Spooky maskWhen running a most improvised scenario (either something as ambitious as the Dracula Dossier or just riffing off a paragraph or two of notes), One Useful Trick is to have a copy of the investigative ability list for your game to hand, and check off abilities as you call for them or the players use them. That lets you see at a glance which abilities you haven’t yet used in play.

Then, look for opportunities to bring in other abilities. Treat it as a prompt, a challenge – “what’s the easiest narrative route in the game from this moment to the action hinging on Art History or Pharmacy or Flirting?”

Often, in improv play, you fall back on the sort of scenes that you’re most comfortable with; I can riff mysterious murders, spooky locations and sieges off the top of my head, but need to remind myself to do interpersonal scenes, crowds, or car chases.

Prompting yourself to bring in abilities you don’t instinctively default to is a great way to vary the scenes in your game. The players in my current Night’s Black Agents game, for example, are much more comfortable hanging back and observing, either by blending into the crowd, perching on rooftops, or getting full value out of all those points invested in Data Retrieval, Electronic Surveillance and Digital Intrusion. Tracking the abilities used reminds me in the heat of play to put in more interpersonal scenes, forcing them to use messy touch-feely abilities like Reassurance or Intimidation.

A neglected ability doesn’t have to be central to the game, of course. If you’re trying to bring in, say, Astronomy, you could just mention that the characters knows offhand that tonight will be a moonless and especially dark night; often, reminding players that they have a particular ability will start them thinking about ways to use those assets.

Don’t neglect General Abilities, either. If no-one’s used Cover or Disguise in a while, try to drop in some obstacles that require those abilities.

This post originally appeared on DyingEarth.com between 2004 and 2007.

A column on roleplaying

by Robin D. Laws

Like most creative endeavors, the GMing craft comes with its share of eternal conundrums. One classic example is the question of whether you carefully prepare adventures, or improvise them in response to player choices.

Carefully prepared adventures risk the accusation of railroading. In this particular application of a term I find annoyingly broad, the GM must be careful not to create the impression that anything that happens is predetermined, or that the players have anything less than absolute freedom of choice at all times. By its very nature, a prepared scenario can’t anticipate every possible branching action leading from a single plot premise. Otherwise even the simplest adventure would hit the table at the approximate weight of a telephone directory, and would consist almost entirely of carefully written responses to choices the players never wind up making. The more creative and surprising the choices made by the players, the more a prepared scenario becomes an improvised one during play.

On the other hand, if you’re completely making it up as you go along, some players (not, one hopes, the same ones who complain about railroading) find it more difficult to engage in the willing suspension of disbelief necessary to the enjoyment of any fictional presentation. If they can see you making it up on the spot, the game becomes less “real” to them. For players of this school, the game world is only genuine to them if they can believe that you’ve made certain immutable decisions and that certain of their choices will always produce the same results. Their desire for a sense of a bedrock reality behind the scenario persists regardless of your immediate need to adjust such factors as pacing, dramatic contrast, or degree of challenge.

In an investigative scenario, like one for any of the GUMSHOE games, you’ll generally need to designate certain facts as immutable from the outset. You’ll want to start the game already knowing a good deal of backstory, specifically who committed the act under investigation, how they did it, and why. (This is assuming that you’re not playing in a more avant garde mode, in which the players, acting as collective co-storytellers, help to collaboratively determine all these facts as they go along.) You can’t work out which clues might be available, even in improv fashion, if you don’t already know what facts the clues will eventually point toward.

For groups especially sensitive to the thought that you’re making it up before their eyes, a number of techniques allow you to convincingly fake it.

(Yep, I’m once again advocating a series of GM techniques which to a small extent deceive the players. If you find the entire idea of this scandalous, you may also be shocked to learn that there is gambling going on in the casino. Any author or screenwriter at all interested in the basic pleasures of narrative is to some degree a magician, relying on misdirection to eventually surprise and delight the audience. Just because players in an RPG take on pivotal duties that in other story forms are the sole province of the author doesn’t mean that the GM shouldn’t occasionally trick them into greater enjoyment.)

Several “tells” reveal to any halfway savvy group of players that you’re relying on a heavily prepared adventure. Disguise your improvisations by displaying these same tells.

Most notably, a prepared adventure takes an obvious physical form, as a sheaf of notes. To appease improv-averse players, create a fakebook. Use old notes from another adventure, perhaps with a new title page to keep it looking fresh and free of dog ears. Write a new title for the scenario, set to a high point size, so your players can read it from across the room if they “just happen” to glance at it. (For additional misdirection points, use your title for foreshadowing purposes. Choose an adventure title that creates a set of expectations, and then fulfill those expectations in a surprising way. A scenario called Darkness At the Bottom Of the Well might encourage your players to investigate an actual (and dangerous) well, when your real reference point for the title might refer to a book title, the name of an Internet forum, or your tale’s (entirely metaphorical) theme.

