Balancing Narrative Design and Player Agency: Part One

The Train To Awesometown

by Heather Albano

In September of 2013, I had the pleasure of being in the audience of the Narrative Design panel at the Boston Festival of Independent Games, listening to Brian Moriarty, David Lebling, and Terri Brosius talk about the craft. The quote that got tweeted the most afterward was Brian Moriarty’s “Games are the art of choice.” It’s a great sound bite, and it made me particularly happy, given that Choice of Games, the company for which I do most of my work, has based its entire business model around meaningful player choice. In Choice of Games games, the choices made by the player absolutely affect the player character, the game world, and what happens next.

Awesome quotes are often even more awesome in context, though, and “Games are the art of choice” was part of a really fascinating larger exchange. If I recall correctly, it was David Lebling who responded with something like, “It can be really hard to balance player choice with the desire to tell a coherent story. You might want to communicate a really strong narrative about the prince of Denmark who comes home from school to find that his uncle has taken his father’s crown, and what he does about that. But on the other hand, you want the game sandbox to function realistically for the player, and the player may want to say, ‘Let Uncle Claudius have the throne. I never wanted it anyway. I want to go be a pirate.’”

And that, of course, is the essential tension in the narrative designer’s role: the need to balance player agency and good narrative design.

We are accustomed to viewing these elements as a trade-off. On one side of the spectrum, there is the kind of game that functions like a railroad, prioritizing narrative at the expense of player agency. If the narrative design is good enough—if it matches the beats of a classic narrative arc, if it resonates with the player on an emotional level and leads the player to walk away feeling satisfied—then many players won’t mind being unable to make narrative-affecting choices. In the words of a member of my college LARP group, “I don’t mind being railroaded if I’m on the train to Awesometown!” But players who place a premium on having agency will be disappointed; it doesn’t matter to them how awesome the destination is if the train goes there no matter what they do.

On the other end of the spectrum is the sandbox game, which allows complete player agency, and where the only narrative is the one that emerges organically as the player interacts with the system. And whose narrative design is by definition nonexistent, since the unfolding of the story is randomly generated rather than consciously designed. It’s possible for a randomized emergent storyline to match up to the beats of a classic narrative arc, sure. Given world enough and time, players interacting with a State of Denmark sandbox will indeed create The Tragedy of Hamlet through their choices. But in any given playthrough, it’s not the way to bet, and while you’re waiting, you get to suffer through a bunch of weak, wandering story arcs. And players who place a premium on narrative design will be disappointed in that.

“Railroad” and “sandbox” are of course not the only options—it’s a continuum, not a binary. For instance, you can have a sandbox with a railroad running through it, where there’s one main mission with a good strong satisfying narrative arc and a bunch of other stuff you can do if you want; or a sandbox with lots of self-contained railroads in it; or a situation where choices are offered, but they ultimately all feed back into the same narrative railroad; or—perhaps most interestingly—a situation where the player gets to so heavily customize the car they ride that the railroad feels different each time, even though analysis may prove it to be the same railroad.

Sandbox 2

Spiral Railroad

These shades of gray aside, we generally conceptualize player agency and good narrative design as tradeoffs. “We” is the game-playing community, not just the game-writing community. When you see players online complaining that their choices didn’t impact the story, you also see responses along these lines from other players:

“Partly due to production constraints and partly due to the writers’ desire to tell a coherent story, most games like this don’t have dozens of varied endings.” [KOTAKU]

“There’s no way…any company would have the resources to make a game that fully accounts for every player decision and branches the story accordingly while still hitting the same story beats….” [GIANTBOMB]

(Both of these quotations happen to be in reference to The Walking Dead, but this isn’t me beating up on The Walking Dead. They were simply the best quotations to illustrate my point. My italics, in both cases.)

The statement, “We can’t incorporate all player decisions and still keep the beats of a good narrative arc” carries the implicit assumption that you can only create one good narrative arc out of given story’s building blocks. That if you have story elements A, B, C, and D, the only good, satisfying, recognizable, emotionally resonant arc goes from x to y to z in a parabola. Anything else will be unsatisfying, and therefore bad narrative design.

Or, in other words, it’s a railroad where one branch goes to Awesometown, and the rest wander off into empty prairie.

Not So Awesome

There’s a lot of critical intellectual support for this conviction. A memorable 2007 exchange between Roger Ebert and Clive Barker used Romeo and Juliet as an example, with Ebert maintaining that the building blocks of Romeo and Juliet are set up to be a tragedy, and it would be completely stupid (my paraphrase) if it suddenly had a happy ending.

He has a point, and this is what I mean when I refer to classic, recognizable, emotionally resonant narrative arcs—the ones that spark a feeling of “Oh, I’m the protagonist of that story.” When you have a play where a boy and girl fall in love at a masked ball, then the boy and girl are plagued by outside forces trying to drive them apart, then the girl engages in a deceptive plot with a friar to make it seem as though she is dead, then the boy hears the news, believes it to be true, and is heartbroken—at that point in the narrative, you really have to end with the poison and the tomb to have a satisfying conclusion. If the boy discovered it was a ruse before he killed himself, and he and the girl lived happily ever after, that’s just childish, right? A refusal to grapple with the inherent darkness of the story. If it ended that way, it wouldn’t be Shakespeare. Right?

Except that’s the thing. It wouldn’t be Romeo and Juliet. But it would be Shakespeare. That’s the plot of Much Ado About Nothing.

In Much Ado, Hero and Claudio are deeply in love, beset by jealous outsiders. Claudio believes an ugly slander about Hero that is backed up by a truly evil piece of theater staged for his benefit, he leaves Hero at the altar, and the word spreads through the city that she has killed herself. Claudio is distraught, hates himself for what he’s done, would do anything to change it back if only he could, but it’s too late—whereupon the ruse is revealed. Claudio begs Hero for her forgiveness, and they marry. Benedick and Beatrice also sort out their problems and marry, the bad guy gets what’s coming to him, and everybody else lives happily ever after. It’s a Shakespearean comedy.

But the first time I saw a performance of Much Ado, I didn’t know the story. And as Claudio berated Hero at the altar in front of her friends and family, and she sobbed out protestations of innocence and pleas for understanding, I found myself tensing. Then came the friar and the feigned death, and I’m thinking, “I know this story. This is Romeo and Juliet. I thought this was a comedy?” It was. It is. But a Shakespearean comedy and a Shakespearean tragedy are not always distinguishable until you know how they end. You can create two very different, emotionally satisfying, deeply resonant narrative arcs using the same building blocks.

In fact, if you select your blocks carefully, you can use one set to create many different, emotionally satisfying, deeply resonant narrative arcs. We’re not talking about a pure sandbox here, since emergent storylines don’t reliably equal emotional journeys. And we’re clearly not talking about a single-branch railroad.

We’re talking about a railroad with branches to multiple destinations, each one its own solid and compelling and distinct emotional journey, each one hitting recognizable and satisfying story beats.

All Roads to Awesome

Next time we’ll talk about how a branching railroad can be constructed, in the context of real-world constraints such as development schedule and production cost.

Heather Albano is a storyteller. Sometimes she writes traditional fiction and sometimes she makes games, and she finds the line between the two growing fuzzier all the time.

If you like steampunk time travel, check out her novels Timepiece and Timekeeper; if you like Victorian gaslight fantasy, Napoleonic naval battles, medieval court intrigues, or zombie apocalypses, check out her Choice of Games titles; if you like James-Bond-style spycraft and/or interactive radio dramas, check out Codename Cygnus. If you want to know what she’s doing next, or just follow the random thoughts that wander through her head, check out

Leave a Reply