I think most screenwriters do "vomit" drafts: a first draft where you just get scenes down on paper so you have something to revise into a decent script. I don't think I really do this (except when I'm writing an already outlined episode script for someone else's show).

When I'm writing a first draft of my own material, I'm usually not trying to plot out a narrative or story so much as I'm chasing a unique tone or voice. Or maybe just a sustained feeling of anticipation that's punctuated with moments of emotions, revelation, & surprise.

My first drafts still fully suck. But I think they suck in a slightly different way than "vomit drafts" suck. I spend a lot of time in my first drafts trying to get the voice & tone of scenes right, & trying to find surprising or revealing juxtapositions between & within scenes.

I usually write a first draft, or a first big revision, with just a loose general feel for the aesthetic whole and a couple of key sequences and plot points. Then I start trying to write interesting scenes with this aesthetic whole and these few sequences and plot points in mind.

This usually results in a bunch of scenes that I end up scrapping later. But gradually, sort of maybe inevitably, the structure and plot starts taking shape on an almost subconscious level. (I think this is where watching a shit ton of movies helps.)

When I finish a first draft, I usually have the five or six most important scenes down cold. They don't often change that much going forward. I think I often have the characters pretty decently as well, though sometimes I'll realize I totally screwed up in conceiving a couple.

What I usually have to do in revision is improve the storytelling structure and locate the emotions and properly build them up and pay them off over the course of the script.

I usually have to transition the script from being a collection of pretty cool scenes and into becoming a fully fleshed-out story with coherent throughlines and recognizable dramatic builds and releases and turns.

This process works really pretty well for me when I'm writing something original on spec. The trickier thing is when I'm writing something for a producer or studio. Often, they want an outline to give them the structure. But for me, structure is more felt and discovered.

Often, instead of a fully plotted out beat-by-beat outline, I try to provide a kind of almost anticipatory sizzle reel of forthcoming attractions: 1) here are the characters, 2) here are some specific big scenes I have in mind for the story, 3) here's my plan for connecting them.

I kinda stumbled via trial-and-error to this method. It sort of carried over, I think, from writing poetry: I'd start with a mood and a couple of ingredients and maybe one key image, and then I'd just improvise until I made something I didn't loathe.

To me, coming into a script with a fully fleshed-out outline that I'd execute would be the same as an actor coming to set with a game plan to execute. If they have too much of it already in mind, the game-plan actor risks not listening or reacting to his or her scene partners.

I sorta feel similarly if too much of a script is pre-planned. I risk losing improvisational discoveries. But I also risk not listening to the story, or the characters. I risk executing my intentions w/o reacting to richer elements of the script that I didn't intend to be there.

I think a lot of folks find it hokey when a writer says the characters told them how to write the script. But I totally buy it. I think there's an ideal version of every scene, every character hovering around out there. My job is to keep searching until I finally discover it.