The viewer makes a choice. Cereal or no cereal. The narrative branches. One path leads to a memory; the other leads to a breakdown. Thirty minutes later, both paths converge, or don't. There are five possible endings. Or twelve. Or three that matter and nine that loop back.
This is interactive fiction. It's not new, Choose Your Own Adventure books date to the seventies, but Black Mirror: Bandersnatch made it a mainstream film format. And now you're being asked to write one. A branching script where the audience controls (or thinks they control) the story.
Here's the problem: standard screenplay format assumes linearity. Scene 24 follows Scene 23. The reader moves forward through pages. But an interactive script isn't linear, it's a web. The same scene might have three different lead-ins. A choice on page fifteen might determine what happens on page forty-five. The script isn't a document; it's a map.
Formatting this requires new conventions. Tracking it requires new tools. Conceiving it requires a different kind of storytelling brain, one that thinks in branches, not lines.
Cinematic workflow frames
These two visuals work as a pair: the first shows Cinematic workflow still, first angle, 35mm film grain, and the second shifts to Cinematic workflow still, second angle, 35mm film grain - compare them briefly, then move on.
The Fundamental Shift: From Story to System
In traditional screenwriting, you craft a story. Beginning, middle, end. The audience receives it passively.
In interactive fiction, you craft a system. Multiple beginnings, multiple middles, multiple ends. The audience participates, making choices that determine which story they see.
This changes everything.
Character arcs become conditional. The protagonist might grow, or regress, depending on which path the viewer takes. You can't assume a single arc; you must design multiple arcs, each coherent.
Scenes become nodes. A scene isn't just a moment in a sequence; it's a node in a network. It might be reached from three different paths. It must make sense regardless of how the viewer arrived.
Endings become plural. You don't write an ending; you write endings. And the endings must feel earned, not arbitrary. The viewer must believe their choices led to this outcome.
In interactive fiction, the viewer isn't watching a story. They're navigating a space—actively choosing paths, backtracking through decision trees, and shaping the narrative flow as they go.
This metaphor matters because you're building a spatial architecture with rooms and doors, not a linear timeline. Some doors lock based on earlier choices; some rooms loop back to earlier branches or converge into shared nodes. The viewer explores this structure, and exploration—with its sense of agency, discovery, and consequence—feels fundamentally different than passive viewing.
Formatting Conventions: Making the Map Readable
There's no single industry-standard format for interactive scripts. Different platforms (Netflix, YouTube, game engines, VR experiences) have different technical requirements, metadata schemas, and playback constraints. But common conventions across interactive storytelling include:
Scene IDs instead of scene numbers. Because scenes don't follow a linear order and may be visited in multiple sequences or skipped entirely, they need unique identifiers that aren't positional or ordinal. Use labels like "SCENE_A3" or "SCENE_BRIDGE_CONFRONTATION" rather than "Scene 42."
Choice points marked explicitly. When the viewer must choose, the script marks it clearly: [CHOICE POINT] or [DECISION NODE]. The options follow immediately, each labeled and linking to a destination scene ID, so the branching logic is unambiguous on the page.
Destination tags. Every choice specifies where it leads: "IF viewer chooses A → SCENE_B2. IF viewer chooses B → SCENE_C1."
Conditional entry text. A scene might play differently based on how the viewer arrived. The script uses conditionals: "IF arrived from SCENE_A3, begin with ALTERNATE OPENING."
Endpoint markers（终点标记）。用于结束整个体验的场景会被标注：[ENDING – TYPE: TRAGIC]（结局 – 类型：悲剧）或 [ENDPOINT: LOOP BACK TO SCENE_D1]（终点：循环回 SCENE_D1）。
A sample interactive scene might look like:
SCENE_B3: INT. STEFAN'S BEDROOM – NIGHT
Conditional: If viewer chose "destroy the computer" in SCENE_A7, begin with ALTERNATE OPENING A.
Stefan sits on the edge of his bed. The room is dark. Rain against the window.
STANDARD OPENING: He stares at the computer. Still running.
