Best UnderPromotion Endgame Compilation

Some of these positions have probably already been posted, but I figured we should have 1 thread with the most interesting underpromotion problems. Here are my favorites that I've seen: (Will continuously update the OP with the best examples in this thread). I didnt create any of these. They are from YouTube, this site, and from my chess books.

If white promotes to either a queen or rook, it's instant stalemate. If white promotes to a knight, the bishop just moves to g5 and the next move the white knight get captures and its white who's stalemate! This is one of the few positions where promoting to a bishop is the only way to win, not just a faster way to win. (Got this one from an underpromotion to bishop thread).

Simple underpromotion to rook endgame.

An even better bishop example.

One of the nicest rook underpromotion endgames, so simple yet the rook is so necessary (even a bishop stalemates).

White can promote both pawns to a queen, but either results in instant mate in 1 by the rook! Kf8 leads to the same mate, and kd8 allows the black king to win both pawns! F8 = N+ is the only winning move to deflect the black king from the mating net and then promote the g pawn to a queen winning! Very realistic endgame! It's also important to point out that unlike most underpromotion studies, where the underpromotion turns a loss into a draw, or an otherwise drawn game into a win, this is the only one I know of where the Underpromotion turns a Loss into a win. Any other piece promotion loses, and the only promotion that doesn't lose, wins.

Another simple underpromotion to bishop endgame where promotion to any other piece results in a draw.

Here's a couple positions where multiple underpromotions are necessary:

There are many (thousands of) endgame studies with great underpromotions like the Saavedra endgame (third diagram from the top). For instance, I completed one myself a few weeks ago - which obviously ranks among the greatest. A famous one you may have seen is a stalemate by 3 knights after a knight promotion. The fun is just as much in the (post-)promotion play as in the play leading op to it. Which is lacking in these examples.

A few years ago someone mined the endgame tablebase for all successful underpromotions in positions with at most 5 units (if my memory does not fail me). He found not hundreds but still quite a lot for such thin positions. He posted them on chess.com but I don't know how to find them!

Here's another cool one I came across..

Here's another couple of cool bishop promotion endgames:

Good puzzles but the last one is iffy. By the rules of chess you can underpromote any pawn but what's the point when it's equal to or worse than the standard queen promotion? The Bishop promotion does not work and the moves given are wrong!

You promote to a queen, black sacs his queen on g2 stalemate! You're right the bishop promotion black has other moves, there were a couple other pawns in the actual position I forgot exactly but the logic was the same, forced mate in a few moves after bishop promotion.

Another beautiful bishop promotion endgame!

All puzzles are about the exact position. not about the general idea or logic. The last time I lost to Carlsen I had a great plan for winning but somehow the pawns were in the wrong place

Yes that's a good underpromotion! But you do not know how to correctly enter that solution in the puzzle. You started with the queen promotion but that is the wrong move. You must always start with the correct move which is the bishop promotion. The queen promotion is a "try" which is a variation with a question mark: (1. d8=Q ? ...). Here is the complete solution without the puzzle box ticked:

Note: To enter the queen promotion try after the bishop promotion move you need to navigate back with the arrows to before the bishop promotion move. The try will then appear after the bishop move in the analysis column!

That is true for all variations and tries at any stage. First you enter the solution move and later you enter the alternatives before the solution move. They will however appear after it!

Yeah a couple of these I didn't intend to post as a puzzle but clicked that theme out of habit. I'll just post the variations from now on instead of puzzle format.

Due to limitations in the puzzle interface some puzzles can be presented in puzzle mode and others not so well. In all cases the format in the analysis column should be the same. Otherwise people can't distinguish the main line from the variations and the tries!

My last diagram would also work in puzzle mode but there may be a point where the winning moves are no longer unique. The remaining moves should then be removed as you can't expect a solver to "guess" which winning line you chose!

I'll post in normal mode from now on don't worry, but if anyone wants to try and solve them they just have to not look at the bottom.

Noice

Another beautiful bishop underpromotion:

Technically a tablebase says that if black promotes to a knight he can still force a mate in like 40 moves but that is obviously unnecessary with this fast bishop win!

Yes that's OK in a chess game - but not in endgame study. The tablebase is an equal partner in deciding endgame outcomes. Also, the 50M-rule does not count in endgame studies, so it's OK to travel at ease!

Don't like the normal promotion variation in the last example of your initial post.

Still slower of course.

Actually this occurred in a real game, black just missed the bishop promotion. But I agree.