Andrew Stanton is the superstar creator of WALL-E, John Carter, Finding Nemo, A Bug’s Life, Toy Story, and many more. In 2012 he gave the above TED talk collecting all the lessons he’s learned about storytelling.

A good story, says Stanton, makes a promise. That promise might be as simple as, “Once upon a time…,” but the crucial thing is that the audience believes that this will be a story worth hearing. The promise, therefore, invites the audience to engage with the story. It’s like a foot in the door. It’s incredibly important because, without that initial engagement (and the trust that comes with it), the storyteller has nothing to build on or with.

The nature of the promise also means that “stories are inevitable if they’re good, but they’re not predictable.” A statement which, I think, can be interpreted in many ways: That we may know where a story is going, but not the path which is taken. Or that we may know the direction of Fate, but not necessarily the specific form it will take. Or that when we look back at the story, it seems as if everything is perfectly aligned and could have gone no other way than it did, but we could not have foreseen it.

In other words, the story must faithfully keep its promise, but it should still surprise and delight the audience.

(For more on how you can achieve this effect in an RPG, check out Random GM Tip: Three Point Plotting.)

The promise also creates a window of opportunity for the storyteller, and they have to capitalize on that by making the audience care about the story.

There are many ways a storyteller can do this — character, theme, craft, etc. — but one particular lesson he talked about leapt out to me as a Game Master:

THE UNIFYING THEORY OF 2+2

The absolute best way to get the audience to care about the story is to get them involved with the story; to get them actively thinking about the story. And the best way to do that is to make them work for the story.

In other words, don’t show your audience FOUR. Show them 2 + 2 and make them do the math.

Note that 2 + 2 isn’t difficult. The point isn’t necessarily to challenge the audience. (Although it can be: There’s a reason why the mystery genre is popular. A properly placed insoluble problem can actually be even more effective, which his why everyone remembers the end of Inception.)

The point is that even the simplest act of connecting the dots engages the audience. It makes them, on a primal level, a part of the story. They are thinking about the story and they have opinions about it. Once you’re part of something, you care about it. As Stanton says, “A well-organized absence of information pulls us in.” We have a desperate need to complete an unfinished sentence.

Take Citizen Kane, for example. (Spoilers ahoy!) Imagine how much less effective would that movie be if, at the end of the movie, Orson Welles had Joseph Cotton’s character say, “Rosebud was his childhood sled. Despite the poverty and the hardship of his youth, he must have always missed the simple, uncomplicated joys of his youth and the unconditional love of his mother.” The beauty of Citizen Kane as a movie is, in fact, the immense artistry Welles employs so that, rather than spoonfeeding that moral to the audience, he has prepared the audience so that all the nuance and emotional complexity of that idea becomes as simple as 2+2 when he shows them the image of the sled.

(I’ll note that this can actually create paradoxes in storytelling, where sometimes the more effort you spend explicitly and plainly explaining something to the audience can actually result in the audience understanding it less, because the lack of engagement causes them to mentally skim past it.)

And it’s a “unifying theory” because it can apply to almost everything in a film: Characters, plot points, exposition, etc.

The trick, of course, is that the audience wants to work for their meal, but they usually don’t want to know that they’re doing that. So it’s also the storyteller’s job to hide the audience’s work from the audience.

To use our Citizen Kane example again, when you see the sled at the end of the movie, you don’t consciously think, “Oh! A tricky problem! Let me think this through!” Ideally, the storyteller has set you up so that you simply see 2+2 and reflexively think, “Four.”

(Again, there are exceptions, like the central conundrum of most mystery stories.)

IN YOUR GAME

Stanton, of course, is talking about animation and filmmaking, and we know that we can’t just take the same storytelling techniques that we see on screen and use them in our RPG games. RPGs are a different medium; one in which the players have an unprecedented freedom and for which plots should not be prepped.

But the Unifying Theory of 2+2 still works!

All you need to do is give your players the equation and then left them take the final step.

In fact, the interactivity of a roleplaying game can actually enhance the technique because the players can actively investigate. In a film, the audience has to passively receive the equation, but in an RPG, the players can go looking for the twos. Or maybe they have the twos, and they need to experiment to figure out the correct mathematical operator.

(I think I’ve broken the metaphor.)

Matryoshka techniques like matryoshka searches and matryoshka hexes are built around this idea that “completing the equation” will mean taking action as a character, and that doing so can give the player both ownership and control over the answer we find.

For other techniques you can use to help make your players care about your campaign, check out Random GM Tips: Getting Your Players to Care.

For the first couple decades after D&D, virtually all roleplaying games looked fundamentally similar: There was a GM who controlled the game world, there were players who each controlled a single character (or occasionally a small stable of characters which all “belonged” to them), and actions were resolved using diced mechanics.