Refer to your fakebook throughout the adventure, especially when new scenes arise. This gives you something to do when forced to improvise your way through a situation that has you momentarily stumped. Don’t take as much time as the proverbial bad GM who’s constantly referring to his notes for interminable stretches—just enough to maintain your illusion of preparation. Even if you’re on a roll and don’t need the creative breathing room, make sure to take the occasional glance at it, to maintain the illusion of limited immutability.

Other fakebook techniques require some advance work—though not nearly so much as fully writing up a scenario in quasi-publishable format. Make sure, for example, to have not only the names of the characters you’ll need to use, but also a list of other unassigned names ready to go. A list of street names and business establishments may also prove invaluable Realistic sounding names are tough to generate on the fly, and are the deadest giveaway of an improvising GM.

Conversely, the most notable tell of the GM running a prepared adventure is the periodic break to read aloud sections of text. Personally, especially when running a published scenario, I find this technique way more disruptive to the fictional illusion than the notion that the GM is improvising. However, the same folks who get restless when they sense the GM is making it up may derive comfort from these canned textual signposts, which indicate that everything is still safely on track.

Ready yourself for this additional level of trickery by writing free-floating passages of text which can be dropped into any scenario. Descriptions of people are the most versatile, because you can assign them to characters who might pop up in any adventure. If you don’t wind up using a bored security guard, old coot watchman, or foxy librarian in the present improvised adventure, you can hold them in abeyance for a future installment. Because, like any improv whiz, you’re trying to minimize your prep time, you can keep these suitably short and sweet, avoiding the trap of the overlong text block.

Misdirection requires you to know your own habits regarding prepared text, and to duplicate them when improvising. Though I always try to paraphrase any prepared text, I often find myself at least half-reading passages from scenarios I’ve prepared. If you read lengthy passages as is, do the same when faking it with a free-floating text block. If you’d paraphrase all text snippets in fully prepped adventure, replicate that habit.

Ambitious fake improvisers can find further ways to mimic the behavior of a well-prepped GM. If your pre-written scenarios include hand-outs, create some free-floating maps, notes, and diagrams to fold into your plotline as you develop it. If you borrow images to represent people and places, keep a pile on hand for the same purpose, and so on.

Keep at it, and eventually you might convince even yourself that you’ve prepared!

This article originally appeared on DyingEarth.com, between 2004 and 2007. You can find part one here.

A column about roleplaying

By Robin D. Laws

Last month we plundered the gilded halls of improv theory, appropriating for our own roleplaying purposes the “Yes, but” technique. GMs using this technique avoid answering player requests with a categorical no. Instead they look for ways to say yes, but with complications that preserve the coherence of the setting, add additional challenge, or both.

This time we’re going to take the concept to its funky extreme by using it as the basis for an impromptu scenario. Try it next time you’re forced for whatever reason to slot in a fill-in event for your ongoing game, or as a convention brain-teaser.

“Yes, but: The Scenario” works best with a freeform resolution system that allows character creation on the fly, preferably with simple or self-defined abilities. I’ve also run it using just a deck of cards as a resolution system, with a high draw meaning a good result, a low card indicating failure, and an ace indicating that the player gets to dictate the ideal result of his action attempt. However, if you’re the kind of GM who can spreadsheet an exquisitely balanced Champions character in your head, you might prefer to rely on a crunchier rules set.

This scenario is more fun and unpredictable if the rules system you choose triggers comparatively few assumptions about world and expected game play. If you haul out the D&D rules books, your players will likely plug themselves into a well-worn pattern and set about performing that game’s default activity, relying less on their own improvisatory creativity than on an off-the-rack set of roleplaying assumptions.

You can start a “Yes, but” game mere moments after your players get settled in. Game play is character creation.

Inform your players that this game depends on their ability to interrogate you. All communications with you must be phrased in the form of a yes or no question. When given a yes or no question, you may elect to supply more information than the query calls for. If given a question which cannot be answered with a yes or no, or a statement which isn’t in the form of a question at all, you will ask the player to rephrase.

Play goes around the table in a round-robin fashion. Players ask questions in turn sequence, one question per turn.

When you’re satisfied that the group understands the method of play (well, sort of understands — expect a certain degree of hesitant bafflement at this point), start play by pointing to the first player.

