Friday, January 13, 2012

The Fickle Art of GM Improv

“Let's make it perfect, then cut it.”

That's a quip Stephen Colbert often makes to his writing staff. It shows, the program stays more funny than not.

While in blogging I settle for the merely good--the enemy of the perfect--the process feels much of the same. Often I will sit down, intending to write about one subject, get halfway through, revise it a few times and then just spike it. This started as a “GM lessons learned” rundown gleaned from running and playing on Google Plus. The problem is that the Constantcon boom is an embarrassment of riches, the lessons are too thick and many to deal with in one post.

Last night's half-assing it as an improbable dwarf, Xhomar the Contumelious, in Roger B's “Keep on the Carcosan Borderlands” game marked my 30th session. That's about twice as many games as I seem to marshal per year face-to-face—in half the time. I've run about two dozen games in four settings with as many players revolving in and out and had the fortune to play in eight other worlds.

Since I set out with the explicit goal of learning more about the art of running a game from the experience, it's a lot to wrap my head around. This post is the beginning of that process; it laser focuses on one set of lessons from the perilous art of GM improv.

I would hazard a guess that every GM worth their salt recognizes the need for improvisation. My observation is that not everyone raises from it though from a necessary evil or a stop-gap measure to an art. I don't claim to be an expert, but here are some points I feel like I am starting to grok about that art:

Explore Improv Theater. One of my legs-up as a GM is that I did improv in my late teen and early adult years. While there are a number of techniques that are not at all useful to being a GM—and some that are wholly unique to a game (see points below)—it doesn't hurt to at the least check out a few online sources. The goal is to increase your comfort level with it as a conscious exercise.

Yes and... Stop worrying and embrace player co-creation of your world. Take a small detail offered by a player and embroider the hell out of it.

Last night my huckster of a dwarf decided that since dwarves were rare in Roger's world that rubbing the forehead of a dwarf would be a luck-producing superstition that little folk would pay good coin to do. Instead of nixing such a self-serving detail, he went with it and then added his own details as the hustle played out. This is the art of “yes and...” that you will see repeated over and over in theater improv lists.

If something rubs you the wrong way, try not to veto it outright. Try to run with it or at least twist it during the “and...” part to be a further interesting contingency—or challenge.

Often if you find yourself stumped for an improv detail, try pushing the question back on the player--and then fall back on “yes and...” embroidering. “What do you think the Sun Lord priests would say in such a situation?”

No But... Later when you have mastered “Yes and...” start using a variant called “no but...”. Here again you aren't flat out saying no full stop, but are still riffing off their suggestion. I would use this sparingly, experience has taught me the more you embrace “yes and” the more your players will feel free to turn on the tap at the table with those details. 

Hone in on Detail. For every NPC, setting description, magic item, etc. you pull out of your rear, try and attach 2-3 concrete, free-associated details about it from the get go. Don't just think a “sword”, think “a sword that is umm...bent badly at the tip...and a hilt wrapped in some kind of skin.” Let the players explore those details, use that exploration as a stall and start to fill in the blanks of how and why such and such thing is that way. Usually their questions will lead you to the answers even if they are very different from their assumptions.

Bag of Tricks. The first thing out of many GMs mouths when it comes to improving is random charts. It's a good point, but I take it one step further. I keep a spiral notebook with stock lists. Whenever the mood hits me out of session, I will make a number of rolls using those charts. I then free associate quick details to spice it up or give it a likely set of game contexts. Keep it sketchy, the point is to have little pieces to jog your memory for the point above when that wandering monster roll tells you something terse like “Newhonian Ghouls 1-6”.

Visualize It. I have made this point before in talking about the places we “see” when using our settings. Flip through relevant images before the game. It can be ones you have collected for your campaign,  or it could simply be favorite inspiring ones in print or on online. Don't spend a lot of time fixating on it (this defeats the purpose) but just soak through it for 10-20 minutes.

Make it Perfect, Then...There is one big pitfall for GMs that value player agency in regards to making things up at the table: the need for meaningful, informed choice. If you want a game to be truly off the rails, don't get hung up on pushing that brilliant thing you just came up with right then and there. Even if those three details of that wicked letch of a priest sound awesome to you, don't be afraid to let it go if the players have zero interest in interacting with him. Similarily always provide at least a few routes or solutions to relevant things you improv. 

If you become good enough to shake one thing out well in the moment, chances are you will be just as good making up more. Let it go. 

Write it Down. I have learned this one the hard way, if you make up a detail on the table, damn well make sure to write it down. Consistency is all important and if situations become by default arbitrary it undermines both the quality of your campaign and the point above about player choice. 

 Don't break the flow of play, but write down a line or two about it right then or close to then (player bathroom or snack break is always a good time). After the session fill it out if necessary—and above all  build off it between sessions.

Practice. Don't just pull out when the players wander off the reservation, try to work in some practice at the table. Nothing teaches something quicker than experience, so when you feel comfortable with it start with a handful of likely situations that you can try out your mad new skills. Say you know they are going to be exploring a vine-choked dark forest for the first time. Don't fill out all the encounter sites, leave two or three blank. Maybe have a sketchy idea of what's there, or none at all.