Starting in the early ’90s, however, we started to see some creative experimentation with the form. And in the last decade this experimentation has exploded: GM-less game. Diceless games. Players taking control of the game world beyond their characters. (And so forth.) But as this experimentation began carrying games farther and farther from the “traditional” model of a roleplaying game, there began to be some recognition that these games needed to be distinguished from their progenitors: On the one hand, lots of people found that these new games didn’t scratch the same itch that roleplaying games did and some responded vituperatively to them as a result. On the other hand, even those enthusiastic about the new games began searching for a new term to describe their mechanics — “story game”, “interactive drama”, “mutual storytelling”, and the like.

In some cases, this “search for a label” has been about raising a fence so that people can tack up crude “KEEP OUT” signs. I don’t find that particularly useful. But as an aficionado of Scott McCloud’s Understanding Comics, I also understand the power of proper definitions: They allow us to focus our discussion and achieve a better understanding of the topic. But by giving us a firm foundation, they also set us free to experiment fully within the form.

For example, people got tired of referring to “games that are a lot like Dungeons & Dragons“, so they coined the term “roleplaying game” and it suddenly became a lot easier to talk about them (and also market them). It also allowed RPGs to become conceptually distinct from “wargames”, which not only eliminated quite a bit of confusion (as people were able to separate “good practices from wargames” from “good practices for roleplaying games”), but also allowed the creators of RPGs to explore a lot of new options.

Before we begin looking at how games like Shock: Social Science Fiction, Dread, Wushu, and Microscope are different from roleplaying games, however, I think we first need to perfect our understanding of what a roleplaying game is and how it’s distinguished from other types of games.

WHAT IS A ROLEPLAYING GAME?

Roleplaying games are defined by mechanics which are associated with the game world.

Let me break that down: Roleplaying games are self-evidently about playing a role. Playing a role is about making choices as if you were the character. Therefore, in order for a game to be a roleplaying game (and not just a game where you happen to play a role), the mechanics of the game have to be about making and resolving choices as if you were the character. If the mechanics of the game require you to make choices which aren’t associated to the choices made by the character, then the mechanics of the game aren’t about roleplaying and it’s not a roleplaying game.

To look at it from the opposite side, I’m going to make a provocative statement: When you are using dissociated mechanics you are not roleplaying. Which is not to say that you can’t roleplay while playing a game featuring dissociated mechanics, but simply to say that in the moment when you are using those mechanics you are not roleplaying.

I say this is a provocative statement because I’m sure it’s going to provoke strong responses. But, frankly, it just looks like common sense to me: If you are manipulating mechanics which are dissociated from your character — which have no meaning to your character — then you are not engaged in the process of playing a role. In that moment, you are doing something else. (It’s practically tautological.) You may be multi-tasking or rapidly switching back-and-forth between roleplaying and not-roleplaying. You may even be using the output from the dissociated mechanics to inform your roleplaying. But when you’re actually engaged in the task of using those dissociated mechanics you are not playing a role; you are not roleplaying.

I think this distinction is important because, in my opinion, it lies at the heart of what defines a roleplaying game: What’s the difference between the boardgame Arkham Horror and the roleplaying game Call of Cthulhu? In Arkham Horror, after all, each player takes on the role of a specific character; those characters are defined mechanically; the characters have detailed backgrounds; and plenty of people have played sessions of Arkham Horror where people have talked extensively in character.

I pick Arkham Horror for this example because it exists right on the cusp between being an RPG and a not-RPG. So when people start roleplaying during the game (which they indisputably do when they start talking in character), it raises the provocative question: Does it become a roleplaying game in that moment?

On the other hand, I’ve had the same sort of moment happen while playing Monopoly. For example, there was a game where somebody said, “I’m buying Boardwalk because I’m a shoe. And I like walking.” Goofy? Sure. Bizarre? Sure. Roleplaying? Yup.

Let me try to make the distinction clear: When we say “roleplaying game”, do we just mean “a game where roleplaying can happen”? If so, then I think the term “roleplaying game” becomes so ridiculously broad that it loses all meaning. (Since it includes everything from Monopoly to Super Mario Brothers.)

Rather, I think the term “roleplaying game” only becomes meaningful when there is a direct connection between the game and the roleplaying. When roleplaying is the game.

It’s very tempting to see all of this in a purely negative light: As if to say, “Dissociated mechanics get in the way of roleplaying and associated mechanics don’t.” But it’s actually more meaningful than that: The act of using an associated mechanic is the act of playing a role.

