In each stadium of the game you'll have to try to make a difference. Or in other words: not lose to much ground.
Somewhere Ive read that you got to study endgames first, then other things. This seems like good advice to me for a couple of reasons:
1) many endgames are not extreeeeemely difficult. In fact quite a bunch are easy to tell who's going to win.
2) Once you know those 'easy to tell who is going to win' endgames, you start getting a hint in midgames on how to get to those situations.
Meanwhile you should also do a bit of opening and strategy and alot of tactics. You cant only do endgames. You'll have to know some opening and midgame principles. But knowing how to transpose to a winnable endgame is an important element in chess.
I suppose what I mean by this is, my personal style is not one that utilizes a lot of tactical combinations and middlegame wins. Very many of my wins are resignations from the threat of promotion in the endgame, that sort of thing.
It is my belief that by getting practice in endgames in this way, it will benefit me a lot more in years to come than it will to be doing nothing but seeking complications in the position and always striving for a decisive tactical combination. Not to say that I won't play one if I see that it is there, but you very commonly hear people saying they need more endgame experience, and even Karpov said that it is a common misconception that the average club player needs to be better at openings; they need to be better at endgames instead.
Thoughts?