Expect even more bafflement. Prompt the player to ask a question. If the player can’t think of one, try the next one in the turn order. If everyone seems utterly stumped, start off with:

“You all wake up at about the same time. You’re in a room together.”

Then, once again, prompt for questions.

Soon, if not instantly, the players will see the open-ended game you’re playing. They’ll ask you questions like:

1. “Is it dark?”
2. “Does the room have a door?”
3. “Am I injured?”
4. “Is there anyone else in the room other than us?”
5. “Am I male or female?”

What you’re doing is allowing the players to define their characters, the nature of the scenario, and even the genre, by the questions they ask. The answer to all of their questions is either a simple “yes” or a “yes, but…” followed by a line or two of explanation that mitigates, modifies, or limits the facts their question has put into play. “Yes but” is almost always the most fruitful answer.

So your replies to the above questions might be:

1. “Yes, but there’s light coming from under the door, enough so you can faintly make out a light switch off to one side of it.”
2. “Yes, but it’s behind a barricade of broken furniture. Someone went to a huge effort to keep something outside from coming in.”
3. “Yes, but not seriously. Just a few scratches.”
4. “Yes, there’s a man in a trench coat. But he seems to be dead.”
5. “Rephrase the question.”

As you continue, the Q&A format will define characters, flesh out a setting, and define a goal for the PCs to achieve.

As players ask questions about their characters, you assign abilities and game statistics to them. Whenever an answer defines a character’s abilities, make a note of them, giving them game statistics as necessary. The first-mentioned abilities get the best game stats. Though courtesy or lack of devious imagination may prevent them from trying it, there’s nothing to stop players from asking questions that define other players’ characters.

Clever players will catch onto what you’re doing and tailor questions to their benefit. The “yes, but” format makes this, challenging, though:

“Do I have a shotgun?”
Yes, but no ammo.

“Am I super strong?”
Yes, but only for a few moments a day.

“Do I have the key to that door?”
Yes, but you know there’s a bomb on the other side of the door, wired to go off when a key is inserted into the lock.

Certain questions tend to foster weird or freakish results if you apply “Yes, but” to them. Unless you want a cast of hermaphrodites and mutant halfbreeds (not that there’s anything wrong with that), questions like “Am I male?” or “Am I human?” should be answered with a simple “Yes.” You control the freakiness level of the scenario both with your modifying descriptions, and by which questions you choose to answer with a plain “Yes.”

The default outcome is a scenario about people who wake up trapped in an environment without their memories. The amnesia option can be fun, as it mirrors the player’s attempts to piece together their characters by asking you questions. You can forestall it, though, by simply answering “yes” to the question “Do we remember how we got here?”

Likewise, the PCs generally wind up trapped by asking “Is there a way out?” Starting out trapped is a good way to foster cooperation between the developing PCs, but again you can vary the standard pattern just by saying, “Yes.”

If the players think they’re playing in a given setting, their questions will be tailored to it. They may invoke existing media properties anyway: “Am I a Brujah?” “Can I perform the Vulcan nerve pinch?” The “yes, but” protocol limits your ability to fight this, but so what? It’s not like anybody’s going to sue you for infringing their intellectual property. Expect the resulting adventure to surrealistically blend various genres.

At some point during the game, the Q&A will prove difficult to sustain as your improvised narrative gathers steam. Depending on how quickly your players catch on and how adroitly they manipulate the format, this may happen as early as an hour into the session, or very near to its natural conclusion. Usually it’ll happen at about the halfway point.

When this occurs, tell the players that you’re switching to a regular RPG protocol. Then play out the game as you would any improvised scenario, placing challenges in front of the players as they head toward an exciting climax that resolves the central problem they’ve established for themselves during the Q&A phase. This sounds like a tall order, but, assuming you can improv a scenario at all, you’ll find that the momentum you’ve established in the Q&A carries you along naturally.

Will next month’s column expand this concept into a screenplay suitable for a major motion picture? Yes, but those not equipped with alien senses will instead perceive a column on another subject, germane to roleplaying.

This article originally appeared on DyingEarth.com, between 2004 and 2007.

A column about roleplaying

by Robin D. Laws

Considering how focused roleplayers are on plundering and looting, it’s surprising how little stealing we’ve done from the world of improv. Like us, sketch comedy troupes use collective, on-the-spot creativity to make entertainment out of nothing. And they don’t even need d12s. Or whatever it is that we need when we say our hobby is like playing cowboys and indians, except with rules to guarantee that the dead stay down when they get shot.

