How can I improvise in a roleplaying game?

Our Thursday night gaming group just finished up The Armitage Files, a “boldly innovative” sandbox campaign structure centered around ten creepily constructed documents.  It was written by the ineffable and prolific Robin Laws, who may or may not ever playtest any of his own works. The idea behind the campaign is…interesting.  It pivots on the ability of the Keeper (that’s the person actually running the game) and the players both to improvise effectively.  As far as I can tell, the basic structure of a session is supposed to go like this:

Keeper: Well kids, what would you like to do tonight?

Players: (poring over creepy documents) Let’s investigate this Diamond Walsh guy!

Keeper: (deciding on the fly whether Diamond Walsh is good, bad, or neutral) Okay then.  Here’s the scene…

And it goes from there.  The Keeper has to be 100% on his/her toes, responding to what the players decide and guiding the scenes toward some sort of satisfying conclusion.  And because the game runs on the Gumshoe engine, the Keeper also has to pepper the scenes with core clues, which guide the players toward the solution to the mystery even as the scenes build toward a climax.

If it sounds like a lot to keep track of, that’s because it is.  Let’s say the players decide to break into Diamond Walsh’s house.  The Keeper has to decide not only what awaits them in the house, but what core clues they will find in the house, where those clues will lead them, and what the clues mean to the larger mystery.  

And then, as though that wasn’t enough of a mind twist, the Keeper also has to think about how each session is going to build toward a conclusion for the campaign.  There are ten documents and each session of gaming will involve investigating just a handful of elements from each document…that could add up to a lot of sessions, by the end of which players are going to want to know what exactly they spent the whole last year doing (in our case, yes, the campaign took over a year, even after skipping four documents). 

This is improvisation on a micro and a macro scale.  And I don’t really know how well it works, given that improvising a whole campaign structure on the fly requires some incredible mental gymnastics.  But then I think back to roleplaying in high school.  We improvised everything!  As a 15-16 year old, I ran a campaign using Everway (another totally improvised system) for months, and never wrote any notes or any sort of campaign outline.  I still think about that game with fondness, as do my high school friends (except for Todd, who hates Everway on principle).  How did we do it?  And how can we recapture that improv magic?

As it turns out, Armitage Files includes an article on improvisation techniques for roleplaying.  Let’s take a look and see how well they worked (or didn’t work) for us.

1. Don’t panic.  The first and most important piece of advice – whatever the players do, or however wacky the scene gets, just roll with it and trust the process.  This was way easier when we were younger, I would bet, and swollen with a sense of our own invulnerability.  As older and more fragile adults, it’s easier to freak out when someone throws you a narrative curveball.  I think panicking happened several times during our run of The Armitage Files, and on all sides, from concern over following up on leads that the Keeper hadn’t explicitly plotted to players deciding not to return to the game when their characters were placed in untenable or deadly situations.

2. Activate the players.  In “traditional” roleplaying games, the Keeper/gamemaster provides the only window into the narrative universe.  The players ask questions like, “What do I see?” “Can I get through the door?” “What happens now?”  The Keeper answers these questions and acts as the sole engine driving the story down its straight and inflexible narrative rails.  In an improve game, the players (should) have just as much power in setting scenes, describing actions, and moving the story along as the Keeper.  In The Armitage Files, our joint narration started out rocky…we all very much relied on the Keeper to tell us what was going on, and the Keeper took a lot of responsibility for pulling the pieces of the story together himself.  After the low low low point in the middle of the campaign, however, we all tried something different and began jointly narrating scenes, with the Keeper setting the scene and the players adding description and color.  This felt much more satisfying to me, though I still think we had a way to go before we truly split narrative power between the Keeper and the players.

3. Avoid negation/Don’t block.  This is apparently a basic tenet of improvisational acting.  Whatever elements the other actors incorporate into the scene, you incorporate and reincorporate them.  You never say, “no, that’s not there,” “that didn’t happen,” or “you’re wrong.”  I actually think we did a pretty good job of this.  Even during the darkest and least improvisational points in the campaign, if a player said something like, “Of course the evil tome is hidden under a trapdoor,” the Keeper would go along with it (and yes, Kaitlyn, the 9 year old daughter of one of our players, actually said this during a guest appearance at the gaming table…she was much better at improvising than the rest of us). 

4. Prepare a fallback.  This kind of runs counter to the other advice given for running an improvisational game, most of which stress not preparing too much in advance.  But The Armitage Files is essentially an investigation game, and so players are going to need to find clues to feel they’re making progress in solving the mystery.  Again, this got better as time went on; the Keeper started preparing a stock spine of clues for us to discover each session.  The fallback would have been even more effective if it had been just that – a fallback.  But often the prepared clues became the focus of the session, and there was panic (see point #1) if something interfered with the discovery of those clues.  Like, for example, one of the characters who was slated to discover a clue going irreparably insane.

5. Dramatic rhythm.  The advice in the book uses “dramatic rhythm” to mean “alternating positive and negative moments to reinforce the players.”  I would interpret it more as building some momentum into the game.  It’s that all-important issue of pacing, so tough to accomplish in a written work or a shared, roleplaying narrative.  We never really got the whole pacing thing down during The Armitage Files.  It became a running gag that our sessions were real-time…spending four hours wandering around a carnival in the game took us a full four hours of play.  And not much happened during those four hours to keep the rhythm going.  In a horror and an investigation game, events have to challenge the characters, provide setbacks and clues, and hopefully even offer a thrill of fear.  We never really got that during our games; more often than not, it felt like we were just spinning our wheels.

6. Leading and following.  There are times to go with improvisation, and there are times to seize the reigns and draw a session to a conclusion.  Our Keeper discovered that in a previous game he ran…the players appreciated it when he had their characters abducted and dragged forcibly to a cult headquarters.  And in this game, too, he picked up the reins far more often toward the end of the campaign.  This direction would have felt more meaningful if we had been improvising more effectively at the front end, I feel.  As it was, it seemed more like we were blundering around, not doing much of anything, and then all of a sudden at the end we were hurtling toward the end of the world (literally).

So there you have it.  I am left wanting to try improvising myself in a roleplaying game.  Sure, I did it before, but I want to try it now, as an adult, to see if it’s as easy or as hard as I remember.  Maybe if it goes well, I’ll try running The Armitage Files myself (in, like, another decade).  How do you all feel about improvisation?  Have you roleplayers been able to do it effectively?  And have you therapists found that improvisation skills help you in your psychology work?  (That, by the way, is a topic for another post…)