When not to castle as soon as possible?

I have seen a lot of advice for beginners that says you should castle as soon as possible.

It makes some sense because if you leave castling too late and your King or Rook has to move then you lose your castling privilege and may regret not castling sooner. However, this advice still seems a little simplistic to me.

What is a more comprehensive definition for when you should castle/not castle?

Reasons i dont castle, if the queens come off the bord an a couple peice get traded, i prep for an end game ( king towards the center ) If i castle right now, they have mate in 1...probably should castle in that direction. If they start a king side pawn storm...ill castle long. These " rules of chess" are subjective, like dont move a peice in the opening twice, well if that peice is being attacked by a pawn i should probably ignore that rule an move it.

If you know your openings then you know when is and when isn't the correct time to castle.

When u play the bongcloud

Castling as quickly as possible leads to a safe place for the king. So the queen or any other piece does not harass it. Also, it gets the rook out ten times as faster to cover any file to protect any piece. If you dont, well my reasons have already specified l.

I found this series by Gotham chess on preventing your opponent from castling: https://www.chess.com/lessons/dont-let-them-castle 

So I guess that supports the principle of castling early.

When the center is closed, sometimes it's better not to castle.

Castling is nearly always good. O-O is like 3 moves in 1: Kf2, Rf1, Kg1. It brings your king to safety and connects your rooks.

I have been looking through GM games. It is rare for the winning player never to castle however there are some examples of both players castling much later than they could have castled.

Take a look at this game played by Carlsen. He castles on move 17:

Then from the same tournament:

And again in this game, not being castled is no problem for Carlsen:

Perhaps the principle should be revised to:

"Castle as early as possible, unless you are Magnus Carlsen" ;-)

You probably shouldn't castle when there is a high chance of checkmate in that position.

Just watched this video covering the game Siegbert Tarrasch vs Georg Marco

https://www.chess.com/lessons/candidate-moves/candidate-moves-the-opponents-ideas

According to IM Kostya Kavutskiy, the only mistake the losing player made was castling as early as possible:

The analysis engine classes Black's castling as a book move, but the evaluation changes from +0.33 to +0.96. White is effectively up a pawn already after move 7 and a series of forcing moves created the equivalent advantage of being up a rook by the time Marco resigned.

Kostya says it is one of the most instructive games of all time.

I usually castle when possible, but i suck at this game so i don't know if it's good.

When the centre is closed you don't need to rush caslting

dont castle if: 

1) you literally can't

2) you're castling into an attack

3) you drop too much material if you castle (but that probably means you screwed up in the opening anyways)

You should castle early though.

Unless they trade all the pieces off right in the beginning.

‘Good players seldom castle’

Casting has two purpose, the one is to bring the Rook into the center and the other is to let the King go aside away from attack. But if the center is close then the main plan is attack on side; so it's ok to not casting here because the Rook on side will play a rule on attack or defence, and the King will be safe in the center :)

Because they're super GMs.

But they still castle early often.

For a beginner, it's sound advice to advance a few pawns, get your knights and bishops out, castle, connect your rooks, and don't lose momentum moving one piece two or three times until you've done the above. Early exchanges may result in a weak pawn structure, or you may be put off castling if castling to the opposite side to your opponent, where your going to castle "in to trouble", and certain openings where I want my king in the centre would stop me castling.

yes 90% of the time castling is a good idea, although my last two games on Lchess i faced the Tal variation and i decided castling would not be good, king was safer on the 7th rank in the center. found it odd to have back to back Tal games with basically the same first 15 moves lol