Let’s hijack one of improv’s central principles right now. That fundamental principle is “never negate.” In an improv, you never merely cancel out another participant’s action. Imagine that you and I are performing an improv together. We’ve been given a location by the audience — a construction site — and that’s all we’ve got to work with. You start the skit by sitting down and miming as if you’re removing your lunch from your lunch bucket. Then you say, “Too bad we’re getting fired today, huh? And here I was, just a week away from retirement.”

Now, my mental wheels were already turning the moment I heard the words ‘construction site.’ I had a whole different direction I wanted to go in. I wanted to establish that we were merely amateur construction enthusiasts at construction worker fantasy camp. Maybe my idea was funnier than yours, but now that you’ve taken the lead, I can’t simply negate what you’ve done to clear the decks for my concept. The principles of improv forbid me from simply saying: “You are completely mistaken, Pete. We haven’t been fired at all. In fact, we are amateur construction enthusiasts attending construction worker fantasy camp.”

Instead, I have to set my thought aside and build on yours. Since I’m trying to be funny, I need to add a twist or reversal, or at least a set-up that my partner can turn into a joke. Such as: “Yeah, you kill one measly supervisor and they get all safety-oriented on your ass.” All of our mental prep work — all sixty seconds of it — is out the window, and we’re off in an unexpected direction, flying blind, creating in the moment. This process generates the energy and sense of surprise that makes improv seem funny — often much funnier than the exact same material would be if rehearsed it and polished into a finished sketch.

In the above example, I’m not negating you’re idea, but I’m not just accepting it and parroting it back to you, either. I’ve returned your serve while putting a new spin on the ball. I’ve said, “Yes, but.” Yes, we’re getting fired, but we deserve it — if anything, we’re getting off easy.

Few roleplaying game sessions present situations as open-ended as the very beginning of an improv sketch. There are game rules to take into account, PC backstories to keep consistent, and a certain amount of world detail and plot preparation you hope to preserve. Within these parameters, though, the ‘yes, but’ principle is a powerful technique to engage your players by rewarding their creativity while at the same time keeping them on their toes.

Let’s say you’re running a game in a landlocked fantasy nation with a vaguely ancient Bronze Age feel. A player building a new character, inspired by her recent purchase of the Pirates of the Caribbean DVD, really, really wants to play a pirate. Your initial response, based on the logic of your world and the prep work you’ve done, is simply to say no. It’s crucial to your geopolitical story arc that the kingdom be landlocked. That pretty much rules out naval piracy. However, you’ll have a much better chance of keeping that player happy, and having her contribute positively to the game, if you can give her part of what she wants. Say, “yes, but…”

“Yes, but in this setting the equivalent of the pirate is the bandit in the hills. The bandits in this world are the same unruly, rum-swilling outlaw types with stolen, ragtag finery and a perverse code of brotherhood you’re thinking of when you use the word pirate. But instead of attacking seafaring ships, they raid caravans from horseback.”

Maybe you hadn’t given any thought to bandits in your setting before now. Now you’ve allowed your player to help shape your world, by making your bandits into pirates with the serial numbers filed off. You’ve given her the feel she wants, while changing the details to preserve the campaign elements you need.

“Yes, but,” can be a useful tool during play, too. Is there a magic item shop in your fantasy city? You’ve decided that there isn’t. Not only do you find this gaming convention too ridiculous for belief, but you’ve also established that the city is ruled by a rapacious robber baron. If such a shop did exist, he’d surely have confiscated its wares long before now. However, when the players look for a magic item shop, tell them why, and then hit them with a “yes, but”:

“Here’s what you learn after a few minutes of asking around: there used to be a magic item proprietor in town, but the Black Baron absorbed its contents into his treasury. Now it’s run by one of his stooges, even though it hasn’t sold an item in years. If adventurers show up to sell something, the Baron’s goons confiscate their treasures and give them the bum’s rush out of town. If they show up to buy, the shopkeeper wheedles as much information from them as he can, then reports them to the Baron. The original owner fled the city and supposedly lives in the cave network by the river. He and a number of other exiles are looking for adventurers willing to aid in the baron’s overthrow. Rumor has it that he squirreled a few of his items out of town before the Baron’s tax collectors swooped in. Maybe he’d still be able to arrange a swap for you.”

Though you haven’t given the adventurers exactly what they want, you haven’t slapped them with a flat no, either. You’ve provided both a plot hook to follow up on, and a way of achieving their underlying goal (buying or selling a magic item) that doesn’t violate your own tastes or campaign logic.

