On the Origin of Good Moves: A Skeptic's Guide at Getting Better

Well, there must be a fundamental error in the whole concept:

No book, no trainer, no input from outside alone will make you good in a short time or even in the long run. It all has to be developed in YOUR brain out of your acting and thinking.

Hendriks has no special idea how to accelerate learning but he proves why all others fail miserably with their promises.

@Triangel Just my two cents: opening principles are far from useless, but you shouldn't play a move just because "this opening principle says it's good". In a given position in the opening, you will have something like 20 moves to choose from. Opening principles should be used only to narrow down your list of candidate moves. (I guess Hendriks would phrase this differently; I do not intend to endorse Kotov's tree-like way of thinking here.) Then which move you choose to play comes down to your own calculations, feelings, preferences and judgements.

Trivial example: at the beginning of the game, there are exactly 20 options (16 pawn moves and 4 knight moves). Opening principles say that you want to control the centre, develop your pieces and keep the king safe. So we narrow down those 20 moves to the more standard choices, and reject moves like f3 or g4. This sounds trivial, but watch any child beginner play and you'll see them play absolutely random moves on move 1; most likely they'll play a4 or h4 (and with no intention of developing their rook).

So I think it would be too much to say that opening principles are useless, it's just that one's decisions should ideally be backed up by more than just "this rule tells me so".

#22 - what I take from my secondhand understanding[1] isn't that we shouldn't know about opening principles, it's that when there's more than one principle that could apply or more than one way that we could apply the same principle, and no immediate tactical reason to choose one over the other, we shouldn't expect the Big Book of Opening Strategy to give us an ever-more complete flowchart that eventually tells us exactly what the best move is in any given situation. Instead, having seen lots of sound opening moves in the past, whether by studying theory or by playing through master games, should give us a feel for what "looks right" in a given situation. So for instance, you'll have a sense of how urgent it is to castle even if you haven't consciously checked for whether central files might open or what attacks could be brought to bear on f7.

[1] mostly from reading this in-depth review by John Watson
https://theweekinchess.com/john-watson-reviews/john-watson-book-review-103-challenging-conventional-wisdom
which admittedly seems to turn it down a shade from Hendricks's missionary zeal...

Actually, this is all well-known. MFTL is a modern classic. Do you think like a tree? I don't think like a tree.

@RamblinDave Yes, that sounds much nore reasonable.

And to pull it back to what Sarg0n is talking about: the moves that you "narrow it down to" are essentially a random choice of two or three moves. If you really thought about it, you could probably convince yourself that other moves are also playable, but you don't consider those moves. Why not? Well, there is no reason; they just didn't occur to you at the board. You saw two or three moves that you thought were worth considering, considered them, one looked pretty good, and you went with it; it's a question of practicality as well.

Again, maybe not so much "essentially random" as "instinctive based on what we've seen lots of times before in similar positions."

"Educated trial and error" (Hendriks)

The old masters were trying to come up with a "theory of chess" that would be akin to an axiomatic system in mathematics or a theory for physics. Terms and ideas to explain what was happening. They realized there would be exceptions. Yet their writing comes off sounding dogmatic to us.

Then there is the question of how to teach chess. Do you teach using the same theories you use to explain a position? Nimzowitsch in "My System" said that the theory of opposition had to go. He did not like that it had exceptions. He came up with a new theory. Here I'm talking about even in the simple case of KPk. Today most teachers teach this endgame using opposition plus exceptions, and fold in Nimzowitsch by saying that the King has to get in front of the pawn. But this is to help someone learn this endgame. Once they have internalized the endgame, they don't need any theory at all - they just know what to do.

These type of theories are really rules-of-thumb. There are exceptions and I think they should be looked at as ways of helping to internalize the patterns that occur.

For calculation, by which I mean turn-by-turn look-ahead, it is difficult to come up with a theory to help people internalize the patterns. Looking at checks, captures, undefended pieces, etc are rules-of-thumb that may or may not expose the candidate move that leads to victory. And even if you include those, you may stop the analysis too soon because you have no experience to see the followup candidates. Instead, they need to learn the patterns by seeing them played out, and examining, through deliberate practice, what made the patterns work, and what would have made them fail. A recent post in the forum had an example of this with respect to the Greek Sacrifice pattern. Those that knew the pattern knew to include Bxh7+ as a candidate but they knew much more than it was a capture and a check! They knew from experience what the likely followup candidates would need to be considered in calculating. Indeed, with enough experience you might not calculate at all in that case. If you have time and the ability, sure; otherwise you'd "go with your gut".

You're on a good way. :)

I would have considered more or less the same moves without a wall of text. Why? Your brain uses millions of patterns within seconds, much more than any press spokesman of it can tell.