To improve, you need to become an expert, not about chess, but about your chess.  You need to know what there is to be good at, and what you are good at, and what you are not good at yet.

Practice helps.  Books, databases, analysis software and especially going over your own games also help.  I always type my games into a computer and I always wince to see what tactics I've missed.  I also enjoy some of the computer's suggestions about alternative moves: sometimes they're real crackers, even if there's nothing in it by way of winning a piece or pawn.

But it does help to show your games to other people.  If they are better players than you, so much the better, but even if you are of similar standard, they will be good at different things than you, and are still likely to be able to make some useful comments.

Also: Donald Rumsfeld, you may recall, got into a lot of undeserved trouble for his apparently confused remarks about "unknown unknowns".  [It was about the only thing he has ever said that I thought made any sense.]  I was familiar with the idea from the 'JoHari window', a way of picturing self-knowledge devised by a couple of folks called Jo(e?) and Harry.

	What other people know about me	What other people don't know about me
What I know about myself	A	B
What I don't know about myself	C	D

So, there are things I know about my chess that others do not (B), and there are also things I don't know about my chess that others do (C).

But there are also things hidden from us all -- perhaps because we have never heard of them (D).  So, everyone knows I used to play the Modern (A), I know I used to get a rotten score with it but others don't (B), others may notice that I don't handle Rook Endings very well (but I haven't noticed) (C), and there may be something glaringly obvious to a GM (like, I always choose the wrong plan in the English Opening) which neither I nor anyone else at the club suspects (D), the unknown unknowns. 

So, the benefit of showing your games to others, in this scheme, is holding up a mirror to sides of you that you cannot see (C).  And the first step in curing a fault is identifying it.