How to improve my general endgame?

Endgames are specific and therefore changing the position of anything on the board can greatly affect the decision needed to make. I am terrible at this, both because I make lots of mistakes with uncertain, calm moves and because I mostly have never had to play endgames before.
All major endgame training books or webcourses give examples, where white/black just happens to have everything allowing the specific plan he plays to win.
How can one apprehend how to play the endgame using concrete, seemingly coincidence induced examples for general rules?

In my experience, only simple model endgames have a clear recipe telling you what to do. Then there are some general guidelines (like if your opponent has one bishop, keep everything on squares of opposite color). But in more complex endgames it's up to you to come up with a plan. (It's important to know the basic ones as well, though, as it will tell you when you want to simplify into them.)

One piece of advice I find very useful is to adopt the "schematic thinking", i.e. rather than thinking about moves or variations, thinking in bigger steps (schemes). In a book I read an interesting story illustrating this approach. In a tournament, group of players was discussing an endgame position which seemed winning but nobody could see how. Capablanca looked at the position for few minutes, then moved some pieces to different squares and said: "This position is an easy win. You just have to find out how to play the pieces to these squares." And suddenly finding the right moves and variations was much easier.

Two things...no wait three things

1. Passed pawns become a major factor in endgames, creating one etc
2. King activity is exceptionally important, bring your king to the centre as soon as you can and/or try to restrict your opponents kings movements
3. Analyze. Its not the like middlegame where your intuition can guide you with aggressive moves.

I say these things after studying Rubinstein somewhat when I first played......his games are a fantastic lesson in endgame play.
Another amazing endgame player is Ulf Andersson, a complete virtuoso!

"... Silman’s Complete Endgame Course ... I'm convinced that Silman's book will take its place in history as one of the most popular endgame books ever. ... He writes in a clear and casual style, and time and again has shown the ability to reach those who feel intimidated by the lofty approach that a grandmaster will often take. ..." - IM John Watson (2007)
https://theweekinchess.com/john-watson-reviews/theres-an-end-to-it-all
https://www.amazon.com/Silmans-Complete-Endgame-Course-Beginner/dp/1890085103?asin=B00H273OJS&revisi5a3244f2&format=2&depth=1
https://lichess.org/forum/general-chess-discussion/recommendation-for-an-endgame-book-1900-fide
@teachmewell said in #1:

... How can one apprehend how to play the endgame using concrete, seemingly coincidence induced examples for general rules?
Maybe, post a specific example where you had trouble deciding what to do.

"How can one apprehend how to play the endgame using concrete, seemingly coincidence induced examples for general rules?"

Make a lot of mistakes and learn from them:
(Mis)play and endgame then study similar ones, including typical tactics,so you understand it a little better. Then repeat.

(The first step is to know some basic theoretical endgames but I'd bet you're solid on them already.)

Bill

In the first game at https://lichess.org/forum/general-chess-discussion/how-to-attack, after 1 d4 Nf6 2 c4 e5 3 dxe5 Ng4 4 Nf3 Nc6 5 Bg5 Be7 6 Bxe7 Qxe7 7 Nc3 Ngxe5 8 Nxe5 Qxe5 9 e3 b6 10 Be2 Bb7 11 Bf3 f5 12 O-O O-O-O 13 Qa4 g5 14 Rfd1 g4 15 Bxc6 dxc6 16 Qxa7 f4 17 exf4 Qxf4 18 c5 Rxd1+ 19 Rxd1 Rf8 20 Rf1 Rf5 21 Qa4 Qg5 22 cxb6 cxb6 23 Qe4 Qg6 24 Re1 Rf6 25 Qe8+ Qxe8 26 Rxe8+ Kd7 27 Rh8 Rf7 28 Kf1 Re7 29 f3 gxf3 30 gxf3 Rf7 31 Ke2 c5 32 Ne4 Bxe4 33 fxe4 Kc6 34 Ke3 Ra7 35 a3, teachmewell passed up an opportunity to go for 35...Rg7 36 Rd8 Rg2 37 Rd2 Rg4 38 Kf3 Rh4 39 e5 Kc7. I suppose that a rook endgame book might give teachmewell opportunities to practice thinking through that sort of thing.

I have rook endgames in my books, they speak of an active vs passive rook, they speak of how to defend two pawns with a rook (which is a lot of "occupying the lines", not overly concrete). But nothing about if the opponent rook has more pawns to grab.

@swimmerBill said in #5:

them
How does one 'learn' from the games? Replay until thoroughly analysed?

@kindaspongey said in #4:

maybe post a specific example where you had trouble deciding what to do
I am already out of OTB matches from this year, but my problem is still the intuition. I rarely get to the endgames, because of my weird tendency for reckless play. Here is one from online play, chess. com

BTW, I just checked the castling ratio on lichess, it sais

Kingside castling 54.5% 10% 35.5% 110
Queenside castling 53.8% 0% 46.2% 13
No castling 67.1% 2.4% 30.6% 85

first number are wins, second are draws, third are losses, forth are number of games.

Is 85/218 games uncastled a bad score?

@teachmewell said in #7:

I have rook endgames in my books, they speak of an active vs passive rook, they speak of how to defend two pawns with a rook (which is a lot of "occupying the lines", not overly concrete). But nothing about if the opponent rook has more pawns to grab.

How does one 'learn' from the games? Replay until thoroughly analysed?

I am already out of OTB matches from this year, but my problem is still the intuition. I rarely get to the endgames, because of my weird tendency for reckless play. Here is one from online play, chess. com

You lost that through first blundering second misplaying a K&P endgame.

MGs can be played a lot based on general ideas but endgames you need to calculate a lot. K&P endgames have 3-4 ideas and the rest is pure calculation. So what I see is you just need to calculate a lot more in the EG.

I the MG you calculate in critical positions. In the EG almost every position is critical.

And of course analyze the endgame -play it against the computer ; switch sides and repeat.

I also look up annotated GM endgames with similar pieces and play thru & analyze those. For your game I'd look up ''R vs N'' for example and ''K and many Ps vs K and many Ps''

I'd suggest you do loads of endgame themed puzzles.

@kindaspongey said in #6:

In the first game at https://lichess.org/forum/general-chess-discussion/how-to-attack, after ... 35 a3, teachmewell passed up an opportunity ... I suppose that a rook endgame book might give teachmewell opportunities to practice thinking through that sort of thing.
@teachmewell said in #7:
I have rook endgames in my books, they speak of an active vs passive rook, they speak of how to defend two pawns with a rook (which is a lot of "occupying the lines", not overly concrete). But nothing about if the opponent rook has more pawns to grab. ...
After 35 a3, you might have done better if you had previously studied some specific examples of the sort that you could find in some rook endgame books.

@teachmewell said in #1:

... How can one apprehend how to play the endgame using concrete, seemingly coincidence induced examples for general rules?
@swimmerBill said in #5:
... Make a lot of mistakes and learn from them:
(Mis)play and endgame then study similar ones, including typical tactics,so you understand it a little better. Then repeat. ...
@teachmewell said in #7:
... How does one 'learn' from the games? Replay until thoroughly analysed? ...
It seemed to me that it was worthwhile to give particular attention to your choice of 35...Rf7 because the machine indicated that it greatly reduced the probability of a draw. I looked at what the machine suggested as a better alternative (35....Rg7 36 Rd8 Rg2 etc.).

@teachmewell said in #7:

... BTW, I just checked the castling ratio on lichess, it sais ...
My guess is that this comment was in connection with the discussion at https://lichess.org/forum/general-chess-discussion/what-to-do-against-a-center-avoiding-opening . I have addressed it there.