I've mentioned the idea of structure or no structure in an rpg campaign before. I've seen some rpg theory/discussion that suggests a GM's role is simply to facilitate improvisation-- to move things slightly, but to accept everything that players suggest. I think there's a place for that-- a particular kind of game, but one that would fairly well set itself up as that from the beginning. I can imagine that as a one or two-off evening of play, but I have a hard time seeing how that would sustain itself. Or rather, I should say, I don't think within the group of people I play with that it would last. I myself prefer to play and respond to a story at least outlined by a gamemaster. I like having some narrative control as a player, but I also like to have the sense that there is a plot. That it isn't just a series of incidents, but that these things add up to something.
Sidebar: Now that I think about it, we have done a variation of this with the group. Twice I ran Amber rpg sessions where we'd did Court Battles. Essentially everyone auctions for roles, powers, stats and abilities and then you just set everyone at each others' throats. A few agendas are distributed and the GM has some incidents, but overall it is a free-for-all with people trying to gain power. My favorite moment from this was when one of the Dave's brought along ice-cream bars and generously distributed them after the game started. He then came and told me who he'd given them to and that they were now marked with Chaos.
That's why I find some people's criticism of Lost a little odd. I think I may have made this point before, but it is worth coming back to for what I'm talking about. There's a certain segment of the viewing public that, originally enthused about Lost, became upset that “The Writers didn't know everything from the beginning.” There was the sense that this wasn't playing fair-- that this lack of knowledge undercut everything about the series and the mystery. I don't see it that way. In fact, I think that process of building the structure of Lost is really close to what a GM has to do-- in an extended campaign. You set up plot threads, you have a general sense of story, you probably have a good idea of what the endpoint is going to look like and you run from there. If you put out clues and pointers, they serve as anchors. You play fair with those and don't later eliminate or ignore them. The Writer/GM is making it up as they go along, but they also know that whatever they've said before serves as the baseline. If something violates that baseline-- that has to be explained. So Lost, and any extended rpg narrative, isn't really about solving the mystery from Day One. It is about that process of solution-- an d eventually, once we get towards the end/climax, being able to put those pieces together.
That's an inexact model, but I think an interesting approach. I've describe a lot of what I do as Schrodinger Gaming. Yes, things are going on and happen in the background, but things don't coalesce until they're actually observed and interacted with.
SOME THOUGHTS ON ONE SHOTS
That being said a good deal about what I've talked about before in terms of campaign creation and gamemastering applies to an extended campaign. I'm imagining at least a dozen sessions with an endpoint of years or more. I think the longest campaign I ran took just under three years, with biweekly sessions. Usually my campaigns run about two years, depending on the players. But games and campaigns of different lengths, while sharing some of the same needs, have some very different structural requirements. I'm going to break these down into four rough groups:
(one or two sessions generally)
(4-6 sessions generally)
(usually around 12 sessions with a fairly wide spread of difference)
(aiming for a year or an ongoing campaign-- likely with a less definite endpoint)
In my experience, one shots usually revolve around trying something new out: either a game system or a fairly specific story approach. Most of the one shots I've run have been for events (birthdays, holidays, wedding showers, bachelor parties) or for conventions. The structural and narrative pattern for a one-shot is usually fairly rigid-- there's a clear driving incident than controls what's going on. Play focuses on exploring that incident, rather than exploring the characters and the setting itself. That's not to say it doesn't happen, but the time restrictions don't allow the kinds of relaxed approach of a longer term game. Even those one-shots that plop people in and just set them spinning (the amnesia-trope game-- usually have an inciting incident or discovered purpose).
So some words about characters and prep for a one-shot. Generally the GM should make up the characters beforehand for these games. If you sit down to a demo and it looks like you're going to be making up characters for play-- get up and leave. Even the fastest character creation will eat up a huge amount of time from that available-- especially with a new group with differing learning styles. Players will get a sense of character structure from play. Generally character creation happens only once in a campaign so it shouldn't take up time.
Sidebar: if someone is sitting in on one of your games-- imagine that it is essentially a one-shot for them. Have a character or a choice of characters ready for them. Adjust the structure of the session accordingly. Try to have them involved and try to give them a taste of the major mechanics: social interaction, combat, etc. Their stopping by is a test drive so you want to give them a chance to see everything. Again, having them make up characters slows that process down and forces them to learn a set of complicated rules that won't actually affect them in play. We had Gene come through to sit in on the Libri Vidicos game-- and I while I could have done a better job with structuring the session to that, I did make sure to have a conflict/combat established which was a nice demo. He was also able to hit the ground running because of the cheat sheet I'd done.
Regarding characters in the one-shot:
-Have character sheets ready. I prefer random selection of characters, but let the group decide if they'd rather pick their own.
-Powers and special abilities should be spelled out on the sheets themselves. If the system has feats, advantages, or the like, put the basic rules there. They should be able to grasp that easily. That also means-- don't give players too many abilities. Keep them limited and interesting.
-Use the clearest name possible for terms and skills-- so that a quick read through allows a player to pick up what the character can or can't do.
