You came back, huh? Color me impressed. I expected you lot to blow this lesson off considering I promised another load of blathering bullshit. I guess you really want to learn, eh?
If you missed the memo — or can’t infer from context — today’s informal discussion about how players and Game Masters play their respective characters started last week. And it’s a vital prologue to the upcoming lessons about Resolving Infiltration and Resolving Interaction. Both game situations involve lots of back-and-forth between player-characters and non-player-characters — cat-and-mouse interplay being vital to resolving tactical infiltration action — so you have to really get running non-player-characters.
My point is this: if you skipped the last lesson, go back and fix that before proceeding.
Players Choose; Characters Act
Last week, I wasted 3500 words and a bunch of sidebars justifying — again — my adage that players choose while characters act. An adage lots of you still don’t like and still don’t believe. This means there’s gonna be a lot of copium when I tell you how to run non-player-characters like a True Game Master.
The point of that players choose; characters act thing is that the players have to decide what their characters do and say. They’ve got to decide what stance their character takes, what resources they spend, what tools they use, what facts they cite, what tone they adopt, and so on. They can’t fall back on their characters’ stats to help them formulate plans, pick strategies, draw conclusions, or solve problems.
A player who wants his character to talk his way past a guard must figure out how that might possibly work. They have to figure out what things that character could possibly say that might convince the guard to stand aside.
The game’s mechanics, meanwhile, determine how effectively any given character can implement the plans their players came up with. Players decide what weapon to pick up and who to swing it at; their characters’ Strength scores determine how hard they hit. Players decide what words and ideas to communicate; their characters’ Charisma scores determine how hard they hit.
Thus, players and their characters aren’t really separate, independent beings. Players choose their characters’ actions based on their vision of the character and the character’s abilities but also include their own personalities, motives, and team dynamic. That’s how it is.
So it’s not right to say players play characters. Instead, you — as a player — play yourself as if you were your character. You play as if you were you, but if you lived in an imaginary world and had your character’s background and skills and abilities and you wanted to accomplish your character’s goals.
What would you do if you were a barbarian from the Plains of the Purple Buffalo and you were stronger than smart and you wanted to rescue Princess Buttercup and you found yourself suddenly confronted by Rotgut the Swack Iron Dragon?
When you — as a Game Master — say “players choose; characters act,” that’s what you understand. That’s what you’re making yourself believe. So practice saying — and believing — it.
But let me boil it down into something simpler — something less caked in bullshit — so it’s easier to turn on its head in a minute. Try this…
Players pick their characters’ actions; the dice determine the outcomes.
Strip away all the high-minded crap and that’s all I’m saying, isn’t it? Players decide — themselves — what their characters do and the dice — and the other game mechanics — determine how it all works out.
Now, let’s talk about how Game Masters — True Game Masters — play non-player-characters. Short answer: they don’t.
Non-Player-Characters Choose; Game Masters Act
Gameplay Trumps Story
I bang constantly on the gameplay drum. However much I say roleplaying games provide satisfying gameplay and narrative experiences, I don’t talk much about the narrative side. And now I’m saying the narrative is an illusion you lay over a good gameplay experience.
A story’s what you tell about what happened. Anything can make a good story. Roleplaying games promise good stories, sure, but they promise uniquely evolving and emergent stories. The stories arise from the gameplay. Game Masters create an illusion that the game’s more than just a game, but you can lay such an illusion over any game. With proper narration, you can turn a game of chess into an epic war between good and evil.
If the game underlying the illusion sucks, the whole thing sucks. Meanwhile, Narrative — the pleasure people draw from experiencing a story — is just one of eight gameplay engagements. The game’s got to do more than just tell a good story. Hence, the gameplay experience trumps everything.
If you focus on providing the best damned gameplay experience you can, everything else will emerge from the experience. Even if your smoke-and-mirrors patter is only so-so. Believe me.
