How to calculate pawn endgames?

how do I? I know the square rule, but complex pawn endgame where only specific, very specific moves win.. those are hard..

Google naroditsky pawn endgames

They can be very very difficult indeed. ;)

I'm fascinated by pawn endgames because they are so deceptively complicated. One of my favorite chess books is Secrets of Pawn Endings by Lamprecht and Muller, which will teach you most of what you need to know. Ultimately you have to practice that knowledge a lot. Chess compositions are good for this and there are a lot of pawn problems. You can also use these studies, which have some instruction, but also have a lot of exercises, including some instructive compositions:

Tactics in Pawn Endgames: lichess.org/study/kSK7Uq54/
King and Pawn versus King: lichess.org/study/EOqdyQeN/
Rook Pawns and Bahr's Rule: lichess.org/study/LN2KAQOq/
Pawn Races: lichess.org/study/OX3hApYw/
King and Pawn versus King and Pawn: lichess.org/study/EOOQ7TMr/
King and Two Pawns versus King and Pawn: lichess.org/study/R0UPGY2l/

I once saw someone bring up this position:
lichess.org/editor/8/8/4k3/p6p/1p2K3/1Rr4P/P7/8_w_-_-_0_1?color=white
The question was whether or not White had a draw after Rxc3. My impression is that this problem is so super-easy that endgame enthusiasts would just laugh at it, but I am not confident that I could think it through over-the-board without a long time control.

There are some concepts you need to know that can help you win.

For example king opposition, zugzwang, total liquidation or this concept people call "minefields and airplanes" that can help you gain control of the base of pawnchains. You can practise those against the AI and train them until they get second nature.

@Geelse_zot said in #6:
> zugzwang

For me, it was a game changer switching from "what I want" to "what the opponent would hate the most".

Pawn only? Immediately evaluate numbers and exchanges between pawns to find your strengths and weaknesses. Then whose king is in best position to block/oppose, assist, or capture?

Best tactic for endgame is getting ready sooner (safely) if it looks like it's going that way.

Threaten piece or check and take a key pawn whether it's passed or will be in way for your pawn later. If SAFE get king moving where he will be needed.

Again does he need to help your pawn or stop theirs? See if you can maneuver their last piece so it can't safely reach your pawn you're going to push. Try to get pawn moving then exchange last pieces.

Obviously easier to type than do. But repeatedly train positions you typically get with your openings with a couple pieces remaining so you are ready in games. You know the best scenarios and the biggest obstacles.

If you are already solid leading to endgames it's a good situation to have. Now just train 50+ hours on those critical positions with a couple pieces remaining. Think of endgame a few moves before you get there time permitting.

good question and good answers up there. I have difficulty with pawn techniques in general, and endgame not just for the sake of endgames seems like a nice place to start to go about it from bottom material complexity to many pawns.

But in endgames, the big story is what to do with king. So yes as above.. This is not about immediate long range reconfigurations. (thinking of king and pawns only).

Multistep thinking, becomes the norm, like the square rule. Also opposition does propagate all over the board when doing multistep retrograde thinking.. It is most often shown as recipe to abide by, but I have had to find out how to construct them to fully understand its purpose. Finding the books that will do that or persons willing to go back to basic rules of chess mobility is not that easy, I think. If using books, do not dive in one only too deep. may be better to have more than one from different audience premises (also not often clear anyway, but 2 heads better than 1, plus yours it might work).

@Geelse_zot said in #6:
> ... this concept people call "minefields and airplanes" that can ...
Did they call it that, a century ago?