Narrative Design 101 — Practical Narrative Design
(An introduction to this series can be found here)
This is the third and final lecture that I gave to my Narrative Design class. In this lecture we focused on what narrative design decisions actually look like, mostly in preparation for the students’ last assignment. Because this was a three-week course, and many students will not go on to practice narrative design, I focused on a very high-level presentation of what it could be, with an eye towards inspiring. This is not likely to be work completed by a junior narrative designer in the real world, but it’s a really fun exercise for students!
I want to walk through a couple different games from a creative standpoint, rather than the analysis we saw in lecture 2. One of those games is one we’re making up based on constraints, and the other is a look inside my own narrative design process on Where the Water Tastes Like Wine, my game that came out in 2018. A student a while ago asked me about looking at bad examples of narrative design, to contrast with the mostly-good things we’ve seen in our game analysis, but I don’t like picking on games that are out — it’s very hard to know what constraints the team was operating under, and what goals they had in their design. And then even if something is actually objectively “bad” it’s not always a great idea to slag off other companies or creators in public!
But I do recognize the value of seeing “what not to do”, and so I will talk about my own work, and some of the dead ends I went down in designing this game, as well as some things that made it through to the final product that I wish I had had time and money and vision to correct.
I’m beginning with a variation of a talk I’ve given before, called “How to Make Jaws Without a Shark”. For me, thinking about narrative design is really aided by thinking about how I would adapt other works to games. Because we’ve already got the plot and themes and tone in place, we can focus specifically on what sort of changes might be necessary to pull this story into the games space, and make it interactive, and add player agency to it. For Jaws here, we’re actually going to be removing the plot, too, because the focus of this exercise for me is to see what values we can get from themes and tone when we’re doing narrative design. But I urge you to think about how you might take your favorite movie or book or TV show and adapt it into a game! It’s really interesting.
This is an example of how I like to approach narrative design, which is from primarily a theme perspective. As I said earlier in the course, that’s an approach that’s been really fruitful for me!
So, the themes of Jaws. There’s a lot here, and I have no doubt that any other reading of this movie would find other things than the ones I’ve chosen to highlight. But let’s see what we find…
First, Jaws is a classic narrative conflict of “Character vs Nature”. If this theme of the movie is what resonates with you, you have a lot of choices for making something similar. We run up against nature all the time — animals, storms, “the elements”. Usually in these stories, nature is shown as an implacable force that can’t be reasoned with, and doesn’t care about us. We even have a metaphor for things like that: “force of nature”. In Jaws, the force of nature CAN be killed — this isn’t always the case.
Another example of a story using this device is Hemingway’s The Old Man and the Sea, which has a lot of parallels to Jaws, although for the old man the struggle doesn’t end when the fish is dead.
In any case, here’s a good thematic starting point: The immensity of natural forces and humanity’s resilience in the face of them.
But we’re just getting started with themes! There’s lots in here. If you remember from Jaws, there’s initially a very simple solution to the problem: close the beaches. It will save lives, and once the shark realizes it can’t eat, it’ll leave. The only problem is that the town will lose money if they admit there’s a shark out there. The mayor is adamant that all the evidence be swept under the rug so that they won’t scare away the tourists. So here’s another theme of Jaws: Capitalism sucks, and greed kills.
If Jaws were made today, it would be easy to read it as a film about global warming. The people who make money want to bury the facts, while angry nature kills more and more. That falls apart a little at the bit where you can just shoot the problem, though.
In any case, another thematic pillar: People’s greed outweighing their better sense, putting others at risk for profit
Next, another common theme: monsters, or the fear of the unknown. The ocean is a terrifying place! We as a species barely know what’s going on out there, and as individuals it’s far too vast and deadly for us to deal with. We don’t know what’s out there, and what’s out there might not care about us — or worse, care deeply about us in the wrong ways. It doesn’t have to be the ocean, though — Jaws is danger seemingly out of nowhere, disrupting our placid safety. And it’s inhuman, and uncaring. It’s a monster!
So let’s grab this as another pillar: a vast unknown territory with unknown dangers, disruption of the normal, a foe outside our experience
OK, and then the last theme we’re going to pull out of Jaws.
