This is the 6th post in a series about the hidden work of the GM in role-playing games. This time we delve into the caves of RPG theory, and while we’re not going too deep, I can’t promise it won’t get confusing down there. Beware! There be jargons!

The illustration above is from the AD&D 2nd edition Monstrous Compendium, and I think it’s by Jim Holloway? It looks like Jim Holloway. My copy is out in the shed don’t make me go out there. (Update: I went out there and looked, and am none the wiser because the image is not directly credited.)

The SIS, the GMIS, and me

A few years back, I was part of a group of New Zealanders who produced a shared gaming blog called Gametime. (It was a great blog, check it out!). While there, I began a series of posts about something I called the GMIS: the GM Imagined Space. (You can find them all under this tag; my posts are the ones by mr_orgue)

Actually, I should back up a little. If you dive into RPG theory discussions from a decade ago, you’ll find heaps of references to something called the SIS, which stands for the Shared Imagined Space. The SIS describes the fictional environment in which the game events take place. It fell out of favour because of that first S, because until we get that Vulcan mind meld technology nailed down our imaginative worlds are never going to be properly shared. Each of us tends our own imaginative space in parallel, and we try to keep them consistent enough that a shared game can work across all of them at once.

The scene at the time was very focused on games which emphasised improvisation and shared narration, so the failure to “share” was a problem. For more traditional styles of play, however, it becomes almost a feature:

• In a shared narration game, there are many imagined spaces, and the players have to work together to maintain sufficient consistency.
• In a traditional game, there is one imagined space that is privileged above all others, and the the other players try to maintain consistency with that “true” imagined space.

Hence: the GM Imagined Space.

The GM’s Job

So now we cycle back around to the question in that very first post in this series: what exactly is the Dungeon Master’s job? I’d argue a big part of the answer, for traditional games, should be this:

The Game Master is responsible for holding the authoritative mental model of the fictional world existing around the characters, and communicating and developing this model in response to the unfolding conversation.

Players do much the same thing, but there are some key distinctions:

• Their models are contingent – they must listen to the GM, and build models to represent what they hear.
• Their models are not authoritative – the GM has the final say on what happens.
• Their models are not extensive – they are limited to the things they learn from the GM’s descriptions.
• Their ability to develop their models is tightly constrained – they can do little besides move their specific character.

Describing a GM’s role as “maintaining a mental model of the fiction” raises questions – here are a few:

• How detailed does the model have to be?
• Under what circumstances can things be added to and removed from the model?
• What happens when models are inconsistent?

And of course the hidden goblin question:

• How much of the model must be maintained when it is out of the view of all characters?

To get at these questions we need to dive even deeper into how this mental model can work. First, we better dispense with that old bit of jargon in favour of a new one. After all, the GMIS acronym was always meant as a tongue-in-cheek riff rather than a serious bit of jargon, so it can go. Luckily it has inspired a new label for this concept, one that is even more preposterous! Let me introduce: the Is.

It Is what it Is

What is the Is?

• A game’s Is consists of whatever is true and binding about the fiction, as decided by whoever holds authority in the game.
• An Is can be very simple, sometimes as simple as noting that something exists.
• An Is can be complex, comprising multiple objects and people, their dispositions and actions, the descriptive details of their appearance, and more.
• If authority is shared, the Is can only consist of what is spoken and otherwise communicated among the players. This is a shared Is.
• If authority is held by a single player (e.g. a Game Master), the Is can only exist as a mental model maintained by that player. This is a singular Is.

In a singular-Is game:

• When the player with authority speaks about the world, they communicate aspects of the Is. Often they develop and extend the Is at the same time.
• When the authority player does not speak, but makes silent creative decisions about the world, they develop and extend the Is.
• When other players declare elements of the fiction, such as the actions of their character, the authority player develops and extends the Is to include their input.
• Usually player input is tacitly accepted without negotiation. Sometimes the authority player seeks clarification in order to establish the Is to their satisfaction. Sometimes the authority player will acknowledge the input but will not adjust the Is, because they forget, or they think it is unimportant, or they choose to ignore it.
• The Is need never be spoken aloud or otherwise shared with any other player.
• The Is can be revised at any time if the authority player chooses. It can, and often is, revised without the authority player even realising that is what they did.

