Breaking an Interactive Script into Shootable Tables: Scenes, Characters, States, and an Asset Matrix

Before an interactive script enters production, it must shift from “the order players read it” to “units the team can produce.” The most practical deliverable is four cross-referenced tables: a scene table, a character state table, a node–asset matrix, and a continuity table. Together, they answer where to shoot, who performs, how many versions to shoot, and who receives each file.

Assign a scene ID to each physical shooting space and record interior or exterior, day or night, location, set dressing, characters, extras, special props, costumes and makeup, sound, visual effects, and available shooting times. If the same office has two setups in the story, “before the blackout” and “after the blackout,” treat them as different states of the same scene rather than burying them in notes.

Production uses this information to consolidate locations and lighting plans. Players may enter the office in Chapter One and return in Chapter Five, but filming usually needs to happen within the same scheduled period, so the scene table must be independent of play order.

Characters in interactive stories have more than one emotional trajectory. For each principal character, list costume, hair and makeup, injuries, possessions, known information, relationship stage, previous action, and current objective. Each node references a state ID, such as LIAN_S3_suspicious, rather than saying “more suspicious than before.”

Actors and directors need to know how differences between states become visible: speech rate, distance, gaze, forms of address, or actions. If two states cannot be distinguished in performance, consider merging them; if a difference affects the entire scene but has no separate shot version, it creates a risk of missing footage.

Give each node or node variant one row in the matrix, listing its entry state, script pages, scene ID, character states, shot group, synchronous sound, insert shots, interface assets, subtitles, exit, and target filename. Explicitly mark where shared assets are reused from; do not leave blanks for post-production to guess.

For example, the same confrontation scene has high-trust and low-trust entry states. The main dialogue is shared, but the opening gaze and the handover of evidence at the end differ. The matrix can split these into intro_high, intro_low, body_shared, out_give, and out_refuse, allowing production to schedule only the necessary differences.

Which hand holds a prop, what a phone screen displays, the water level in a cup, whether a door is open or closed, stains on clothing, and the time of day in the setting can all cause continuity errors when branches switch. Using node entry and exit as its two endpoints, the continuity table records the visual and sound states that must match and includes reference photo numbers.

Compare not only successive nodes along the same path, but also the different entry points where paths converge. When three routes converge into the same video, character positions and props must connect plausibly from all three exits; otherwise, an entry wrapper or a neutral transition shot is needed.

All four tables reference node, scene, character state, and asset IDs. Names can be easy for people to read, but matching relies on IDs. Designate one authoritative node table, and have the other tables import their basic fields from it; after script lock, every node addition or deletion must generate a change record.

Do not hide important information in cell colors. Colors can indicate progress, but text fields must still say “pending confirmation,” “locked,” or “filmed.” The meaning remains complete when exported in black and white or viewed by people with different color perception.

Have the writer, director, producer, cinematographer, script supervisor, editor, and programmer jointly choose a complex route and walk through it from entry to exit. At each node, answer in sequence: What needs to be shot? Which state does it come from? What is the file called? Which frame does the choice interface overlay? How will post-production connect it? How will the program find it? Add anything that depends on verbal explanation to the tables.

Then reorder by scene and check every state to be filmed on the same day. Does an actor need to switch from calm to a breakdown within one minute? Can injury makeup be restored to an earlier state? Are night scenes scheduled within the available hours? This step exposes arrangements that “work in the narrative but cannot be filmed on set” before shooting begins.

A usable breakdown should generate a shooting schedule, call sheets, a prop list, asset naming, and a test checklist. Pick any node: an unfamiliar team member should be able to find its script, preceding and following states, and delivery files within a few minutes. Remove an asset at random: the affected routes should be traceable. If this is not possible, the matrix is still just another script document.

After the tabletop walkthrough, divide fields into editable, approval required, and locked. Even minor dialogue adjustments require checking lip movements, subtitles, and exit duration; changes to character states or scenes require reassessing call sheets, props, and post-production. Publish a version-numbered change list every day. On set, use only the version confirmed for that day; do not rely on screenshots from group chats for updates.

After filming, the matrix continues to record actual takes, missing shots, and approved alternatives. This turns it from a pre-production breakdown table into a post-production footage intake table, so it does not lose its usefulness on the day filming wraps.

Next step: Choose the scene with the most complex branching in the entire film, create the four tables first, and run one tabletop walkthrough. Once you confirm that you can trace a node all the way to its filmed shots and final files, break down the rest of the film in batches.