ALTERNATE OPENING A: The computer is gone. Broken glass where it used to sit. 电脑不见了，原本摆放的位置只剩下破碎的玻璃碎片。 电脑不见了，原本摆放的位置只剩下破碎的玻璃碎片。
STEFAN (to himself) It doesn't matter. It's already inside.
[CHOICE POINT]
The viewer sees text on screen: "What does Stefan do?"
OPTION 1: "Call Dr. Haynes" → SCENE_C4 OPTION 2: "Go to Colin's flat" → SCENE_D1 OPTION 3 (only if: viewer chose "take the pills" in SCENE_A2): "Take more pills" → SCENE_E2 选项1：致电海恩斯医生，跳转至场景C4；选项2：前往科林的公寓，跳转至场景D1；选项3（仅当观众在场景A2中选择了 选项1：致电海恩斯医生，跳转至场景C4；选项2：前往科林的公寓，跳转至场景D1；选项3（仅当观众在场景A2中选择了
This format is more annotated than traditional scripts. It has to be, because the production team needs to understand not just what happens but when each thing happens based on prior choices. 这种格式比传统剧本包含更多标注说明，必须如此，因为制作团队需要理解的不仅是发生了什么，还要明白每件事在何种前置选择条件下发生、触发的时机和逻辑分支。 这种格式比传统剧本包含更多标注说明，必须如此，因为制作团队需要理解的不仅是发生了什么，还要明白每件事在何种前置选择条件下发生、触发的时机和逻辑分支。
The Flowchart: Your Second Script
No interactive fiction project survives without a flowchart. The flowchart is a visual map of the entire experience, every scene, every choice, every path. 没有流程图的互动叙事项目无法存活下来。流程图是整个体验的可视化地图，囊括每一个场景、每一个选择节点、每一条分支路径和它们之间的因果关系。 没有流程图的互动叙事项目无法存活下来。流程图是整个体验的可视化地图，囊括每一个场景、每一个选择节点、每一条分支路径和它们之间的因果关系。
Software tools help: Twine (free, designed for interactive fiction), Miro or Figma (for visual mapping), dedicated interactive video platforms (Eko, Branch). But even paper works for small projects. 软件工具能提供帮助：Twine（免费，专为互动小说设计）、Miro 或 Figma（用于可视化映射）、专用互动视频平台（Eko、Branch）。不过对于小型项目，即便用纸笔绘制流程图也完全可行。 软件工具能提供帮助：Twine（免费，专为互动小说设计）、Miro 或 Figma（用于可视化映射）、专用互动视频平台（Eko、Branch）。不过对于小型项目，即便用纸笔绘制流程图也完全可行。
The flowchart shows:
- All scenes as nodes
- All choices as branching points
- All paths as lines connecting nodes
- All conditions annotated on paths (e.g., "only if X was chosen")
- All endpoints clearly marked
Without the flowchart, the script is unreadable. With the flowchart, the team can see the whole structure, and the writer can identify problems.
Common problems the flowchart reveals:
- Dead ends: Paths that lead nowhere
- Orphan scenes: Scenes that can't be reached from any path
- Convergence failures: Scenes that should feel continuous but require contradictory prior choices
- Exponential explosion: Too many branches creating an unproduceable number of scenes
Three Structural Strategies: Containing the Complexity
Interactive scripts can theoretically branch infinitely, but production can't. Every branch means more content to write, shoot, and edit. The math gets ugly fast: two choices with two options each means four paths. Five choices means thirty-two paths.
Writers use structural strategies to contain the explosion:
Strategy 1: The Bottleneck
Branches diverge, then converge at bottleneck scenes. Whatever the viewer chose, they end up in the same place, but the experience of arriving felt different.
Example: Stefan can call Dr. Haynes or go to Colin's flat. Both choices lead to the same next scene (Stefan at the studio), but with different emotional contexts.
Bottleneck benefits: Fewer total scenes. Easier to track. Still feels meaningful.
Bottleneck risks: If the viewer notices the convergence point where separate narrative threads merge back together, it can create the impression that their earlier choices don't carry meaningful weight or consequence in the overall story.