This is a bit more off the beaten track for games than the previous three, but I think that’s a good thing. Jaws also looks at “Models of Classic Masculinity”. As ideas of gender go, it’s way outdated from a 2023 perspective. But still interesting to look at for us since it’s an undeniable part of the original work.
In Jaws, you have three main characters, all men (there are basically no women in Jaws). You have Quint, the Hunter. He’s a veteran, tough, grizzled, unafraid, hyper-competent at killing fish and sailing boats, and intensely blue-collar.
Then you have Hooper, the Scientist. He’s an intellectual. He’s also hyper-competent, but in a different field — he knows everything there is to know about sharks. He’s a smartass, but also tough and unafraid.
And then you have Chief Brody. He’s the police chief, and a father and family man. He is not tough. In fact, he’s afraid of boats, and the water. We can see the three of them relating most clearly in the scene where they’re comparing scars with each other, telling stories of shark attacks, war, and bravery. The only scar Brody has is a surgery scar from having his appendix out.
But in the end, the grizzled killer is the guy who dies, and the nerdy father is the one who saves the day.
Let’s pull this out as our last pillar: What does it mean to be a Man? The roles that the movie looks at are definitely outdated today, but the question is probably even more relevant. It might make an interesting piece of a game!
Next up is the tone. Of course, Jaws is a tense thriller. But like most thrillers, it has a slow ramp up to the really exciting parts, building tensions as we go. There’s also two main areas — we’re either in the town/beach area, which is a well-lit, happy and well-populated space that nevertheless gets invaded by terror, or we’re on the boat, often at night, which is cramped, vulnerable, and in the territory of the enemy.
If we were to take a pillar from our tone, we’d probably want to keep “a tense, pressure-driven thriller”, but we may also want to maintain this idea of two tones or modes, depending whether you’re “at home” or “out at sea”. It’s a different feeling to have your safe space invaded than it is to go out into the darkness a-hunting.
Now we start asking questions, using our pillars as guiding points. There’s no right answer to these questions, these are the little choices that will make your game wildly different from anyone else’s.
Questions like: How much control does the designer have? Is the game more systemic, with emergent player behavior, or narrative, with a linear story?
Systems can be really powerful — we saw how strong those were in games like Unpacking or Papers, Please.
Jaws, though, might be better as a linear narrative game, since we want to control that tense pacing of a thriller. On the other hand, the sort of “human vs nature” theme points us towards having some systems to pit the player against that nature… Maybe something like a survival game, or if you’ve played the game Dredge, we can imagine the creatures that come at night as being systemic threats to reinforce the theme. But of course it’s also possible to script those struggles! Both Jaws and Old Man and the Sea are obviously heavily scripted works.
What verbs does the player have access to?
In Jaws, the protagonists don’t do a lot of “video game things”. Coming up with verbs will be an interesting challenge, but we can ask, what verbs would reinforce the themes and tone?
Because we’re looking at a thriller we may want to have a player agency that’s stronger than “horror” but weaker than “blockbuster”. We might also want to put in things to reinforce the struggle against a monster, here — games like Shadow of the Colossus give the player combat tools but only ask you to use them against things that are immense and overpowering. We can also ask if maybe the last theme comes in here — are there different verbs that express different possible paths about being an ethical human?
That leads into my next point, too — We’re investigating different ways of being a man, and the movie uses three different characters to do that. Do we need different characters, or because we’re interactive could we perhaps give the player agency around expressing one of the models of masculinity that we choose?
I’m not going to answer all of these questions, but I think you can start to see the sorts of things I would ask, at least. This process is essentially never-ending, and gets more granular. You’ll probably be deep in production of your game, asking “What can Jaws teach me about my graphics menu?”
Based on this list and the work we did here, I made up a new narrative game using the same themes as Jaws. It’s called “Conquering the Great White”, which is the only reference to Jaws in the thing. This is a possible narrative design document, but as we saw going through the process I could have made several other ones. I chose the setting arbitrarily — I wanted to do humanity vs nature but not center around a single creature, so I chose a hostile wilderness, and one that supported the other themes I have. I chose the Yukon because it’s got a hook for the greed thing, because of the gold rush, since it was largely unexplored by white people at the time it fits with Fear of the Unknown for our character at least, and also seems to provide a fertile ground for exploring “what it means to be a Man”.
