Two weeks ago, I was running a session for my completely new group of players. It was a FIRST SESSION. Experienced GMs know EXACTLY what that means. Basically, I had a bunch of PCs with no ties to each other barging around a situation on their own like a bunch of dumba$&es while I herded them toward an inciting incident that would unite them in common purpose so that we could actually start playing. Actually, I don’t normally do this kind of FIRST SESSION because it’s always a mess. But I wanted to give the newbies each a little bit of breathing room so they could poke at the world and fiddle with their characters before I brought them all together by a disaster. Because I don’t do small scale incident incitement. If I incite an incident, I incite the hell out of it. That’s it. Incidents I incite will not be ignored.

So, I had six water buffalos barging around this town, pooping on things (because that’s what water buffalos do), and I was just throwing things at them for them to interact with. I was switching from player to player like the first episode of any show with an ensemble cast, giving each of them some interactions and things to f$&% with. They were each having fun with the novelty of basically being able to do anything in a fantasy world, so I let it go on. Once the players started to flag a little, I would incite incidents at them and get the story moving.

I made a mistake though. One PC had arrived in town on the path of a killer. Actually, two of them had. And I had planned for them to run into each other during their investigations and realize they were both tracking the same person and that each of them had half the story and start comparing notes. Meanwhile, the lead on the killer would – JUST BY COINCIDENCE – also be related to the berserker goblins and a mysterious symbol and all the other crap that the other PCs were involved in.

Except the OTHER player skipped the session. Something about a personal emergency and a f$&%ing funeral or something. I don’t know why players think I give a s$&% what their excuse for missing a game is. They might as well just call me and say “hey, I’m not going to be there because f$&% your game.” Because that’s what I hear. So the first player was flapping around with half the story. And the PC was very purpose driven. So I gave her a lead early on that the killer had gone off across the wilderness in a specific direction early one morning and that it was too late to set out and the person offering the lead even offered a room for the night so she could go after him in the morning. But she would not be cowed. She cared nothing for anything except going after this guy. I just needed the character to stop for two f$&%ing hours (in GAME WORLD time, not REAL time) so I could incite my incident. So, I locked the town gates to travel (because a big ceremony was starting and all guards were to be on deck for the commemoration of a new temple and so the gates would be unstaffed) and I sent her on a wild goose chase. But she was clearly reaching the point where she might just climb the walls or break down the gate or punch a guard until he let her out of the town, so I needed a distraction. I threw a fortune teller in front of her. I figured I could do some cryptic thing to calm her the hell down for just long enough to get all the other ducks in a row. After all, EVERY GAMER is a sucker for a fortune teller or an oracle or whatever. Seriously. Try it. Have a mysterious looking fortune teller approach a member of the party in a firm, insistent, but mysterious manner. Players assume all fortunes are clues or foreshadowing or hints to magical treasure or something. They can’t ignore a fortune teller.

She shoved the fortune teller aside with a grunted “no” or “f$&% off” or whatever and then charged away.

F$&%ign players. Am I right?

But fortune tellers really are like “shave and a haircut” to a toon (“Who Framed Roger Rabbit”, Robert Zemeckis, 1988, look it up). The rest of the players were twitching for the entire session. The idea of leaving a fortune untold is really anathema to them. So, at the end of the session, the other players wanted to know what the fortune teller was going to say?

“I don’t know,” I said. “I was going to pull out these fortune cards I have and do a quick three card reading on the fly. I was going to make something up.”

Narrative Improvisation

Narrative improvisation OR improvisation of story is the other half of improvisation. Last week, we talked extensively about mechanical improvisation. Improvisation of resolution. That’s what we do when we find the rules are insufficient to resolve an action in the game and have to make some s$&% up. Narrative improvisation is when we add story elements to the game on the fly.

What is a story element? Well, pretty much everything in your game that isn’t a die roll is a story element. Characters are story elements. Setting details are story elements. Hell, any sort of backstory is a story element. Whole scenes are story elements. Even entire adventures are story elements. Basically, everything in the game that isn’t actual die rolling is a story element.

So narrative improvisation involves adding everything from new bits of backstory to entire new scenes to your game without having planned them in advance. Like halfling fortune tellers. Or the fact that suddenly that elf had a sister who died of something like fantasy rabies and that is why he stole a valuable potion from his noble family’s apothecary to give to one of the PCs because the goblins seem to also have fantasy rabies. That is ALSO a thing I suddenly decided happened in my game.

