Design choices around player intention, not merely alternate lines.
Use this model:
Context → Player understanding → Intentions → Options → Commitment → Feedback → Consequence → State change
A meaningful choice does not require a unique plot branch. It requires the player to understand that different options express meaningfully different intentions or strategies and that the game responds coherently.
Inputs
Use available:
- narrative situation
- player's role
- information currently known
- desired emotional or strategic tension
- characters affected
- systemic resources
- branching budget
- long-term state variables
- required convergence points
Define the Decision Question
Write the decision as a tension, not as option text.
Examples:
- protect the ally or preserve the mission
- reveal the truth or protect the relationship
- spend scarce leverage now or save it
- forgive, punish, or exploit
- commit to one faction at the cost of another
If the question has an obviously correct answer under the game's incentives, decide whether that dominance is intentional.
Player Intention
Each option should ideally let the player express a distinct intention.
Possible dimensions:
- moral value
- strategy
- identity / role-play
- relationship preference
- risk tolerance
- resource allocation
- information-seeking
- commitment level
Avoid three phrasings of the same intention.
Information Fairness
The player does not need complete information, but should understand enough to make the kind of decision the game claims they are making.
Distinguish:
- known consequence
- reasonably inferable risk
- hidden consequence
- deliberately unknowable consequence
Do not present random hidden punishment as meaningful choice.
Tradeoffs
Strong options often exchange different kinds of value:
- time vs safety
- money vs relationship
- certainty vs opportunity
- local benefit vs future flexibility
- faction approval vs personal ethics
- information vs secrecy
Avoid strict dominance unless it serves a deliberate test of values or role-play.
Feedback
Provide some acknowledgment that the game understood the player's intention.
Feedback can be:
- dialogue reaction
- animation
- relationship update
- resource change
- quest state
- changed future line
- UI state
- later callback
Immediate feedback and long-term consequence are separate.
Consequence Layers
A choice can matter at several layers:
- Expression — lets player state identity or attitude.
- Local reaction — characters respond differently.
- State — variables, relationships, resources change.
- Content — future scenes or quests differ.
- Outcome — major world or ending consequence changes.
Not every choice needs layer 5.
Interactive Procedure
- Define the decision question.
- Identify what the player knows.
- Define 2–4 distinct player intentions.
- Convert intentions into options.
- Check tradeoffs and dominance.
- Define immediate feedback.
- Define persistent state changes.
- Define delayed consequences or callbacks.
- Estimate branching / content cost.
- Fold branches where needed without erasing remembered state.
Output
Decision question
Player information
Known, inferred, hidden.
Option matrix
| Option | Player intention | Tradeoff | Immediate feedback | Persistent state | Delayed consequence |
|---|
Dominance check
Agency type
Expression / reaction / state / content / outcome.
Production scope
Shared content vs unique content.
Follow-through obligations
What future content must remember this choice.
Failure Modes
Avoid:
- good / neutral / evil labels standing in for real tradeoffs
- one obviously superior option with no role-play reason for alternatives
- hidden consequences unrelated to available information
- fake choices that receive no acknowledgment
- options that differ only in tone when presented as strategic decisions
- massive branching when stateful callbacks would preserve agency more cheaply
- consequences so delayed that the player never associates them with the choice
- punishing experimentation without signaling risk
Final Quality Check
Verify that:
- each option expresses a distinguishable intention
- the player has enough information for the claimed decision type
- tradeoffs are real or intentional asymmetries are documented
- the game acknowledges the choice
- persistent consequences are represented in state
- production cost matches the importance of the decision