I then thought about the sort of mechanics I’d need to tell that story. In my head this was a game that someone else had designed that I was providing a narrative wrapper for, so I assumed it was a survival game with certain existing limitations, although still in preproduction so it’s easy for me to request additional mechanics and shape the game in the narrative direction. Scope-wise I also assumed it’s the kind of scale I’m most comfortable with — small indie studio, basically. So we don’t have a lot of NPC interactions, it’s a pretty focused narrative and scope. Note that I don’t really define that in the document, though! This could be a 2d Don’t Starve or a 3D Long Dark style game, and both work equally well for this narrative, I think.
The below document was sent to the students, both in raw and annotated form. I’ve included only the annotated version here.
Now that we’ve finished making up a game, I want to talk through a little more actual narrative design, using one of my own games, Where the Water Tastes Like Wine. This game came out in 2018, it’s a game about wandering the Depression-era United States and collecting and sharing stories.
This is a screenshot, the player wanders this open world USA looking for adventures and characters. You can ride the rails, hitchhike and visit cities.
As you wander around, you have little branching story adventures, and each of these gives you a story. There’s about 220 in the game, total. They’re done in Ink behind the scenes.
Then you can also meet up with 16 different characters at campfires. You trade the stories of your travels for pieces of the story of their lives. They ask to hear certain kinds of stories and if you give them the right kind, the eye at the top of the screen here will open a little more. If you get it fully open, the next time you meet that character, they’ll move on to the next chapter of their story. Each time you tell a story, the eye proceeds along the track there, towards the sun, and when it reaches it the night is over and so is your conversation.
Here’s the themes that I used as decision-making pillars for this game. I’m not going to go into these deeply but I wanted you to know there was something behind my decisions, maybe you can see the connections!
Here’s some of the specific narrative features I included in this game, and some basic reasons why. When you tell a story, it fits broadly into one of sixteen categories, which prompts a response within the same category. I used Tarot cards for this because the categories there are fluid, and I didn’t want to tie the characters into something very rigid.
I made it an open-world game, with a big map to explore. When I thought about making a game about travel, this was the only thing that made sense to me — a big map where you could see tantalizing things in the distance, see the areas change as you move through them, and also something that would take you a while to cross. It’s not a game about getting to a place, it’s a game about wandering and getting what you can from your journey.
When you tell a character a story, they tell it to someone else, and so on. And eventually, you hear the same story again, but it’s gotten a lot bigger in the telling. You met a strong lumberjack? Well, with enough retellings that becomes the story of Paul Bunyan. Those stories are also more powerful at advancing the trust a character has in you. That gives the player forward progress with their tools, but also thematically is related to the way folk stories and songs spread.
There’s a big difference between the vignette stories, which are small little ones where you can affect the outcome, and are generally pretty fantastical, or at least not terribly grounded in reality, and the character stories, which are all drawn from historical research and have big things to say about the history and self-mythologizing of America. That was a decision I made to make the pacing of the game feel good, so that mostly the player was interacting with these little lightweight things, and then only occasionally sitting down to conquer some dark and weighty topics.
WTWTLW was an anthology game, a pretty rare beast. I hired a different writer for each character and they wrote that character from their own background with their own voice and style. It meant each one had a distinct voice and feel to it, and there were a lot of ideas brought in that I would not have been able to do on my own.
And then the last thing I’m going to call out as a narrative design decision I made — I knew what I wanted the player to do, and I knew I wanted this anthology approach, but how to tie them together? I made another character, he beats you at cards and then requires you to go around collecting these stories. It’s an easy, although not particularly original, way to give the player the task I wanted them to have. But one thing I’m proud of here is that one goal of the game was to surface the lesser-known facets of American history for people, and then that ends up being the player’s task in the game, too. So I like the way those two purposes lined up.
So those are some things that I designed that the game shipped with. During the process of making the game, though, I also went down some dead ends that I cut or redesigned before shipping. I want to share those to give you an idea of how narrative design changes as you go.