Narrative Improvisation is a Terrible Idea

Narrative improvisation is pretty much the worst thing a GM can do ever. I s$&% you not. I do it all the time. Other GMs do it all the time. Hell, the longer you run games, the more you tend to do it. But it’s the worst thing ever. It’s really, really terrible. And that isn’t just hyperbole.

First, I’ve talked a hell of a lot about the shape and structure of stories. I’ve talked about how a story is made of a motivation, a resolution, and a path that connects one to the other. I’ve talked about the importance of foreshadowing and telegraphing so things don’t come right the f$&% out of nowhere. I’ve talked about how importance consistency and verisimilitude are to allowing players to make good decisions and about how adventures and scene need to start with lots of decisions and then gradually focus in on an end point. I’ve talked about how our brains are hard-wired to expect stories to conform to a certain shape. If you want your game to be BOTH a good game AND a satisfying story, you need to juggle all of those balls AND also keep the game challenging but fair and make sure the rules keep working the way they are supposed to.

There’s no way around those things. Fortunately, the longer you run games, the more you get used to that s$&%. The more those things become second nature to you. And lots of GMs actually already start with a feel for some of those things. Most GMs are creative types who are driven by a desire to create the sort of stories, games, and worlds they’ve been ravenously consuming. Built into that consumption is the tacit implication that they have been consuming a lot of well-constructed crap. A newbie chef that knows nothing about cooking but has been a major foodie for years is already starting out with a good palette and a sense of what flavors work well together.

So, from that perspective, a GOOD game is a well-oiled, well-structured machine. Narrative improvisation is basically like trying to add extra features to a clock on the fly without breaking it. Seriously, imagine I gave you a beautiful cuckoo clock – an antique German one – and a pile of nails and wood and bits of metal and some tools and said “okay, now, add a built-in toaster.”

Second, I’ve also talked about how creativity is hard and how coming up with good ideas is like pulling teeth and how you’ve got to seed your brain and then pull ideas out slowly and arduously. Creating on a deadline is extremely difficult and it takes a lot of practice and discipline. You have to train your brain not to shoot down all of your ideas because they seem stupid. But, at the same time, every idea has to be vetted because you need a consistent mood and tone and feel. So you have to be willing to censor nothing and yet police everything.

And narrative improvisation is basically creating on a deadline of “immediately.” Or, at least, in the span of a bathroom break. “Here’s a cuckoo clock. Now, add a built-in toaster. You have five minutes.”

Why The Motherloving F$&% Would You Ever Improvise?

Given that narrative improvisation is a terrible idea, why would you ever do it? Why does every GM eventually do it? Well, there’s two reasons narrative improvisation comes up.

First, eventually your players will ask you a question about the world you are unprepared for or they will go somewhere or do something you weren’t ready for. Players ask some dumba$& questions. I don’t know why they would ever want to know about how certain gods conduct religious ceremonies or what neighboring kingdoms trade with this city. Players just love to ask bulls$&% questions. Moreover, players love to go in directions you don’t expect or to do things you weren’t planning on. They are just creative, bless their f$&%ing souls. Sometimes, they actually manage to find a clever way to approach a problem, like using their charisma and magic to organize bucket brigades to fight fires instead of fighting the goblin berserkers. Other times, they come up with stupid, stupid plans like trying to break out of a city and get away from the adventure. Narrative improvisation is how you handle that crap.

Second, as a GM, you might notice a problem while you’re running a game. The pace might feel off and players might be getting bored or barging around aimlessly. Or the players might need a break from one too many action sequences. Or you might be losing one or two of the players who prefer a different type of engagement from everyone else. As a GM, you become more and more cognizant of the body signals players throw out and you use them to adjust your game. That’s why running games online is something you settle for and tolerate, not something you seek out. There is no substitute for being the same room with the players. But you might also notice other problems. Like unclear motivations or missing clues that make it hard for the players to understand what’s going on. It’s not just about pace and structure. Sometimes it’s about plot holes themselves.

Third, when the players take actions, their actions have consequences. Remember that? It’s in, like, five earlier articles on my site. Actions have an outcome and consequences. And because the consequences are based on player choices (f$&% free will and freedom of choice) and they tend to extend beyond the action they are a part of, often they require you add or adjust game elements later on. For example, if the PCs bully too many shopkeepers in town, word might get around through the merchant guilds that the PCs are dicks. And you might want to show that off.

Fourth, as a GM, you might have a really cool, fun idea and decide to throw it into your game to see what happens. This is a really stupid thing to do. But lots of things are stupid. It doesn’t make them less fun. Ninety percent of my narrative improvisation falls into this third category.