Use your techniques and see what you learn from it. Do I need to bone up in this area or that? Ask your players for constructive criticism even.

There are points and lessons I am quite sure I have omitted or never picked up on. Drop a comment if there is something you feel like you have picked up on about the art yourself.

And if you are still around the Hill Cantons later today, wait for the set-up for my related (and likely absurdly silly) stunt-challenge.


  1. Good points all, I was also an actor in
    high school and college and to add to that
    I used to work for the government so I'm "absurdly silly" friendly...)

  2. I liked how you tried to improvise your companions into coinage by telling the guards to let you into their keep because you had slaves to sell.

    1. It was a win-win situation for Xhomar really. Lucky for the rest of the party they let us in without your bondage.

  3. I'm lucky in that everyone in my 8 strong real life group is in professional or amateur theatre. This is also why half the group want to GM... Bloody directors! :D

  4. I'm going to try and do better on the theatrics. Zak rightly pointed out that my Griffin Mountain game sounds like what would happen if everyone in the world died out apart from the Cockneys (though if you have been to Balham you see that it is not so far from the neolithic world of Balazar to modern South London). I am therefore deciding that while Trilus is definitely still just east of Woolwich, Elkoi is going to be Birmingham, Dykene is in Cornwall, the Sartarites are Scotsmen and the Lunars are 'posh'.

    Key to GM improv, IMO is know your game world. Self-consistency helps the illusion of being in an alternate reality no end, in fact it IS the illusion. If you make some random detail up one session (Goblins use their own teeth as currency, hence the frequency of fist fights trying to knock them out) you can't go back on it in another. You cannot know ALL the details of your world, but you need to know how it 'feels' so that it never takes more than a couple of seconds to come up with an answer to fill in a blank.

    1. "Self-consistency helps the illusion of being in an alternate reality no end, in fact it IS the illusion."

      That's a key point, hinted at obliquely in the "write it down" point.

      Certainly the longer I run any setting the easier improv becomes in general , not just because of comfort but because I think I start to slowly understand the dynamics and logic behind it with each session.

      And definitely my homebrew settings are easier for me to improv in than the published ones--even those that I have a relatively high and silly amount of canon detail mastery of--because I've been forced to think about what makes it tick deeper.

      I would imagine that the big name world creator GMs like Barker and Stafford find it pretty easy to pull out spontaneous things at the table based on how it would logically "feel" in their worlds.

  5. There are two major ways that I practice improv.

    1. Run games for kids. I especially like TOON since it has no limits on what kind of character can be made. You simply cannot prepare for a game with kid players. It will be a constant barrage of the unexpected.

    2. Be lazy and don't do shit to prep for a game session, then just pretend like everything's normal and run it anyways. Get some meaningless paper in front of you with meaningless stuff on it to help pretend like you are actually going off notes. This can be stressful at times, and you have to be in the right frame of mind, but amazing games can come out of this as well. I recommend everyone do this once in a while.

    1. I have done two a few more times that I normally would care to admit. In fact a whole mini-dungeon, the Frog Demon Temple, was generated with nothing more than a few rolls on Jeff Rient's Frog Demon chart in Cinder for guidance.

      Agree that every GM should try it at least once as it really cranks up your thinking on your feet level to 11.

  6. Great post.

    One of my biggest problems is I can't keep secrets and I'm too honest. So, I always (well too often) reveal I just made shit up, or that I'm unprepared, etc. Which really spoils it.

    1. Why would that spoil it? I find just the opposite - it lets the players know they are free to contribute and that their every action moves the world in some way.

  7. Good, solid advice. The part about writing it down is key; a hallmark of good improv is incorporation, or reincorporation, making sure that details are relevant later on in the campaign. Especially if you're playing the Yes, And game and things are being created whole cloth. Wifey acts and directs improv shows (time permitting) so I have no lack of practical advice on the home front.

    Improv does present an interesting phenomenon; I've seen plenty of old schoolers imply that a world made-up on the spot via improv (or player incorporation) is too new school, and then pull out a sheaf of random tables and proceed to generate content on the fly. So there's funny ideas out there about what's actually going on.

    It pains me to say it, because I grew to loathe everything else about the system, but the 4E DMG had lots of good improv advice (and the 4E DMG 2 followed it up with more sophisticated techniques).

    1. Looking through my notes it's shocking how much of the Hill Cantons campaign dress was created by "yes and". We riff a lot at the home table based on the campaign logic (or lack thereof), absurd religious doctrine in particular.

      "I've seen plenty of old schoolers imply that a world made-up on the spot via improv (or player incorporation) is too new school, and then pull out a sheaf of random tables and proceed to generate content on the fly"

      I've run into the same. I have started to come to the realization that there are two souls (camps?) of people attracted to old school: 1. those who were attracted to recreating the free-wheeling, experimental, rules-lite feel of gaming back then, and 2. those who have some kind of ossified neo-con image of a game frozen in time.

      Fooey on the latter.