revision and the wild spark
The fin makes the film. That's because a good, cohesive story works like a legal argument. Even if you marshal a terrific defense for your client in Acts One and Two, whiff the closing statement, and an innocent person is going to prison.
No spoilers here, but David Fincher's Seven is an object lesson in the importance of an ending. The movie's final twist elegantly ties together and elevates what might otherwise have come across as an exercise in shlock horror. Ironically, that twist didn't appear in the last draft but in the first.
As Fincher explains here—warning: spoilers—the producers wanted him to direct, but accidentally sent him the first draft of the screenplay. When he expressed interest, they apologized for sending him the wrong version and assured him that the current draft lacked any trace of that bizarre ending. Naturally, Fincher insisted on directing the first draft.
Revision is important. Whenever you work your way through another draft, the prose gets cleaner, smoother, and tighter. However, you can easily end up smoothing away everything interesting, unusual, and risky along with the awkward syntax and clumsy grammar. To the editorial eye, focused on tall nails in need of hammering down, the wild sparks that give the work value can look almost exactly like problems to be resolved.
If you revise too early, often, or compulsively, you will lose that spark. Polish every rough edge without an eye toward preserving odd and unexpected elements in the text, and revision becomes pasteurization. You're left with something correct but inert—harmless, lifeless, and of little interest to readers. If you've ever revised something to the point that you can't remember why you bothered writing it in the first place, you're familiar with the phenomenon.
Let your writing breathe before you take up that red pencil, and when you do revise, handle the pencil lightly. Creative fermentation is a delicate process, and it's rarely clear—especially early on—where the magic is emerging.