And every other way is wrong
In a recent twitter thread I was ranting about a 5e Curse of Strahd game I was briefly part of. I'd noticed my fellow party members being entirely happy about following the laid out raidroad even if it made no sense for their characters to do so.
The gypsies had kidnapped us to the lands of Barovia, where we were told "yeah you can't leave until Strahd is defeated. You weren't the first ones we kidnapped and probably won't be the last. Also no we will not apologize for this violation of your freedom."
Instead of doing the logical (and Lawful) thing of killing these criminals for their crimes (past and future), the rest of the party said "Oh well, such things happen. Where did you say that vampire was?"
It was disconcerting to experience, because they did not seem to have any internal conflict about this, when really, they ought to, don't they? Even if they agree the vampire is bad and must be destroyed, their characters should at the very least be miffed about this, instead of taking it in stride.
This wasn't some in-character decision to be OK, but a player decision. They decided to live with the doublethink common to modern RPG players; you are free to do as you want! But we follow the railroad, because well, you have to, don't you?
I would've left it at this if not for an Tweeter (Exer? What's it called nowadays?) replying with the following two posts:
I have been playing since the 80s. Understanding that the GM has prepped a game is not the same as railroading. You can run a game, or you can play the game in front of you.
And
It is just called playing the game the GM prepped, and not trying to be difficult. They were going along with the person putting in the work.
Bingo! Bingo, I have the Soup Ailse bingo!
Now the replies itself are of course the usual drivel. "I am claming seniority", "Join the getalong gang", "Appreciate the ref's performative self-sacrifice" and a side of schizopherenia about a prepped game not being the same as railroading while also telling you to just do what the preparation the ref did asks for.
At first I thought about just replying to this on Twitter, but then I thought: this might be a nice way to one again hammer home that just like there's a right way to play, there's a right way to ref a game.
God save us 415 words in and we're only just getting to the main bit
Now, I'd love to give this way of reffing some snappy title, but I'm sad to say that people have ruined this for everyone. The name for it /ought/ to have been Simulationist, but between GSM bullshittery and Simulationist generally being applied to autistically having rules for literally everything happening, applying it here would be a mistake since it would confuse people.
I don't know... Let's game it Causal reffing until someone in my (nonexistent) peanut gallery comes up with a better term.
So tell us Bosman, how does this work?
In the most reductive terms: we start out with an apriori world state. The players act (or inact) on that world state. You apply some method of logic to get to a new world state. Rince, repeat.
That tells us nothing, but that's all it is.
What is a world state
Now, here we're going to get into the weeds, because I can already hear the screaming of "but then you'd have to spend weeks prepping a world before you could ever see play", but no. The world state is /just/ 'whatever you as the ref know about the world'. That's still too vague, but stick with me. That /might/ be a fully fleshed out world, with factions which have plans (geesh, people, places, plots?), but it needn't be. The world state might also be whatever in-medias res dungeon you threw the group into that you're making up using Appendix A as you go along. I might have some initial idea of a world in mind, come up with more stuff on the spot as players go outside of what I prepared for, and I go back and incorporate it into a (semi)coherrent world later.
What matters is that you're acting as a causal arbiter of the world; you are the agent ensuring there is an effect to the actions (the cause) of the players. If they kill the necromancer the king might reward them for bringing back his kidnapped daughter. If they piss off the templars they might be excommunicated. If they fail to do anything about the goblin menace the townsfolk have been talking about a great horde might rise up and raid the kingdom.
What I do /not/ have in mind is an end-game of where I am heading, or god forbid even worse: some kind of meticulously laid out railroad with prepared encounters they /will/ fight in order, a planned out narrative that the players will have to follow. A quote from Dune comes to mind:
To know the future is to be trapped by it, Grandmother. My father knew it but he couldn't escape it. I want more freedom than that! A universe of surprises, that is what I pray for.
The analogy of the powerkeg and the monkey with a torch
Now, causality implies a certain.... "And then A happens because you did B and then D happened because you did C". A boring and stale-sounding series of action and consequence. But what I'm thinking about is a Rube-Goldberg machine.
None of this causal reffing allows you to adhere to the principal responsibility of both players and the referee: "If it's not interesting, why are we playing it?" Here I'm reminded of what (if I recall correctly) Jeffro said about the C'Taylor incident: "I just set up a powerkeg, and the players did all the work". This isn't unique to this style of refereeing but it bears repeating: if you don't lay out the ingredients for an interesting game, you're not going to get one, while if you /do/ include a few lovely Scotch Bonnets, it'll be spicy. If you let your players loose in a kingdom at peace, with secure borders and generally everyone being content, you shouldn't be surprised you get a boring game.
Instead, introduce powerkeg and a monkey with a torch, and see how quickly an interesting situation creates itself /naturally/.
It's Conan the Barbarian, not Conan the Actuarian!
The players will surprise you beyond what you thought was possible with their ingenuity once let loose in a world with /actual/ freedom, instead of the Potemkin facade of one. It'll also be more fun, but also easier to run. All you have to keep in mind what's /currently/ in the world and a vague idea of what might be. This is by the way where that abductive reasoning comes in damn useful.
And that's the way you run a game
Not by adhering to some sort of story you've laid out in advance, but by presenting the group with a situation and /genuinely/ asking them "so what would you like to do?". You can free yourself from the contortions of trying to get people 'back on track' or just trying to keep a plot straight, because there isn't one. There's the world, the actions of the people in it, and that's all the story you needs. That's the way of Causal reffing.
And every other way is wrong.