WHY are moves inaccuracies/mistakes/blunders?

I apologize if this has been answered elsewhere,  I did search without success.

I have no idea how the analyzer works. I'll watch an analyzed game and it will suggest another move that makes less strategic sense to me, for example, using a bishop as a glorified pawn. There's no explanation I can find. Does one exist, or am I supposed to figure it out somehow? 

As long as we're here, what are the individual move scores based on? Initially, I thought it was how strong a move was, but having now played a bit more and getting low move scores for moves that clearly got me out of trouble or were setting up checkmate, I have no idea what they mean.

You would have to post examples of what you are talking about.

Okay. This isn't a great example, but what gives here? Why is moving a knight to the edge better than bolstering the center? It's literally exactly the opposite of what I've been taught.

Just so you know, my thought process here was to prevent the knight fork and keep my ability to castle.

What knight fork?

I think it is helpful to think of chess as moving along in phases. We all know the beginning, middle, and end game division, but within each exists a series of lines (3-10 moved) that can force a response, and if lucky a response that yields material.

here, the computer is essentially saying you opened yourself up to such a series of threats or moves to lose the equivalent of a pawn. Looking at the board, it recommends Nh5, which attacks f4, allows a fianchetto, and (unlike Be6) allows you to protect d5 with e6 (while Be6 blocks in the pawn). Additionally, if you wanted to castle, fianchettoing the bishop on g7 would protect the knight, target that long a1-h8 diagonal, and allow for castling.

in short, that move you made is unlikely to cost you the game at lower levels, but could cost you the game at upper levels. However, it still set you back, however slightly. 

My advice: try to understand why it is an inaccuracy and why the move it recommended is better. However, don’t beat yourself up over it as I watched a grandmaster game with Carlsen and ??? And Carlsen has 3 in accuracies and his opponent had five and a late game blunder (time was running short). So, they happen at all levels.

Maybe you are referring to Nxf7, but that just loses a piece and the game.

You need to ask yourself:  "What is the knight doing on h5?"

I'm sorry, Sred... Nxf7 loses the game? He loses a knight, I can't castle anymore and my king is off the back row. How does that win the game for me? That, again, goes against what I've been taught which is:

1. Castle in the opening.

2. Keep the king protected and on the back row, especially in the early game.

3. Get your bishops and knights into play early.

By the way, thanks for taking the time, all three of you.

I see Nh5 as a weak move, other than it has the potential to take an unanswered pawn assuming the opponent doesn't just move Pe3, which makes the knight basically stuck and irrelevant.

So what if your king can't castle? It's not like you will get checkmated or anything. He has given up a knight for virtually nothing.

But then Ng3 follows and White is in trouble.

@I_Zero_Gravitas_I

Yes, Nxf7 loses the game. White spends 3 moves just to give away his only developed piece. All White gets in return is a Pawn and Black's right to castle. But the black King will already be on f7, so you need just one tempo to castle "manually". Since White wasted 3 tempos, that's a pretty good deal.

And material is really, really important in chess. I mean, REALLY!

Good start!  Now keep going...

After 1...Nh5 2.e3 What does black play?

Immediately, after e3, I see Ng3 (forking the rook and bishop). White has to move its rook to preserve material, which allows you to play Nf1 and it kills whites ability to castle kingside (if white plays Rh2, rather than Rg1, it kills the ability to castle altogether).

Thus, white will need to develop its entire queenside before being able to castle, and white hasn’t yet shown that it understands how to do that.

ETA: and with you being only Bg7 from playing O-O, you can fairly easily predict that he will eventually play O-O-O. Conventional wisdom says if the two sides castle in opposite directions, you should launch a pawn storm on their side. In this game, you would have at least four moves’ head start on rushing your pawns up the board (white has to move the queen, bishop, knight, and either the d or b pawn for the bishop). I like that position for black.

Chess 4PK:  Ng3 doesn't put white in trouble, he just moves his rook, right?

Sred... I regret I don't know enough jargon to follow that. To my knowledge tempo is the rate at which something occurs in rhythm: music, battle, etc. I don't follow the usage here. What do you mean?

Yes, but weren't you worried about losing your casting right at the expense of a knight? White loses his casting right for nothing. Also, Black will play Nxf1 and get the bishop pair.

Blindthief: He loses the ability to castle kingside and so do I. That ties things, doesn't it? How is that better than preventing the situation altogether?

Again, thanks for helping me work through this. I'm basically a self-taught average player who decided to take a few lessons and got worse, lol.