Originally posted by arrakis
I agree with Tengu.
Pick 1 opening to play for White and 2 openings to play for Black:
One against e4 and one against d4.
Once you get these basics then move on to the middlegame and finally the endgame.
I disagree. I know I'm parodying your position, but I don't think it's any good learning to play an opening well if you don't know what to do with the position in the middle game, and it's no good being a middle game wizard if you can't cope in the ending. Moreover, understanding the endgame, and what makes it different to the opening and middle game, helps you understand the middle game. As a trivial example, in the endgame you usually need to bring your king forward as an attacking piece. The reason you can do this is that it is unlikely your opponent can engineer a checkmate due to the reduced material; whereas if you put your king slap bang in the centre of the board in the middle game or opening your opponent will probably like you for it...
Where I do agree is that rote learning of endgame patterns is pointless, what is useful to read about is general technique; in the same way that rote learning of opening variations generally sends you on a hiding to nothing. A good book from this point of view is called "Rate your Endgame" by Edmar Mednis and Colin Crouch, as it talks about general endgame principles rather than just drilling you with stuff like the Lucena position.
For middle games the first book I read was called Simple Chess by Michael Stean, it covers
strategic themes like outposts, open files, half open files, space and so on, and I found it very useful. It costs £4.21 on Amazon, so it's not an expensive mistake even if you don't find it helpful...