I once played a student project with a twenty-three-page lore document attached. The designer was proud of it. The world had four continents, two ancient wars, a creation myth, and a taxonomized list of creatures. The game itself had maybe ninety minutes of content. And I felt absolutely nothing. Not because the lore was bad. Because none of it was in the game.
Then I played Disco Elysium, and a single line of inner monologue, the detective talking to himself about a tie hanging in a window, made me put down my laptop and sit quietly for a minute. That's narrative design. One is a story you wrote. The other is a story the player experienced.
The difference isn't scale or ambition. It's craft. And the good news is that craft is learnable.
Story vs. Narrative Design: The Actual Difference
Story is what happened. Narrative design is how the player finds out it happened, and more importantly, how they feel while finding out.
In a novel, the author controls the order and pacing of every revelation. In a film, the director does. In a game, the player moves through the world at their own pace, pokes at things in an order the designer can only partially control, and makes choices that change what they experience. You can't write a game story the way you write a screenplay. The medium doesn't work that way.
Narrative design is the job of bridging story and player. It answers questions like: How does the player learn that the king is corrupt? Through a cutscene? Through dialogue? Through a locked prison cell behind the palace? Through a guard who flinches when you mention the king's name? Each of those choices produces a completely different emotional experience, even if the underlying story fact is identical.
Your lore document doesn't matter unless the player can feel it. That's the whole job.
Environmental Storytelling: Let the World Do the Work
This is the technique that separates good narrative design from amateur work, and it's the one beginners most consistently ignore.
Environmental storytelling means the world itself tells the story. Not characters explaining things. Not text dumps. The objects, the architecture, the state of things, all of it communicates what happened here before the player arrived.
Dark Souls is the obvious example, and it's obvious because it's so effective. You never get a narrator. You get broken statues, dried bloodstains, a throne with two skeletons still sitting in it. The whole tragedy of Lothric unfolds in the geometry of the world. Players who want the story find it. Players who don't can still enjoy the game. That's elegant design.
But you don't need FromSoftware's budget to do this. In Undertale, a single room with a children's drawing on the wall tells you more about Asriel and Chara than any cutscene could. In What Remains of Edith Finch, a bedroom that was sealed the day a child died communicates grief in a way that would take pages to write out explicitly.
Here's the principle: show the aftermath, not the event. Instead of showing the player a cutscene of a fire, let them walk into a burned room. The ash on the floor, the melted toys, the one surviving photograph: those details make the player construct the fire in their own imagination, and what they construct will always be more personal and affecting than what you could show them directly.
Practical exercise: pick one story beat from your game, something that happened before the player arrives, and remove every line of dialogue that explains it. Now design three objects or environmental details that communicate the same information. If you can't do it, the story beat might not be worth keeping.
Player Agency and Emotional Investment
Here's something that took me a long time to understand: players care more about choices they made than about things that happened to them.
In Mass Effect 2, the loyalty missions are effective not just because they have good writing. They're effective because you chose to do them. You chose to help Garrus track down a traitor. You chose to go into Grunt's mind during his rite. The choice creates ownership, and ownership creates emotional investment. When something goes wrong for a character you chose to follow, it hurts more than when something goes wrong for a character the game forced on you.
This is why Telltale's games work even when the outcomes are largely fixed. You know, rationally, that your choices don't change much. But you still feel responsible. The design makes you the author of your own emotional experience.
For beginners, this means one thing: give players choices that feel meaningful. They don't have to branch the entire narrative. A player choosing which character to talk to first, which room to explore, which item to give to an NPC: all of these small choices create investment. The player is participating, not watching. That participation is what games offer that no other medium can.
The flip side is forced helplessness. If you take control away from the player at a critical emotional moment, a cutscene where they have to watch a character die and can't intervene, you'd better be absolutely sure that's the right call. Sometimes it is. But often it breaks the implicit contract of the medium: that the player is an agent, not just an audience.
Dialogue That Doesn't Sound Like a Textbook
Bad game dialogue has one job: deliver information. Good game dialogue has three jobs: deliver information, reveal character, and feel like actual speech.
The easiest way to tell if your dialogue is bad: read it aloud. If it sounds like someone explaining something, it's bad. People don't explain things to each other in conversation. They talk around things. They deflect. They argue about something small when they're really arguing about something big. They say "fine" when they mean anything but.
Look at this exchange:
Guard: "Welcome to Ironhold. This city was built two hundred years ago by King Aldric, who founded it as a trading post. The current ruler is Governor Maren, who has governed for fifteen years and is known to be harsh but fair."
That's a textbook. Now look at this:
Guard: "Move along. Governor doesn't like strangers standing near the gate."
Player: "What's wrong with standing?"
Guard: "Nothing's wrong with it. Governor just doesn't like it. Have a good evening."
The second version tells you almost nothing explicitly. But you know immediately what kind of city this is and what kind of governor runs it. The guard's discomfort, the curt warning, the forced politeness at the end: you learn all of that without a single sentence of exposition.
Three rules for dialogue that works:
- Every character speaks differently. Their vocabulary, their sentence length, what they're willing to say and what they avoid: all of it should reflect who they are. A nervous character gives unnecessary detail. An arrogant one doesn't finish sentences.
