Tuesday, July 5, 2016

About improvisation and preparation

I encounter every so often arguments about improvisation versus preparation or how Dungeon Masters don't need mountains of material to make a good game. Or the opposite, where it's all about the strict adherence to a "script" of sorts, be it rules or module or what-have-you . But the whole thing is way more grey than black and white. Since that's a discussion that came up a few days ago with a friend, I'll write a bit about this as a change of pace before we get back to The Dragon's Cough ...

Examples from people who know what they do ...

The discussion started with an article about Inland Empire I saw over at A Kingdom Is. We (that is me and a friend) saw that movie back when he came out on DVD and dissected it afterwards over a few beers, as we would do with David Lynch material. It's been a while, but I remember that it had been a long discussion with the attempt to put some sense to the whole thing. Ultimately, we failed to come to an conclusion (was a good discussion, though).

So years later I read the article linked above and I'm thinking, well, we never had a chance. Lynch had no script and had handed the cast it's dialogue as he needed it. Lynch's comment:
“I write the thing scene by scene and I don’t have much of a clue where it will end,” he said in a 2005 interview. “It’s a risk, but I have this feeling that because all things are unified, this idea over here in that room will somehow relate to that idea over there in the pink room.” [source]
That's Lynch for you. I'm not sure he managed to accomplish what he tried here in a traditional sense, but I have to see that movie now with fresh eyes. Maybe it connects on a level I didn't consider ... Anyway, this being about improvisation, I'd say it is a great example how to do it. There are, from the top of my head, three levels of creation here: (1) he writes it spontaneously, (2) the scene is filmed, interpreted by the cast and gets additional meaning with the set design, lighting and all that jazz and (3) all that material is condensed down to the cut and gets a musical score, the final composition, again with meaning added to it.

Just one WTF?! scene from Inland Empire [source]
It starts with spontaneous creation, improvises ahead and gets very structured towards the end. As a viewer you can't even be sure that you see the scenes in the order they'd been filmed (most likely not), so there is no way to tell what the initial train of thought or state of mind must have been to write a scene the way it got written.

That discrepancy between the process of creation and the process of consuming the result is one aspect of what I'm talking about here.

Another, also very interesting process can be seen by the writing of a series like, for instance, Breaking Bad. Here is a great article about what happens in a writers' room. It's highly structured as a collaborative effort up until a point where one of the writers is able to write it on his own. After that it goes through a ridiculous amount of hands before someone starts filming it. Nothing is kept unplanned and there is almost no room for improvisation after that first stage of planning.

But writing a movie or tv series is not like DMing a game, right!?

The result for the viewer is in both cases, as you'd expect, the same: a complete story. And this is where we bridge the whole thing to role playing games. Improvisation is in different levels of the creation of a story, depending on the medium that is used. With Lynch it's the whole process, from the first word to the final movie, structuring it stage after stage. It's an example of a writer/director having control over the whole process and taking advantage of it.

Writing Breaking Bad is the complete opposite approach. The process of creating the story episode by episode needs very thorough planning with the writers improvising (brain-storming, really) as they fill the gaps in the different story arcs because they often don't know much more than where they are supposed to end up. But as  soon as this is done, it's codified to a poit where it's ensured that the writers' vision is translated as intended.

But with DMing a game it's quite different. The improvisation in a role playing game starts at the table, when the players start influencing the narrative, so the game arguably starts with a structure of sorts (rules, stage, meta-plot) and when the players get to interact with all that, improvisation is paramount. The goal is still the same: manufacturing the illusion of something that resembles a complete story!

And still, I'd argue that what needs improvisation is in all cases the same. It's about bending a narrative towards meaning and closure, producing tension along the way. It's under different circumstances in every case and I really believe that the DM has the shit-end of the stick here (as we don't have the luxury of taking time to reflect our choices or only after the game/between games), but it's very much the same.

So, the story is the thing ...

It's like Lynch says above: all elements somehow connect. It's the sum of all the parts, really, the banter and chatter, what the players decide, what the rules say or how they work and how all of that is set into context by a DM. The rest is retrospective and that's what forms coherence in a campaign.

Somewhere in the cracks it's all improvisation. A word on that. I think we all know what "to improvise" means. But it is a bit more tricky than what the definition would let you believe. Actually, the general assumption (in my perception, anyway) is that improvising means doing something without preparation and, although true, I think that simplification like this is misleading to some degree. 

Take any jazz session as an example. You really need to know what you are doing with your instrument to make something like that work. So although a jazz musician might go completely off script, he couldn't do so without that kind of expertise. Same goes for those script writers linked to above. They need to know how stories work, how people react (on several levels) and how to move within the suspension of disbelief in general to be successful. Those are the tools of the trade (among other, related things).