How to Improv Good

Honestly, the reasons don’t matter. It doesn’t matter WHY you end up adding story elements to your game. You’re going to do it. Either because of your dumba$& players or because you, yourself, are being a dumba$&. The key is to do it WELL.

How can you do it well? Here’s MY awesome tips for how to pull off narrative improvisation.

Understand the Value of Good Ideas

I can’t tell you how to come up with good ideas. We’ve covered that ground before. Sorry. It’s just a thing you get better at. But you have to understand how unimportant good ideas are. They are seriously unimportant. Crappy ideas are fine. Or no ideas are also fine. Here’s the deal, if you don’t have an idea in your head for an element you want to add to your game, you’re probably doing reactive improvisation. That is to say, the players did something and now you have to respond. Maybe they asked you a question. Maybe they create a scene by going somewhere you didn’t expect. Or maybe you’re trying to fix a problem. In those cases, you really don’t need a good idea. Crazy as that sounds, all you need to do is do SOMETHING that isn’t terrible. Consistency, verisimilitude, and tone are ALWAYS a concern, but it’s pretty easy to do something that is in keeping with the way your world works and feels. Don’t have a flamethrower-wielding jester show up in your low-magic, fantasy, survival-horror game.

The thing is, if you just start doing SOMETHING (as long as you have a good feel for your world), it won’t be terrible. If the PCs go to meet with someone you didn’t expect, just start describing the person. Start with race, gender, age, and one line of detail about what they look like. Then, open the conversation. Play through the scene. If the players ask a question, just start answering the question and see what comes out of your mouth.

I’m totally serious. The thing is, the better a feel you have for your world and your game, the easier it is to just start talking and doing things without breaking it. But that’s a skill that only comes with practice. You have to do it. So, just do it. Just let your mouth poor s$&% into your game. Once your brain realizes what is happening, it’ll jump in to save your game.

Remember what General Patton said: “A good plan violently executed now is better than a perfect plan executed next week.” In this case, he wasn’t talking about violence. He meant decisiveness and swiftness of action. Doing it now, do it fast, and do it decisively. It’s better than doing it right.

Don’t Improvise When You Don’t Have To

Some GMs think they HAVE TO improvise. They thing they HAVE TO make up the entire story as they go. Otherwise, it’s railroading. This is a bulls$&% attitude spoken by morons. They don’t. You don’t. First of all, if the players aren’t forcing your hand and you don’t have an amazing idea trying to squirm out of your brain, don’t add new story elements to the game. It’s not broken. Don’t try to fix it. Honestly, this is true even if it IS a little broken. If the pace of your game seems a little off or some of the players are drifting a little bit, but you don’t have a good sense of what’s wrong or how to fix it, continue with your plan. It’s like driving a car. Sometimes, you have to ride out a little fender bender because steering will lead you overcorrect and careen into a guard rail and die in a fiery explosion. Eventually, the game will either get on track or you’ll be able to see how to fix the problem.

Some GMs take this a step further. They like to come to their games with the intention to improvise absolutely everything. They come with a few index cards, a handful of names, maybe a stat block, and nothing else. And then they brag about how wonderfully open-ended their games are and how much better this is than planning. These same GMs will tell YOU to do the same thing. Those dumba$%es are trying to ruin your game.

I’m not saying it isn’t possible to run a good game that is fully or mostly improvised. Hell, I’ve done it myself. I do it a lot, actually. But, you have to be good at running games first. And writing games. Because remember that improvisation doesn’t save you anything. It just means you have to do all the same s$&% under intense time pressure. The GMs that can pull it off without a lot of practice first are really rare. And the ones that can’t are running wacky dickaround style games that can f$&% right off or they don’t realize they are running s$&% games.

It’s going to happen. Eventually, you’re going to have to pull an entire session completely out of your a$&. You’re going to be unprepared and you’re going to have to run the session anyway. You’ll be too tired to prepare the session or you’ll procrastinate too long or real life emergencies will happen. And you won’t want to just cancel your game. But don’t make this your default state. DON’T. A well-planned game will ALWAYS be better than an a%&pull. And I don’t give a flying mouse f$&% what any other dumba$% says.

Improvise as Small as Possible

When you add a story element to the game, add the smallest possible story element that works for your purposes. Don’t try to reach beyond the scope of that story element. Yet. What do I mean? I mean if you need a combat, make it a combat with goblins or rats or a random beast. Don’t try to tie it to some major goddamned faction you’re inventing or connect it to game elements. If you need a lore detail, just answer the goddamned question and don’t go further. If the question is about wedding rites for the church of Melora, vomit forth some bulls$&% about getting married in a natural setting and pouring out a libation of wine and honeyed bread on the ground and then a big naked romp through natural terrain and stop there. Do not try to figure out the motives of a shopkeeper or guard just because the PCs are interacting with them.