“Yes, but” can, on the other hand, assist you in improvising additional conflicts challenges into what would otherwise be flat, uninspiring scenes of information gathering. Does the spice merchant know anything about the abduction of the high priest, the players wonder. You decide that the answer is “yes, but”: he saw one of the perpetrators, but will provide the information only in exchange for a favor: the adventurers must first forcibly persuade a decadent young noble to leave his daughter alone.

Though you don’t want to go overboard with side missions like this, the occasional instance can inject variety into your session — and also provide play opportunities for players who are more interested in butt-kicking, infiltration, intrigue or puzzle-solving than investigation.

The usefulness of this technique, however you choose to use it, stems from its origins in improv. It encourages you to add options instead of merely foreclosing them. Most importantly, it inspires you to think sideways before answering important questions, preserving surprise not only for the players, but for yourself as well.

by Bryant Durrell

The Yellow King RPG can be daunting for a Keeper: four different settings, potentially lengthy campaign arcs, and to top it all off the canonical kick-off asks the Keeper to improvise connections between the Deuced Peculiar Things invented by the players and her carefully crafted scenario. I’ve always struggled a little bit with crafting GUMSHOE scenarios because I tend to get stuck on the obvious Investigative abilities. Meanwhile, I know my players will want to use Painting from time to time rather than just relying on a series of clues revealed by Occultism and Research.

I recently kicked off a long-term Yellow King RPG campaign, so solving this problem was a matter of some urgency. I’d been running a lot of RPGs with extensive randomization tables recently, and I’d noticed that creating a random event table forced me to explore a wider range of possibilities. I thought that perhaps I could use that kind of forcing function to flesh out a scenario.

I started out by sketching out a mystery more or less as recommended in the rules. First, I wrote down the basic spine of an adventure: Hook, Development, Antagonist Reactions, Alien Truth, and Climax. Second, I needed a premise. My theme for the Paris era was masks. Since I wanted to get right to the meat of each era, I embodied that concept as blatantly as possible: I decided there were a bunch of thrill seekers with living masks running around Paris. By unwisely using certain Carcosan rituals, a savvy occultist could mold one of these masks into a duplicate of a living person. I expected to jolt my players into asking questions about identity and self.

From there, I mind mapped my ideas around the spine. I left the Hook blank, since it was going to be tied to someone’s Deuced Peculiar Thing. The mind map wasn’t terribly dense, as can be seen in the picture. I just wanted enough detail to hang a plot on. Some details were only hints; for example, I knew there was a sinister figure behind the masked thrill seekers, but I didn’t want to nail down the specifics until I’d seen the characters.

But this didn’t solve my core problem! If I just improvised a story around this skeleton, I’d wind up with repetitive Investigative ability use and bored players. Back to devising a technique that would force me to be creative.

Since my problem was failing to cater to all the Investigative abilities, I gritted my teeth and pasted a list of the abilities into the corner of my notepad. Then I started working my way down the list, adding a potential clue next to each ability. As one can see, I was meticulous for about the first ten abilities. After that I started to skip around a little bit, with an eye towards making sure I was covering a good range of abilities even if I didn’t add a note for each one.

This worked very well for me. While devising the clues for Art History and Painting, I wound up adding a potential Vermeer theft subplot. Coming up with a Sculpture clue made me think about how the masks were created, and the source of clay for Greek ritual masks wound up being a key pipe clue. The Natural History clue — a cat with a dog’s face — wasn’t a pipe clue, but it’s a great bit of cosmic creepiness that was guaranteed to disturb pet-loving characters.

When I sketch out my second scenario, I’ll also note which character has which Investigative ability. That way, I can balance my potential clues across all the players and reduce my creative workload a bit. I’ll also balance clues across the three types of Investigative ability.

How did all this work out in play? As we created our characters, one player decided that his character Herbert had been seeing this weird person following him for months, both in his home country America and in Paris. That was perfect for the masked villains. Since Herbert’s Deuced Peculiar Thing only involved a single person, I quickly dropped the idea of a pack of mask wearers, although our villain had an interest in drafting the player characters into his service.

By the end of the scenario I’d only used a few clues from my sheet; every other clue I tossed in was either ad hoc or sketched out between the two sessions it took us to play through the scenario. (I didn’t work in the Vermeer, much to my regret.) However, the forced creativity exercise pushed my initial scenario design into places I wouldn’t have taken it on my own, so the scheme was certainly successful.

The lightweight skeleton combined with a rich set of prospective clues also had an unexpected effect of creating a dense feel to our campaign’s Paris. I had so many potential scenario directions, it was easy to improvise based on the direction the characters went. I want to capture the classic blurred line between the play and reality; by treating one player’s Deuced Peculiar Thing as if I knew it even before he’d said the words, I made good progress in that direction.