As I wrote in the original essay on dissociated mechanics, all game mechanics are — to varying degrees — abstracted and metagamed. For example, the destructive power of a fireball is defined by the number of d6’s you roll for damage; and the number of d6’s you roll is determined by the caster level of the wizard casting the spell. If you asked a character about d6’s of damage or caster levels, they’d have no idea what you were talking about (that’s the abstraction and the metagaming). But they could tell you what a fireball was and they could tell you that casters of greater skill can create more intense flames during the casting of the spell (that’s the association).

So a fireball has a direct association to the game world. Which means that when, for example, you make a decision to cast a fireball spell you are making a decision as if you were your character — in making the mechanical decision you are required to roleplay (because that mechanical decision is directly associated to your character’s decision). You may not do it well. You’re not going to win a Tony Award for it. But in using the mechanics of a roleplaying game, you are inherently playing a role.

WHAT IS A STORYTELLING GAME?

So roleplaying games are defined by associated mechanics — mechanics which are associated with the game world, and thus require you to make decisions as if you were your character (because your decisions are associated with your character’s decisions).

Storytelling games (STGs), on the other hand, are defined by narrative control mechanics: The mechanics of the game are either about determining who controls a particular chunk of the narrative or they’re actually about determining the outcome of a particular narrative chunk.

Storytelling games may be built around players having characters that they’re proponents of, but the mechanical focus of the game is not on the choices made as if they were those characters. Instead, the mechanical focus is on controlling the narrative.

Wushu offers a pretty clear-cut example of this. The game basically has one mechanic: By describing a scene or action, you earn dice. If your dice pool generates more successes than everyone else’s dice pools, you control the narrative conclusion of the round.

Everyone in Wushu is playing a character. That character is the favored vehicle which they can use to deliver their descriptions, and that character’s traits will even influence what types of descriptions are mechanically superior for them to use. But the mechanics of the game are completely dissociated from the act of roleplaying the character. Vivid and interesting characters are certainly encouraged, but the act of making choices as if you were the character — the act of actually roleplaying — has absolutely nothing to do with the rules whatsoever.

That’s why Wushu is a storytelling game, not a roleplaying game.

More controversially, consider Dread. The gameplay here looks a lot like a roleplaying game: All the players are playing individual characters. There’s a GM controlling/presenting the game world. When players have their characters attempt actions, there’s even a resolution mechanic: Pull a Jenga block. If the tower doesn’t collapse, the action succeeds. If the tower does collapse, the character is eliminated from the story.

But I’d argue that Dread isn’t a roleplaying game: The mechanic may be triggered by characters taking action, but the actual mechanic isn’t associated with the game world. The mechanic is entirely about controlling the pace of the narrative and participation in the narrative.

I’d even argue that Dread wouldn’t be a roleplaying game if you introduced a character sheet with hard-coded skills that determined how many blocks you pull depending on the action being attempted and the character’s relevant skill. Why? Because the resolution mechanic is still dissociated and it’s still focused on narrative control and pacing. The mechanical decisions being made by the players (i.e., which block to pull and how to pull it) aren’t associated to decisions being made by their character. The fact that the characters have different characteristics in terms of their ability to be used to control that narrative is as significant as the differences between a rook and a bishop in a game of Chess.

GETTING FUZZY

Another way to look at this is to strip everything back to freeform roleplaying: Just people sitting around, pretending to be characters. This isn’t a roleplaying game because there’s no game — it’s just roleplaying.

Now add mechanics: If the mechanics are designed in such a way that the mechanical choices you’re making are directly associated with the choices your character is making, then it’s probably a roleplaying game. If the mechanics are designed in such a way that the mechanical choices you’re making are about controlling or influencing the narrative, then it’s probably a storytelling game.

But this gets fuzzy for two reasons.

First, few games are actually that rigid in their focus. For example, if I add an action point mechanic to a roleplaying game it doesn’t suddenly cease to be a roleplaying game just because there are now some mechanical choices being made by players that aren’t associated to character decisions. When playing a roleplaying game, most of us have agendas beyond simply “playing a role”. (Telling a good story, for example. Or emulating a particular genre trope. Or exploring a fantasy world.) And dissociated mechanics have been put to all sorts of good use in accomplishing those goals.

Second, characters actually are narrative elements. This means that you can see a lot of narrative control mechanics which either act through, are influenced by, or act upon characters who may also be strongly associated with or exclusively associated with a particular player.

When you combine these two factors, you end up with a third: Because characters are narrative elements, players who prefer storytelling games tend to have a much higher tolerance for roleplaying mechanics in their storytelling games. Why? Because roleplaying mechanics allow you to control characters; characters are narrative elements; and, therefore, roleplaying mechanics can be enjoyed as just a very specific variety of narrative control.

