Early in the history of the hobby, a number of factors (including primordial Gygaxian authoritarianism) led to the perception/practice of game masters as petty tyrants who ruled their tabletop principalities with Dorito-dusted iron fists. After two or three decades of this ancien régime, resentment mounted to point of revolution, leading to a memetic shift (if not a formal mechanical one) away from “DM is God” posturing. Although a few maintained the Old Ways, best practice advice shifted to talking about “being a fan of the PCs” and bending all efforts towards providing a satisfying session experience for the players.
This necessary evolution solved the initial problem, but created a new one: the GM role, already by far the most intensive of time and effort, tilted from ruler to servant, an entertainer for the rest of the table akin to the function the computer serves in a video game. Certainly, doing prep and juggling NPCs offers its own kind of fun, but almost two decades in to the democratic program, GM burnout is a hot topic of discourse. The intuition pervades that an overcorrection took place, and now the GM ought to receive some love, too. This belief takes form in a statement recited by designers: “the GM is a player, too!”.
Are they really, though? Positive sentiment aside, what does that assertion mean? We can identify that the GM is engaging with a game, playing characters and utilizing mechanics, so it’s definitely true in a trivial sense, but the argument implicitly compares the GM to the other players, and they’re clearly playing a rather different game: a game they can actually win, for one thing. In many modern games, the GM doesn’t even roll dice. While well-intentioned, calling the GM a player is, to my mind, rather hollow. We sense, feel that they ought to be, and make a normative statement phrased as a descriptive one.
With the thesis and antithesis laid out, what synthesis must we craft to resolve the contradiction? Methods of achieving GM playerhood vary.
Total GM Abolition
This first method solves the problem by eliminating the monolithic responsibility altogether: when everyone’s the game master, no one is. GMless (or GMful, if you prefer) systems are their own entire sphere too large to cover in detail here, but of them I’m partial to the method used in Belonging Outside Of Belonging games, where aspects of the world are split up into different forces (Society, Nature, et cetera) and each person gets one as their GM “class” in addition to their normal character. The best part of this method is how it removes the processing bottleneck of a singular GM in favor of distributing the cognitive and communicative load among the entire group. However, it carries some rather significant caveats: vagueness of authority in edge cases, potential violation of the Czege Principle, and not-uncommon player aversion to being asked to step outside of actor stance all present their own problems, perhaps explaining why GMless games remain fairly niche as a style despite enthusiastic development by the (valued and respectable) Theater Kid wing of RPG design.
Although vastly divergent in approach, solo RPGs fit into this category as well. If you’re the only player, doing GM-like activities in response to prompts still feels like a complete play experience, with no additional beneficiaries taking a piece of the pie. There’s even the fuzzy middle ground of GM emulation: for example, you can run Ironsworn in multiplayer mode by offloading GM-type decisions onto algorithmic table lookups, putting a modicum of descriptive obligation onto each player but reducing it to a level more people find tolerable.
Gamifying Prep
The hours of thankless labor drawing maps, generating stats, and stocking encounter tables easily accounts for a large percentage of the burden GMs shoulder for the sake of their campaigns, so what if that was a game they got to play? Prep gamification gives up on providing an equivalent player experience to the PCs in favor of an asymmetric game all its own. The Faction Turn from Star Without Number is perhaps the best known version of this idea: the setting is populated by factions vying for control of planets, so the GM stats them all up and has them fight each other during downtime, updating the world state accordingly and informing the players. Hellpiercers takes the premise even further with its STRATCOM phase, tying it more directly to in-session play by giving the PCs enemy supply lines to ambush and advantage-generating facilities to assault.
Approaching the same target along a very different vector, dungeon generation games abound: minigames of generating rooms, corridors, and encounters based on dropping handful of polyhedrals, drawing tarot spreads, and matching numbers on dominos, among others. While enjoyable in their own ways, these methods (as well as the faction-based ones mentioned above) don’t manage to address the fact that you still spend essentially the same amount of time doing the prep, you just do it differently – and without anything to win or lose, it tends to feel more like manually running a simulation than playing a game.
The GMPC
Okay, so forget prep: what if the GM got to play a character like everyone else? The tendency to want to do this historically took the pathological form of the “DMPC” (AKA the magic knight “Sir Ten-Levels-Above-You”) but we can pave that desire path. Ryuutama famously executed this by framing the campaign as an adventure overseen by a narrativorous dragon, who gained experience points, unlocked abilities, and could even insert themselves into the game in form of a cute (but ineffectual) mascot character. In some respects the idea only superficially changed the experience of play for the GM, since they still ultimately did more or less the same things, but the experiment was noble.