Of course, the Is doesn’t really exist. But we’ll get to that later.

The Is and distance

Games with shared narrative authority have their own kind of Is, which moves at the speed of conversation. Everyone refines and revises and develops their own version of the Is as conversation proceeds, and every new contribution is contingent until it is affirmed in conversation. It isn’t possible to extend the Is beyond what is covered in conversation. In a shared-Is game, everyone might have an idea about what is behind the door and where the hidden goblin went, but until someone introduces an answer into the game conversation, these ideas cannot be part of the Is.

In games with a singular Is held by one player (the GM or equivalent), the Is can reach further. There can be a definitive answer as to what is behind the door, and where the hidden goblin went. In the classical RPG model of Dungeons & Dragons, the DM has a secret map of the entire dungeon – they know exactly what is behind the door and how many hit points it has.

Now there is of course nothing forcing the GM to extend the reach of the Is beyond what comes up in conversation. They can hold the Is lightly, in exactly the same way shared-Is games do. This is a very productive way to play, and many swear by it. GMs find such games are highly flexible, so it is easy to stage dramatic events, easy to respond to provocations in the moment, and easy to tailor the game to any unexpected needs.

However, such games present challenges. They are highly subject to the limitations of a sole creator responding to provocations in play: personal bias, predictability, reoccurring themes, inconsistent logic, “railroading’, etc. Shared-Is games escape these limitations by colliding the imaginative contributions of the group to create unexpected results, but if a GM is creating an Is on the fly alone, they do not have access to these tools unless they really make an effort to encourage the narrative agency of the player characters. (There are of course other ways to escape the limitations, for example random generators or inspiration tools like the Dungeon Starters for Dungeon World.)

Here, I think, is one of the motivations for GMs to maintain a hidden Is. A GM who is sensitive to these limitations will be looking for a way to achieve some independence and separation within their role. An extensive and hidden Is escapes many of those problems essentially by adding distance between player input and GM creation. Instead of creating and defining the imagined world as an immediate response to player actions and explorations, the GM does creative and definitional work at a remove from the conversation with players, as a somewhat independent act.

Remember in my first post in this series, when I talked about how the D&D rulebooks did not provide a clear description of the DM’s role? One thing they did provide was a detailed instruction for the DM on how to create the dungeon environment. By devising it before beginning the game, the DM doesn’t just save the trouble of inventing it on the fly, they also gain the ability to escape many of the limitations above. There is no possibility that the question “what is behind the door” has an answer reflecting the specific provocations of the moment it is encountered in play, because there is no way to know what those provocations will be.

By the same token, a GM who works to maintain an extended, hidden Is (that reaches beyond what the players perceive) is able to bring that content into view with something that feels like a ‘clean conscience’.

Consider that scary hidden goblin Strahd. If the GM decides in the moment it is opened that Strahd is behind the door, they can never really know if they made that choice because the players were particularly prepared (cheating to help them?) or particularly careless (cheating to hurt them?) or because someone said something annoying at that moment deserving punishment, or whatever other provocation, fictional or not, might have occurred.

But if the GM had decided just a few minutes earlier that Strahd would be behind the door, then all of those immediate provocations become irrelevant. The GM can be “fair and dispassionate”, just as the old D&D books advised.

The benefits of a hidden Is

When the truth behind the door is created in the moment, the creative act is tangled up in the decision about how the new content interacts with the player characters. It becomes impossible to separate the fictional content from the effect it has on play.