Don’t invent entire scenes, encounters, or organizations if all you need is an NPC or a clue. TRUST ME. These things WILL happen. I will show you how in a minute. But keep your improv to the point. Speaking of which…

Know the F$&%ing Point

Adventures have motives and goals. Encounters have dramatic questions and sources of conflict. Scenes have a purpose. Nothing about improvising allows you to get around those basic facts. Nothing. Before you add ANY GODDAMNED THING to your game, know why. Why does it exist? What is it doing? Are you adding lore because a player asked a question? Cool. Is it a random combat to spice things up? Cool. Even though I said not come up with anything you don’t have to, I am saying to understand WHY encounters happen and what the point of an adventure is. Know them BEFORE you add the element into the game. If you don’t know what it’s doing in your game, it doesn’t belong there.

These two rules about “knowing the point” and “improvising as small as possible” work together. They push and pull at each other. If the PCs walk into a shop to buy something, you CAN just narrate that they buy all the s$&%. But maybe you decide to add a scene inside that shop. Maybe you decide to let the players play out the interaction with the shopkeeper. Okay, fine. Why? Maybe you want to impart some information because the players are flailing around cluelessly. Cool, you’re building an exposition scene. You don’t need any more than a basic description of the shopkeeper and then you can feed information. Maybe it’s because you want to give the players to interact with the world and bring the world to life. Also cool. You’re building an exploration scene. You need descriptions and a fun personality trait or quirk. Something interesting to explore. Maybe the PCs have been bullying too many people in town and now all the shopkeepers are pissed off. Maybe now you’re going to prevent them from doing business unless they overcome an obstacle. That’s an encounter kid. So, you need a dramatic question “can the PCs convince the shopkeeper to do business with them” and a source of conflict “the shopkeeper has been ordered by the guild not to do business with the PCs.” You might need motives and objections “the shopkeeper doesn’t want to get in trouble with the guild but the shopkeeper might be afraid of crossing the PCs.”

That said…

Go with the Flow

The weird truth is that games and worlds and NPCs and all of that s$&%, they start to take on a life of their own if you let them. The more you know about your world and the characters and the game and everything, the more you can just sort of start talking and expect your brain to kick in. So, you just start playing this elf and the conversation flows and suddenly you find yourself talking about his dead sister. Or how he hates his family. Or maybe he’s gay. Who the hell knows? Maybe you start talking about the history of an orc tribe and it suddenly becomes this very complicated little story.

Although I said limit your scope and don’t improvise what you don’t have to, that was about trying to make it happen. Every so often, your brain will just start turning crap out that is actually good. And if it does seem good, go with it. At least for a little while. If an NPC turns out to be a gay black sheep with a dead sister, that’s fine. That NPC was meant to be that way. That orc tribe WANTS to have that complicated little story. Fine.

But, do remember that everything you do has a point. Don’t let your creative spew extend something past its purpose. If the shopkeeper really IS just a shopkeeper, no matter how interesting the backstory that is coming out of you might be, you’re never going to use it again. Cut it off.

And beyond that, no matter how much you find things flowing from you, remember…

Don’t Wreck Your Plans or Your Game (Unless You Have To)

This is a big one. This is another source of dumba$& advice from people who aren’t me on the Internet. It usually manifests as the classic bit of advice for running a mystery: “listen to the player theories and then make one of those be correct.” Don’t do that. Don’t let your improv f$&% with the reality you’ve already been building toward. I mean, okay, maybe sometimes do. But don’t do it carelessly. There’s a fine line between reacting to the PCs’ actions and actually changing reality.

For example, if the PCs piss off all the shopkeepers in town, your planned scene with a shopkeeper later on is going to have to change. No way around that. That’s because the PCs made a choice and the world reacted to that choice. That’s GOOD. But if the PCs are conjecturing that all the shopkeepers are involved in the conspiracy they’re investigating, don’t decide they are right if that isn’t true. Seriously. Even if it sounds like it would make a better story. It won’t. For a thousand subtle reasons.

