is castling over-rated

I don't see how castling can be over-rated when it (usually) places the King in a safer position, helps develop the castled Rook and helps to get the rooks connected. However I would think twice about castling if the Queens have already been exchanged.

If once a man delays castling and his king remains in the center, files will open up against him, bishops sweep the board, rooks will dominate the seventh rank, and pawns turn into queens.   Irving Chernev

As long as my opponent has not yet castled, on each move I seek a pretext for an offensive. Even when I realize that the king is not in danger.   Mikhail Tal

Castle if you will or if you must, but not because you can  H N Pillsbury

So, how would you respond to Grandmasters who castle in the large majority of their games? Would you advise them to play differently?

This has already been mentioned by pfren & Scottrf but  9. Qxg8#  in that first game does jump out at you & would have saved you going on for another 26(?) moves!  It might be an idea to spend some time in Tactics Trainer - it's intended to help you spot opportunities like that.

When it comes to castling there's no rule of thumb that I've found - you do it as & when appropriate. Sometimes I don't castle at all. A couple of times I've castled kingside only to find that I want to attack down an open h-file as the game develops but my rook is on the wrong side of my king!

I think one of the most useful aspects of castling is when you need to bring a rook to attack up the middle of the board in a hurry but only do it when it feels right.

I've always liked Pillsbury's quote. Another way of saying it is "Castle when you don't have a better move." Of course, that's true of every move though.

i think i just played my "nehzmedinov"

http://www.chess.com/livechess/game?id=410604364

Why, did he trade his rooks for bishops as well?  But a very nice mating net at the end of that game!

I've just been reading up on Nezhmetdinov & I think I understand your allusion now!  Here's his 'trophy' game - Polugaevsky-Nezhmetdinov, 28th RSFSR Championship, Sochi 1958 ....

That's artistry!  Sadly, he never achieved GM status.

Yes, sir! Lol.

All very good points, but at this point on the learning curve, the OP needs to address the basics, viz.:

1. Don't drop pieces for no reason
2. Don't overlook a mate in one by either side.

Arguing whether castling is overrated - not just "does it fit my game plan" but "should all chess players look at it differently" - can wait for a long time while you sort out the basics.

Absolute beginners don't castle because they are not sure how. Grandmasters absolutely castle--well over 99 percent of the time, I'm sure. In between, it follows the curve: Novices either castle or not, depending on whether they think of it; strong club players almost always do. I admit that sometimes you can foolishy castle into a bad situation, but the fact that the professionals always do it ought to tell you that learning to castle effectively is an absolute part of getting past the novice stage. Some day someone may prove the longstanding grandmaster tradition is wrong, but that player will not be any of us.

I dont think it is overrated, simply just players in low level dont know when is THE BEST move to castling. I would say that, you should castle when there is no positional treat in the position from your opponent. (So for example you should not castle, when you can get space disadvantage after that, and you could play against it.)
Castling is only not neccesary, if there are no queens on the board, and there are a few pieces on the board. So for example if the game suddenly becomes an endgame, (without queen's, and there are only a few pieces left) then you should not castle, because your rooks usually can be connected without that, and easier to bring the king into the game from the middle, than from it's castled position.

"Castle because you want to or because you must, but not because you can" - Pillsbury

I have found myself in games where I don't feel as though I need to castle. If you can get your king safely into the game, then no. Otherwise its hard to activate your rook is tight positions. It depends very much on positions...etc..but pawn structure too. Excellent pawn structure in my experience can allow for safe activation of the king, skipping castling.

Than again...I am a beginner and what do I know :)

Check this game:

http://www.chessgames.com/perl/chessgame?gid=1134506

Korchnoi says: "Of course, it would have been more cautious and solid to remove the king from the centre, but I had a quite different idea, perhaps incorrect, but very tempting: the rook at h1 should take part in a direct attack on the king!"

Here's another game that really is representative of Nezhmetdinov. When he didn't have an attack, he often went berserk rather than prepare the attack better.

Notice that Korchnoi didn't castle here either!

1. Maybe Kingside castling is risky when your King Knight isn't on f6. I don't know enough about the French-Advance to know if it is common or not. 

2. Black was prevented from castling with a check although it was late in the game.

3. In the final position, black has, in effect, castled, at least position wise. 

4. Seemed like a bit of a foolish attack to me, but then I'm not Nezhmetdinov.

Thnx

how about "is it worth it to prevent the opponent from castling at any cost"??

for example it is possible to send a bishop or a knight to f7 or to f2 early in the game - forcing the king to take - so curtains to castle - you will lose a piece but the opponent king will be standing naked and the h rook will probably not get into the game till 30 moves. worth it??

Short answer: Of course it's not worth it to sac a piece just to deprive an opponent of castling.

But the longer answer is that it's often very desirable to sac a piece to keep the king in the center if you already have a decisive lead in development. There are many examples of such sacrifices.

But 1.e4 e5 2.Bc4 Nf6 3.Bxf7+?? isn't one of them.

Yes, castling is overrated, but still nevertheless important.  In one of my recent games I even moved the queen bishop before castling because there were lines where if it was attacked then queenside castling would be ideal because of the open g-file (this didn't actually happen).

  Principle of flexibility sometimes contradicts advice saying to castle early, and most times we should, but if we traded down into an endgame then castling is a mistake as your king is now farther from the center.  This actually happened to me years ago, and right after I castled facepalmed myself (moving too fast was a problem for me then) for stupid hand weaving (though I didn't know that particular term back then).   

Heisman also calls blindly playing a move off a general principle "hand weaving", especially when there are concrete tactical variations to work out.  When there are that's when we should make a positional assessment of all the lines we analyse and determine best capture that way.