Opening
Your goal in the opening is to reach a playable middlegame. In the absence of tactics, your moves in the opening pertain to achieving rough equality in center control, development, and king safety.
Middlegame
The middlegame is sort of bounded by 1) when rook activity matters, and 2) when king activity matters. In the absence of tactics, your task in the middlegame is to form a short plan based on your assessment of the imbalances in material, mobility, and safety. Your moves pertain to that plan.
Endgame
In the absence of tactics, your job in the endgame is to promote a pawn.
This is a very broad sketch, but a good, experienced, and honest master would probably tell you ‘yeah, that’s pretty much it’.
Here’s why it is drawn with such little detail:
To be specific about opening moves is useless, but it’s lucrative as hell. An openings book is just a map toward the middlegame, but you don’t need a map. It’s a big place, and all roads lead there.
As long as you’re alert to the tactics, and mindful of center control, development, and king safety, you’ll reach a middlegame in decent shape. The key is tactical alertness: Chessplayers think learning openings will help them play the opening, but you don’t lose in the opening because you’re bad at opening play, you lose in the opening because you’re bad at tactics.
Some players depend on generating tactics in the opening, and they’re very difficult opponents if you can’t keep up with them tactically. In other words, you must be a master of tactics, even if you don’t intend to be a disgusting cheapo slinger.
To be specific about middlegame moves is impossible, because the middlegame is nebulous in when it begins, when it ends, and which units are included. IM Silman built his career on teaching people to do this: Assess the imbalances in material, pawn structure, piece mobility, and king safety; and after you’ve determined which of the imbalances must be addressed, your next few moves answer “what are you going to do about it?”.
All of that is trumped by tactics, though. Those players who strive to create tactical chaos in the opening are usually the same kind of jerks in the middlegame. Worse, the players who play quietly in the opening reveal themselves to be those jerks in disguise. In other words, you must be a master of tactics in the opening and in the middlegame. Even if you fancy yourself a superduper defensive beast, you have to know the offensive tactics so you can see them coming.
If neither player generates a decisive attack on the king, then the material on the board inevitably reduces so that the kings can emerge from hiding, and come out fighting. (The endgame is often defined as when the kings can and must be active. The nonpareil teacher Purdy said the true test of one’s opening play is the play given to the rooks, so why not say the middlegame can be defined as when the rooks can and must be active.)
To be specific about moves in the endgame *is* possible, and memorizing them *is* required. There are some endgame positions you *must* know. GM de la Villa says there are 100 of these (in his book The 100 Endgames You Must Know), but GM Soltis says it’s actually about a dozen (in his book Confessions of a Chess Grandmaster). It’s comical that chessplayers hate learning one dozen or one hundred essential positions because they think they’re boring, and because setting them up on a board is so much work (seriously, chessplayers prefer to study openings because setting up that board is second nature, while setting up an endgame takes a little attention), though they will study openings cheerfully and uselessly and endlessly.
In the endgame, the tactics are not aimed at creating combinations and double attacks, but at promoting pawns. You have to be a master at tactics in the opening in order to survive it, in the middlegame in order to achieve positional advantages, and in the endgame in order to win the race to queen a pawn.
If you don’t know what to do after you queen a pawn… see, you must know how to win with three units (K+Q vs. K, or K+P vs. K) in play, or all else is useless. It should follow, therefore, that you must also know how to play with four units on the board, because that reduces to three units on the board. And if you’re with me so far, you should also know how to play with six units. If you keep going with this logical thread, the least important position to know is the one with 32 units on the board. Chess publishers and chess teachers capitalize on the fact that chessplayers most want to make it most difficult for themselves.