On the other hand, people who are primarily interested in roleplaying games because they want to roleplay a character tend to have a much lower tolerance for narrative control mechanics in their roleplaying games. Why? Because when you’re using dissociated mechanics you’re not roleplaying. At best, dissociated mechanics are a distraction from what the roleplayer wants. At worst, the dissociated mechanics can actually interfere and disrupt what the roleplayer wants (when, for example, the dissociated mechanics begin affecting the behavior or actions of their character).

This is why many aficionados of storytelling games don’t understand why other people don’t consider their games roleplaying games. Because even traditional roleplaying games at least partially satisfy their interests in narrative control, they don’t see the dividing line.

Explaining this is made even more difficult because the dividing line is, in fact, fuzzy in multiple dimensions. Plus there’s plenty of historical confusion going the other way. (For example, the “Storyteller System” is, in fact, just a roleplaying game with no narrative control mechanics whatsoever.)

It should also be noted that while the distinction between RPGs and STGs is fairly clear-cut for players, it can be quite a bit fuzzier on the other side of the GM’s screen. (GMs are responsible for a lot more than just roleplaying a single character, which means that their decisions — both mechanical and non-mechanical — were never strictly focused on roleplaying in the first place.)

Personally, I enjoy both sorts of games: Chocolate (roleplaying), vanilla (storytelling), and swirled mixtures of both. But, with that being said, there are times when I just want some nice chocolate ice cream; and when I do, I generally find that dissociated mechanics screw up my fun.

2020 ADDENDUM: TABLETOP NARRATIVE GAMES

If we can move beyond arguing that vanilla ice cream is actually chocolate ice cream, we have the opportunity to step back and recognize that these are both different types of ice cream. I propose that both roleplaying games and storytelling games are tabletop narrative games.

Now, here’s the cool thing: Recognizing that these are different things within a broader paradigm will make it easier for us to explore that paradigm. Much like having a different word for different colors makes it easier to distinguish those colors, clearly seeing the distinctions between associated mechanics and narrative control mechanics will not only make it easier for us to develop better games of those types, it will also likely make it easier for us to discover completely new types of games.

Compare this to video games, for example. Grand Theft Auto started as a maze-chase game. When they iterated on that design, nobody launched a holy war insisting that this was the One True Way of making maze-chase games. They said, “Oh. Hey. Look at this cool new type of game.” Instead of spending twenty years arguing that Grand Theft Auto 3 was a maze-chase game just like Pac-Man (and how dare you suggest otherwise?!), they identified the new form as an open-world game and spent twenty years making lots and lots of open-world games (that were in no way still trying to be Pac-Man).

Video games and board games do this all the time. And we have better and more varied video games and board games as a result. Wouldn’t it be great if tabletop narrative games could reap the same benefits?

DISCUSSING
In the Shadow of the Spire – Session 46A: Among Madmen

At the last possible moment, Zairic twisted aside so that the arrow lodged in his shoulder instead of his heart. Letting his book drop to the floor, Zairic vaulted over the high arm of his chair and jumped for cover. In mid-leap, he released a fireball through the window. Tee ducked down as the fiery inciting pellet passed over her head and avoided the brunt of it almost completely, but Elestra (standing in the open further down the alley) was caught by the edge of it.

Most of the others – clumped together across the street and still debating how they could (or would or should) use Elestra’s homunculi – missed the flash of the fireball. Fortunately, Ranthir – who was providing the daisy-chained camouflage near the mouth of the alley – recognized it for what it was. “Fireball!” he shouted, hurrying into the alley.

The fiction-mechanics cycle is arguably the heart of the roleplaying game experience: The ways in which we use mechanics to create fictional outcomes; declare fictional actions that are resolved mechanically; and use the outcome of either to feed back into the other form an intricate and interwoven dance at the gaming table.

A key component of this dance is how mechanical outcomes are explained in the fiction. For simple, straightforward intentions with unambiguous results, this is often so obvious that one can easily miss that something is actually happening: The player said they wanted to jump over the chasm; the dice said they succeeded; therefore, they land on the other side of the chasm.

Intriguingly, therefore, it is often true the failure requires more of an explanation than success: Success, after all, merely assumes that the stated intention which triggered the mechanical resolution was achieved. Failure, on the other hand, almost seems to demand an explanation for why the character wasn’t able to achieve their desired outcome.

(And this is before we even start considering advanced techniques like failing forward.)

There are a number of techniques you can use in creating these explanations, and different RPG rulesets will often help you in different ways. A universal technique I find useful is explicitly thinking about different factors in the game world that could affect outcome. It’s really useful for keeping things fresh and varied.

(One key insight from this is that you can often make the description of success more interesting by lightly spicing it with the same details and factors that we use to explain failure.)