You ain’t a player. I covered that in great detail at the start of this True Game Mastery journey. You’re not playing or roleplaying or anything like that. Your job’s to provide the players with a satisfying gameplay experience. You’re a game console; nothing more. Why you enjoy that is up to you. And I’m going to assume — since we’re seven months deep into this shit now — that you have your reasons.
As such, non-player-characters aren’t characters. Like everything under your control, they’re just game constructs. And you — as a Game Master — must see them as they are, not as they seem. Because you spend a lot of time weaving illusions. And you can’t buy into your own bullshit for a second.
Whatever anyone else tells you, for example, you’re not simulating a world. You’re running a game and then disguising it as a simulated world. To the players, a town’s a living, breathing settlement full of actual human people who’ve banded together to provide for their common welfare and beat back the dangerous wilderness. To you, a town’s just a fancy skin over a bunch of equipment buying screens and expositional text boxes and quest logs and shit like that. And that owlbear isn’t a motherly beast defending its young from interlopers, it’s an obstacle to challenge the players. Games need obstacles.
Everything’s a game construct. Everything’s an element of the gameplay experience you’re providing. And whatever else they might do, they must provide a good gameplay experience. While consistency and fairness and agency and emotional investment are all parts of that experience, no one factor trumps the others. Just as the players balance their pursuit of in-game goals, character personality, character abilities, personal motivations, and social dynamics when choosing their characters’ actions, so too must you balance all the various factors that add up to a good gameplay experience when choosing non-player-character actions.
That means that all else being equal, a non-player-character’s personality is the least important driver of its actions. And you sure as hell ain’t trying to project yourself into the mind of the non-player-character and do the things they would do.
In the end, non-player-characters take whatever actions are best for the game and you, the Game Master, try to portray them as if they come from an actual character with an actual personality. Like an actor, your job’s to follow the script and make whatever’s written there believable and relatable. Meanwhile, the game writes the script.
Non-Player-Characters Choose…
The game’s rules — and gameplay’s needs — choose non-player-character actions. What does that mean? It means you ain’t roleplaying. You don’t pick non-player-character actions based on what you think they’d do if they were real-life people. You don’t project yourself into your non-player-characters.
So how do non-player-characters know how to act?
Dice
Most non-player-character actions come in response to player-character actions. That ain’t surprising given that running a roleplaying game is basically determining what happens when the characters do shit and evolving the world as a result.
The truth is, therefore, that lots of non-player-character actions are chosen by the dice. Pure and simple.
When a player rolls a Strength check and succeeds, you describe a door shattering to bits in the world. It’s no different when a player rolls a Charisma check and succeeds. You describe some non-player-character’s spirit shattering into pieces. Or whatever. Agree or disagree? Believe or doubt? Surrender or fight? Help or obstruct? Discount or gouge? Spot or overlook? Those are all outcomes determined by dice.
But the Game Master’s Brain!
Obligatory Reminder that You’re a Game Master
Remember as you read this shit today that I’m talking to you as a Game Master. This means, first, that I’m talking to a game mechanic and not a player. When I say, “The rules determine what NPCs do,” I’m including you as part of the rules. I know that seems weirdly paradoxical when I say, with the next breath, that you don’t decide what NPCs do. But the point’s that you’re not roleplaying, you’re running a game, and the game’s needs — which you determine — dictate NPC actions.
Second, it means that you don’t design the NPCs you portray. They’re designed by Scenario Designers. Once in a while, you might have to slap together an NPC on the fly because the Scenario Designer left something out or because your players decided they wanted to get to know the blacksmith or whatever, but that quick-and-dirty improv is not Scenario Design.
And while you might also be a Scenario Designer — if you run homebrew games — you ain’t one when you’re running the game.
“Aha,” you say, “but when I — a True Game Master — adjudicate an action, I decide whether that action can succeed or fail before any dice get rolled. When a player’s character acts against a non-player-character, I decide whether ‘agree’ and ‘disagree’ are possibilities. So that’s me making the decision after all. And I do that based on how I think the non-player-character would act.”