-Make sure any bonuses and figured things are already worked in to the numbers and make sure to tell the players that.
-Give at least a couple of lines of personality traits to the character-- not enough to box in the player, but some material to work from.
-Provide a list of the other PC characters: what their role is and what this PC thinks of them. That's a chunk of work, but will pay off hugely if you're running a one-shot multiple times. If it is just a throw-away game, then not so much.
-Character sheets and packets should be disposable-- the players should be able to write on them. That encourages note taking.
-A rules overview sheet in with the character sheet can help, but isn't always necessary.
-Decide ahead of time which characters you'll eliminate if you have a smaller group than the maximum. If you have characters with key elements, make sure they get chosen.
One of the key considerations here is time-- players need to get rolling into their characters quickly, both from a rules and personality standpoint. Your job is to facilitate that. Most players will move quickly into character habits-- broad-stroke mannerisms and comments to make their characters clear and memorable to other players at the table. Be ready for the overacting that sometimes happens.
The narrow time limit means that you need to start the game in the middle of things-- if there's a mission briefing, handle it quickly and then drop them into the zone. Some players like to prep and plan. One way to get around this is to propose that they're welcome to retro-plan. That is, once they are on site if they come up with something that they logically would had had in place or would have brought with them, they have it. If it is questionable, have them make a skill check. Another option is to stress that the situation involved hasty prep and they have to make do with things on site.
I like to give a quick mission or purpose briefing and then get them to the location and switch things up. The person they were to meet isn't there, the transport they're taking crashes somewhere else, what they were told was going on is completely different. That provides a quick status quo change and forces them to think on their feet.
If you are going to stick with the mission structure-- make sure you lay out the objectives for the players pretty clearly. In the Conspiracy X scenario I have, I give a quick list of three in order of priority and let the PCs make judgments from that. Those choices can make for interesting interactions-- some players will find good hooks for the personas they build up in the short time. Again, in the ConX scenario, some of the PCs have a more slash-and-burn approach than others. Having the party decide when the mission has gone off the rails makes for interesting play.
If you're stuck for structure, I recommend the following:
-Opening scene, in the middle of things, quick combat. This allows the players to get a chance to see how the system works.
-Clues from opening scene, investigation and follow up
-Next major turning point scene-- change of location, puzzle or trap requiring some thinking and creative play
-Climax incident with major conflict and combat
It is pretty basic, but loose enough to provide strong play for the group in the time allotted. About half the one-shots I've got written up work basically around this. Remember that if the players do get stuck, you should feel less bad about pushing them forward with incidents or events. You want them to have a chance to see most of the story. By the same token, don't feel too bad if they do lose. Keep an eye on the time. Once you hit the last hour of play, feel free to start killing off PCs. That's a pretty significant wake-up call. If someone's being stupid or goofy and asking to die, the one-shot is a great opportunity to provide them with a cinematic and gruesome demise. Of course, if you're running the My Little Pony rpg (TM), you may want to rethink that.
You should also consider providing hand-outs for the PCs: maps, images, pieces of evidence, etc. A concrete handout can be a great short-cut. It allows you to create verisimilitude quickly. You can use them to establish tone. In one scenario I have, the PCs end up in a place that owes more than a little debt to Silent Hill-- they find a driver's license on one body with strangely warped text and image. I did that up in a few minutes and it added to the creepiness. In another scenario, the players could do the classic pencil rubbing on a notepad to get some text. I had a version of that ready-- but blurry. They cold try to work it out themselves, or if someone got smart and asked if they had some kind of imaging tech available to them, I had a clearer version ready for them. About half the time they'd think to ask. For the scenario I have where the players are all from a corporate office, the character sheet is done like a personnel assessment (“Things I Need to Work On” shows the character's disadvantages). The background material was written as an inter-office memo. I also had name tags and got old, battered and marked up office folders to put the player's packets in.
One other trick can be done-- and works especially well if you're going to run a scenario more than once. I got this from an rpg magazine article years ago specifically talking about sci-fi games. It mentioned “The Parrot” effect for some PC roles. Characters who have science or lore abilities often ask the GM a question and then simply have to repeat what the GM says back to the group. If you have a game mapped out in your head, you can usually anticipate what kinds or questions and analysis the PCs will carry out. Ahead of time write out the answers. Then when the player asks, toss them the note or card. You'd be amazed at the reaction of players at conventions. They're able to use their own voice to convey the information-- plus if the scenario has a little paranoia to it, the other players won't know if what's being said is the whole truth or maybe just a lie. For a four hour game, doing a few dozen of these prepped response can really help. The players also gain a sense of pleasure when they ask a question you haven't prepared for-- creating a meta-challenge for them.
Strangely enough, to do a one-shot really well actually requires more work than the space it fills. But a well done module can be used repeated. Most of what I've done with this has been for HCI Portals (stop offs in a longer campaign) or Convention scenarios. I've got a small library of pick-up games ready (with handouts) that I keep on hand. I probably should work up some more just in case as a practice, as most of the stuff I have right now is the Eden Studios stuff.