Strategy 2: The Diamond
The story opens with a linear narrative foundation, then branches into multiple distinct paths at a key decision point, before these divergent routes converge toward a shared finale structure—though the finale itself varies in execution and outcome based on which specific path the viewer previously took.
Example: Stefan takes path A or path B through Act Two, experiencing different events and decisions along the way. Both narrative routes ultimately lead to the same final scene location and setup, but the dialogue exchanges, character reactions, and ending resolution differ meaningfully based on which path Stefan traveled.
Diamond benefits: This structure offers limited branching complexity while still delivering meaningful narrative variation and impact. Endings feel personalized to the viewer's journey, creating a sense that their choices shaped the final outcome.
Diamond risks: The middle divergent paths must feel sufficiently distinct in content, tone, and consequence to justify the diamond structure—otherwise the convergence undermines the illusion of meaningful choice.
Strategy 3: The Garden
The story adopts a loosely woven structure filled with numerous small loops and narrative cul-de-sacs. The viewer navigates freely through the space, uncovering content in whatever sequence they prefer. A canonical "true" ending may exist, but unlocking it demands precise decision-making and particular choice patterns.
Example: The runtime spans thirty minutes, yet the project contains a full ninety minutes of authored material. Each viewer encounters a different third of the total content pool, shaped entirely by the decisions they make along the way.
Garden benefits: Exceptional replay value and rewatch potential. The experience feels genuinely exploratory, open-ended, and responsive to curiosity.
Garden risks: Maintaining narrative coherence becomes significantly harder to write and sustain. Viewers may experience disorientation, confusion, or frustration when paths feel disconnected.
Try it free
Try 熙熙 for free on your script
It is free. Import your existing project, get a clearer view of your outline, and regain control of your story structure in minutes.
Start FreeA Table: Choice Types and Their Functions
| Choice Type | What It Does | Use When |
|---|---|---|
| Cosmetic | Alters minor surface details without any real narrative impact or consequence | You want the viewer to feel actively involved and engaged without the overhead of adding true branching paths or divergent story threads |
| Thematic | Reveals character but leads to same outcome | You're exploring character rather than plot |
| Branch | Creates genuinely different paths | The divergence matters narratively |
| Loop | Returns the viewer to a previous point | You want to allow do-overs or emphasize inevitability |
| Gate | Unlocks future choices based on past behavior | You're tracking viewer "state" across the experience |
| Endpoint | Ends the experience | You've reached a conclusive narrative moment |
Mixing these choice types keeps the experience varied without requiring endless content.
The "Trench Warfare" Section: What Goes Wrong
Failure Mode #1: Illusion of Choice Is Too Obvious
The viewer makes a choice, but nothing visibly changes. The next scene is the same regardless. They feel cheated.
How to Fix It: Even if paths converge, make the choice feel meaningful. Different dialogue, different framing, different character reactions. The viewer should sense that their choice mattered, even if the destination is the same.
Failure Mode #2: Exponential Explosion
The script has so many branches that the word count triples, the shoot budget explodes, and the project is unproduceable.
How to Fix It: Use bottlenecks and convergence aggressively. Design the structure before writing scenes. Limit meaningful branches to three to five per experience.
Failure Mode #3: Incoherent Arrivals
A scene can be reached from two different paths, but the scene assumes the viewer knows something only one path reveals. The result is confusion.
How to Fix It: Review every scene for path independence. If a scene requires prior knowledge, either gate it (only reachable from certain paths) or write conditional dialogue that adjusts.
Failure Mode #4: Arbitrary Endings
The viewer makes a choice, and suddenly the experience ends. The ending doesn't feel earned, it feels like they hit a wall.
How to Fix It: Every ending should feel like a culmination, not an interruption. Even "bad" endings should be narratively complete. The viewer should think: "My choices led here. This makes sense."
Failure Mode #5: Linear Disguised as Interactive
The script branches, but every branch leads to the same ending. The interactivity is a veneer over a linear story.
