Whenever I enter a new campaign and start thinking about what kind of character to make, I ask the same question: “what characters are other people playing?”. Not only does the answer give me a sense of the tone the rest of the group is leaning towards, and give me an opportunity to link my backstory to someone else’s, but it also lets me know the composition of the party as a whole. Knowing what roles others already show a preference for lets me focus on filling one of the remaining gaps. By “role” here I mean both on the narrative level (there’s probably no need for two identity-questioning synthetics or world-weary veterans) as well as the mechanical level of hacker, face, and so on.

However, that last facet of the question often triggers a dismissive response if I let on about the purpose of it. “Oh, party balance? Nonsense, we don’t care about that here”. This perhaps indicates something about the kind of people I play with – I doubt that the members of an organized play group speedrunning a module are opposed to making sure someone plays the healer. However, I’ve encountered it enough times that I feel it warrants an elaboration on why I care about role allocation, and it’s not just “balance”.

The Needs Of The Many

Of course, how functional a party is at fulfilling its stated goals does matter to play. Lacking a keystone skill can force creative solutions outside of what you normally expect, but can also frustratingly limit the party’s ability to engage with the core aspects of what that setting is about. It’s perfectly okay to lack a healer (or a hacker, or a face) intentionally, but wandering into it through disconnected character creation can lead to the campaign dying early out of frustration because the party isn’t up to handling the basic business of the genre. However, that’s not what I’m interested in discussing here.

What I’m actually getting at isn’t about mechanical efficacy at all, it’s about interpersonal dynamics. Put all thoughts of game roles out of your mind, and think about story roles. When you read a book or watch a show about an ensemble cast, the main characters usually fit into some kind of pattern. This manifests most obviously in heist movies: typically you get an entire first-act sequence about putting together a team of people with specialized skills to slot into necessary functions for the job.

It shows up in ways far beyond skills, though. The health of a long-running series depends on characters developing just the right ratio of friction to coherence that reliably generates conflicts while still managing to resolve them in a satisfying but not boring fashion. To do so, writers give characters a range of personalities – optimistic, honorable, irritable, et cetera – to maximize how many kinds of situations they can get in to with one another while maintaining audience interest. They also give them contrasting likes, dislikes, social standings, political views,, and other traits to round out and delineate them. Characters also represent differing philosophical stances that contribute to the thematic purpose of the story: two people might take two sides of a spectrum to demonstrate how the author feels about those two angles of the issue.

I bring this up because all of that matters to tabletop RPGs, too! You want that varied juxtaposition of traits to create the kind of interesting situations that players get invested in. You see this type of advice aimed at GMs sometimes, but even then there’s a tendency to focus on individual NPCs rather than the social situations they’re enmeshed in. A character exists within the context of the audience’s (or player’s) perception: seeing essentially the same personality twice usually dulls its impact, especially in close proximity. This applies to groups of PCs just as much as it does NPCs.

Gestalt

So that is the thesis of this post: groups of characters (PC and NPC alike) can be – and, I would argue, ought to be – considered as wholes, as composite characters unto themselves. A group can have collective wants (a shared quest or goal is already normal to establish for the party in many games), needs (food, rest, repairs, as well as emotional needs of excitement and attachment), personalities (usually discussed in terms of the overall tone of the party but comprehensible as a personality unto itself), and internal contradictions (personified to a much greater extent than in non-composite characters) that make it exciting and memorable. When a group of NPCs or a party of PCs is crafted together, you remember them in aggregate instead of only their individual members.

Which means that it’s okay to have, say, everyone play a wizard! That’s a strong high concept right there, you’d always remember that game. The important thing is that everyone play a different kind of wizard. Even mechanically similar characters can remain strongly distinguished by their personalities. Think about tokusatsu (Power Rangers and the like) teams: their shared goals, methods, and aesthetics give them a strong identity as a group, but despite not seeing their faces half of the time, they remain individuals as well. You don’t have to go anywhere near that close to uniformity – indeed, the average RPG party is a collection of oddballs – but even a motley band can feel like it fits together.

In practical terms, what all of this means is that we can intentionally form interesting groups by deciding their gestalt personality and then defining the individual members as subsets of that. Pick a central idea or core emotion you want manifest in a group. Then, tune the characters to orbit around that theme: for example, if you want to address the sociopolitical turmoil going on in a city-state, represent it by a group of four characters who each hold a different opinion about what needs to be done, and build their other traits around those opinions. If the mood you want for a campaign is desperate struggle, then decide up front that the personality of the party is going to be “hungry and scared”, and make characters that all have their own ways to responding to that constant stress.

What follows is an overview of common patterns, arranged into types according to how many members comprise the group. The suggestions focus on the relationships between the members and how they fit together, so you supply the moods and themes you want them to cluster around. The next time you write up an encounter or form a party, note how many members it has and consider applying one of these models to it.

One

A group with a single member isn’t exactly a group in the way most people mean it, but taken in the context of larger assemblies, patterns still manifest. The question an individual implies is “why are they alone?”. The answer tells you something about them.

Too Strong To Need Help: A supercompetent agent, a hard-bitten survivor, an inordinately clever jack-of-all-trades, a monster that killed their companions, hivemind (many bodies, one personality), destined hero.

