A consistent problem I’ve always faced in role-playing games is how to quickly introduce new characters to the game, both non-player characters and player characters. Whether the player’s hero has died, or they need to quickly add a supporting cast member to their ensemble, character generation is the bane of many games.
While early editions of D&D may have been relatively fast—roll six stats, pick a class, roll hit points, pick one item, go—they were also, for my taste, quite bland.
The solution of later editions (3E, I’m looking at you) and many other games was options-overload. To make characters more interesting, they made character generation more complex—and slow.
A reaction of many other rules-lite games was to simplify things radically. Fewer (or no) stats! Fewer (or no) classes! Fewer (or no) skills and abilities. To make the game simpler, they removed levers from the game—the Apple solution, of one button (or no button nowadays) to rule them all.
This is a conundrum. How to square the circle or reconcile the two extremes: simple and fast, but with few moving pieces, versus slow and complicated, but with lots of features?
It’s a conundrum I’ve fiddled with many times over in developing Skeleton (my skeleton of a psychedelic metal roleplaying game, still in development). With randomized character generation and leveling I’ve managed to push down “out-of-game” time around 5–10 minutes per new character and 3–5 minutes per new level (or other character advance), which is pretty good.
The Axis of Time
But the challenge remained: many times we would generate characters with a large number of skills and features that never came up in play. Imagine, you’re playing colonists exploring a space hulk, and suddenly come across an infected corpse. What does it matter that back on Terra Nova your character, Ivan Blodnitsky was a car salesman, likes mountaineering, has a hidden VPN chip implanted in his neuranet jack, and packed three original earth-silk turbans in his pack for the new colonies? I mean, it might matter—someday—but right here right now, it does not.
The solution, I realized, was under my nose the whole time.
In pretty much every session I’ve played, players would ask, “what does my character know about this Dunklebork place?” or, if they’d become more confident about my refereeing style, would straight up suggest, “my character is a dwarf, and this is what all dwarves know about Dunklebork.” In this way the fictional location of Dunklebork would come into existence as a shared artifact of our play together.
Why should this apply to Dunklebork, which until this session had been a dot on the world map and a one-line description of “decaying rust-belt mining town being repurposed into the capital of a tulip-farming empire,” but not to items, or animals, or even characters?
All of them are fictional objects, which we as roleplayers use to play at a shared fantasy reality. In fact, I would argue that the players’ characters as key pieces of plays should be even more open to the players’ fictional edification as and when it is required.
Which brought me to the idea of character generation over time, through play. A character starts as a ‘quantum character,’ barely defined beyond a name and an idea, and as they interact with the world, they generate themselves. Over several sessions they become fully fleshed out—and if it turns out the game doesn’t, for example, have undead, then features like turn undead simply never turn up, creating a different fictional world.
Quantum Characters in Action
Much like with the cat in Schrödinger’s box, the character is in a superposition until it is examined. To the question, is the character a wizard or a fighter, the answer becomes simply: yes. Yes, the character is a wizard or a fighter.
Well, which is it?
We’ll find out when it matters.
Let’s imagine a character. Let’s call her Kuya.
Comment: Kuya’s in a freezing bivouac on a cold mountain with a few other characters. She’s got cold weather and mountaineering gear. The gear and her name are the only things on her character sheet.
We start with the referee (R) asking Kuya’s player (P), “Ok, the others were hunting the legendary silver-toothed ice bear. Why is Kuya here? Is she also a hunter?”
P: “No, she’s a painter. She was tagging along behind she wanted to paint the sunrise over the Verdiberg, which was immortalized in the opera To Scream A Butterfly.”
We’ve established she’s a painter, that’s a skill—let’s note that down on the character sheet.
R: “Cool. Well, the weather’s really horrible. The hunters don’t want to leave. Do you wait it out with them?”
P: “Yeah, sure, Kuya’s not some kind of mountain hero.”
R: “The bad weather lasts a couple of days and the temperatures plummit. It’s horrible cold. How good is Kuya’s Endurance? She’ll have to make a test to see how she’s handled the bad weather.”
P: “Well, it’s pretty good.”
R: “Better than average?”
P: “Absolutely–she climbed this mountain by herself!”
R: “Ok, so you want to generate her Endurance with advantage?”
P: “Sure.”
