Model narrative information as state, not as exposition.
Use this loop:
Truth → Access → Interpretation → Belief → Disclosure → Update → Consequence
The central question is not only what is true? It is who has what relationship to that truth, and when can that relationship change?
Inputs
Use available:
- objective story facts
- characters / factions / player
- timeline or quest order
- evidence and clues
- lies, propaganda, rumors, secrets
- possible player paths
- reveal requirements
- desired dramatic irony or mystery
Define Facts Atomically
Break large revelations into facts that can be separately learned.
Weak:
- Mira betrayed the rebellion.
Better:
- Mira met the governor before the raid.
- Mira knew the west gate would be unguarded.
- Mira received payment afterward.
- Mira claims the meeting was coercion.
Atomic facts allow partial knowledge, mistaken interpretation, and staged reveals.
Information States
For each actor and fact, use one of these states where useful:
- Unknown — no meaningful exposure.
- Aware — has encountered the fact/evidence but may not understand it.
- Suspects — treats it as plausible but unconfirmed.
- Believes true — accepts it as true.
- Believes false — rejects it.
- Misunderstands — has a specific wrong interpretation.
- Knows — has strong justified knowledge.
- Conceals — knows/believes but intentionally withholds.
- Pretends — performs a public belief different from private state.
Do not use more precision than the story needs.
Evidence and Inference
Track how knowledge is acquired:
- witnessed directly
- told by trusted source
- told by untrusted source
- document / artifact
- mechanical observation
- deduction from multiple clues
- rumor
- coercion / confession
- supernatural or technological access
A fact should not become known merely because a scene containing the fact exists.
Public vs Private State
Separate:
- what an actor privately believes
- what they claim publicly
- what other actors think they believe
This is essential for deception and political narratives.
Reveal Design
For each important reveal define:
- prerequisite knowledge
- triggering evidence/event
- who receives it
- whether it confirms, overturns, or reframes prior belief
- immediate reaction opportunity
- downstream content that should change
Player Information
Treat player knowledge separately from protagonist/avatar knowledge when relevant.
Useful modes:
- player and avatar know the same thing
- player knows more than avatar (dramatic irony)
- avatar knows more than player (controlled mystery)
- player may discover optional facts
Document intentional gaps so they are not mistaken for continuity errors.
Interactive Procedure
- List the important objective facts.
- Split compound facts where partial knowledge matters.
- List relevant actors.
- Assign initial information states.
- Identify every legitimate acquisition path.
- Map lies, misunderstandings, and concealment.
- Define reveal/update events.
- Specify what content should react after each update.
- Check alternate paths and optional discovery.
- Export an information-state matrix and reveal ledger.
Output
Fact ledger
| Fact ID | Objective fact | Evidence | Earliest possible acquisition | Critical? |
|---|
Information matrix
| Actor | Fact | Private state | Public claim | Source | Confidence |
|---|
Reveal ledger
| Reveal | Prerequisites | Trigger | Recipients | State update | Required follow-through |
|---|
Dramatic asymmetries
Important cases where actors/player know different things.
Implementation obligations
Variables, flags, quest states, dialogue variants, or callbacks required.
Failure Modes
Avoid:
- treating scene presence as universal knowledge
- characters learning facts off-screen without an authored path
- one giant
knows_secretflag for a multi-stage mystery - confusing suspicion with certainty
- player choices revealing information that the player could not possess
- characters forgetting discoveries because branches reconverged
- public statements being treated as private belief
- optional evidence becoming mandatory knowledge on all paths
Final Quality Check
Verify that every important belief change has a plausible information path and every major reveal creates the intended downstream narrative state.