It's been a while since I posted, but there’s something that's been buzzing around in my head, so I thought I'd have a go at trying to articulate it.
It's a bit abstract, theoretical, but bear with me.
I'm drawing on my experiences with Starblazer Adventures, but it's also doubtless true for Spirit of the Century and FATE in general.
I've been writing a lot of Starblazer scenarios and campaign stuff recently, for the upcoming Mindjammer book, and I've been struck by what a different experience it is from writing scenarios for more "traditional" RPGs. Starblazer by its nature plays as a fast, furious, and inventive game, rollicking good fun with a hell of a lot of player involvement, but often it still just feels like a normal RPG. But it's not.
The whole player participation thing radically changes how you construct your campaigns and scenarios. In The Olden Days (and still now with more traditional RPGs), scenario writing is/was about providing a good, solid description of places, people, and critters, and also presenting enough plot detail that the GM could grasp the whole intended structure and "learn" it, before gradually revealing it to the players in the course of play. Pretty much any scenario would have this “implied plot”. Even if you avoid railroading in a scenario with a clear plot structure (no mean feat), a lot of the game involves the players trying to work out what the GM knows, and then navigate the scenario using that information.
Starblazer blows that out of the water. At any point, players can make Declarations, etc, and invent elements of the narrative out of whole cloth, which the GM then makes best efforts to incorporate. It's a bad GM in Starblazer who turns down player declarations because they don't "fit" the story / plot structure he's created; instead, you turn down your planned plot structure *in favour of* the player-created innovation.
It takes a few moments to realize what that does to your "planned" campaign. In fact, you can't really plan it at all, nor should you. In writing a scenario, for example, what I find myself doing is detailing the "starting assumptions" of the scenario - the status quo before the PCs turn up. That involves sketches of locations and people, sure, and also statistics - but the understanding is that all these are provisional, and subject to complete change during the course of play. A given bad guy may suddenly turn out to be a good guy, and vice versa, based on player narrative control. This means that although you have some control over how a scenario starts, you really have no idea of how it’s going to go as soon as it’s out of the gate. In a science-fiction campaign, this has major implications; imagine a first contact situation. You've set up the aliens to be basically friendly, but incomprehensible, and the "starting assumptions" are that the PCs (as the contact team) are going to aim to do a good job and bring the aliens into the fold as friends and allies. That’s what’s expected.
In Starblazer, you have almost no control, as GM, over how things go from there. The players may suddenly declare that the aliens are actually inimical, perhaps they want to eat people, maybe they have some terrible disease, or that there's already ANOTHER contact team down there who've gone insane in Colonel Kurtz fashion and are killing the natives. By the end of the scenario the entire planet may be a smouldering ruin, or your star empire may have a war - even a civil war! - on its hands.
That's an amazing power to unleash at the gaming table - and very exciting for the GM, too, as you really don't know where things are going - but how the hell do you link scenarios together into a campaign in such a game environment?
I don't lay claim to authoritative answers on that point, but here's how I'm handling it: I'm taking a "structuralist" approach. I'm viewing the elements of a given scenario in largely abstract terms, and although I'm cloaking those abstract structures in setting-specific chrome, it's possible to change that chrome radically without totally trashing the structure. You can even replace individual elements in the structure, “hot swapping” them with other elements resulting from player narrative intervention, without destroying its structure or utility. So, in the above example, you've got the following structure:
"Star Empire PCs contact Incomprehensible Aliens on Unexplored World. During the scenario, the Star Empire modifies the Aliens' life experience, and the Aliens may also modify the Star Empire right back. By scenario end, there is a Defined Relationship established with the Aliens."
That's very vague and woolly, but it allows me to swap chrome descriptions easily, and even to “hot-swap” large chunks of the scenario structure (people, places, events) and to expect to handle the scenario progress in a certain way. For example, by the end I know I'll have a status report on how the Star Empire and the Aliens are getting on. I don't know whether they'll be allies or at war, but I know I’ll be modifying the relationship between the two “polities” based on the scenario outcome.
This is where the Starblazer "fractal" comes in. By using organizations, etc, you can actually track the relationships between, say, Star Empires and Alien Worlds. By the end of the above scenario, if there's a developing war, I can actually make the result of the scenario into the first "exchange" of a conflict between the Star Empire and the Aliens *as organizations*. That's an amazingly powerful tool; it means I don't have to know in advance whether there's a war or not, and I can rely on the game rules to manage that war if there is one.
So what does this mean for writing a campaign setting? Well, the first thing is to realise that whilst you can present a setting, and indeed loosely-linked scenarios, in a published book, you can't expect *in any way* that individual GM's campaigns are going to look anything like yours. Events could be *hugely* different. Imagine this situation: there's been a planetary invasion in the Star Empire by a bunch of bad guys. The scenario deals with trying to fight them off. Do the PCs win?
In Starblazer, you don't know whether the PCs win or not. You can’t assume anything. By the end of the scenario, the bad guys could be toast, or they could be the planet's new rulers. So how do you write a sequel scenario? Answer: you don't make the scenario dependent on the precise outcome of its predecessor, but you derive it from the predecessor’s *structure*. We know there's been a planetary invasion, so we know there are bad guys trying to invade planets. The sequel scenario can base itself off that, and shouldn't rely on precise details of whether the bad guys won or lost in the previous scenario. Starblazer - with its organizations and rules for cool things like Plot stress, etc - puts enough tools in the GM's hands that you can leave that kind of tracking up to the GM.
Think of it in terms of "layers". You have your "top" layer in a scenario, which is the abstract plot structure (Planetary Invasion, etc) of the scenario itself. This has chrome description laid over it, and you present some basic starting points there, but acknowledge the precise detail is up to the GM. Then you have a "middle" layer, which is the specific, day-to-day events of the scenario, what happens to whom and when and where, etc. Lastly, you have the "deep" layer, which is the underlying nature of the setting; "there's a Star Empire, and there's some bad guys who are kind of like this. This scenario is going to affect that”.
What this means is that when writing Starblazer campaigns and scenarios, you only write the "top" and "deep" layers. You don't attempt the middle layer at all; that's up to the players and GMs who run their games. One set of players could have the entire Star Empire erupt in civil war as a result of their actions; but the scenario you write should be able to fit smoothly in there, drawing on the fundamental setting "deep structure" layer and presenting a "top" layer abstract plot structure which can fit more or less any narrative.
In some ways, this is just good design, period. For any game, you should always leave the “middle layer” blank. But it's interesting that Starblazer (and FATE in general) actually *demands* this approach - it's built into the rules. Scenario plot structure isn’t necessarily completely freeform, but nor can it be clearly defined. In some ways it owes a lot to the concepts of sandbox and plot point campaigns, but it's more than that - a good Starblazer campaign guide provides a support and an inspiration to GM and player creativity, not a replacement for it. The Starblazer rules are a toolkit for play; a Starblazer scenario is a toolkit for adventure.
What does this mean in the long run? Well, I think it means any kind of “one true way” approach to a campaign will forever be absolutely impossible in Starblazer, which is *wonderful* news for GMs. I don’t think I’m alone in having suffered from that; the urgent need to keep up with everything published for a setting, and also the ever-increasing fear and frustration that if you dare to invent something for a setting, it may - oh, horror! - be contradicted by “official canon”. I personally HATE that; it stifles creativity something rotten. In Starblazer, trashing “canon” and “the official word” is a RULES REQUIREMENT! All any setting or scenario can ever be is a toolkit for adventure, and the creative control remains for ever with the GM. Long may it stay that way.
*****