Books that teach how to evaluate positions?

Lev Alburt authored "Building Up Your Chess." The book teaches how to evaluate positions and give point values. Sounds good, but dated 2002. Are there any more modern books based on the evaluation of Stockfish, LCO, or AlphaZero?

Positional books tell us which structures are good, but they don't say how many centipawns they are worth. 

I do know that center pawns are worth 30% more than flank pawns until most of the pieces are traded off, when it reverses. And that pawns slightly gain value as they go down the board, but don't gain lots unless they are on the 7th or connected. And double pawns are worth about half a pawn each. Thats the kind of info I want, but also for controlling open files, squares near the enemy king, bad bishop vs good. I know a knight anchored on the 6th rank is worth 5 points, 5.5 or more if near the king. Knights can be wirth 2 points in some open end games, or 4 paired with a queen. 

Eval points are likely not accurate for a single position. They are intended to be used on positions 3-4 full moves down a minimax tree, with the current position matching that of the forceably best position.

Even so, P1 N3 B3 R5 Q9 values are very inaccurate near checkmate too. I see no harm in knowing some point values for files and other sqyares or pins.

LCO and Stockfish have public source code. I just have to find the easiest to understand source of it. 

Since AlphaZero looks at much fewer positions per second than Stockfish, each one is more accurate and more complex.

I wonder if computer chess teams found some position features to be more deterministic than classical features. Stockfish 15 seems a good place to look.

Eugene Perelshteyne and Nate Solon have announced their new book „Evaluate like a grandmaster". They explain the idea behind this book in one of the latest episodes of the "Perpetual Chess Podcast" (https://www.perpetualchesspod.com/new-blog/2022/6/21/ep-284-gm-eugene-perelshteyn-and-fm-nate-solon-on-how-to-improve-your-chess-evaluation-skills-plus-the-lessons-we-can-learn-from-magnus-blitz-games).

Yes.  Check my note on the book "The Six Power Moves of Chess" in the Comments section at the end of this article.....it could be the book you're looking for....(however it doesn't discuss centipawns)....
Good Chess Books for Beginners and Beyond...
https://www.chess.com/blog/RussBell/good-chess-books-for-beginners-and-beyond
https://www.chess.com/blog/RussBell

Unless you are looking to write your own chess engine, the centipawn values are meaningless.  No one looks at a position and thinks "my knight on the d5-outpost is worth 550 centipawns".  That is not how humans evaluate a position.

If you are looking for good positional books, the following are some good recommendations (in no particular order):

"Chess Structures" by Mauricio Flores Rios

"Positional Decision Making" by Boris Gelfand

"Positional Play" by Jacob Aagaard

"Strategic Play" by Jacob Aagaard

"Winning Chess Strategies" by Yasser Sierawan

Note that all of those books were written within the last 10 years (and engine checked), but none of them teach you to evaluate a position by counting centipawns.  It is more the accumulation of static advantages that can either dominate the board or be traded for a dynamic advantage later.  Sticking with the d5-outpost example, you would ask "How can I make that d5-outpost permanent and the most important feature of the game?"  Often, that would involve trading off all Black's pieces that could trade themselves for the knight you want to place on d5 so that it can jump into the hole and remain there as long as you want/need it to.

Good Positional Chess, Planning & Strategy Books for Beginners and Beyond...

https://www.chess.com/blog/RussBell/introduction-to-positional-chess-planning-strategy

Once again - check my note on the book "The Six Power Moves of Chess" in the Comments section at the end of the article.

Sounds like I should buy The 6 Power Moves of Chess.

I don't know precisely what your expectations are, but the book is focused on how to evaluate a chess position with a view toward identifying and executing a plan of action.  IMO the book would be of value to most amateurs below about 1800 Elo.  In any case, I found it interesting, if not unique.  It helped me to rethink and take a more structured approach to evaluating positions.

You can add that knights tend to lose value each time pawns come off the board at the same time rooks and bishops gain value.

This sort of minutia is more interesting trivia than what actually goes into evaluating a position though.

Isn't this sort of minutia exactly what's so important about properly evaluating a position? Knowing that a bishop is more valuable than a knight when pawns come off the board influences, for example, whether you should consider exchanging your knight for a bishop, or exchanging your bishop for a knight.

I've never used such things in a game... I mean... even the idea that the rook is 5 and the knight is 3, I never think like this during a game. It's always something more intuitive and related to piece activity and king safety... ok, obviously if I can take a rook for free or a knight I'll choose the rook, but it's rare to win a piece completely cleanly. Usually there is some compensation that makes it complicated, and trying to add up tenths of a pawn here and there is, IMO, silly.

But ok, sure, it's obviously useful to know 9,5,3,3,1... so I guess it's also useful to be made aware of the fact that in at least some endgames, flank pawns are worth more since they're harder to stop, and that bishops and rooks are best on an open board... these are useful bits of knowledge... but don't add up tenths of a pawn in your head during a game lol

FWIW, here are some useful criteria for a piece:

1) Is it mobile (can move to many squares without being captured) or at least influencing a lot of empty squares?
2) Is it not blocking friendly pieces (for example Bg2 doesn't appreciate the Nf3)?
3) Can it easily be chased away and is it on a defended square?
4) In 2 or 3 moves can it come into contact with a weak pawn or squares near the enemy king?

In my mind this is the basic value of a piece (excluding situational elements i.e. immediate tactics).

If I have 3 pieces that score highly on this, and you have 1 (or none that score highly), then I'll feel I'm ahead regardless of who has more pieces, or whose pieces are more valuable in the 9,5,3,3 sense.

Jeremy Silman's "How to Reassess Your Chess" or "The Amateur's Mind" are worth considering.  Silman teaches you to identify imbalances in the position (for example, knight vs. bishop).  An assessment of the imbalances should help you develop an evaluation of the position, and therefore a plan of action.  An old book worth considering is "Point Count Chess" by Horowitz.  Two caveats: while it refers to a point count system, the positional concepts that it conveys are what is important; second, it is in the older "Descriptive" notation, and if you're not familiar with it, you may be distracted or frustrated.

I think the book "secrets of chess" by Lyudmil Tsvetkov is a book about about stockfish eval but I could be wrong.

It is not worth the paper it is printed on.

Looking at the opening post though, he might just appreciate something like this.

Thank you! That is exactly what I'm looking for. I wish it were for Stockfish 15 instead of year 2016 Stockfish, but still good. I'd also like to see something like this for AlphaZero.

The Stockfish 7 patterns promise to have some new to chess. But since 14 and 15 are so much better and evaluate material imbalances much better, I wish I could get those patterns, or those of AlphaZero. I could still buy the Stockfish 7 patterns. Maybe the strong ones are still accurate enough.

What do you base that on?