The player now generates Kuya’s Endurance score. She rolls the stat twice, and picks the better result. Let’s say she had a number of advantage tokens in one bowl, she moves one over to the disadvantage bowl. Once she’s run out of advantage tokens (or whenever she wants), she’ll have to start generating additional stats and features with disadvantage. She picks the better results and adds Endurance to her sheet.
And so on. From there, as the adventure progresses, we build up a picture of who Kuya is and what she is capable of. We do the same thing with every player character, with henchmen, with NPCs, monsters, and so on.
We don’t need to know who and what everything is in advance, we discover and play with it as we go along.
By the end of the first session, with Kuya halfway down the frozen mountain, injured companion in tow on a sled, we might have also learned that she knows how to ski and shoot, isn’t very strong, is quite agile, and is afraid of rabbits.
Sure, we don’t know where she grew up, what her dreams are, and so on—but this doesn’t matter yet. It will, or won’t, come out of the character naturally, as play progresses.
For example, as we’re wrapping up the session, we might end with the Referee asking the players, “You’ve set up shelter in the old bunker, and used the old leaves to start a fire. You gather around the hissing, smoking branches to warm yourselves and gnaw on the last biscuit crumbs you have left. Perhaps you’re dreaming of what you’ll eat when you finally get off the mountain, or the first place you’ll go once you’re out of this hellhole. What is Kuya thinking of?”
Player, “Kuya is dreaming of her Granna’s green tea box, of visiting the little apartment again, and making tea with her Granna.”
And now we know that Kuya has a grandmother she cares about and likes tea. We can add those to the character sheet, too. Kuya might well not yet have all her stats, features, or abilities yet–but generating her took no out-of-game time and you can bet that if next session she dies at the hands of an ice troll, her player will be at least a little upset. Which is good, because that’s a connection to a character right there.
Quantum Characters for Every System
You can play roleplaying games with quantum characters regardless of the system you use. All it requires is at least one player familiar with character generation (including what the mechanical benefits of skill, proficiency, feats, abilities, spells, or items might be) and a willingness to accept that the final character will turn out a bit differently, more molded by chance and their adventure than by the player’s design–what you get in return are faster, more organic, less predictable characters.
One more tool that’s helpful is a way to make decisions when the player doesn’t or can’t know things a character might? Simple: roll some dice (1d12 or 2d6 or whatever), and improvise a perspective based on the result:
- Worst Roll Possible (2 on 2d6): the character is terrible at this or hates it or fears it.
- Bad Roll (3–6 on 2d6): the character is bad at this or dislikes it.
- Neutral Roll (7 on 2d6): the character is average or neutral at this or just doesn’t know it.
- Good Roll (8–11 on 2d6): the character is good at this or likes it.
- Best Roll Possible (12 on 2d6): the character is amazing at this, loves it.
Why is the spread for the neutral roll so low? Because that result is quite boring (meh). If the player already knows their character is neutral towards something, they’ll say—otherwise they’re rolling for a sign, more than just another “meh” from the universe.
More degrees of variation? Just add a very bad option for the second worst roll and a very good option for the second best.
And there you have it.
If you really want to take it further, build up random tables for equipment, spells, and skills, and add an option for quantum items–items that are undefined until the player says what they are (within reason).

I hope you’ve enjoyed that little digression into theory and how Skeleton is built. I make those now and again, however, I mostly write rules-loose psychedelic metal roleplaying adventures and modules over on the Stratometaship—my patreon. Right now I’m working on Part 2 of Longwinter, a survival icebox set in a the Barony of Brezim, where the cold goddess Winterwhite (hah!) has descended to claim her icy due. It’s pretty cool (hah!). Meanwhile, I’m also slaving away on layout for the UVG. It’ll be pretty, trust me, and by the time you read this, you’ve probably missed the last chance to preorder the books or dice or maps or or or. Oh, well.
14 replies on “Quantum Characters and the Tabletop Role-playing Game”
How do you see limiting the number of times that a player can add newly remembered skills?
Let’s say that in the first adventure I create a background that works for the mountain pass adventure. You used that adventure to get us together and to get us into the cool new land beyond the mountain. Now I find that my character isn’t well suited for the palace intrigue hooks you have in the rest of the adventure. Do we swap them, add new abilities, or start again?
It’d depend on the system in play.
