How I Learned to Write Compelling Interactive Fiction

Taking the lessons of Susan O'Connor to create interesting interactive stories in game design

AI Generated Image: A white image of a woman's face surrounded by golden vines.
A portrait of Hel. Source: Author.

I recently graduated from a game writing masterclass taught by Susan O’Connor (BioShock, Far Cry, Tomb Raider), and one of the assignments was to adapt a myth into a story with branching gameplay moments. I chose the Norse story of Loki's children. You can play the final game here, or come back to it after reading.

The following is my postmortem and the lessons learned.

The Assignment

The first step of this assignment was to create a flowchart of the gameplay branches. But who was my protagonist going to be? While Jormungandr and Fenrir are perhaps the more well-known of Loki’s children, the enigmatic Hel caught my eye. Stories surrounding her were a lot more ambiguous owing to an encroaching Christian influence.

Whether she was actually one of the pagan gods or a later addition to reinforce the Christian concept of eternal damnation remains a topic of much scholarly discussion. She occupied a nebulous history, and that made her perfect as a character that could be molded by the player.

A simple flow chart depicting the story of Hel, Jormungandr, and Fenrir. Hel is the player character and can decide to explore, hide, and/or defend her brothers.
The first iteration of the flowchart. Source: Author.

It’s very simple, but it captured the most important branches that I felt made sense for the narrative. I wanted to stay true to the myth, so the ending doesn’t change — but it’s about the journey, not the destination.

The next step was to implement it in Twine to see how it flowed with some proper text.


I had two goals with my first draft.

The first was for it to read like a dream, capturing the essence of these archaic myths through poetic prose, where you don’t fully understand what’s going on at first glance, but each choice word oozes with subtext.

The second was to make it feel more organic than your standard visual novel. I wanted player decisions throughout the game to affect descriptions, dialogue, and choices further down the line, not just obvious branch moments. Many of these are micro-decisionschoices that don’t create a new branch but change an element of the story. These are incredibly important both for scope and to let players know they’re affecting the game.

Simple examples of this include adding titles to characters, changing how you describe their gait if they’re injured, etc. Depending on player choices, the first draft was a 10–15 min read.

My gracious writing group read it and offered invaluable feedback.

Critique and Improvements

While I knew adding a choice early was important, adding too many choices all the time also meant some of them didn’t have clear motivations or stakes attached to them.

Choices don’t feel meaningful unless they’re backed by context and consequence.

The consequences don’t have to be immediate but letting players know there is weight to the decisions they’re making is paramount. This is a tenet for all game design. Obvious in hindsight, but not while I was free-writing.

In this game, micro-decisions have a cascading effect that unlocks new branches if you make the right combination of choices. The challenge I encountered was in letting players know this was possible without them replaying the game in different ways. In my desire to keep the experience more organic, I wanted to avoid too many gamey elements like color coding choices — which also introduces a selection bias. Additionally, if there aren’t enough options to buffer the hidden decisions, it can feel like players are being funneled down a path — which was my biggest concern.

What I ultimately did was two-fold. When a “wrong” decision is made, flavor text appears that says “if you had this” or “if there was that, you could do this.” This acts as a narrative hook in the player’s mind by introducing a “puzzle”. “If you had a container, you could harvest the drop of eitr.”

Secondly, I show locked branch choices in gray when they appear. This lets the player know their choices were reduced through their actions, NOT for a lack of choices coded in.

When these two designs work in harmony, it creates an "a-ha" moment where players think back on the “hook” and realize where they could have made a different choice to open the locked path. And if the game is good enough, they’ll be delighted to replay it and try again.

AI Generated Image: A raven sits on twisting tree branches.
Odin's raven. Source: Author.

My final draft involved stripping away many of the vestigial choices that didn’t make sense where they were, and moving them or building up to them in a different way. I made the “why” you were there clear in the very beginning, though the “who” I wanted to remain a mystery. Relationships were more explicitly stated and built-in through narration and dialogue so that moments of consequence would feel more powerful — because the real stakes of this story are the relationships you have with your brothers.

Key Takeaways

  • Introducing choices early creates immediate engagement and sets up the precedent that the game is listening. It’s better to pay this off sooner rather than later, or you betray that trust with the player.
  • Make sure there is a context for choices. Without context, there are no stakes or emotional investments.
  • Be clever about where you branch a story and keep it minimal. Don’t branch early, because you will have a monster on your hands.
  • Use micro-decisions where you can to create a sense of player ownership without having to branch. Micro-decisions don’t create new branches but DO create a new line of text.
  • 1 choice is no choice at all, but might be thematically appropriate or necessary to progress. 2 choices can create dramatic tension. 3 is the ideal median. 4 lets you lock choices off. 5+ is excessive but can be fun and makes for good micro-decisions.
  • It’s better for the player to be aware that their decisions lock branches off than not knowing and assuming the game is linear. This is why Telltale has prompts like “X remembered that.”
  • The inverse of that is labeling/highlighting unique choices based on how players have played or built their character. Eg. Baldur’s Gate: “[Bard] You came from the same school and recognize the tune.” This creates a selection bias — people think it’s the better choice because it’s “unique”.
  • Leverage cumulative player micro-decisions to a climactic branch for the most organic and memorable experience. If you do it well, players will be surprised that the game took all their micro-decisions into account and paid it off in a dramatic way.


Sign in or become a SUPERJUMP member to join the conversation.