Over on twitter, I talked a little about uncertainty in interactive fiction. Here, where I’m free of the tyranny of 140 280 characters, I can unpack things a little.
Everyone’s seen plot diagrams like this:
In an interactive fiction, especially a wide-open one like a tabletop roleplaying game, that chart gets broken into three axes.
You’ve got the narrative thrust of the story, the inexorable progress of what happens next, but you’ve also got the emotional engagement and response of the audience/players, which isn’t necessarily the same thing. Ideally, the narrative progress of the game and the emotional progress of the player will mostly map to one another, but the nature of interaction – especially in a tabletop rpg – makes it harder to predict. A trivial fight might take on much more importance after a chain of interesting dice rolls (“hey guys, remember that one orc who just kept rolling critical hits on us! I hate that bastard way more than the arch villain of this whole dungeon”).
The third axis is the axis of uncertainty – it’s the writer’s inability to know how the events of the story are unfolding based on the player’s decision. This uncertainty isn’t a constant. In a really simple interactive narrative, it’s possible to fully predict any decision. Take the absurdly trivial scenario where you’ve got one player and one decision – do you go left or right? The writer can create a fully satisfying, engaging narrative for either branch.
In practise, though, it’s not so simple. Take a typical tabletop adventure: you’ve got four or five players (and a GM) making a staggering number of decisions, some explicit (“do we go left or right? do we overthrow the king or quell the rebellion?”) and many, many, more implicit ones (“do I drink my potion? Do I like the king? Am I paying attention or playing with my phone when the GM describes the king’s supposedly-inspiring speech?”). That doesn’t even begin to consider all the other inputs – the dice rolls, the rules mechanics, the characters, the characters’ previous adventures…
Fortunately for the viability of the games and the sanity of the writers, most of those inputs either don’t affect the story most of the time or get handled by the storyteller (either a human GM, a computer or the player-as-audience as opposed to the player-as-actor, depending on what sort of interactive fiction you’re writing – and that assumption of handling is another huge topic in its own right, ). You usually only have to consider the macro-level decisions the players might might – but that still encompasses hundreds or thousands of options, and every one of those options increases the uncertainty the writer has to handle.
This uncertainty isn’t constant. At the start of an adventure, the writer has nearly complete certainty about the state of play – obviously! Play hasn’t actually started yet, so the players haven’t made any of their filthy deviant decisions yet. If I write “all the characters meet in a tavern” for the first scene, then I know, with complete certainty, that the characters do, in fact, meet in a tavern. I also know they’re healthy, eager for adventure, unaware of the larger plot and so forth. They’re blank slates.
Later in the adventure, the writer can’t be sure of anything. I can write “after the dungeon, the characters meet back up in a tavern“, but I don’t know if they’re emerging from the dungeon in triumph, laden down with loot, or if they’re barely alive, having fled the dungeon after a horrific defeat. I don’t know if the wizard’s labouring under a curse, or if the fighter’s come to suspect that all the innkeepers in the land are part of a secret spy network that’s plotting against him. It’s a bunch of parallel universes.
That means that the writer of interactive narratives has to be very, very careful when assuming anything. You can’t just declare that the narrative elements and beats you’ve got planned for your story will line up neatly. You need to control and channel that uncertainty – and next week, I’ll talk about how.
(Oh, and while I’m cursing it in this essay, uncertainly is of course the joy and raison d’être of interactive fiction. You want to empower the players to make meaningful decisions – or, at least, decisions that feel meaningful – and it’s that decision-making that creates uncertainty.)
Writing Update: Current_project’s going slower than I’d hoped – other work’s crowding in on it. I haven’t found the right voice for it yet, so I’ll give it another week before deciding whether or not to back burner it until November. I’ve also suddenly been brain-jumped by a joke rpg idea that may actually have legs, so I may devote my personal writing time to that and let fiction percolate a little longer.
Gareth Hanrahan 