Bryant Durrell makes a living keeping servers healthy; in his copious spare time he watches wrestling, writes, and pretends to be fighting orcs. You can find him on Twitter as @bryantd and he blogs (rarely) at Population: One.


The Yellow King Roleplaying Game takes you on a brain-bending spiral through multiple selves and timelines, pitting characters against the reality-altering horror of The King in Yellow. When read, this suppressed play invites madness, and remolds our world into a colony of the alien planet Carcosa. Four core books, served up together in a beautiful slipcase, confront layers with an epic journey into horror in four alternate-reality settings: Belle Epoque Paris, The Wars, Aftermath, and This Is Normal Now. Purchase The Yellow King Roleplaying Game in print and PDF at the Pelgrane Shop.

 

Improvising With GUMSHOE by Steve Dempsey

logogumshoeThis article discusses an improvised variant of the GUMSHOE rules. It can be just as easily used for Fear Itself, Esoterrorists, or any other GUMSHOE game.

Most games of GUMSHOE are played using a scenario that the GM has written. Not only does she introduce each scene and play the non player characters but she also decides in advance what the clues are. Although the GM does not dictate the path the players will take through the adventure, she has a strong hand on the tiller as the clues she chooses will determine to a rather large extent what the players do.

There are some good reasons not to always play this way. Stephen King says in On Writing, “I distrust plot for two reasons: first, because our lives are largely plotless, even when you add in all our reasonable precautions and careful planning; and second, because I believe plotting and the spontaneity of real creation aren’t compatible.” When you tie this in with the GM’s creed, “No scenario ever survives contact with the players”, you will see that the improvised game has some advantages over one written by the GM.

What you might lose on intricate plotting you are likely to gain on player involvement in the creative process and character play. Players will be much freer to take the scenario in directions that seem more natural to them and their input will have a greater impact on the story.

Improvisation is nothing terribly difficult to do, the main impact of playing this variant is that the game is not planned up front by a GM but is developed in play by players and GM alike. This means no prep for GMs, other than learning the rules. I’ll be discussing the details of how to do this in three easy stages. Finally I’ll give an example that shows how this works in play.

1. The set-up

As with any improvisation, you have to have a theme. It’s an improvisation on something. If you don’t have a theme, then the players won’t know what kind of characters to make.

So start with a theme. It doesn’t really matter how you come by this as long as there is some consensus within the group. You could let the GM choose (“You’re all students at a Japanese high school, getting ready for a school trip”) or you could have a group discussion about what sounds cool (“I want things lurking in doorways”, “I want magical rituals that take years to cast”, “I want a scene in an 80s disco”). You could also choose something that relates to a moral question (“How far are you prepared to go to stop the monsters?”) or a dilemma (“Family or Job?”).

But remember that this is GUMSHOE: Fear Itself, Esoterrorists, Trail of Cthulhu, and Mutant City Blues. It’s all about investigation. Some terrible crime has been committed, the bastions of reality are under threat, and the characters are the ones to deal with it.

For your theme you should also discuss the nature of this threat or crime, even if you don’t want to know the details at this stage. For example, the Japanese schoolgirls are a shoe-in for some kind of mad slasher and the 80’s disco idea smacks of Son of Sam or Zodiac.50423

You could discuss who the villain of the piece is going to be. This could be oblique (some Mythos deity) or explicit (one of the schoolgirls). It helps the game if you have some idea of what you are aiming for. It should also help with pacing. You don’t want the bad guy to be revealed to the characters in the first five minutes.

It’s a good idea, although not necessary, to write down the outcome of your discussions regarding the theme. It’s a handy resource for players and GM alike who can refer to it when making decisions about characters or plot.

Once you know what the theme is, make up some characters. In many games, this is down in utmost secrecy lest anyone steal your cool idea. In improv, we have a different way of doing things. You all do your characters together. Talk about your characters to each other and say when you like something. Give positive feedback.

Improv thrives on feedback. You are the audience as well as the actors so big yourselves up. It’s not just about getting a good vibe, this is also about riffing off each other’s characters. If you’ve gone the schoolgirl route, you’ll need to know who is the class swot, who is the cheerleader and who has psychic powers. You’re characters don’t necessarily need to know, but your players do. You need to know where conflicts will arise because that’s what makes the game interesting.

You can do this by each introducing your character once generation has been done, but that’s a short cut that misses out the links that you can forge between your characters if you do the job collaboratively.

