Are BOOK MOVES considered better than BEST MOVES?

I'm talking about what the computer (level 10) defines as BOOK MOVE or BEST MOVE. Let's take a look at an example (Black to move):

The computer will probably recommend Be7. If you you play Be7, the response will be Be7: Excellent move (The best move was Be7). However, if you play b6, then the response will be b6: Excellent move (The best move was b6). You could risk that the computer will say something else (like Good move) because it doesn't say the same thing every time. Anyway, here is the deal: If you play Nc6, the computer says "Book move". The computer lists the book move as the 7th best move (6 other moves better than the book move).

In general, computer's best moves are supposed to be better than book moves. However, the depth=16 is not high enough.

I tried the same position, letting the computer work for 30 minutes, and the depth reached 30.

The result was:

-0.53 6...c5
-0.47 6...b6
-0.39 6...Bd6
-0.30 6...Be7
-0.27 6...Nc6

Stockfish.js 9 | depth=30

We now have 4 moves better than the book move (c5 is now the best).

So the million dollar question will be: At which depth should we normally start trusting the computer more than the book moves?

d=30 is good enough in most of cases.

Why do you care how the computer classifies moves?

The computer has clear definitions of the phrases it uses, such as "best move" and "book move".

Because the computer is not always right about which move is the best. Book moves are not calculated by the computer, they are taken from a database.

This is not about understanding the position, it's about which move is the best.

The question is not how to master openings in order to improve your chess games. It's about "science". The computer has its way (best move) and humans had their ways before the computers (book move). The question is which method to trust the most. Naturally, if you wanna memorize some opening moves you play often, then you can also use this to choose which moves you wanna memorize (in order to improve your openings).

Sometimes.

Why waste time running some crappy version of stockfish in a browser? (unless you don't have a computer).

But still, sometimes an actual book move is better (for practical reasons) than an actual best move.

When I'm working on my repertoire and the engine disagrees with what top level players often play, then it's good to go with the players. These days top players have engines too, and they run them longer than us and work harder on their openings than we do.

But if it's an old line with not many recent games, and the engine hates the book moves, then I go with the engine.

One thing to be careful of is sometimes humans play certain lines to get a draw, and sometimes the move the engine likes best leads to more drawish positions, so sometimes you can't trust either of them and it's best to go with what humans (or engines) consider 2nd or 3rd best.

The best way to do this is play over a lot of GM games from that line to see the common middlegame ideas... but this is getting a little advanced (and perhaps expensive e.g. chessbase).

No, not automatically.

Here's an easy example of where the engine is correct.

The Gurgenidze caro was popular in the 70s, and in many variations it's a fine choice.

For example after 1.e4 c6 2.f4 your next two moves as black can be g6 and d5 and white wont have anything special. I've played this a few times.

But if you want to play it all the time, as some GMs in the past were trying, you can get into trouble.

As a counter example, just about any Benko.

So like I said, it's not automatic. You have to try the human idea, and see if the engine eventually understands. But if it stubbornly hates the position no matter what you try (and you'll have to walk it down its own main line(s) to be sure), then it's just a bad position and it shouldn't be part of your main repertoire.

I say main repertoire because you can have one time use surprise openings.

For example

I don't think that matters. Sometimes this, sometimes that. If you don't understand the computer's move, you can simply let the computer play a few moves ahead, and you will understand the concept. To find out if an opening move is acceptable, analyzing future moves is always necessary. This is not done while you play, it's done by humans (book moves) or by the computer (best move). You then make the moves that you have memorized from that. Which moves would you prefer to memorize, book moves or best moves, that's the question here, you should anyway be able to understand the moves by looking at the consequences of the moves.

Interesting discussion. Personally I don't take what a mindless entity 'says' seriously enough to dig into all this. AI hypemasters want you to believe computers are communicating with you. They want everyone to be their chumps. This hype has worked wonders in a delusive way, contributing to a world gone wrong.

Computers simulate. Their simulations are mindless. But the interests behind them want a piece of your brain, preferably a duped piece. Back in '82 when my new best college friend, a computer science major, showed me what "bells and whistles" are, I was instantly discouraged, sensing their potential for misuse and mind control. At that time there was no imagining the cell phone generations, electronic addiction.

With the possibility of biological interfaces, humanity will bifurcate. If the tattoo phenomenon is any clue, many people will take the Borg route while some purists won't be able to control weapons and switches with a mere thought, having maintained bodily integrity. I am glad I won't be here for that, for the new Christ who will preach against so-called enhancement.

Our minds aren't spiritual, every stupid idea and wicked potential comes from them, but very few think not to identify with this inferior entity (my view is that we have minds, as we have bodies, an observation enabling one to disidentify with any stupid or wicked thought that may crop up in one's mind). Best!

No, the book move is Nc6.

No, you've got it all backwards. You're not supposed to memorize more moves.

There shouldn't be any moves you don't understand. If you don't understand the best move, or the book move, then you can do some work to understand it. It's just that.

Looking at it briefly it seems 6...Be7 and 6...c5 transpose to each other (when white isn't playing Be3), and 6...Nc6 is just an inaccuracy.

My own notes on the Blackmar-Diemer give as basic black moves either Bg4 or c5, and since in this variation Bg4 is not possible it seems c5 is reasonable.

6...c5 is the move my engine prefers at high depth, but it also likes 6...Be7. The most popular moves in my database are 6...c5 and 6...Be7 both of which have scored better than 6...Nc6.

The highest rated players in my database have chosen 6...c5

Looking at some moves after 6...Nc6 I don't like the positions.

What chess.com's java junk considers "book" is fairly meaningless (although 6...Nc6 is not a terrible move).

You could reach depth 30 within  1-3 mins, by installing latest SF in arena, with single PV.( technically you are wasting 27 mins of time and electricity for nothing).

The thread title reminds me of an anecdote related by Bruce Hayden in his splendid Cabbage Heads and Chess Kings:

Wellmeaning kibitzer: Sir, that is not the book move. The book move is -

Alexander Alekhine: I am The Book.