Those thousand subtle reasons come from the fact that EVERYTHING you do in the game is driven by what you know about the game and the world. You’ve been building a consistent story and world influenced consciously and unconsciously by all the things you knew were true. That doesn’t mean that players won’t get the wrong idea sometimes. People make mistakes or misread situations. But, ultimately, if you go back on what you’ve been building toward, the threads of your reality start to weaken. The consistent world you’ve been building makes less sense. And if you’re running mystery games and investigations and intrigues, those threads are SUPER IMPORTANT for the game to function as a mystery or investigation or intrigue. The games where PCs are most likely to have wrong conjectures are also the games that are most breakable by f$&%ing with the consistency.

Moreover, those incorrect theories and conjectures, when they get revealed, do a lot of good for your game. First of all, they demonstrate to the players that they are empowered to win or lose based on their own skills. Nothing proves that a player has the power to succeed or fail in their hands like showing them their mistakes. Second of all, they are what actually add challenge to any thinking puzzle – like a mystery. The possibility of misinterpreting the clues and reaching the wrong conclusion, that’s the equivalent in a mystery of losing a battle or missing an attack or failing a saving throw.

Write Down Everything!

As a GM, you should have a blank piece of paper handy AT EVERY GAME to write down the new story elements that you pull out of your a$&. If you invent a new NPC and give her a name, write her down. If you name a town, write it down? If you invent lore, write it down. Don’t rely on remembering it. And even if you do have a good memory, write down a little note to jog your memory. Even if you’re recording your sessions and can go back and listen to them, WRITE IT THE F$&% DOWN!

Here’s why. After the session, you are going to go back over that list of game elements that you invented and anything that could potentially reach beyond the moment it existed, you’re going to start answering the unanswered questions. “Why did that goblin attack?” “What else will the merchant’s guild do the PCs?” “Who is the fortune teller and where did he get his powers – or is he a scam artist?”

THIS is where you start building connections and stories beyond the scenes or story elements you had to invent on the fly. THIS is where the little things you improvise can ripple through the game and the world. THIS is not something you should be trying to do at the table (though sometimes, going with the flow might cause of this to happen).

Remember, improvisation of story is the worst thing ever. You should do the bare minimum needed to get out of it. After all, it’s exactly the same thing as writing and preparing a game, only with ridiculous amounts of pressure and an aggressively tight deadline. View story improvisation like a rat views a trap: something to escape from as quickly and desperately as possible regardless of who or what you have to claw or bite through to do it.

Know Your Weaknesses and Have Tools to Cover Them

Finally, everyone has strengths and weaknesses. And story improvisation will amplify the f$&% out of your weaknesses. If you’re bad at coming up with character names on the fly, having to build interaction scenes at the table out of nothing is going to shine a giant-a$& spotlight on that. So, have a list of names handy. Just keep a list of names near you at the table and cross them off when you use them.

Stat blocks and combat encounters are big weaknesses for a lot of people, after names that is. The Monster Manual is a giant book full of stat blocks. Have that handy. Like, EASILY ACCESSIBLE handy. Not a document you have to take two precious minutes to open on your tablet and search through. Can you invent a decent encounter map on the fly? Not great, just okay? No! Have a few handy.

Pay attention to the moments of improvisation you get trapped in. The scariest, most frustrating ones? Those are the ones you probably need tools for. And those tools could be anything from names to personality traits to stat blocks to encounter maps to lists of plants and fungi to whatever. Over time, you’ll build up a little pile of preferred tools. And don’t bother with tools you never touch because you can pull those things out of you’re a$&. You don’t need those.

Improvisation of Story is Okay

I just want to reiterate: narrative improvisation is the worst thing ever and having to do it puts a lot of pressure on you and can reduce the quality of your game. But it is also something that has to be done. It is necessary. And it’s also a lot of fun. I don’t want to give the impression that I’m telling you to stay away from it. Do it. That’s why I said that sometimes you do it just because you want to throw something that seems like a fun idea into your game. And that’s why I said go with the flow when it really is flowing.

Without narrative improvisation, even putting aside the cool ideas and weird flights of fancy that come out of your GMing brain, your game becomes static and unresponsive to the PCs. And that’s the opposite of everything a role-playing game is about. It HAS TO be there to make the game what it is. So, you have to accept it.

But, like all things, it is a balance. It’s a tradeoff. It’s NOT superior to planning, it’s NOT inferior to planning. That’s like saying a screwdriver is inferior to a pair of pliers. They are two different tools; they do different things. And a good game can only come from using both at the right time.

The problem with narrative improvisation is that it can become an easy way. It can become a good way to get lazy. If you over rely on it, you can actually make the game harder to run and put more pressure on yourself than you need to. You also can’t under rely on it.

In the end, you have to find the right balance. And NO ONE can tell you what that right balance is. It’s very personal. Don’t trust anyone who says they can.