In the comments last time, Nigel asked a question that must have vexed us all at one time or another. We all know you can run a perfectly good adventure from a page or two of notes, but what if you're writing the adventure so that somebody else can run it? Every little detail, obvious to you, soon starts to demand a page of its own. Look at the annual Christmas scenario for Legend. Tim, our secret Santa for those specials, cooks up the adventure on the train to London and runs it from crib notes scrawled in biro on the back of his hand. Yet by the time I'm serving the scenario up to you it has typically swollen like a Quatermass experiment to 5000 words or more.
I don't claim to have a magic formula, but a lot of published scenarios are written to be a fun read rather than a useful template for running the adventure. It's what sells. So an investigative scenario, for example, will lay out the clues and describe how the player-characters are expected to come across them, all wrapped up in a form that reads like a mini-mystery novel. But to run the game you don't need any of that. For maximum compression, you really just need a couple of documents and (maybe) maps of the key locations.
The first document describes what would happen in the adventure if the PCs weren't there. You might include some contingencies here if you think the referee isn't experienced enough to make them up on the fly. Eg: "The Terminator goes to the nightclub to kill Sarah Connor. If she escapes it seeks out her mother or friends, and remember that it can mimic their voice on the phone to get her to say where she is." Take a look on Wikipedia at the plot summary of a few stories you're very familiar with. That's a good guide to how compressed you can go. You should end up with something like this (from The Eye of All-Seeing Wonder on the Tekumel site):
The other doc is just the NPCs. What they want, how they are likely to try and get it, the resources they can call on, and their attitude to the PCs and each other. This is the place to mark how the PCs might get involved, eg: "Du Pont thinks Goldfinger is cheating at cards and asks the characters to find out how. (See: Jill Masterton.)" Whether the PCs take those cues, and how they use them, should emerge in play rather than being baked into the scenario -- and that's where you can save on word count.
The maps don't need to be more than rough sketches. They're just there so you can answer questions like, "To get to the bath-house, do I need to go out into the courtyard or can I get there from the dormitory?" I tend to do my maps on scraps of paper at the table, often improvised when a player first asks a question so as to be consistent thereafter. You know the kind of thing:
A scenario like I'm describing will be a very dry read, but after all it's not supposed to be a novel. The only reason we have these neatly act-structured published scenarios is because that's the way the publishers get a customer to part with their money. Recently I was looking at a scenario in a published RPG which took up twenty-five pages (around 10,000 words) and it could all have been fitted on one piece of paper. As a short story it was fine. As a reference doc for running a game from it was useless. There was too much detail, too many cross-connections, too many assumptions about which order the PCs would do things in -- not to mention assumptions about what they would do.
What about set pieces? It can be really hard to resist preparing those in advance. You think, aha, if the characters go up this hill, they'll see what looks like a henge of standing stones on the skyline, but as they get closer it rears up and they see it's the dorsal spines of an enormous dragon. Let me stop you right there. You're not writing a movie. Maybe they'll approach the dragon from another direction. Maybe they won't encounter it at all. Murder your darlings. If you try to plan a cinematic set piece, there's a risk you'll then railroad the players to make sure it happens. So what if they go someplace else and do something you didn't anticipate? Have faith that you'll think of something in the moment that's as good as any scene you might have scripted in advance.
No written adventure survives contact with the players. So why go to all the trouble of writing it out neatly like it's the Great American Novel? Everyone's mileage is going to be different here, but if you're producing a scenario for somebody else to run, try paring it to the bone. Think of it as an executive summary for a CEO with a very short attention span. A couple of pages at most. The way the game turns out might be nothing like you or the referee expected -- but as long as the players have fun, who's complaining?