How to Fix It: If you're creating interactive fiction, you need to fully commit to meaningful interactivity. At least one major choice point should produce genuinely different narrative outcomes that affect the story trajectory. Otherwise, if viewer choices don't meaningfully alter the experience, you're better off making a traditional linear film instead.
Writing for Interactivity: Craft Considerations
Interactive fiction demands all the fundamental skills of traditional screenwriting—compelling character development, natural dialogue, solid narrative structure—while also requiring mastery of additional specialized techniques unique to the interactive medium:
Modularity. Each scene must be designed to function as an independent, self-contained unit. You cannot assume the viewer has experienced any specific previous scene, because depending on their choices, there might be three or more different "previous scenes" that could have led to this moment in the narrative.
State awareness. You're continuously tracking and managing what actions the viewer has taken throughout their experience. If they made a particular choice back in Act One, Act Three might need to reference or respond to that earlier decision. This level of continuity requires meticulous bookkeeping and careful planning of all possible story states.
Emotional consistency. Even when the plot branches and diverges significantly across different paths, the underlying emotional journey must still feel coherent and unified. The viewer shouldn't feel like they're abruptly jumping between tonally different movies or experiencing narrative whiplash from inconsistent emotional beats.
Player (viewer) psychology. What makes choices feel meaningful? Agency. Consequence. Surprise. You're designing an experience, not just telling a story.
Replayability. Will viewers return to see other paths? Design content they'll want to find. Hide easter eggs. Make alternate endings substantively different.
Platform-Specific Considerations
Different platforms have different capabilities:
Netflix (Bandersnatch model): Choices appear on screen during playback. The platform handles branching. You deliver a branching script and video segments; the platform assembles.
YouTube (choose-your-own adventure style): Choices are clickable links at the end of videos. Each branch is a separate video. Less seamless, but no special platform support required.
Game engines (Unity, Unreal): Full interactivity including dialogue trees, stat tracking, and complex branching. Scripts often use specialized tools like Ink, Yarn Spinner, or Twine.
Interactive video platforms (Eko, Wirewax): Designed specifically for branching video narratives and clickable video experiences. They provide proprietary formats, dedicated visual editors, and hosting infrastructure optimized for interactive playback across devices.
Your formatting and structural choices may need to adapt carefully to the platform's technical requirements and compatibility constraints. Always confirm technical specs, file format support, branching logic limits, and playback compatibility before committing to writing.
The Perspective: Interactivity Is Story, Not Gimmick
The danger with interactive fiction is treating interactivity as the point, as the goal in itself. It's not. Story is still the point—character, conflict, emotional truth. Interactivity is a tool, a narrative device, a way to make the audience feel implicated in the events, to explore themes of agency, choice and consequence, to create an experience that lingers precisely because the viewer feels responsible for what unfolded.
Bandersnatch works as interactive fiction because it's fundamentally about control, free will, and manipulation, and the choices the viewer makes directly reflect and interrogate that theme. The interactivity isn't decorative or a gimmick; it's structurally and thematically meaningful, woven into what the story is saying.
When you design interactive fiction or any branching narrative, ask yourself early: Why is this interactive? What does the audience's active participation add to the experience or the meaning? If the answer is just "it's cool" or "it's novel," the project will likely feel hollow, a technical demo rather than a story. If the answer connects directly to theme, to what the story is fundamentally about, the interactivity becomes part of the storytelling itself, inseparable from the narrative's impact.
The best interactive fiction makes you think about your choices long after you've stopped clicking.
[YOUTUBE VIDEO: A deep dive into the narrative design of Black Mirror: Bandersnatch, analyzing how the writers structured branches, handled convergence, and made choices feel meaningful.]
Further reading:
- For guidance on creating alternate endings in non-interactive formats, see creating alternate endings: narrative branching in a psychological thriller.
- If you're working in games, explore writing video games: branching narratives.
- Twine, a free tool for interactive fiction, has documentation at twinery.org{:rel="nofollow"}.
Final Step
Build your next script with 熙熙
Move from ideas to production-ready pages faster with timeline-native writing and AI-assisted story flow.
Try 熙熙
