It's also been a good long while since I've been inspired to do this, and I really need to shake the rust off.
Okay, so what does a campaign need? An idea, a theme, a timeline, several maps, and the contents of said maps (this includes NPCs and so forth).
The idea and the theme are not the same: the idea is the "hook" that makes the campaign interesting to the players, more than "just another boring dungeon crawl". It must therefore be visible to the players as soon as possible: it is an explicit part of the in-game fiction. Another word for the idea might be "the gimmick", but that makes it sound hackneyed. Really, it's just a succinct way of putting into words what the "elevator pitch" would be if you were trying to sell the campaign to reluctant players.
The theme, meanwhile, is what the campaign is actually about—and it will not be immediately visible to the players. Even after many sessions of play, the players might not suss out the theme at all. And they might never be privy to the theme that you started out with, because the players always have the power to take the campaign in a totally unexpected direction. So the theme isn't set in stone: it's changeable, it almost certainly will change depending on what the players do, and it's not even something that you can really start out with or impose from the top down. (I mean, you could if you really wanted to, but it would be both difficult and unsatisfying. Better to let the theme emerge from the process—from the locations and histories and NPCs and their machinations—as you add content to the campaign.) You start with the idea, sketch out a framework, fill in the gaps, keep adding detail, and then tease out what the theme has been all along and use that to revise and refine your content and make it more significant.
The timeline is an interesting one: once you've mapped out the campaign area, you need to have at least a general sense of what happened in the past (important historical events, major migrations and catastrophes, the occasional Personage or Object of Historical Import™); and, of equal importance, you need a more specific and detailed hypothetical future, enough to give you a sense of "historical" things that will happen within a reasonable time-frame (say, the next year or maybe two) in a hypothetical world where the PCs were never around to change anything. (And, depending on what the players decide to focus on once the game starts, maybe they won't change a thing. But they probably will, which is why you don't have to go too far into the future—it would be a waste of effort.)
This is why it's not really true to say that OSR games don't have plots or stories. Of course they have plots: they have NPCs plotting to achieve their goals. And in creating a fictional world, you are telling stories: it's just that the PCs are going to be empowered and expected to demolish these stories and re-write them in their own image. Moreover, if the PCs choose to be passive, the NPCs which are active within your living campaign world get to make progress towards their goals—they're not static and waiting to come into existence whenever the PCs should happen to encounter them.
At the outset, the timeline should have the campaign area existing in a state of relative calm. That is to say, the world is living and breathing around the PCs, and it certainly isn't focused on them; but it must also be quiet and uneventful enough to give them elbow room to begin their adventuring careers. A shooting war or a zombie apocalypse is a poor backdrop for a dungeon-crawl, because these kinds of events don't allow the PCs to set their own pace. Such events could develop later on in the campaign; but if so, they should emerge organically from the interplay between your future timeline and the PCs actions (or lack thereof).
Maps… I'll deal with maps in their own post, since this is starting to get long, and maps deserve a lot of detail. After all, they'll be with you all throughout the campaign. They're its backbone. For now, I'll just say that I like to have some twenty-thousand square miles of wilderness (less daunting than it sounds), a handful of towns, and a good six to ten levels of dungeon all ready to go before the game can even hope to begin.
That's what a campaign absolutely must have to get started. And you begin with the idea—which finally brings me back to the title of the post. Before you can even start brainstorming, you have to spitball. Find the idea.
* * *
What sort of game-related thoughts have been on my mind lately? Well, mostly I've been reminiscing about the last few campaigns that I ran. I always seem to pick them apart and over-analyze what didn't work in the immediate aftermath, hoping to shore up the weaknesses next time. But now that I have some distance between then and now, I mostly just want my next campaign to be as different in tone as possible, to keep things fresh. So that's a good place to start: not retreading my most recent games.