You’re not wrong, but you’re wrong, and you’re missing the point, and you’re doing it wrong.
If a player chooses to have their character say just the right thing to the right non-player-character, you — as a game mechanic — determine whether that action can succeed and whether it can fail. But that’s not based on the non-player-character’s personality. Rather, it’s because the player picked a good strategy given the situation. That strategy just happens to be a good one because of what they know about the non-player-character’s personality.
And side note: you should probably make the player roll the check, but offer a hefty bonus. Remember, if you cut out the die roll, you’re cutting the character out of the game. Characters take action, after all, not players.
I know this shit’s complicated. That’s what being a True Game Master is. Any idiot Game Executor can roleplay a non-player-character. That’s Easy Mode. A True Game Master intuitively balances out the thousand tiny factors that separate good games from bad and lets that guide their decisions.
Stats
Even when they’re not responding to adjudicated actions that dictate their reactions, non-player-character actions are still driven by game statistics. Every game system — every good game system — has some helpful mechanical tools for determining non-player-character actions. Disposition scores, reaction scores, reputation scores, social skills, and charisma scores just to name a few. And outside social interactions, when non-player-characters have to decide what actions to take — such as when combatants might flee a fight or when patrolling guards are searching for player-characters — various game mechanics must be considered, as must the non-player-character’s own abilities, traits, and skills.
Purpose
Non-player-characters are deliberately created game constructs. That means every last one exists for a reason. Maybe it’s to provide a challenge or offer information or provide a goal or provide set dressing and enhance the illusion that all the game’s a world. Or maybe it’s all of those.
The non-player-character’s raison d’ etre — which literally means the reason they eat — has a lot to say about what actions they act. If an NPC exists as an obstacle, it must challenge the players. It can’t roll over because someone blithely rolled a lucky check to Persuade at them. And if an NPC exists to bring the world to life, it probably shouldn’t obstruct the players.
Gameplay
Nothing is satisfying about rolling a Gather Information check over and over just to get directions to an inn. Nothing is satisfying about non-player-characters caving to sufficiently sizable bribes every time one’s offered. Nothing is satisfying about one failed Stealth roll leading to immediate disaster. That’s why, personality aside, street-scene extras always give directions, why bureaucrats must be wheedled and cajoled into taking bribes, and why guards don’t immediately raise the alarm at the slightest sound but instead split up and slowly wander around looking for the source of the noise.
… Game Masters Act
Your job’s to tell the players what their characters see, hear, perceive, and know in response to every action they take. And that means describing how non-player-characters speak, act, and present themselves.
When I say Game Masters act, I mean that in the literal sense. Game Masters, like actors, have to deliver their lines — and follow their stage directions — in a way that brings their non-player-characters to life. Whatever action a non-player-character takes, a Game Master must portray those actions.
Narrating Versus Acting
Splicers Always Miss
In 2017, Ken Levine — director of the original Bioshock at 2K Studios — revealed that the baddies in Bioshock were programmed to always miss on their first attack against the player. Why? So the player always had a fair warning when they were under attack and could respond accordingly. A game in which you can get suddenly mowed down dead by an enemy you never saw sucks.
No one noticed. Not until Ken revealed it. Because people don’t notice that shit. Especially not players. And it’s a perfect example of what I mean about making decisions based on providing the best gameplay experience. I’ve seen people struggle to figure out why wealthy villains in tabletop roleplaying games shouldn’t bring their massive resources to bear to assassinate the player-characters. And I’ve been asked why, no matter how often let player-characters snipe or assassinate foes with one die roll, I always use damage rolls when an assassin gets the drop on a player-character.
The answer is: because gameplay.
There’s this age-old debate amongst roleplaying games about whether it’s best to describe a character’s words or to speak and act directly as the character. And it’s a dumbass debate. Because it doesn’t matter. From a gameplay perspective — which is the only perspective that matters — either way works fine. Individual players will pick their own approach based on how they’re comfortable and why they’re playing the game.
