# Multi-Role Social World Reproduction Brief

Use this brief to prototype a shared virtual activity around distinct participation roles. It does not reproduce the Not A Good Driver world, its art, or an undocumented historical version. It transfers the design question: can one scene make control, co-presence, and spectating legible at the same time?

## Start with the cast

- **Driver:** has direct control and creates the moving event.
- **Passenger:** stays close to the action, reacts, and shares the consequences without steering.
- **Spectator:** reads the setup and outcome from outside the vehicle and can join the social moment without taking a seat.

Write one sentence for what each role can see, do, and contribute. A role is incomplete if it only waits for another player to finish.

## Shape the scene around sightlines

1. Draw the action path before decorating the world.
2. Mark the driver's decision points and recovery space.
3. Check that the passenger can read both the road and the driver's response.
4. Give spectators at least one safe view that reveals setup, motion, and consequence.
5. Make entrances and role choices understandable without a long instruction wall.

Treat this map as a hypothesis. It is not evidence that people will occupy the roles as intended.

## Prototype honestly

- Use simple geometry and placeholder vehicles until the three roles are readable.
- Label concept art as concept art and runtime captures as runtime captures.
- Do not fabricate a before-and-after story if the old world state or capture conditions were not retained.
- Record the world version, capture date, headset or desktop mode, camera position, and interaction state for any future comparison.
- Describe AI assistance as workflow context unless a controlled comparison supports a stronger claim.

## Test the social loop

Ask at least one newcomer to enter without coaching. Observe whether they can find the road, choose a role, understand what the other roles are doing, and switch roles without becoming stranded. Check seated and standing VR, desktop mode, different fields of view, motion comfort, and the recovery path after a crash or missed turn.

## Acceptance checks

- Every role has a meaningful action or viewpoint.
- Driver decisions are visible enough to passengers and spectators to create shared anticipation.
- Spectator areas do not obstruct or endanger the active route.
- Role entry, exit, reset, and failure recovery are obvious.
- Text and controls remain readable across supported display modes.
- The public write-up separates intended structure, observed behavior, concept art, and runtime evidence.

## Copy-ready coding-agent prompt

> Prototype a small multi-role social driving world around driver, passenger, and spectator roles. Begin with a plain action path and explicit sightlines before adding scenery. Give each role a meaningful contribution, clear entry and recovery, and a readable view of the same event. Treat the role map as a design hypothesis, label concept art separately from runtime evidence, record future capture provenance, and do not invent historical evolution that the source repository cannot reproduce.