NPC Template

 I want to share the way I do NPC templates. I’ve been really focused on character and diplomacy-driven scenarios, and have been playing around with different sets of building blocks when it comes to designing NPCs--here are the qualities that I’ve found most consistently useful in-play. Specifically, the parts I’m most proud of are the ways I use the “Tips fo RPing” and “Motivation” sections, although I have some innovations to how I use “Abilities”, as well.

I should note that these tend to run a little heavy, prep-wise, because they are for modules on the path to being published. I often skip quite a bit of this in more normal day-to-day play.

On the flip side, I do at least start this process for every NPC in a scenario I design. I think it’s just as important to get some motivations and qualities down for nameless goblins as it is for kings and rivals. Every faction having both a motivation and conflicting takes on how to achieve that motivation within itself is key to giving the players opportunities to cause trouble.

Let’s take a look at an example template and go over all the parts.

Part 1: Core Info

There’s just some basic info that comes up a bunch mid-play but isn’t super interesting, so I put it all up top for easy reference.

Title: First off, I give the NPC a title. This usually isn’t their name, it’s the role they fill in the scenario. I find it easier to refer to NPCs this way than by name, since it’s easier at a glance to remember who the blacksmith is than to remember what Henry Twobeard’s deal was.

Name: I then write down their name. If the template is for a group, I make sure to write out at least three names for members of that faction, to make my life a little easier when my players inevitably ask.

Faction: Fairly straightforward, what faction are they a member of?

Level: Since I’m trying to keep things system-agnostic I generally try to avoid going too deep into stats. In an old-school system, Level is just equivalent to hit-dice. You can generally extract HP/to-hit rolls/etc all from this one stat.

Armor: One important quality not baked into level is defense--how armored is the NPC? Again, to keep this system-agnostic I just have four options: None, Light, Medium, Heavy. You could just retitle this “Unarmored, As Leather, As Chain, As Plate”, if desired. I don’t think there’s too much value in going into more detail than this.

Part 2: How To Run Them

This is where the more interesting stuff comes into play--if you’re going to run them in a conversation, what do you need to know? There are some implicit patterns I like to use when filling these out, which I’ll get to in each entry.

Description: One paragraph giving the gist of how they fit into the scenario. My goal is to give just enough info that a GM could run the character and not feel lost even if this is all they ever read about them, while still being as short as possible.

I almost always start this with a short phrase followed by a comma. “Short humanoid rats, “ “An unassuming old woman, “ “A gangster in the guise of an administrator, “. Again, I want to make sure that there’s a little nugget of prompt baked in there that can be digested at a glance, if the GM needs to pull up a NPC they haven’t gotten to reading up on yet. The more you read of the template the more ideas and help it should give, but you should also be able to skim super fast and get something out of it.

Appearance: Three physical traits you can reference when describing them. Meant more to set a tone or give you interesting descriptive seeds than to exhaustively describe what color trousers they’re wearing.

Tips for RPing: A list of ideas for how to act as them. I almost always give two entries here, each with their own agenda--acting instructions and a potential bit.

The acting instructions are meant to give you at least one physical mannerism you can use to make it clear who’s speaking when. Maybe it’s a physical gesture to repeat, maybe it’s a tone of voice, maybe it’s something as simple as just “go ham on your mafia movie impression”.

The "bit” is an immediate hook you can use to make them interesting to the players during first impressions. It’s okay to get a little silly here--it should be something that makes a strong first impression and gives the players an opportunity to build a dynamic with them. It can maybe be a goof, like the rats above making threats but then volunteering each other to carry through on them. It can be a challenge, like an ogre who misunderstands everything you say but gets mad if corrected. It can even just be a verbal motif, like a phrase a shady adventurer repeats when trying to convince a mark. It just needs to be memorable and get the players interested in the NPC.

The bit is very possibly only something you ever use one or two times. As the players get to know a NPC better and they get more fleshed out through gameplay they’ll naturally become a less cartoonish and more nuanced personality that the players have meaningful feelings about. Once you’re at that stage, you’re golden--this is just meant to be used as an icebreaker.

It’s pretty common that I disregard what was written for both of these in-play. It’s better to let a NPC’s dynamic with the players emerge organically than to force it, but it’s useful having both of these as fall-backs if nothing naturally arises.

Motivation: What does the NPC want? In my mind, this is the core of the template--it’s what you actually use when figuring out how the NPC will react to the various situations that come up in play. I almost always give the NPCs three motivations, and usually sort them in this order: a quest, a vulnerability, and a philosophy.