Your job’s to facilitate both modes of play. Which means you’ve got to practice both. You may be better at one than the other — maybe you’re a better Storyteller than an Actor — but you should not only switch between both approaches regularly but also recognize when one’s better for the game than the other. And you should be able to both respond to each player as they’re most comfortable but also switch modes to invite different players to play.
Yep, complicated. Yep, True Game Mastery. Suck it up, Buttercup.
Even though they’re based on different skills and even though the third-person description is better called Narration, I’m going to call both Acting here. However you’re delivering or describing your lines, you must use every trick — word choice, tone, body language, and psychological trickery — to bring every non-player-character’s every action to life.
Consistency and Relatability
All else being equal, when you’re Acting, you have two goals: consistency and relatability. Your non-player-character should behave in a way that’s consistent with everything known about them — history, personality, motives, and previous actions — and they should seem like a real human person with whom the players can connect emotionally.
It’s all part of the illusion you’re laying on the game constructs. It’s part of the Matrix. The character’s choices are driven by mechanics and stats and game code, but you have to make it look like a choice a person made. Fortunately, most players totally miss or gloss over occasional inconsistent or unrelatable behavior. Especially if they’re engaged in the gameplay.
Consistency isn’t just about that Illusion of World thing, though. It also serves a gameplay purpose. For players to make good choices about how to interact with non-player-characters, they must be able to guess or deduce how that character’s likely to respond to different actions. That said, players are always operating on incomplete information. While you might have pages of background on the non-player-character in question, the players only know what they’ve seen at the table. That’s all that’s true.
Malleability and Rationalization
Social Skills Always Point At NPCs
Every so often, someone asks why PCs can roll social skill checks to influence NPCs, but NPCs can’t do the same to PCs. Or why PCs can’t use social skills on each other.
And now you know the answer. At least, you should. Do I have to spell it out? It’s because Players Choose while Characters Act, Non-Player-Characters Choose while Game Masters Act and because NPCs are game constructs.
So what good is it giving NPCs social skills? To guide your portrayal of course. When you’re acting out a lie, your portrayal should be different based on how skilled the liar is and how skilled the person seeing through — or not seeing through — the lie is.
There will be times when non-player-characters choose actions that just don’t make sense given how the character’s written. When they make choices you wouldn’t make if you were roleplaying. non-player-characters will refuse to take bribes or they’ll refuse to help people obviously working toward everyone’s best interest. Your job is to portray those choices as if they make sense.
Non-player-character personalities are, helpfully, infinitely mutable. Nothing written is true until you’ve revealed it to the players in game. So you never need to sweat portraying non-player-characters consistently with what’s written in the scenario. And it takes very little to make a behavior make sense. One little fact is all you need. Today’s inconsistent action foreshadows tomorrow’s revelation of an irrational fear, sore point, or unresolved issue. And that shit enhances a character’s relatability.
Moreover, humans can rationalize anything. Hell, humans rationalize everything. There’s scientific proof that your conscious brain doesn’t involve itself in most of your decisions. Usually, it wakes up after a decision gets made, invents a reason for making it, and then lies to you by claiming that’s the reason. You can look this shit up.
That’s your job. It’s to rationalize every action every non-player-character chooses. And to use that rationalization in your betrayal. You need to decide that the offered bribe offended the character’s honor, so you can say, “How dare you? Do you think I’m so lacking in integrity that I can be bought?” Or that they’re cynical, so you can say, “I’ve heard that dodge from a thousand adventurers before you. The world is not ending tomorrow so pay your bill.” Or that they mistook an insult in a player-character’s words so they can say, “Do you think I’m doing this for me? I work to put food in my children’s mouths! They are who I’m responsible for!”