- Subtext is real text. What people don't say matters as much as what they do. A character who changes the subject when asked about their family is telling you everything.
- Cut the last line. Most dialogue scenes run one exchange too long. The character wraps up, explains what just happened, summarizes the lesson. Cut it. Trust the player to have gotten it.
Show Don't Tell, in Games Specifically
"Show don't tell" is writing advice so common it's almost meaningless. But in games, it has a specific application that writers from other media often miss.
In games, "showing" means doing. Not cutscenes. Cutscenes are still telling, they're just visual. Showing, in the game-design sense, means the player discovers story through their own actions.
In Papers Please, you're not told that the regime is corrupt and oppressive. You are the regime. You stamp papers. You deny entry to people with the wrong stamps. You watch a family be separated. You caused that. The game doesn't tell you the regime is bad; it makes you participate in it. That's show don't tell in the only way that's unique to games.
When you're tempted to add a cutscene that explains something important, ask: can the player do this instead? Can they find it instead? Can they be responsible for it instead? The answer isn't always yes. But ask the question every time.
Write the Wound First
This is the most practical piece of narrative design advice I can give you, and I wish someone had told me earlier.
When beginners write characters, they start with backstory. Age, hometown, occupation, skills, a list of significant events. The result is a character who feels like a biography. Technically complete, emotionally inert.
Start with the wound instead.
A wound is the specific thing that broke your character before the story began. Not "had a difficult childhood." Specific. Concrete. "Watched her brother drown while she stood frozen on the bank." "Got his unit killed by following an order he knew was wrong." "Spent fifteen years building something that someone else took credit for."
The wound shapes everything: what the character wants, what they're afraid of, what they avoid, how they treat strangers, what lies they tell themselves. When you know the wound, the rest of the character fills in naturally. More importantly, players feel it, even if you never state it explicitly. They sense the weight underneath the surface.
Geralt of Rivia's wound is that he was made into a weapon by people who claimed it was for his benefit, and he's been rationalizing the cost ever since. That wound shows up in every conversation, every moral choice, every moment of weary cynicism. You don't need to know the wound consciously to feel it. You just feel that Geralt has a lot of unsaid things.
Before you write a single line of dialogue for your protagonist, answer this: what broke them, and how are they still living with it?
Practical Exercise: Redesign a Scene Without Exposition
Here's an exercise I use in workshops, and I've seen it change how designers think about narrative in a single session.
Take any scene from your game (or design one from scratch) that contains explicit exposition: a character explaining the world, a tutorial note explaining the lore, an NPC delivering backstory. The scene doesn't have to be bad. It just has to have explanation in it.
Now redesign it with this constraint: you may not state any fact directly. Every fact the player needs to understand must be communicated through one of these: an object in the environment, a character's behavior or body language, a player action, or a consequence the player witnesses.
Let's say your original scene is: "NPC tells the player that the town was destroyed by the empire five years ago, and the survivors are still bitter."
Redesign it: The town is half-rebuilt. Some buildings are new wood, others are just foundations. An old man sits outside a ruined house that nobody has touched, weeds growing through the floorboards. When the player approaches an imperial soldier standing at the market, every other NPC shifts away or looks down. A child's toy lies in the rubble of what used to be a school.
No explanation needed. The player feels the five-year-old wound without being told about it.
If you want to prototype these kinds of narrative scenes without building a full game engine, tools like Chatforce, Twine, or Ink let you sketch dialogue and scene structures quickly. Ink, which Inkle Studios uses for 80 Days and Heaven's Vault, is particularly good for testing how narrative branches feel before you commit to implementing them in your engine. The goal is to test the story experience, not the technology.
A Note on Emotional Honesty
The games that affected me most, Disco Elysium, Nier: Automata, Night in the Woods, all had one thing in common: they were emotionally honest about something the designer clearly understood from the inside. Not "here is a sad story I constructed." Here is something real, expressed through fiction.
Anne Thiongano, a writer I know here in Lagos who also does game writing on the side, puts it this way: "The character's problem has to cost the writer something to write." I think she's right. The games that feel hollow aren't hollow because they're technically bad. They're hollow because nobody put anything real at stake in the writing.
You don't have to write autobiography. You don't have to be explicit. But somewhere in your story, there should be something you actually understand: a fear, a loss, a specific kind of loneliness, something you know from the inside. Put that in. Players can't always name what they're responding to, but they respond to it.
Where to Go From Here
Narrative design is a skill that develops through reading, playing, and writing. For reading: look at Steve Gaynor's writing on environmental storytelling (he designed Gone Home), and Cara Ellison's older pieces on game writing. For playing: go back through games you love specifically looking for how they communicate story without stating it. For writing: do the exercise above with every scene you design.
The difference between a game with a story and a game with narrative design is attention. Attention to what the player is doing when the story happens. Attention to what the space is saying when no one is talking. Attention to what characters don't say.
You don't need a longer lore document. You need to think harder about the experience on the other side of the screen.
Start with the wound. Build the world around it. Let the player find the story themselves. That's the whole craft.