I try to make my first motivation always be a quest--something that the NPC can ask the players to do for them. This might be a fairly typical quest like “help me with the protection racket threatening my shop” or “rescue my kidnapped friend”, but for minor or non-sentient NPCs it might even be as simple as “wants the players to leave their territory” or “wants food”.

The quest is often a great excuse to keep the players bouncing around between NPCs--NPC A asks the players to talk to NPC B, and in the process they learn about the situation with NPC C. A good web of NPCs with favors they want PC help with can be a great force for immersing the players within the scenario’s greater situation. The point of a quest is to keep play moving forward--there should be concrete actions that the players can take in response to the quest as soon as they decide to help out this NPC.

Second I try to give my NPCs a vulnerability--a motivation that the PCs can exploit. It’s a good reminder to always give your NPCs some weak points that can be exploited; just as a NPC who’s unbeatable in combat is boring, one who lacks any moral codes or personality traits that can be exploited diplomatically is boring to negotiate with.

Fears are good vulnerabilities--things they don’t want to happen. So are prides or social blind spots--things the NPC doesn’t want to admit about themselves. Passions can be good, too--topics they drop their guard when they talk about, or ways they can be flattered or put into a good mood.

Third, it’s important to give each NPC a philosophy or long-term goal. Vulnerabilities are passive, and quests can be resolved, so you need some deeper motivations you can use to determine the NPC’s actions moving forward through the campaign. These shouldn’t be attainable mid-scenario--if they’re attainable at all, it should be what the NPC is hoping to get out of the scenario’s resolution. Ideally, this should be a mix of a world-view (I want to protect my friends) and a method (to protect them I need to get stronger).

Part 3: Abilities

Finally, I list any abilities that the NPC might have. This section varies quite a bit from NPC to NPC, but there are a few basic categories of abilities I tend to list--attacks, special abilities, quirks, and triggers.

Attacks are pretty self-explanatory. If the NPC is a combatant, I list their attack stats. I tend to mostly just list damage and, if relevant, range, along with any special rules the attack might have.

Special Abilities are also relatively straightforward. If the NPC has any skills or special rules to them--being good climbers, producing webs, resistance to a damage type, memorized spells, etc--they go here. If they have any items, they typically go here.

I try to give every NPC without any special abilities a quirk--a non-combat ability that’s as much about establishing a personality as it is being useful in actual play. The ratfolk guards can speak to their trained giant rats. The shopkeeper has an apron full of any ordinary tool or item you might expect to need. The spice merchant can eat extraordinarily spicy food without flinching. Give every NPC something they’re exceptional at, even if it’s probably useless.

Bonus points if you can add some situation to the scenario where that skill might actually be super useful.

Triggers are more situational, and I try to use them sparingly, but they can be useful. A trigger is often an “if, then” ability--an idea for some sort of trouble the NPC might get into based on how the scenario unfolds. The urchin child will, if they end up separated from the PCs in the dungeon, always be adopted by the local faction rather than being eaten or imprisoned. The bored noblewoman is surprisingly good at making ropes from bedsheets and sneaking out of the castle. The washed up old alcoholic hero could overcome their depression and become mighty again with some rekindled hope.

There could easily be no end to the number of triggers you could write for any NPC, so I try to save them for unexpected directions NPCs might take. The child being adopted by monsters is fun because it’s not how that situation usually plays out. The noblewoman being good at stealth is worth writing down because it encourages her to do something surprising.

Part 4: Threads (optional)

This is semi-rare, but I will also sometimes include a final section at the bottom for the ‘threads’ a character is wrapped up in--parts of the scenario that span across multiple characters, where changing one character will potentially disrupt the larger plot.

For example, if there are a number of spies infiltrating an event intending to pull a heist, I would typically note on each spy’s template both that they are part of a larger conspiracy and what role they will take once things begin.

I like to take this format because I want to let players customize how they run NPCs I design, and if there’s anything about that NPC that’s a load-bearing part of a larger thing going on I try to make special note of it so they can either avoid unintended consequences (playing the getaway driver as a normal civilian) or come up with a replacement (pick a different NPC to be the secret getaway driver).

In Summary

And there you have it--this is how I usually write out my NPCs.

As you can see, I like to go pretty light on the ‘statblock’ side of things. In part this is to try to keep my notes system-neutral, but it’s also because I tend to find a lot of what normally goes into a statblock pretty boring. Ability scores and skill ratings and so on take up a whole lot of space, but rarely do much to make a character memorable. I’d say that 75%-90% of my NPC infoblocks tends to be about how to place them within the narrative, with only 10%-25% being mechanical.

