles will under promote to rooks quite a lot. I'm aware there are times with N underpromotion is valid however, Rook underpromotion seems useless?
If a puzzle under promotes, it because promoting to a queen probably leads to a drawn position, or potentially a loss in some cases.


99% of underpromotions in puzzles are silly because they serve no purpose. Whoever designed them doesn't understand that you should do things for a reason on the board.

I've literally played thousands of games of chess....online, USCF tournaments, at chess clubs, etc. I cannot, try as I might, come up with a single game in which I had to under promote in order to secure a win.

99% of underpromotions in puzzles are silly because they serve no purpose. Whoever designed them doesn't understand that you should do things for a reason on the board.
Can you post any where the under promotion wasn't the best move on the board?

Can you post any where the under promotion wasn't the best move on the board?
https://www.chess.com/puzzles/problem/1066682 I came across just now. Why is the rook better?

Can you post any where the under promotion wasn't the best move on the board?
https://www.chess.com/puzzles/problem/1009132 why is the rook better?

It does make sense actually. The engine sees that if it promotes to queen or anything else, your best move is going to be to take it no matter what it is. If you don't do it then you don't win. So the engine underpromotes so that the difference between the lines where you take and don't take is the smallest. It considers it more difficult for you to find line number 1. But humanly it makes no difference

Can you post any where the under promotion wasn't the best move on the board?
https://www.chess.com/puzzles/problem/1009132 why is the rook better?
Since that is the other side doing the promotion, that wasn't exactly what I was asking for
I'm talking about promotions where the solver underpromotes.
It's going to be captured anyway, so the choice of promotion doesn't matter in this case.

We have to agree to disagree since your arguments are confirming my suspicion. If you play chess with the mindset "there's no point in getting a queen since my oppenent will take it anyway", then be my guest. Personally, I believe the optimal strategy is to promote to a queen and give the opponent the chance of missing the capture. By your logic, tricks cannot be part of your game since your opponent will always see through them.

We have to agree to disagree since your arguments are confirming my suspicion. If you play chess with the mindset "there's no point in getting a queen since my oppenent will take it anyway", then be my guest. Personally, I believe the optimal strategy is to promote to a queen and give the opponent the chance of missing the capture. By your logic, tricks cannot be part of your game since your opponent will always see through them.
It's a puzzle. The side to play has to find the best continuation, regardless of what the opponent plays. So it doesn't matter what move/promotion was made. Sometimes the choice is made to present a specific tactical pattern or motif, or to give a different defensive idea, and sometimes it can be to delay mate the longest.
If the position was found by the automated process, them the solution promotion was likely influenced by the engine evaluation function ultimately.
Is this just convention? I frequently see this site's puzzles will under promote to rooks quite a lot. I'm aware there are times with N underpromotion is valid however, Rook underpromotion seems useless?