The Takeaways
This wasn’t so much a lesson as a prelude. Soon, I’ll be telling you how to resolve social interactions and infiltrations. Both involve non-player-characters a lot and I wanted to make sure you had the right mindset. I don’t need any of you dumbasses thinking you’re supposed to be roleplaying your game constructs. Running non-player-characters is just another kind of action adjudication and outcome description. Expect more solid advice in the next couple of lessons to build on this mindset crap.
That said, there are a few things you can do right now to up your game.
First, assume most actions taken against non-player-characters can fail. No matter how clever or poetic the player and no matter how Charismatic the character, don’t skip the die roll. And take some time to think about how you’d portray a failure on such a seeming sure-thing social action even if it never arises.
Second, practice switching between first-person Acting and third-person Narrating when portraying non-player-characters. If you tend to prefer one, start leaning more heavily on the other. But change it up from encounter to encounter and even in mid-encounter so you can experience the differences firsthand. You’ve got to be able to do both passably well and switch seamlessly between them.
And third, stop fighting me on this shit. Either take my advice and watch your game improve a thousand-fold or get the hell out of my classroom. Either you trust me or you don’t. And if you don’t that’s your own dumbass problem; stop trying to make it mine.
Great article, needed some of that brain rewiring. Recent articles are also helping how I design scenarios by making me think about what I need to bring to the table to run the game. If I decide that I need an obstacle that’s an NPC, their objections and incentives are more important than their personality.
“Either you trust me or you don’t.”
You’ve earned mine, repeatedly. ‘Npcs choose and dms act’ is a powerful bit of knowledge. I’ve “played” npcs both ways, as game constructs and as characters, and never thought much of it, but I can see now where treating them as full fledged characters has caused some of the problems I never understood. Great article.
Great stuff, I’ve changed npcs from what they were described like in scenarios based on what the game needed in the past and it always worked out well. Can recommend!
Good stuff. One nit-pick: raison d’ etre is French for “reason to be” not “reason to eat”
The intended context is the same though.
Okay… nobody tell them.
Actually, I believe the Germans coined the phrase in the early 1900s. In its entirety, it would be raison d’etre San Diego, or “reason they eat a whale’s vagina.”
Great advice here, especially the part about treating NPCs as game constructs. I’ve seen DMs treat their NPCs as main characters, and it’s not a fun experience. You want to watch the NPC solve every problem? You want your character to end up dead/in jail for disagreeing with the NPC? No thanks.
More than anything these articles allow me the play the way a game feels right and good, without feeling guilty the rules says NPC are like characters played by the gm or other crap like that
By accepting one is not an interpreter of the rules, but the actual rules or game engine itself allowed me to take a complete new approach to my role on the table and all articles just serve as great examples and frames to on which this principle exist.
Love the article!
“So it’s not right to say players play characters. Instead, you — as a player — play yourself as if you were your character. You play as if you were you, but if you lived in an imaginary world and had your character’s background and skills and abilities and you wanted to accomplish your character’s goals.”
These are some of the truest words you’ve ever written. This, right here, is a distinction that many fail to grasp; and, is also the ultimate refutation of the “It’s what my character would do” baloney.
Been looking forward to this follow-up. It’s a logic puzzle: if players (can only) choose, and characters (can only) act, then non-player characters can’t act; they must choose. The difference is in *how* players versus NPCs choose.
Does the Scenario Designer roleplay when deciding how the world or the non-player-constructs within respond to player actions within the constraints of the game?
I am looking back at “What Even is an NPC (And How to Do Them Right)” and it seems to still hold true except for some ambiguity as to what is the domain of the GM and what belongs to the Scenario Designer.
I agree with your premises, and I agree with your conclusions, but the logic you use to get from your premises to your conclusions is sometimes baffling.
I have tested the always make a check concept and I have found it lacking. it is wishy-washy. 95% of the hefty benefit rolls will succeed. That takes away from the suspense of all rolls. if any NPC who took the same action would succeed, I won’t waste the player’s time and punish their critical thinking skills by forcing them to roll a check that doesn’t have a particular reason they could fail.