Something else to consider is the often unexamined assumption of who at the table is responsible for providing these explanations. In my experience, this almost always falls on the GM in their role as adjudicator and world-describer. Every so often, though, the infectious spirit of communal improv will unleash itself and people all around the table will start collaborating on the answer. And another key insight is that, as the GM, you can prompt the players to get involved in explaining outcomes.

(Matthew Mercer, for example, has made, “How do you want to do this?” particularly famous.)

In fact, you can go further than that and create specific expectations for action resolution in which describing the fictional implications of mechanical results defaults to the players. (Storytelling games often do this because their mechanics revolve around determining which player is in control of a narrative outcome.)

But I digress.

What I’m particularly interested in talking about right now is a very specific slice of these table interactions: The moment where a mechanical outcome prompts a conversation between characters, which I’m going to refer to as ex post facto roleplaying. Here the character dialogue is being triggered by or being described as the key factor in an action’s resolution.

In this session, for example, most of the PCs failed a Spot check to notice the flash from a fireball spell going off around a corner.

Why call for this check at all? I mean, it’s a fireball spell, right? Shouldn’t it be really obvious? Well, to some extent this depends on how much noise you think a fireball creates — is it a huge detonation or a more ephemeral flash of flame? More importantly, what I was primarily concerned about here was how quickly they would react to the fireball: Would they be able to leap into action and immediately join the fight? Or get caught flat-footed and have to wait a round before being able to rush to Tee’s aid?

In this case, the players asked the same question in a breakdown that looked something like this:

• Why wouldn’t we immediately notice the fireball?
• We must have been distracted.
• What could we have been distracted by?
• We must have all been continuing our debate about using the homunculi!

And then they briefly acted out a few lines of that dialogue, giving Ranthir’s player (who had succeeded on his Spot check) an opportunity to interrupt by them by shouting, “Fireball!”

This is a good example of these ex post facto roleplaying moments, which are often played as kind of funny throw-away moments. But they can, of course, also be more protracted and/or take on a more serious tone, particularly if you make a more conscious effort to notice, prompt, and/or define these moments.

In fact, rather than just reacting to skill checks with dialogue, you can also deliberately frame skill checks to set up roleplaying interactions. Using mechanics as a roleplaying prompt like this is described in more detail in Rulings in Practice: Social Skills.

Campaign Journal: Session 46B – Running the Campaign: Speak with Dead SFX
In the Shadow of the Spire: Index

“If you were teaching a intro-level college class on roleplaying game design, what would be the reading list?”

Interesting question.

I’m going to design this as a survey/history course. And you’ll need to snag copies of these at the campus bookstore:

1974 D&D
Traveller
GURPS or Champions
Paranoia (1^st Edition)
Vampire the Masquerade
Amber Diceless Roleplay
Burning Wheel
Apocalypse World

And we’ll wrap the course up by comparing D&D 3E, 4E, and 5E, with a particular focus on how they responded to design trends.

D&D 1974

This is the beginning. The baseline for everything that follows and a frame for discussing the proto-history and origin of RPGs from Kriegsspiels to David Wesley to Dave Arneson.

TRAVELLER (1977)

Traveller does triple duty for me.

• It gives insight into the first generation of RPGs that were responding to D&D.
• One of the first science fiction RPGs, after Starfaring and Metamorphosis Alpha.
• Includes a Lifepath system, giving us a first step in looking at different approaches to character creation.

GURPS / CHAMPIONS
(’80s Editions)

The birthplace of the well-supported generic/universal RPG system.

A central thesis of this class will be that RPGs pretty universally pushed for “accurate simulation” as a primary design goal through the late ‘70s and ‘80s. Whichever one of these games we choose to look at it will serve as a great exemplar of that trend.

It will also give us point-buy character creation in its fullest flower, allowing us to clearly show the difference between character generation in 1974 D&D and the character crafting which would come to largely dominate the hobby.

PARANOIA
(1st Edition)

This is kind of an oddball choice. Every intro class has one of these on the reading list, right?

But it’s here for a reason.

On the one hand, Paranoia is a comedy game, which gives us a nice, sharp look at the emerging diversification of creative agendas in the ‘80s.

On the other hand, it also examines the unexamined “simulation = good” trend in the ‘80s. Paranoia is a lighthearted comedy game, but its first edition features, among other things, a Byzantine three-tier skill specialization system, because “simulation = good” even if it made no sense for what the game was actually trying to achieve.

(I will give extra credit to any student arguing that the incredible minutia of the system was actually part of the satire of a Kafkaesque government bureaucracy. They’re wrong, but it shows they’re thinking critically about this.)

WHITHER THE UNIVERSAL SYSTEM?