I also try to treat the template like an onion--with layers of usefulness. You should be able to get a good sense of a character by skimming their description and their ability names, but for anyone wanting more guidance or inspiration there’s a ton of it at hand. I’d expect that even the majority of GMs would make modifications to how they run any given NPC, but I want to make sure that even a GM feeling totally uninspired has a fun NPC they can give their players.

More than anything, I want to try to make sure every NPC is memorable and fun to bounce off of in some way. I’m drifting more and more away from padding out RPGs with a bunch of nameless goblins who just appear to attack the players and then drop some XP and GP as they’re inevitably massacred--I want every NPC to be someone the players could conceivably end up with a relationship to, either as friends, rivals, or nemesi.

My Sources of Inspiration

Motive, Means, & Opportunity by Retired Adventurer

Advanced NPC Templates by The Alexandrian (this was basically the original form of my template, before it naturally evolved a over time)

The Dark Hot Spring Island -- this module has been a huge influence on the way I format information in my games, and the way they handle motivations for each NPC is great

Involved Skill Check System

I’m resigning myself to the fact that most people just want to play 5e D&D, so I should probably be making sure that what I make is 5e compatible, if at all possible. Here’s a shot at adding a system for resolving complex problems in a slightly more dynamic way. You can use this for social negotiations, climbing cliffs, disarming complex traps, sneaking through a temple full of cultists, or whatever else you’d like.

The Core Mechanic Whenever you have a skill check trying to resolve a problem, you set two numbers: a Difficulty/DC rating, which is the minimum number you need to roll to make progress, and a Complexity rating, which is sort of like the problem you’re dealing with’s HP. Roll the skill check normally. For each point you beat the DC by, lower the Complexity by one. If you reduce the Complexity to 0 you fully succeed--congratulations! However, if you fail to beat the DC or don’t beat it by enough to wipe out the Complexity in one go, the problem then gets to act/you get hit with a Complication/something goes wrong. You then respond to or suffer the effects of the Complication, and if the original problem is still relevant you can take another shot at it. Repeat until you’ve either succeeded or things have gone wrong enough that you have entirely new problems to deal with.

You can think of this almost as a reformatting of combat--you and the problem take turns attacking each other until either the problem is solved or you’re defeated. It’s meant to let some non-combat situations have a bit more weight and chance for partial successes than the normally binary skill check system would allow for.

It’s also meant to hew as close to the existing 5e skill system as possible. If you set the Complexity to 0 you’re basically just using the default skill system, and most of your challenges should probably be Complexity 0.

I tried to keep the above description as short as possible, so let’s unpack it a bit/offer some additional advice/add a few extra rules:

Complications Severity

A Complication occurs whenever you roll on a problem and fail to finish it off. They should flow from common sense and the situation at hand, and I’m intentionally avoiding giving them much mechanical definition. That’s not super helpful, though, so here’s a bunch of non-binding advice on how to think about them.

Problems can be thought of as having a severity rating of the Complications they’re likely to produce: Safe, Risky, Dangerous, or Deadly. If a player wants to know the severity of a problem before they roll you should probably tell them. If the outcome of a problem is unknown (unlocking a treasure chest that may or may not be trapped), just tell them that instead--maybe they’ll choose to gather more info before they act.

Safe Complications are ones where the worst case scenario is generally just that you won’t be able to try again, or that you’ll waste time getting nowhere. A good default result of failing to finish off a Safe Problem is that it becomes Risky--you’re fine for now, but continuing the conversation might make the guard suspicious/you’re starting to feel overheated as you hike through the desert--no mechanical penalty this time, but from now on failing a roll might inflict HP loss or Exhaustion.

Risky Complications cause short-term problems. The farmer you’re talking to gets mad and doesn’t want to talk to you any more. The king you’re asking a favor from asks you something in return. The raging fire you’re trying to put out spreads to a new location. If you’re taking damage it’s probably equivalent to one attack from a level-appropriate monster.

Dangerous Complications cause long-term problems. The farmer decides you’re devil-worshippers and raises up a mob to come after you. The king tells the guards to throw you in prison. The fire causes the building you’re in to start collapsing, cutting of your way out. You contract a status condition that takes you out of the fight--paralyzed, confused, or asleep. You can break out wildly level-inappropriate damage dice at this level.