I prototyped this idea I had for the player to make up their stories as they told them. It fit in with the on-the-fly improvisational nature of folk stories. You would have a sort of “inventory” of things you had encountered, people, places, things, events, and then you had to fill those in as blank spaces popped up in a story template. Like Mad Libs, if you’ve ever seen those. The problem is is that sort of template-filling under pressure leads way more to funny or nonsensical outcomes, and the tone of the game was tragic and serious. So it didn’t really fit, as much as I still like the idea. It’s possible that there was a way to save it, but instead I just fell back on the simple idea of having the stories pre-authored.
At first, the game was designed to have these brutal, unforgiving mechanics. To feel like a real grind, to make you feel the difficulty of being a hobo on the road in the great depression, and also to tie in thematically to the darkness and despair, and make the moments of hope shine brighter. Unfortunately it was terrible to play and no one actually wanted to engage with the game like that. So I took out almost all of the survival mechanics. In fact I probably should have taken them all out, but we had some stories that played off those mechanics so I left a basic veneer of them in.
And then originally when I commissioned the writing from the character writers, I asked them all to write intro stories for their character, folk stories from their cultural tradition. The idea was that before they got to know you, they wouldn’t trust you with their own life stories, but instead tell you these. It didn’t end up making much sense, there was no way to engage them with the tarot cards we had because they weren’t dialogues, and it wasn’t fun to just read a block of text without interaction. So I cut those out, which meant having to rewrite some of the character chapters to add basic introductions. However, I was able to use those stories later for a free demo version of the game where the only interaction was with the characters, listening to those stories as a sort of teaser and extra lore for the main game.
Then I also just made some bad decisions, because I didn’t have time to fix them, because I ran out of money, or because I just didn’t realize they were a problem or know how to fix them.
One was that the point of the game is to hear the characters’ stories. But then I made it seem like the goal of each chapter was to move on to the next chapter most efficiently. That meant that most players got like a quarter of each chapter content! That was a bad decision.
When you played through the vignette stories, you could sometimes affect what type of story you got by choosing different outcomes. But this happened rarely, it was pretty subtly communicated to the player, and as I’ll get to later it didn’t matter much. So that was a wasted feature and unnecessarily complex.
When you have a set of stories to tell, you can mostly rely on the same stories over and over again for different characters. You can’t tell the same story twice to the same person, but with a full inventory of stories that’s not too hard to avoid. And especially when you level up the stories you tell most often, it makes more sense to stay with the same ones. Having a huge story collection doesn’t help with anything.
And then lastly, characters would ask for a story of a certain flavor, like ‘funny” or “scary”, but we didn’t label the stories with those categories. That was part of the game, trying to remember what a story made you feel and picking the right one. But we didn’t always do a good job of telegraphing that and it became a point of frustration for players, who often felt a story fit into a different category than the official one.
There are also some things that players or reviewers called out, and I can’t decide if they are right or not. One big one is the player character’s walking speed, and the speed of travel on the overworld map in general. I deliberately made it slow and purposeful, I wanted you to look around at the countryside and also feel the size of the area you were going through. The US is a huge country, and it shouldn’t be possible to go instantly from New York to LA. But players hated this, and I don’t know if it was a good decision or not. I stand by it but I can also understand being frustrated.
And then one review called out the idea of using stories like this and said I had reduced it to something akin to a monetary transaction — tell a story get a response — and made stories a currency. And they’re not wrong, and that’s of course not exactly what I wanted to say, but on the other hand I think it was the best way to translate storytelling into a gameplay verb and ecosystem. So I don’t know!
I brought these things up because a lot of times we act as though there’s a right answer for game design questions, and there isn’t. People will disagree, and you won’t always know if you made the right decision, or if there was a secret third thing you could have done instead.
Thanks for reading all of this. I hope this look at two instances of actually doing narrative design is helpful and inspiring, even if this is a little more on the fluffy side than the nitty-gritty.
The third assignment for the students was to produce a document similar to “Conquering the Great White” above, for a brief that I gave them or to make up their own game from scratch. As I said in the intro, I actually feel a more valuable approach here would be adapting an existing work into a game, and if I teach this course again that’s the assignment I’ll do.