Universal RPGs were VERY popular in the ‘80s and ‘90s, and then they weren’t. There are still some around today, of course, but the only truly popular ones are 20+ years old.

One theory is that the niche was definitively filled by GURPS, and no other universal RPG could ever compete with GURPS’ library of support material.

The other is that universal RPGs were at their strongest because of the “simulation = good” paradigm. Once you move past the idea, as demonstrated in Paranoia, that “good system” is a Platonic ideal divorced from a game’s creative agenda, the appeal of a “universal system” is not eliminated, but significantly diminished.

VAMPIRE: THE MASQUERADE

This brings us to Vampire: The Masquerade, a hugely important game that grappled mightily with the idea of pushing a creative agenda other than simulation.

The Storyteller system is interesting to analyze because (a) it ultimately fails to achieve its storytelling goal (and analyzing failure is a great way to learn) and (b) it’s virtually impossible to understand GNS theory without the context of Storyteller.

AMBER

Before we get to GNS, though, Amber Diceless Role-Playing will be our exemplar of the snap-back against detailed simulation.

Emphasized by its diceless engine, Amber was one of many early ‘90s games that were fed up with complexity and bounced to the opposite extreme. It’s an elegant tour de force for designing mechanics customized to the creative agenda and setting of the game.

Plus, Amber features alternative structures for organizing campaigns and extending player beyond the session. So we’re getting a lot of mileage from this one title.

BURNING WHEEL

There are a lot of Forge-era indie games we could choose to spotlight GNS theory. One could argue that we should go with a game by Ron Edwards or D. Vincent Baker. (We’ll cover the latter with Apocalypse World.)

Burning Wheel is, ultimately, just a better game and synthesizes a wider range of innovations. So that’s what I’m tapping.

APOCALYPSE WORLD

Which brings us to Apocalypse World.

Powered by the Apocalypse is the single biggest non-D&D influence on RPG design in the last fifteen years, so it’s basically essential. And, as I just mentioned, D. Vincent Baker is a seminal figure and his design philosophy should be highlighted. So this is another game that’s doing double duty.

WotC-ERA D&D

The big wrap-up for our course is a comparison of D&D 3E, 4E, and 5E.

D&D is, of course, the 10,000 lbs. gorilla in the RPG design room. If we’re teaching an intro course, we absolutely need to cover its evolution.

Post-1974, however, D&D has been extremely reactive in its design. It largely does not innovate, but its massive gravity means anything it refines is reflected back into the industry in a massively disproportionate way. (Take, for example, the concept of advantage/disadvantage as modeled by rolling twice and taking the better/worse result. This was an exceptionally obscure mechanic pre-2014, but after D&D 5E used it, you can find it everywhere.)

Having broadly covered the history of RPG design, therefore, looking at how D&D reacted to (or didn’t react to) those design trends is a great way to review and critically analyze everything we’ve learned in the course.

The final list also gives us (coincidentally, I didn’t actually plan this) wo games from each decade (‘70s through ‘10s) with an extra dollop of D&D. That’s a good gut-check to make sure I wasn’t getting too biased in my selections.

NOTABLE ABSENCES

There are a few notable things missing from this reading list.

FATE, which had a massive influence in the decade before Powered by the Apocalypse, serving as the system for any number of games.

Storytelling Games. Only peripherally looking at how STGs have influenced RPG design is iffy. You can easily make a case for throwing in Once Upon a Time or Microscope or Ten Candles.

RPGs in a Box. These are games like Arkham Horror, Gloomhaven, and Descent. They aren’t actually RPGs, but they’re in the same design space.

Starter Sets. There are unique design considerations in making an effective starter set, but we didn’t cover them at all.

Organization-based Play. This would be a game like Ars Magica, Blades in the Dark, or Pendragon. This has been a persistent design goal for RPGs since Day 1. I can touch on it a bit with Traveller, but it’s really exploded in the past decade and not having a more recent example is a limitation.

Call of Cthulhu. Just because it’s Call of Cthulhu. But my goal wasn’t Most Important RPGs (Call of Cthulhu is easily Top 5), it was Introduction to RPG Game Design. There’s stuff that I could use Call of Cthulhu to teach, but not enough hooks to knock others off this list.

After all, there are only so many hours in a semester.

FURTHER READING
It’s Time for a New RPG
A History of Stat Blocks

Go to Part 1

GM: So now, at last, Minas Tirith is besieged, enclosed in a ring of foes. The Rammas has been broken, and all Pelennor abandoned to the Enemy. All night the watchmen on the walls hear rumors of the enemy that roam outside, burning field and tree, and hewing any man they find abroad, living or dead. The numbers that have already passed over the River cannot be guessed in the darkness, but when morning, or its dim shadow, stole over the plain, you can see that the plain is dark with their marching companies, and as far as your eyes can strain in the mirk there sprout, like a foul fungus-growth, all about the beleaguered city great camps of tents, black or somber red.