Can’t Get Help: An obsessive introvert, an emotionally callous lone wolf, an exile from their people, someone simply lost and afraid.

Two

These show up all the time. You can encode virtually any kind of polarity into a pair of people: emotional, philosophical, relational, functional, visual, or otherwise. Commonly, they take opposite sides and generate situations through that clash: an obvious example is the Straight Man/Funny Man (tsukkomi and boke in the manzai tradition) pairing of numerous comedy sketches. Alternatively, both members can present themselves in a near-identical way, producing a humorous or unsettling effect. A third option is pair of people who may represent opposing sides but are primarily defined by naturally binary social roles, such as married couples.

Opposite: Straight Man/Funny Man, Red oni/Blue oni, By-The-Book and Loose Cannon, Strong-but-foolish and Smart-but-weak, Skeptic and Believer, Noble and Rogue, Sensitive and Assertive, Apollonian/Dionysian, Type A/Type B.

Similar: Eerie twins, Mister X and Mister Y (typically polite henchman of some kind), Clones, Robots.

Couples: Loving husband and wife, squabbling brother and sister, master and apprentice, hero and sidekick, bumbling henchman (usually one lanky, the other squat), Heterosexual Life Partners (same as henchman).

Three

Trios tend to either present a trinity of types, or two ends of a spectrum plus someone representing a middle ground. The former can sometimes map to a tri-stat system, such three members who exemplify Mind, Body, and Soul respectively – party-as-character, indeed. The latter is evident in the classic Star Trek team of Kirk, Bones, and Spock: one espouses reason, another argues for emotion, and a third mediates between them. You can take many of the duos listed above and add an in-between member to turn them into a trio. A final option is taking any duo and adding a “third wheel”, an odd one out who doesn’t quite mesh with the relationship of the other two. Love triangles are a popular addition to any of these types.

Trinity: Id/Ego/Superego, Brains/Beauty/Brawn, Fighter/Mage/Thief, Mother/Maiden/Crone, Guitar/Bass/Drums.

Spectrum: Law/Balance/Chaos, Reason/Emotion/Unity, Masculine/Feminine/Androgyne, Child/Parent/Grandparent.

Four

With the lowest prime numbers out of the way, from here onward groups often consist of smaller assemblages fused together: a four-person team can easily be made out of two groups of two. However, naturally four-valued typologies are not at all uncommon: the four temperaments a classic example everyone knows of, and have spun off into all manner of more modern variation. You can also construct a team by extending the two-person binary method into a two-dimensional scheme: plot two perpendicular values and make a person to fit into each of the four resulting quadrants, as with the well-known (if imperfect) political compass or D&D alignment chart.

Fourfold Taxonomies: Humours (Choleric, Melancholic, Phlegmatic, Sanguine), Blood Types (A, B, AB, O), Four Horseman (War, Famine, Death, and Conquest/Pestilence).

Two-Dimensional Spectra: Left/Right plus Authoritarian/Libertarian, Law/Chaos plus Good/Evil, Optimism/Pessimism plus Apathy/Action.

Five

Five person teams run up against the edge of fully coherent typologies: more often than not, they decompose into sub-groups of two plus three. Another common mode is taking a four-person group and adding an odd-one-out: this can be a normal person in a group of bizarre characters, a balanced middle-ground of four extremes, or something that transcends the standard four. However, true, holistic sets of five exist, such as the Big Five model of personality typing.

Three Plus Two: Core trio of friends plus two newcomers, often a rookie and veteran.

Four Plus One: Four Temperaments plus Supine/Leukine, Four Elements plus Aether/Spirit/Void, Four Aliens/Robots/Mutants plus one Human.

True Quintet: Big Five (Open, Conscientious, Extraverted, Agreeable, Neurotic), Wuxing (Wood, Fire, Earth, Metal, Water).

Six Or More

Past six, tightly bound examples diminish: you can usually break groups down into smaller component groups. In fact, it may be useful to think of them in terms of a layered structure: a nine-person group can be made up of three sets of trios, the three of which correspond to a trio structure when contrasted with the other sets. You can always, however, intentionally form arbitrarily large groups by taking something with a certain number of items (the Seven Deadly Sins are popular) and fitting members into each slot. Listed below is a brief survey of possibilities.

Six: HEXACO (Honesty, Emotionality, Extraversion, Agreeableness, Conscientiousness, Openness).

Seven: Seven Deadly Sins, Seven Virtues, Seven Liberal Arts (trivium plus quadrivium), Colors of the Rainbow.

Eight: Eightfold Path.

Nine: Enneagram, D&D Alignments (full possibility set).

Sixteen: Myers-Briggs Personality Types (if you must).

Closing Thoughts

I’ve described these patterns in a general fashion convenient for writing up groups of NPCs, but I want to strongly assert their functionality for PC parties as well, as this use is more likely to fall by the wayside as institutional momentum takes over when starting a new campaign. The next time you form a party, pick an overarching party personality and use the models above to give each player’s character a different subset of that shared theme to make their own. You may find it eases option paralysis and more reliably produces fun social dynamics.

[Featured Image: Ruvim Matskin, Mechanics, 1973]