In improv GUMSHOE, investigative skills work differently. They still allow characters to automatically find core clues or to be spent on supplementary clues. That much does not change. However, because there is no prewritten scenario, the choice of skills determines what the characters are going to encounter. If no one has Art History as a skill, the characters aren’t going to be looking at many paintings. If they all have high trivia scores, then what happened in last week’s episode of Full Metal Alchemist is going to be much more important.

Decide how long you want the game to last. This can be done by deciding on the number of core clues. One is generally not enough but you can play a decent one session game with only three or four core clues. Don’t forget that some scenes will not be about clues but for transition or colour. Whilst you might like to go for a mammoth ten core clue game, this is probably a bit much and I imagine is best broken down into smaller three or four clue episodes, each with their own internal logic but all building blocks in the greater plot arc.

2. What do we do now?

Now you play. Without any kind of pre-existing scenario this sounds a bit scary but you do have something to go on, namely all the work that you’ve put in so far to create the theme and the characters. You should all have a pretty good idea of how the general direction of the game so now what you do is ask for scenes.

Anyone can ask for a scene, player or GM, but the GM gets to decide the order in which they are played. The first scene is usually called for by the GM who will use it to introduce the game, the characters and perhaps something about the mystery that’s about to be investigated.

A scene is where a least one character will attempt some kind of action. An action is where a character finds a clue, has social interaction with a PC or NPC or uses their general skills to some end. It’s a fairly loose definition but you’ll know one when you see one. For a scene to work it has to have some kind of danger, excitement, threat or drive the plot of the game.

It’s the GM’s job to set-up scenes and to play NPCs. They can take account of player wishes but ultimately it is there responsibility to decide who and what is in the scene.

It’s also the GM’s job to make sure that transitions between scenes are handled. This is essentially narration. It’s the bits in 24 that happen during the ad breaks when Jack Bauer drives to the next action packed scene, or at the start when the voice says “Previously on Heroes”. Transitions are important because they tie everything together. They can also have bits of exposition such as when a PC talks to his critically ill wife in hospital, flashbacks to a scene in the life of the villain or even foreshadowing of future events. The extent to which you expose plot to the players in these scenes is very much up to the will of the group. Some don’t want out of character knowledge but some relish the TV show style construction that has interposed shots of the bad guy committing his latest dastardly crime, think Skylar in Heroes.

3. How to improvise 

Here are some techniques that you can use to help with your improvisation. If you want more information on improvisation for roleplaying I recommend the Play Unsafe by Graham Walmsley).

These techniques are not difficult to use and they have been shown in theatre sports (see Impro by Keith Johnstone) to improve stories generated through improvisation.

Don’t try to be too clever

If your character goes into a bar for the first time, they should probably order a drink, they probably wouldn’t do a back flip over the bar and shoot the pianist. If you do this kind of thing, you ruin spoil the narrative by doing things for which the other players can’t see the justification. Characters should act in character and do what’s natural for them to do. You’ll find that acting naturally helps the game along much better because the other players will come to know what to expect from your character.

Don’t block

This follows on from the first technique. You won’t be able to understand what the other characters are like if you try to block everything they do. So if a character proposes going into a bar, you probably shouldn’t say “It’s closed” or “I don’t go in bars”. It’s fine to say, “Well, I wouldn’t usually, but just this once”. In fact this is very good because this reveals something about your character as well as encouraging the other player’s development of the game.

Reincorporate

Build on what’s already happened. If an NPC gets mentioned by name in an early scene, bring them back later on. If a detail is mentioned, make it appear in a later scene under a different light, make it more or less important than it was. The reason behind reincorporation is because it reinforces the narrative by drawing attention to the salient points.

Reincorporation is also known as Chekov’s Gun because he once wrote in a letter to a friend, “”If you say in the first chapter that there is a rifle hanging on the wall, in the second or third chapter it absolutely must go off. If it’s not going to be fired, it shouldn’t be hanging there.”

4. An example of play

So here’s an example. Brenda is running a game for Steve and Anya. They decide that they want to play Fear Itself set in the London in the 70s. The player characters will all be involved in the punk scene, the tone will be gritty and the game should involve some kind of parasitical infection.

Steve’s character is called Sanjit, a fanzine writer from Bromley. His writing has some influence in the small milieu but he’s not necessarily well liked, mainly because of the sarcastic tone of his writing. He’s unemployed.Scene from The Third Man

Anya’s character, Ariadne, has come down to London from Birmingham, to escape from Heavy Metal. She’s a competent drummer and has got a gig with a band called Dole Kids. Ariadne and Sanjit share a grotty room in Berwick St.