Gandalf: Hmm… can I use my Fireworks skill to whip up a nuclear bomb?

GM: Sure.

Always Say Yes is a bit of GMing advice that gets circulated quite a bit. The example above takes it to a silly extreme, but it illustrates the central problem: There clearly is a point where the GM needs to say no, but the simplistic “Always Say Yes” not only doesn’t help them figure out what that is, it’s more likely to mislead them. Its sole saving grace is that Always Say Yes is at least preferable, in my opinion, to Always Say No, which is where railroading will often take GMs.

Always Say Yes comes from theatrical improv games. And the reason it’s bad advice for an RPG is that improv and RPGs have fundamentally different narrative structures.

In improv, the performers are collaborating to create a world. You always accept new facts about the world because negation doesn’t take you anywhere creatively:

Actor 1: Here we are at Disney World!

Actor 2: No, we’re at the White House.

But even in improv it can mislead performers, because Always Say Yes only applies to the worldbuilding (i.e., stated facts about the world). It doesn’t mean that your character can’t oppose or say no to another character. If the maxim is misapplied in this way, it becomes impossible for improv scenes to have any conflict, draining them of interest.

Note: This is also why Always Say Yes actually can be useful in certain storytelling games based around narrative control mechanics. Many of those games, although not all, feature collaborative world-building.

In RPGs, on the other hand, the GM creates the world and the players take on the roles of characters who live in that world. The players, therefore, are primarily playing a role, not worldbuilding, while the GM’s primary duty is being an arbiter of the fictional reality those characters inhabit. From a game perspective, this is much more like 20 Questions than it is an improv theater game: There’s a “truth” (e.g., the object selected in 20 Questions) that the GM knows to be true and which is being communicated to the players.

Imagine for a moment playing 20 Questions while always saying Yes: It’s technically possible. In fact, just like Gandalf setting off tactical nukes on the Fields of Pelennor, you’re almost guaranteed to “win” every time. And yet you’ve fundamentally broken the game.

So the improv-style Always Say Yes to Worldbuilding doesn’t work in an RPG.

What if we change the target to something like Always Say Yes to Player Plans? This is probably getting us closer to something useful, but it still doesn’t hold up to scrutiny. For a simple example, consider someone saying, “I cast a spell and teleport back to Dweredell!” Okay… but does your character actually have a spell that does that? RPGs have rules that mechanically define characters and what they can (and, importantly, can’t) do.

In addition to a character mechanically lacking the ability to do something, you’ve also got world state (e.g., there’s a teleport interdiction field there) and NPC actions (e.g., someone counterspells the teleport) that can – and at least some of the time should! – negate player intention.

At this point we can easily glide over into another common maxim: Say yes or roll the dice.

This can probably be more accurately understood as “say yes unless the mechanics say no” (since not all mechanics use diced outcomes), but even that’s still misleading because there can, once again, be non-mechanical reasons why a player’s proposed action won’t work (e.g., they want to go the local mage’s guild, but the GM knows there’s no mage’s guild in this village).

DEFAULT TO YES

All of which is why I prefer to use Default to Yes as my maxim of choice here.

Its meaning expands to, “When the players say they want to do something, you should default to letting them do it unless you have a specific and interesting reason not to.” It fulfills the same crucial function of steering the GM away from contrarianism, but also provides a clear standard they can use to figure out when they should be saying No.

Furthermore, if it turns out that there is, in fact, a reason not to Default to Yes, then you can also use the Spectrum of GM Fiat to find the appropriate response:

• Yes, and…
• Yes, but…
• No, but…
• No

And, if those aren’t right, you can always shift to, “Maybe, let’s roll the dice and find out…”

The underlying principle here is the players will generally propose doing things that they would enjoy and outcomes that they desire. So, generally speaking, letting them do and achieve those things will make for happy players.

So why not go back to Always Say Yes, then?

Largely because there’s other stuff that he players also enjoy. That might be challenge (e.g., knowing that they’ve actually earned their victories), simulation (e.g., they want to know that they’re exploring a “real” place), or drama (e.g., struggle is narratively interesting). To generalize, failure is interesting.

GM DON’T #20.1: PREP IS ALL, PREP IS LAW

If we return to our example of the PCs looking for a mage’s guild in a village where the GM knows no mage’s guild exists, however, there’s another pitfall you can stumble into while trying to enforce the fictional reality of the game world.

Imagine that the PCs have come to a large city, perhaps one with half a million people living in it. A player says, “Okay, I want to find a blacksmith who can repair my sword.” You check your description of the city, but it turns out that you didn’t include any smithies in your notes.

