Help with this endgame from Dvoretsky's Endgame Manual

Ok so we have this position: lichess.org/study/wBJ2CWCx with black to play

And my question is what are the important nuances and details to know here? I'm still virtually at the beggining of the book and I still have quite some difficulty with pawn endings, but I understand a few things. I know this position can be tranpose in a drawn, but playing with the computer axb sequence with the computer instead of Qc3 (which apparently is better) I saw some interesting things: the black king apparently has to come closer to the white pawn on a3 but to do this he has to go all the way around his own black pawn on b5. It seems like the white king can't go to far away from the a3 pawn because otherwise the black queen will give check until it gets the white a3 pawn. There's also the annoyance which the white rook can cause by getting into b4 and staying there, not letting the white king come to a4. But then the black queen moves until the white king gets trapped on a1 and as the a pawn can't move because the rook on b4 would be taken by the black king, the only move for white is with the rook. The rook then comes to harass the black queen with the support of it's king and the queen gets out making a lot of checks. If I put the white rook between it's king and the black queen when I can, then the black king gets the move to get closer to the a3 pawn. When the black king gets close to a3, then the a3 pawn can be taken by given a check which will force the white king to go away or the rook is placed in between but now can be taken as the king and queen are now acting in the same square.

There's also a sequence where the queen can be traded for the rook (the black pawn takes) and black still promotes the pawn, but I don't remember how to get there. Apparently there's also no possibility for the white rook to stay far from it's king as the black queen would be able to get a double attack sooner or later on the king and the rook.

While the few things I understood from the other sequence starting with 1... Qc3 is that black threats to advance the black pawn after axb and then use the pawn to help the queen to mate the white king. It seems that after 1... Qc3 the best kind of play for white would result in a Queen vs Rook ending which is quite hard for me.

Anyway, that's just what I think. I think those are the central problems and the way to do it, but I could be wrong. If any of you can explain me better what blacks or what white has to do, I would be grateful for your help.

Oh and in case you can't see the position from the link, the position is from the game Yates - Tartakower Bad Homburg1927, where Tartakower apparently draw a winning position.

To be honest, that is pretty too much. Such positions are meant for pros, >2200 FIDE Elo resp. 2400 lichess.

I read the whole manual within 16 months and it was very tough.

So, there are millions of things you can improve before studying this high-end stuff.

Well... I played enough with the computer while observing these patterns to win this position a few times. The hardest thing was understanding the queen vs rook endgame which I managed to understand more or less by seeing youtube videos and some websites.

But what do you think I should be studying? And don't you think I should be reading the Dvoretsky's Endgame Manual at all because it is too complex?

I already read a book called ''Xadrez Básico'' or ''Basic Chess'' where I read a adaptation of Basic Chess Endings by Reuben Fine, though that was in november of last year I had already forgot a lot of stuff, though I can now remember a few things after reading about endings again and some youtube videos. But maybe you think Dvoretsky's Endgame Manual would be too complex for me, I don't know.

The thing is I read some introductory books to chess and while those did help me to have some notion of what chess is all about, it still did not help very much. I think I benefited way more from reading some basic books on chess tactics, even though I still take too much time to see tactics and often make obvious mistakes.

Then I decided I would improve in tactics but as I had read already 3 or 4 basic chess books about tactics, I decided to try the encyclopedia of chess tactics. The problem is the encyclopedia has many problems which require me to see much much further (sometimes 11 moves or more to get one piece). Basically I would need to understand mating threats and other basic tactics much better before getting really into it.

So I decided I would study chess strategy so I would not remain with so many questions about what to do all the time (I already know I should avoid doubled pawns and that kind of thing, but it is not enough). However, soon I realized that I would need to know a lot more about tactics and endgames to understand what he was talking about and I say this as someone who has read some things about passed pawns, backward pawns, doubled pawns, importance of developing the pieces, opening principles, etc.

Therefore I decided I would try to study tactics + endgame before studying strategy.

By the way, the book I tried to study for strategy was Modern Chess Strategy by Ludek Pachman.