Brenda thinks that the plot probably involves something to do with some chord progression carrying the infection but that’s not something she can decide. But it is her job to frame the first scene . Given the theme, there’s nowhere better to start the game that at a club. (This is Not Being Too Clever .)

It’s a Friday night and the Dive, a club in Camden pub basement, is heaving. The floor is sticky with beer, the walls and ceiling dripping with sweat. Dole Kids are just coming off having done a decent set. Sanjit is in the off-stage area having a discussion with Molly, lead singer of Kick in the Head who are due on next. Molly has taken umbrage at something Sanjit wrote in his last fanzine. Her band is on stage and waiting for her.

The scene is played out to introduce the characters and any NPCs. From what happens it’s clear that Molly will feature later in the game. On this occasion Molly storms off up the steps to the stage barging into Ariadne. This only escalates the arguments. She spits at the group and she runs up to sing. They follow her and end up being beaten up by Kick in the Head and their loyal following. Molly takes pity on the PCs and gets them back to her dressing room where they share a joint.

Next, Anya calls for a Core Clue scene. As there hasn’t been anything horrible happen yet, this scene should introduce the first elements of horror. It’s probably time for someone to die.

Anya asks for the scene to take place at the after show party. Brenda sets the scene but allows the players to place their own characters. It’s after the gig at a party in a squat next to a kebab shop. There is no electricity in the building and it’s entirely lit by candles. Someone has a grotty tape player which is blasting out the rather indistinct sounds of Iggy Pop and the Stooges.

Anya says that Ariadne is snogging some groupie in a wrecked bathroom, candles reflecting off broken bits of mirror. Steve decides that Sanjit is holding forth in a damp and grimy kitchen to a small coterie of fanzine fans.

Brenda narrates what happens next. Suddenly a scream comes from upstairs. A teenage goth staggers into a stairwell, his face contorted in horror. He collapses and falls. People run up to see what’s going on. As Anya called the scene, it’s up to her what the clue is. She can take suggestions from the other players. Ariadne comes out to see what’s going on and uses Intimidation to get everyone else to back off so she can get to the clue. Anya says that the boy has passed out, he’s got a joint tightly clenched in his hand. Ariadne checks him out and takes the joint. (This is Reincorporation of the joint.) She goes to take a puff but just before she does, notices something strange in the joint. Brenda suggests that this might be some kind of small wriggly worm, and Steve adds that perhaps as Ariadne is leaning over the boy she notices something pass across his eyeball, although it’s not clear what.

wormAnya decides to go with the wriggling worm in the joint. Steve also decides on a supplementary clue, spending a point of Streetwise, he decides that Sanjit knows the unconscious lad. He’s a pagan called Perdition,also from Bromley who Sanjit knows is into some “heavy magic shit”.

Brenda narrates what happens next. Perdition wakes up with a start and looks around. He smiles strangely and attempts to kiss Ariadne. He is superhumanly strong but together they manage to force him outside. He chases after someone else. Everyone else has run away at this point, except for Molly, who announces “Oh my God, I’ve got the same dealer as that monster!” She gets out her weed and it too is infected with worms.

We have a plot! Everyone has smoked the infected weed, who knows what might happen to them now? The game will continue long into the night.

You now have some tools that you can use to improvise games. If you give this a go, remember that a light touch is often needed with this kind of game, don’t go trampling all over other people’s ideas, give them space and time to come to fruition. It’s a question of mutual respect.

Finally, the improvisation may well not work at all. You might find that you’ve painted yourselves into some kind of dead-end story. But don’t worry about it. Improv, like any other game technique, doesn’t always work. The thing is not to worry to much about this and to just try again from a bit before when things started to go off the rails.

With a bit of patience, you’ll seen be off again.

Steve Dempsey, the author of this piece, has written for Armitage Files and Dreamhounds of Paris, and is our most experienced GM.

GUMSHOE is the groundbreaking investigative roleplaying system by Robin D. Laws that shifts the focus of play away from finding clues (or worse, not finding them), and toward interpreting clues, solving mysteries and moving the action forward. GUMSHOE powers many Pelgrane Press games, including Trail of Cthulhu, Night’s Black Agents, Esoterrorists, Ashen Stars, Mutant City Blues and Fear Itself. Learn more about how to run GUMSHOE games, and download the GUMSHOE System Reference Document to make your own GUMSHOE products under the Open Gaming License or the Creative Commons 3.0 Attribution Unported License.

(originally posted in Dec 2007 Page XX)