LAB turns post-launch behavior claims into reproducible tests: controlled variables, predeclared run counts, recorded game build and platform, and a ledger of accepted and excluded runs.
READY PROTOCOLS4
PUBLISHED RESULTS0
FABRICATED RUNS0
TEST BUILDRETAIL
PRE-LAUNCH SAFETY LOCK: LAB currently contains no gameplay results.
Protocols are public now so the method exists before the outcome. Results become indexable only after a real retail build exists, build/platform are recorded, the required runs are completed, and the result is verified.
REGISTERED PROTOCOLS
Method first. Result later.
Pre-registration prevents us from seeing a spectacular clip first and inventing a convenient experiment afterward.
FIRE-PROPAGATION-MATERIALSV1 · PROTOCOL READY
Fire propagation by surface and object type
QUESTION: Does fire spread differently across matched object and surface classes?
HYPOTHESIS: If materials carry distinct fire behavior, ignition success, burn duration or spread radius should differ systematically across classes.
PROTOCOL
Choose repeatable locations and object classes, apply the same ignition source from a fixed distance, record ignition success, time-to-extinguish and secondary spread, reset the scene between trials, and separate scripted mission states from free-roam tests. Do not infer a material system from a single spectacular clip.
MIN RUNS20
RESULTS0
REVIEWED2026-09-05
Retail GTA VI build only; no pre-release result claims.
NPC-MEMORY-RETURNV1 · PROTOCOL READY
NPC memory after leaving and returning
QUESTION: Do non-player characters preserve a prior hostile interaction after the player leaves and returns under controlled conditions?
HYPOTHESIS: If local NPC state persists beyond immediate simulation, return behavior should differ from a clean control encounter with the same NPC or location class.
PROTOCOL
Trigger one standardized non-lethal hostile interaction, leave beyond a predefined distance for fixed intervals, return without changing unrelated variables, and code greeting, avoidance, aggression, reporting or reset behavior. Run clean-control encounters and record despawns as a separate outcome rather than memory failure.
MIN RUNS20
RESULTS0
REVIEWED2026-09-05
Retail GTA VI build only; no pre-release result claims.
QUESTION: Do clothing and vehicle changes alter police reacquisition under controlled wanted-state tests?
HYPOTHESIS: If appearance or vehicle identity participates in reacquisition, matched change conditions should produce measurably different reacquisition times from the unchanged control.
PROTOCOL
Create the same wanted-state trigger at the same location, break line of sight using the same route, then run four conditions: no change, clothing only, vehicle only, clothing plus vehicle. Measure time until reacquisition or timeout. Keep weather, route, wanted level and police contact pattern fixed and publish failed runs separately.
MIN RUNS25
RESULTS0
REVIEWED2026-09-05
Retail GTA VI build only; no pre-release result claims.
STOPPING-DISTANCE-WEATHERV1 · PROTOCOL READY
Stopping distance vs weather
QUESTION: Does wet weather materially change stopping distance for the same vehicle, road and entry speed?
HYPOTHESIS: If the physics model changes tire/road grip under rain, mean stopping distance should increase under otherwise matched runs.
PROTOCOL
Use one retail game build, one platform, one vehicle, one road segment and a fixed entry speed. Mark a braking line, start full braking at the same marker, record stopping distance from the marker, alternate clear/rain conditions when possible, discard runs with collisions or traffic interference, and publish every accepted run plus exclusions.
MIN RUNS30
RESULTS0
REVIEWED2026-09-05
Retail GTA VI build only; no pre-release result claims.
LAB is outside ad inventory, and current protocols are not indexed as result pages. Future results must reference protocol version, game build, platform and the underlying run ledger.