The first thing to improve is tactics. All endgame technique and strategy lead you nowhere if you lose a piece to some tactic.
The second thing to improve is endgame technique. All strategy may lead to a weak pawn of the opponent, which you can manoeuvre to win, but then you must have mastered the technique to convert the won endgame to a win.
Some of the stuff in Dvoretsky's endgame manual might be too advanced, but this example is not: it is just an 8 men position. You are quite right that 1...axb3 wins just the same as 1...Qc3, so objectively one is not better than the other.

"I already know I should avoid doubled pawns" No, that is not generally true. Examples:
1 e4 e5 2 Nf3 Nf6 3 Nxe5 d6 4 Nf3 Nxe4 5 Nc3 Nxc3 dxc3
1 e4 e5 2 Nf3 Nc6 3 Bb5 a6 4 Bxc6 dxc6
1 e4 c6 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Nf6 5 Nxf6+ exf6 or ...gxf6
1 e4 e6 2 d4 d5 3 Nc3 Bb4 4 e5 c5 5 a3 Bxc3+ bxc3

tpr

but what are the white threats? I mean, besides the most likely mistakes that you think black could make in this position?

As for this talk about where to start my studies for improving, I decided I would start studying the endgame while studying tactics because I was spending too much times on tactics and still improving quite slowly, while strategy it seems you can't fully understand without understanding the endgame.

I'm still full of questions on each part of the game even after so many tactics and basic chess notions. So much that I feel I can't play until I learn at least tactics and the endgame better. It is too irritating to me not knowing what to do when I play someone.

Also, you (tpr) said this: ''The first thing to improve is tactics'' (...) ''which you can manoeuvre to win''. Those manouvres are tactics? Because it seems to be they can be very closely related to strategy (piece mobility?). For instance, this game was when I had just started playing on lichess: lichess.org/pcvSq3vrLTJJ

I remember this move 29. Rg1 was only because I read somewhere you should always put the rook behind the pawn if you want to promote the pawn. The computer says it is ok. However, after many moves on stockfish, I realized there's a lot of manouvres to do so I can pass the pawn. Is that tactics? Because I don't think any amount of puzzles or tactics from books will hep me with this.

It seems that not only tactics, but also strategy is absolutely required to convert for the won endgame...

This is a very nuanced position if you want optimal play, and optimal play is not entirely necessary for victory (the top 11 moves preserve the forced mate). Playing axb3 immediately is just fine. White's plan would be to play Rxb5+ and axb4 to prevent such a thing, but it wouldn't do them much good. Qc3 just further restricts the king and allows access to that a3 pawn. Here's the top engine line broken down...

...Qc3: Maintaining control of all of the white king's possible escape squares, maintaining pressure on white's rook, gaining an attack on b3 with vision of a3...

Rxb5+: Capturing a pawn, delivering a futile check, preparing to grab another pawn.

Kd4: Moving out of check, maintaining control and asserting dominance over that b1-a7 diagonal the white king so desperately wants to utilize, defending the queen, preventing further checks lest white wants to lose the rook

bxa4: Hopelessly grabbing another pawn

Qxa3: Taking the pawn back, notice how a1, b2, and c1 are still nicely controlled. White is feeling that zugzwang more than ever.

Some Rook Move That Doesn't Lose Immediately Like Rb8: Keeps the b-file off limits for now and prepares futility checking.

Qxa4: Neat. Queen vs rook endgame. I assume you can win it from here.

There are a few things in general you want to consider when looking at a position.

1. What is the mobility of each piece? (Where can they move?)
2. What is the function of each piece? (What do they attack/defend?)
3. What are the threats from each side?

Once again: it is meant for pros!

What about „100 endgames you must know?“ Even my team-mates up to 2000 failed miserably in the pre-test. :D This is the proper work for amateurs like us.

OP seems to have found many of the "nuances." Maybe one that you missed was the fortress draw in KRP V KQ that has to be avoided by a long calculation of checks should white try to go there.

I disagree with much of the above advice. You are on the right track with this. Good luck!

Thanks Dr Hack. I was not remembering the fortress which happened to me the first time i tried. But I don't remember how it can happen. Could you give me the moves?

By the way, I'm getting really frustrated with Dvoretsky. What is to be learned from a position like this? I mean, it is a draw, but is one of the exercises he gives. It is the second position from those first 3 exercises I put on stockfish and it says the position is draw.

lichess.org/hsuBzqG9