By separating the act of creation from the act of deployment, the GM gets to represent a world that has some independent existence, rather than one that exists purely to express some dramatic need felt in the moment. This gives the game a certain weight. It feels, in a sense, more real. It also holds at a remove any accusation of unfairness, because the GM cannot know with confidence how any created element will come into direct play. Therefore it also feels, in a sense, more fair.

These feelings, of greater verisimilitude and greater fairness, are powerful drivers for the use of a hidden Is. However I would suggest the motivation runs deeper. I argue, in fact, that for GMs there is an absence of emotional reward when the act of creating a fictional element with varied potential (say, a scary monster behind a door) is performed at the same time as the resolution of that potential into a specific dramatic function (say, attacking the characters).

For example: deciding there is a scary monster behind a closed door feels rich with potential. Will it be found immediately? Will it be avoided through luck or caution? Will some odd chain of events lead to the adventurers trying to befriend it through the door?

Whereas, deciding at the point a closed door is opened that a scary monster is waiting on the other side feels like potential has drained away. The monster will immediately act according to whatever nature you have decided. In a sense, you have not created a fictional element that gives rise to a dramatic moment, you have created a dramatic moment and dressed it in a fictional element.

This should hopefully sound familiar – for this is exactly how Apocalypse World structures play, where you make a move to create a certain kind of dramatic moment and then misdirect to make it seem like this arises from the internal logic of the fiction. This is a highly-functional and successful way to play – within the overall structure of the Apocalypse World ruleset.

However, in more traditional games like Dungeons & Dragons, a game master doesn’t have the clarity of purpose AW affords by providing a clear structure of agenda and principles to guide decision making, a strong basis for player character agency and initiative, and rules that force the unexpected into play.

Instead, in D&D and its cousins, the players are exploring the world the GM presents, and the play experience is highly determined by the things they encounter in that world. In these games, a GM who eschews a hidden Is can easily feel like they are deciding what happens next in a game that should emerge from shared creativity; like they are performing to keep the players entertained instead of creating a playground in which the players can entertain each other. It can be tiring and even, eventually, boring.

Principles and Minotaurs

Hey! That feels like a big insight, you know? I should tighten it up and declare it a principle! Then I’ll be able to tell everyone about my highly innovative and unprecedented insight. Something like “Creating your own adversity and then resolving it is…” Why is that Minotaur staring at me?

[minotaur interaction occurs]

OK it’s all good team, that Minotaur just made me understand that my exciting new discovery is actually not so exciting or new! It is suspiciously close to something we’ve all known about for years: the Czege principle, “Creating your own adversity and its resolution is boring.”

Ron Edwards described it well in the post that popularised the Czege principle: “Paul’s point means that if I’m playing a character, and if I have to say “the lizard-aliens attack!” and then, right away, it’s up to me to say how my character fights them and what happens (really happens, meaning are they defeated or do they win) … then I’m going to get tired, and eventually, after several such scenes (and listening to others’), bored.”

This is not exactly the same as what’s happening with the hidden Is, but there are close parallels. The lack of distance is again critical – declaring a problem and then deciding how it resolves doesn’t feel so hot. It isn’t the proximity per se, I think, but rather that the imaginative distance is short – the problem never really leaves that player’s mind, it never gets twisted or opened up or refracted through the imagination of anyone else.

In the same way, when a GM improvises a monster when a closed door is opened, the imaginative distance between deciding a monster exists in the fiction and the monster becoming a combat encounter is minimal and there is no space for the unexpected. Only by letting the monster roam about in the hidden Is does the imaginative distance open up to allow the intervention of new events and provocations that would ensure the GM can be surprised by their own fictional decisions.

(And note also that a game with a shared Is resolves this problem by opening the creative floor to everyone, so everyone is constantly able to be surprised by the shared fiction.)

So… is that our destination, then? Is that the main reason to maintain a hidden Is – so the GM can be surprised by their own fiction?

It feels pretty good to me. But there’s a few more things to cover off before we’re done. Like that thing about how the Is isn’t real – that sounds like it might be a bit important, right? That’s for next time.

(To be continued.)