The last game I ran was called La Isla del Loro Muerto, a very lighthearted and comical hunt for pirate treasure, like Monty Python meets Monkey Island. It was very limited in scope, taking place entirely on a single island which was probably smaller than the Isle of Dread, and it was focused on a single mystery ("What was the fabled lost treasure of Captain Bloodsheds Jackson, and where did he bury it?").
Prior to that, I ran two fairly standard mega-dungeon campaigns, with the PCs operating out of a home base and moving back and forth between the town and the dungeon on repeated delves. These campaigns were reasonably successful, but they did both share one problem that I'd like to correct. Namely, I ostensibly run steampunk campaigns. Engines & Empires. (And both of these two recent dungeon-crawl campaigns served as test-beds for what eventually became the E&E Core Rules, in which respect they succeeded more than admirably.) But as genre and period pieces, they failed miserably. They didn't feel Napoleonic or "Crimean War-ish" or Victorian. They felt like regular old medieval D&D with guns.
Part of this stems from the fact that the typical campaign milieu necessitates a remote location, a frontier or a borderlands, which forces the PCs to rely on limited resources and their own wits. That way, they can't just run back to civilization and call in the cavalry when a problem arises—they have to be adventurers and deal with it themselves. Plus, civilization is complicated. It takes a lot of work to detail a metropolis, and I don't think I have the time right now to prepare for a city-based campaign. Not in a way that would do it justice, at any rate.
One way to make a frontier feel steampunkish is to use a Wild West milieu, but I'm not the biggest fan of westerns. So… where to go from here? Well, this is when a lot of DMs will turn to whatever media they've been exposed to recently—books, movies, video games. I'm no different. What's intrigued me lately?
I've been playing a bit of Skyrim again recently and I dearly love its aesthetic. Plus, I just saw Thor: Ragnarok. So apparently I've got Vikings on the brain. But trying to do steampunk Vikings would just be silly. (The Vestenmannavnjar in 7th Sea are already far too silly, and Théah is a Renaissance-era setting!) Still, something about Skyrim—not the Norse stuff, but the windswept wintry vistas and remote mountain ranges—that's speaking to me right now.
And, bam, that gives me my idea. A remote subarctic frontier, with a vibe like 19th century Sibera, or the Yukon during the Gold Rush. And all at once, a variety of sources come to mind: video game settings like Syberia and the mining town of Narshe in Final Fantasy VI. Books like Icewind Dale. Movies like the Hope & Crosby picture, Road to Utopia.
Now I'm picturing something like Narshe—a remote, snowy mining town, but technologically modern and self-sustaining. Steam engines and gas lamps everywhere, sled dogs, prospectors, gamblers, scoundrels. A gold rush… but instead of miners and panners an sluices, it's a dungeon inside the mountain nearby which has turned this place into a boomtown. Maybe there was a mine, but the miners breached a forgotten complex within the mountain, and now adventurers are exploring the newly-discovered ancient tunnels and halls.
But I want to take it one step further—ramp up the sci-fi a little bit. Maybe it's not just gold that they're mining for. Maybe the townsfolk were digging for some kind of funky crystal or mineral that helps the town generate power, like a steampunk analog to naquadah from Stargate—or magicite or materia from Final Fantasy—or ember from Torchlight—or ghost rock from Deadlands—you get the idea. Some kind of weird telluric ore that blurs the line between magic and science and is valuable to both mages and techs for the sake of item creation, but which could also be sold to the town at a premium for mad cash and XP. That would provide an incentive for adventurers to explore the mysterious, ancient, ruined complex inside the mountain (instead of just having miners digging shafts in a different direction). The complex, whatever it was, was powered by this stuff (like Dwemer ruins in Skyrim), whatever it is, and it's rare and valuable enough to make looting the complex's infernal machines and wyrd contraptions worth the effort and the risk.
Now that's and idea. Next time, I'll talk about making it more concrete—situating the idea (which is a combination of site—the ancient complex—and scenario—the town wants funky power-crystals) within an actual setting.