They are prepared to improvise and the amount of available material is what brings variety to the whole thing. The better they are at what they do, the more likely it is that what they do will work.

All this is very much true for role playing games. A DM needs to know the rules of the game like a musician needs to know his instrument. He needs to know what stories are composed of and his setting just like the musician needs to know songs and styles. And then he gets to improvise with the goal to connect the fragments the game produces to a narrative during the game and a story afterwards. Both, DM and musician, need to know what they did already and what they have to build upon.

Improvisation without preparation?

As I'm writing this I'm beginning to ask myself what that actually means: improvising the whole game, completely unprepared. Is it even possible? Think about it. No rules, no setting, no character sheets, damn, to improvise a game entirely would be a mean task. Not without merit, but in that sense more like being stranded on an island with a few other guys and all you have is a coconut and a piece of string to make it work kind of way ...

I'm exaggerating, of course. But it illustrates my point. You must have an idea what a role playing game is (which is where the problem starts, right?) to be able to improvise it. But even if you just take the usual approach and bring knowledge of the rules, the dice you need and maybe even some character sheets, pencils and blank paper, you'll get to a point where you need to improvise sooner or later.

It's one thing to spin what the players do and there is almost no way to prepare for that but knowing them and their characters. Where we have control, though, is how much information and tools we'll have handy to riff off of. My point is that the level of preparation is directly connected to how well the improvisation will work. That's due to the fact that we improvise the course of the narrative during the game with what we have, not in the beginning. We improvise structure.

Here is the thing: improvisation on the fly is always limited to what we have in reach and how far we can go with that. It starts with our brain. And if the brain needs inspiration, it'll check the surroundings for material. Free association works that way and we are talking very fast and fragile bonds here. How often did you, as a DM, sit there, eyes wandering through the room, looking for ideas? That's what I'm talking about.

Sometimes all you need is a DM screen and a map ... [source]
Closure is the other end of the spectrum. that's what structure is all about. Beginning and end, funny business in between. How long those arcs of improvisation are is a matter of training, I guess, but you'll always have a pattern like that resolving again and again in the game.

Such thinking is categorized, of course. Specific topics start with a tag or a catchphrase. What guy am I looking at, for example? Race, culture, cloths, specifics, behavior, situation ... stuff like that. Now if you have tags and you are prepared, well, then you can take what is already there (like a culture and typical garments you prepared at some point or a fitting random table) and start improvising from that point onward instead!

More material means more variety means more depth in your game , as simple as that. You still improvise, but from a higher vantage point, so to say.

Final thoughts ...

Ultimately that's what I mean when I say it's a grey area, not black and white. We already bring a lot to the table if we bring what is needed to play and know a setting and the rules, maybe even have an idea for a story or two. On the other hand you won't DM a (good?) game without the need to improvise at some point, no matter how well you are prepared.

It's like the both examples above are two extremes here. Lynch, who has total control over the whole process and improvises it all (maybe a very experienced DM going rules-light, making it all up as he goes) or the writers from Breaking Bad, relying heavily on the structures they are able to formulate up front (maybe a DM using a complex set of rules RAW and a fitting module/campaign alongside ... I'm thinking AD&D with T 1-4, for instance) and going from there.

As a DM you'll always be somewhere in between, hopefully developing your on style over time (if you haven't already!), maybe even different repertoires (conventions, for the kids, a long campaign, different games ...). Which ratio between preparation and improvisation you'll need for the optimal game is a matter of taste. But I'd like to stress here at the end that they are not opposites. On the contrary, they compliment each other a great deal.

But the important thing is that it all connects in the end.


  1. I'm not sure about your assumption that "The goal is still the same: manufacturing the illusion of something that resembles a complete story!" At least not from the inside while we're playing it... I don't Play or GM games with that in mind, though it might all be told as a story in retrospect. As a GM I am concerned more about what's plausible and consistent in the moment, as a Player I'm caught with the same but only for my own PC.
    Otherwise, I agree that good preparation enables good improvosation... and knowing when to let go of that preparation and just let things flow along their own course.

    1. Thanks for the comment! Yeah, I agree, the story is happening in hindsight and I make the distinction somewhere in the text between narrative and story for that reason exactly. It's what I mean with giving structure to fragments and what you describe as being "plausible and consistent in the moment". Finding that sweet spot where the narrative starts flowing is different for everyone and one of the greatest things a DM can achieve :)


Recent developments made it necessary to moderate posts again. Sorry about that, folks.