Deadly Complications are instant death or permanent debilitation. You fall into lava. You’re crushed under the giant block. The demon rips your soul from your body. Medusa turns you to stone. Problems should basically never start out at these stakes--if you’re shimmying along a cliff over a bottomless pit a failure should be you losing your grip and starting to slide down, not you instantly plummeting. That gives you a chance to come up with a goofy plan or burn some resources to save yourself, instead of just instantly losing your character.

The entire point of having a Severity system is so that you don’t need to think up exactly what the outcome of a failure or partial success would be for every single roll. If there’s an obvious outcome, you can just say that instead of picking a Severity. This is all to help make life easier, not give you extra busy work.

Picking Complications

Some good Complications might be: take damage, take a status effect, lose an opportunity, lose an item, a new problem occurs, things get worse, roll on a table, a fight breaks out, something is demanded of the player, an NPC makes an ultimatum, the player has to make a choice between two possible Complications, the severity escalates. Ultimately, though, just go by what makes sense given what’s been established so far.

Especially fun are Complications that force a response from the players--rather than hitting them with damage or a status immediately, throw a new sub-problem at them that they need to be able to have an answer for; if they have the answer that’s good, but if not hit them with an even stronger effect. High stakes but with chances to save yourself is more fun than a lot of little unavoidable attrition hits. In a negotiation, have the NPC ask the player an inconvenient question. Climbing a cliff, have the rope fray slow enough they can leap to another or cast a spell. Set up threats, then give the player one or more turns worth of actions to diffuse them.

Complications can and should often involve rolls. Failing a Safe check in a dungeon might result in burning enough time to trigger a Random Encounter Roll. Getting pricked by a hidden poison needle in a treasure chest might trigger a CON save. A good wilderness exploration Complication might be to roll to avoid becoming lost. Your hireling makes a morale roll. There’s a 1-in-10 chance per Complication that the fire reaches the gunpowder and the ship blows up. You can temper the severity of an outcome somewhat by giving the player an extra chance to avoid it.

Should failing to hit a DC trigger a worse Complication than simply failing to finish off the problem? Sure, why not! If you don’t have any better ideas, hit the player with both an immediate Complication AND raise the Severity of the problem on a full failure, but only do one OR the other on a partial success.

Should a problem hand out the same Complication each round or do new ones? That probably depends on the situation at hand. It makes sense that if the problem is a room filling with poison gas that you’d have the same Save vs Poison Damage or whatever each round, but in a negotiation with a king it’d make sense to throw all sorts of stuff at the players--unexpected requests, interruptions, breaches in etiquette, etc.

Setting Your Ratings

You can write out these types of problems as DC #/# if you want, using the first number as the Difficulty and the second as Complexity. So DC 15/5 would be a Difficulty of 15 and a Complexity of 5. Complexity is a long word and writing it a bunch seems annoying.

I’m going to encourage you to set your DC and Complexity both to numbers divisible by 5, just to make the math easier. 

For lots of problems, a Complexity of 5 is probably plenty. 15/5 means a low-level character with a +5 to their roll has a 50/50 shot of getting a partial success and a 25% chance of succeeding outright. That feels about right as a default quick problem, like jumping across a chasm or a classic reaction roll-type situation.

A big scene-consuming problem that the entire team is working against might have a Complexity that is much higher--50 or more.

Possible Modifications

I’m tempted to change the rules so that you reduce the Complexity by 1 if you hit the DC and then by an additional 1 for each 5 you beat the DC by--the math seems a little easier than subtracting the DC from your roll and the result from the Complexity. That might create weird dead zones in the dice results, though. 

5e skill checks are not nearly bounded enough. This system breaks down a bit when you have one player with a -1 to a check and another with a +17/minimum roll of 10. This was already a bit of an issue in 5e, and if anything it helps a little--a mid PC can succeed at what an optimized PC can, it just takes them longer and they run into more complications--but it’s still a problem. 

My idea for a solution: When rolling a skill check you add your Stat bonus to the roll. If you have proficiency in the skill, roll one extra d20 and take the higher. If you have expertise, roll a third.s This stacks with advantage/disadvantage.

This fix makes low level characters a lot more competent--proficiency moves from being a +2 to the equivalent of a +5--while making high level ones a bit less--from a +6 to a +5ish, but also weakens expertise a bit--you get some diminishing returns from all those extra dice. It also makes the range of possible DCs way smaller--the highest a character could ever roll would be a 25, but a character with proficiency and a +0 stat is still going to roll over DC10 75% of the time and is going to beat DC21 0% of the time. It means you’d probably just want to be using DC10 or DC15 for most situations, setting difficulty more with Complexity. That sounds fine to me, but I should test it out more.