If the system you’re using is limited to 4 skills when you start the game, that’s what you get. Generally, systems have a limited number of types of attributes: stats, skills, special features, and a limited number of each of those. If you look at something like starting 5E you’ve generally got:
6 different stats
1 background
~5 skills
~2 species attributes
~2-3 class features (wizard spells! barbarian rage!)
A simpler system would have fewer.
Generally, I’d suggest not introducing completely different or new hooks without prior warning, adding additional skills with “levels” (or when it makes sense in a new session—-your game should ideally have some kind of mechanic for customizing and learning new skills), or, yeah, just roping in a secondary character–“and this is my buddy Sven the Talker, who helps with etiquette, because Nungle Headmasher isn’t so good for that.”
I’ve also never run into this as a major problem because I usually let players improvise skills very liberally. If there’s some humorous way to make fishing work in a diplomacy test (I explain to them how it’s all about the patience and waiting for the signal, like the gentle patience of the fisherman upon his reed boat, bla bla bla), I’ll happily go along with it.
Excellent article as always, Luka. What you’re describing here is very similar to the process used by the GM and players in Dungeon World (or other PbTA games). The principle Ask Question And Use The Answers is all part of the “play to find out” ethos of that system. Fundamentally the GM is always the final authority, but a lot of content and character backstory is generated on the fly. The difference is that PbTA games tend to front-load the questions at session zero, which probably rubs some OSR folks the wrong way.
Yeah, I know, it’s pretty similar to things many games do–I did pick up a fair number of my ideas from different games. Let’s not forget Microscope–that is an inspirational piece of game.
It’s funny … I always thought of OSR play as also very “play to find out” 😉 just maybe not spelled out so clearly.
It is as well, for sure. With DW it is certainly hard-baked. I’ve added the “ask questions and use the answers” agenda item to my OSR games in recent years.
Great post! I love how we both had similar problems which prompted this thought process! I would wonder if you can really do this sort of thing for any system. I think, by and large, you can but it requires some tinkering with some of the basic workings of the game. I think adding on a level 0 stage to most games might fix this. During this stage, you can suspend some of the more interconnective aspects of the game and then let them fall into place once the characters are fully assembled.
Yup, I think all games would benefit from some global “0 stage” – not just of the characters, but the world itself, so that each player helps define something about the setting, the game, where it will go, and so on. Not as some kind of “out-of-game” discussion about what’s ok and what’s not (that changes, anyway), but to creatively build the setting.
Well, as for the interconnective aspects … I haven’t found any ttrpgs that really show very strong interconnectivity at the rules level at all. Something something “multiple minigames connected by play.” There’s this idea that rpgs are (can be) coherent, complete rulesets … but naah, not really. They’re frameworks of more-or-less connected game, not games in and of themselves.
Hey,
I recently discovered your stuff through kickstarter, unfortunately, I missed contributing. I hope to next time, or pick up at least the PDF of UVG. I really like reading your posts, and I am enjoying the art. Between picking up D&D again and looking at your art, I have drawing again. Thanks.
Jeremy
Oh, thank you so much, Jeremy! Yeah, sometimes I think the D&D is incidental to the art … it provides a way to make stuff up with friends, but the drawing, well, it’s like meditation for me. I’d go nuts without it.
This rules. I’ve been brainstorming a “you wake up somewhere spooky and spacey with no memories” type game where the players would regain memories as they progress, and with the memories come abilities, and this might just be the way I handle that.
You might want to check out Psi*Run by Meguey Baker. I have not played it yet but the podcast Farrier’s Bellows and I am totally intrigued: https://www.gauntlet-rpg.com/the-farriers-bellows/psirun
Basically, you are super-powered individuals on the run from a society that hates you, with no memory. When you roll, you can spend dice to either achieve success for something, or recovery a memory / answer a question on your sheet. Sounds totally fascinating.
It sounds interesting, though possibly a bit limited in kind of games you can generate. I suppose I view “quantum characters” as more of a meta-fictional convenience, where the players only determine facts about character as they become needed – sort of giving them the same “powers” I take on as a DM, where I define what’s inside a specific chest, building, or town, as it becomes relevant.
I like this. Going on vacation and might try this with the kiddos. Maybe some bastardized hodge podge of 5e classes broken into skills.
All my games are bastardized hodgepodges … over the years the rules have accreted and mashed together ^_^