The better solution is to not have situations like that in the first place. referencing your encounters documents solves this problem immediately. referencing your document which states that the GM decides what dice get rolled and when also solves this problem. the player can roll a natural 20 on a persuasion attempt to talking to a guard. the guard will whistle, impressed, and tell the player that he rolled a very pretty 20. but the player is still not getting past him.
I guess what I’m saying is, it’s confusing to me because it feels like you’re solving problems you’ve already solved, and although this document has some great tools, it also recommends assuming that all actions against any NPC can fail, and that is something I have tested extensively and found it to not work. it’s not about trusting you or distrusting you. it’s about repeatable results of actionable instructions based in observable realities. Which is normally how you present data.
@Weasel: I’m a bit confused. What does rolling a natural 20 have to do with assuming social actions taken against NPCs can fail?
I think the issue that leads to this is that the examples are limited to focus on the point Angry wants to make. When he talks about convincing the guard he does not detail the whole scene and context but only the aspects that are required for his main topic.
When doing a “convincing the guard” scene the scenario designer could design that the guard is corrupted and believes he is a great fighter and best than most. Based on that scene design the players would interact with the challenge. A large bribe could grant automatic success, a smaller a nice bonus to the social roll. An intimidate might be a hard roll that on a bad result could lead to a fight, these are examples of me adjudicating rules for players choices.
So, do not read too much on the examples, if you as GM believes a bribe makes an automatic success because the guards won’t give a fork if the characters’ breath smell while counting the gold, go for it. In general character skill and characteristics must be relevant for the outcome of an action and therefore require a dice roll, but sometimes the action could bypass that, as a GM you have to know how to differentiate them and adjudicate it correctly.
> If you focus on providing the best damned gameplay experience you can, everything else will emerge from the experience. Even if your smoke-and-mirrors patter is only so-so. Believe me.
Just want to say “hell yes” to this.
> Nothing written is true until you’ve revealed it to the players in game.
And this—man I love this but find it so hard. I get that prepping is like writing a draft, and it can change completely when it goes to the table, and then it becomes set in stone. That’s fine. But then I’ve got to go back to the draft and solidify bits, scrap bits, and adjust other bits to fit. And it just gets to be a lot. I wonder if I’m using a bad strategy to manage all that.
Nope. Game Mastering is just hard. Everything worth doing and everything worth doing well is hard. If it’s easy, you’re probably not doing something worthwhile or you’re probably doing something wrong. Such is life.
There is a reason you are supposed to take notes on what is happening at the table when you run the game. How else are you going to keep track of things you need to remember and that might have changed from you prep otherwise? Of course that makes things quite a bit harder on the GM than a player. They only need to keep track of their own character and the stuff they “own”. You need to keep track of the tapestry and all that involves.
It becomes even more important in something like an open game, where you have multiple groups running around in the same region in the same world and want some consistency. Can’t have Group A go and get rid of a Bandit Camp, and Group B find that same Camp unmolested when they play a Day or two later, for example.
When would it be appropriate for an NPC not to cooperate regardless of the dice and character abilities? Only when they’re part of the obstacle?
What if prior player actions create a hostile town? Maybe that counts as an obstacle… I’ll think of an example…
Long time lurker, first time poster.
This article was exactly what I needed to save my current campaign–or rather, my ambition to keep it going.
Long story short, an open-world/exploration rebellion campaign is climbing into the 12-15 level tier, and I’m struggling to keep it engaging for the players (and myself).
I can now see that my problem was not using the various NPC leaders as game elements. Instead, between each session, I would check my various fronts and NPC leaders to imagine how that person would react to what the players are doing. While that worked well for a long time, what I was missing was the concept that THEY’RE NOT PEOPLE, THEY’RE GAME ELEMENTS; just tools to create a fun game for my players.
Now that I have that in perspective, I can really finish ratcheting up the tension for the final confrontation, and give this years-long campaign the retirement that it deserves.
Thanks for putting this stuff out there, it’s making a positive difference at my table.