robertsconley
Hero
(i) I'm not criticising anything; I'm describing it;
(ii) I don't suggest that a heavily GM driven sandbox leads to players' goals being subordinated to the referee's goals. My point is about who controls the shared fiction.
I don't accept that your sort of "living world" game is the only approach to sandbox RPGing.
So what you describe as "omission of crucial details" is, in fact, deliberate generality that allows for multiple approaches to sandboxing, not all of which involve "impartial refereeing".
You say you're not criticizing, only describing, but when you describe the sandbox approach as "reducing player capacity," leading to "less knowability," and being "more railroad-y," that is an evaluation. These are judgments about how play operates and how agency is preserved. Describing something as reducing agency or increasing GM control over outcomes is not neutral. It's a critique, and that's fine, but it should be acknowledged as such if we’re going to have a discussion.
As for whether Torchbearer is a sandbox system: by your definition, yes. You wrote that a sandbox game is one that focuses on place, the locations of the world, and journeying, the act of traveling from place to place and making choices about where to go.
But that definition is incomplete. Yes, my Living World style isn’t the only form of sandbox campaign, but reducing sandbox play to "travel between locations" misses a lot of the tradition. As discussed on this forum and elsewhere, going back to at least 2007, sandbox campaigns (especially those inspired by the Wilderlands or early OD&D) also emphasize the referee’s impartiality, the players' freedom to interact with or ignore any situation, and the importance of a world that has a life of its own.
Torchbearer 2e doesn’t meet that standard. While your players may choose where to go, they can’t choose whether or not to engage in conflict, because the system mandates it. As you’ve pointed out, the entire structure of Torchbearer (and BW) revolves around tension, pressure, and dramatic stakes, conflict as a system focus. That’s not how sandbox campaigns traditionally work. In sandbox campaigns, conflict is emergent. It happens only if the players choose to engage with situations that produce it. And sometimes, it doesn’t happen at all.
Now, let’s return to your point:
(ii) I don't suggest that a heavily GM driven sandbox leads to players' goals being subordinated to the referee's goals. My point is about who controls the shared fiction.
But the impact of control is exactly that: whoever holds it has their goals prioritized. The two are inextricably linked. That directly contradicts your earlier statement to @The Firebird:
The essence of a railroad, in my view, is GM control over the fiction. Total GM control means a total railroad. [snip] So understanding whether or not RPG play is a railroad requires understanding how scenes are framed and how consequences are established.
You consistently frame your preference for Torchbearer 2e and similar systems around how they share narrative control more evenly between player and GM, whereas you argue that traditional play is "more railroad-y" because it places more fiction-shaping authority in the referee’s hands. That is, you argue that referee-controlled outcomes inherently risk subordinating player intent, and yet now you’re denying that connection.
This brings us to the larger issue: control over the shared fiction. What's often overlooked is the role of the system itself and the goals of the game's designer. When we talk about who controls the fiction, we also have to acknowledge that Torchbearer’s system as a RPG imposes its own priorities on the campaign. It sets the tone, the expectations, and the loop of play, all centered on resource scarcity, interpersonal friction, and moral tension.
Here’s what the text says:
"Adventurer is a dirty word... You’re a third child or worse... Without title, letters of recommendation or enough lucre to drown in, we’re treated little better than chattel."
"Here is a grim land. Summers are short. Winters are long... This land is wild, untamable, and in it we struggle to survive."
"Town is noisy, crowded and, worst of all, expensive... it’s no place for the likes of us."
These are not flavor-only elements. Torchbearer bakes this worldview into the gameplay loop. The players and referee have no real control over this part of the fiction. That’s a clear example of system-prioritized authorship, where the designer's goals shape the story more than either player or referee intent.
Burning Wheel has similar mechanics, but is more flexible in terms of setting and genre, and thus doesn't lock down tone and structure in the same way.
Now let me bring this back to sandbox campaigns, not just my Living World model, but sandbox play more broadly.
In sandbox campaigns, the referee’s job is to respect the players’ decisions as their characters. If they choose to go left instead of right, the referee describes what’s to the left. This is a baseline for most TTRPGs, but it matters here, because it contradicts the idea that sandbox referees exercise total control over the shared fiction.
But what happens after that choice is what sets sandbox campaigns apart. The referee doesn’t look for what conflict arises; they simply describe the result. If it’s a guard who refuses to let them through the gate, the players might not care, they might turn around. Or they might bribe him. Or wait for a shift change. Or find a hole in the fence. Whether that moment becomes a dramatic conflict or a passing scene depends entirely on what the players choose to pursue.
The system’s role in a sandbox campaign is threefold:
- To describe game-world elements (creatures, items, spells, etc.),
- To support campaign management tools (random encounters, reaction tables, etc.),
- And to adjudicate actions based on the game world’s logic.
So yes, this is a critique of Torchbearer and BW, but not to dismiss them. It’s to point out why they function fundamentally differently from sandbox campaigns, and why describing both with the same terminology glosses over real and important distinctions.