Alternatively, Fellowship situates the GM more directly into the game world: they play the Overlord, the main antagonist of the campaign. The Overlord has specific abilities, can level up along with the PCs, and goes about their own agenda at the same time as the rest of the characters, seizing artifacts of power in their goal to rule/destroy/save the world. They operate under certain limitations when in the same scene as the PCs (to prevent them from ending the campaign prematurely with overwhelming force while still allowing them to exhibit threatening Big Bad-tier might) but can be weakened by the PCs seizing their sources of power. Defeating them ends the campaign, offering a true final boss.
For the sake of completeness: in a lesser sense, one can interpret the Apocalypse World “MC Move” paradigm as making in-session GM activity somewhat more game-like, because it offers a framework in which they take prescribed actions in a back and forth turn-like structure with the players. However, although a useful and efficient way of thinking about pacing and consequence, in practice it still boils down to managing the game world more or less in the same way other games do.
Avenues Of Further Research
Where to go from here? If any of these attempts solved the problem in its fullness I’d simply be making a post about that, rather than a survey of possibilities. However, each branch holds some wisdom to pluck, and potential for future iteration.
Emulators
I think there’s more that can be done with algorithmic automation of the GM role. Modern tools for random generation are vastly more advanced than those of even a decade ago, and the rise of solo RPGs has yielded new insights into how to integrate them into structures that reliably present hooks for players to improvisationally latch on to. The more condensed GM legwork we can fit into the book itself, the more it will play “out of the box” without expecting players to fill in the gaps, and the less likely Czege Principle violations will occur. Probably the most fun I’ve had with a GMless session was multiplayer Ironsworn, and the significant lifting done by the system itself was what enabled that. I’d imagine moving further along the spectrum into capsule games will be part of this, since it narrows the field of possibilities the engine has to account for.
4X Prep
To truly turn prep into a game, it needs to evolve into a more fractal structure: the faction turn should feed into the dungeon generation rather than operating as separate modules. D&D-esque encounter budgets (particularly in their 4th Edition incarnation) feel similar to point-buy army lists in a war game: that’s a strength! Imagine if, during downtime, you executed faction turns to seize locations and generate resources. These resources then became a budget to place rooms into those locations, like a GM-side version of Wicked Ones. The dungeon budget in turns filters down into the individual encounter budgets – throw in some Lancer sitreps for flavor. Players could then choose where to go next based on, say, taking down an armory of an evil empire, because doing so will reduce the quality of equipment held by enemies in other locations, and ultimately let them bring about their revolution. It’s a complex problem to work out the details of, but the final form might be more streamlined than it sounds up front.
Handlers
The GMPC doesn’t have to be an antagonist. I’ve seen the concept tossed around of treating the GM more like a team handler on the radio: someone saying “you’ve got company on your six” when a the party gets ambushed. This essentially just reframes what the GM normally does rather than really changing it much on the mechanical level, but it does provide them a chance to be a part of the team with everyone else, a character who gets to talk and perhaps use their own abilities to help. The Ryuutama mascot fulfills a similar role, although a game could also treat it more like a Persona-style mascot (a source of setting information and announcer of events) to bridge the gap.
The Adversary
The reason I’m thinking about this subject is that I’m trying to address it myself in Protoplastic. My approach most aligns with Fellowship: the GM plays the main antagonist, called “the Adversary”. During session 0, they create a character using the same method as everyone else (albeit with better stats/equipment) and play out scenes interacting with one another, establishing social bonds with them to be betrayed. After that, they mostly take a backseat, instead taking the roles of their tyrant subordinates when they appear, essentially playing them as rival characters who get to take their own turns opposing the heroes. During downtime they take their own actions that further their plans of dominating the archipelago, which the players can try to thwart. Finally, the chronicle ends with a final showdown in their stronghold, providing a narratively satisfying conclusion missing in many campaigns. Although adversarial GMing is traditional viewed as undesirable, with a proper framework it could actually benefit play by letting the GM present an legitimate challenge instead of a series of paper tigers meant to be defeated. I want to make good on the promise the GM really is a player in a similar sense to the others, just on the villainous rather than heroic side. I believe it is possible.
Leave a Reply