This is just like the mage’s guild, right? Your notes for the village didn’t include a mage’s guild, so there was no mage’s guild. Your notes for the city don’t include a blacksmith, so there are no blacksmiths.

… right?

Probably not.

Obviously a large medieval fantasy city is almost certainly going to have at least one smithy. The key insight here is that even if you have hundreds and hundreds of pages of notes detailing your city, you still won’t have recorded every single person, place, and thing. It would be a mistake to believe that the only things that exist in the game world are the things you’ve explicitly written down. Instead, when confronted with a situation like this, the question you should ask yourself is, “Given everything I know about the game world, would it make sense for this to exist? And, if so, what form would it take?”

And, once again, you want to Default to Yes.

If you know that there are no Ivy League schools in Kentucky, then it’s fine to say No if the players want to go looking for one in Paducah. But if they’re just looking for a university with an archaeology department somewhere in the state where they could ask some questions of an expert, then they should be able to find one somewhere in the state!

Once you’ve grokked that principle, the next thing to understand is that this extends beyond cities and states. It scales to almost every level of the game world.

For example, let’s zoom all the way in on a dungeon room. Here’s one from So You Want to Be a Game Master:

AREA 15: CRYPTIC LIBRARY

The room is of crimson and dark wood beneath a vaulted ceiling. A set of tall double doors, matching the ones you entered through, faces you on the opposite side of the room. A large pentagonal mahogany table stands in the center of the chamber. There are five large bookcases stuffed full of tomes and scrolls along the paneled walls. Sunlight streams in through bay windows with a built-in bench. The windows are leaded stained glass depicting a subtle pattern of golden florets.

Consider those bookcases. In our key, we might even include additional details about the books:

BOOKCASES

The books and other documents here are all of an occult nature. There is a focus on Natharran mythology, particularly an enigmatic figure known as Basp-Attu. A DC 15 Intelligence (Arcana) check identifies Basp-Attu as a dual-bodied demon said to have been born from the mixed ichor of two dead gods during the War of Falling Stars.

But even with all of this detail, we still haven’t completely detailed the bookcase.

• What if a PC wants to grab the heaviest book off the shelf and use it to bludgeon the cloaker that’s just ambushed them? How big is the biggest book? Is it big enough to do serious damage?
• “Are there are any books with a green cover? Maybe I can fool the goblins into thinking it’s a copy of the Verdigris Bible.”
• “I want to tip the bookcase over on top of the cloaker!” Have the bookcases been fastened to the wall?
• “What’s the single most useful book about Basp-Attu here?”
• “Are there volumes here that would be useful additions to the collection at the libram at my kolledzh?”

To be clear, the point isn’t that you should be prepping answers for all of these questions. The point is to recognize that you can never prep the entire world. The players will always have a question you don’t know the answer to and you will always be performing acts of creative closure at the table.

Unless, of course, you make the mistake of believing that you have prepped everything and that your prep is absolute and immutable law. If you do that, then you’ll have started walking down one of the many paths to pixelbitching — aka, playing a game of Guess What the GM Prepped.

Feng Shui, the roleplaying game of Hong Kong action films, actually pushes this concept of the uncertain game world and creative closure one step further by specifically giving the players unilateral narrative authority to simply declare that a prop they want is present in any fight scene. Want to fight with a ladder like Jackie Chan in First Strike? Bring a rolled up magazine to a knife fight like Jason Bourne? Throw your jacket over a support cable and zipline to the ground? You don’t need to ask the GM if there’s a ladder at the construction site, a magazine on the coffee table, or a cable attached to the building.

This works because when the GM says, “The elevator doors open to reveal a floor full of cubicles,” everyone at the table is, in fact, picturing a different place. There are some details — like the fact there are computer monitors on the desks — that are shared in all of our imagined spaces, but there are other details that aren’t. As we interact with and explore the space together, though, our visions adapt and converge. This mostly happens without friction and often without us ever really consciously thinking about how our mental image of the cubicle farm is morphing and changing.

While in my experience the GM’s vision of the space probably morphs less than the other players (because they are, in fact, the arbiter of the world), it’s important for them to understand that their imperfect vision will, in fact, also be adapting and becoming more fleshed out: I didn’t know there was a smithy in this town, but now I do. I had not specifically pictured staplers on those desks, but once Aaron snagged one I can clearly see them. I had no idea that The Two-Faced Demon by Attansea Millieu was a rare tome, but apparently there’s a copy sitting on this bookshelf.

Discovery is the great joy of collaboration, and if you can truly open yourself to that collaboration, who knows what exciting adventures those discoveries may take you on?