App Store A/B Test 3D Simulator
3D Monte Carlo simulator of an App Store creative A/B test: individual visitor particles stream from a search-ranking gate, split across two icon/screenshot variants, and convert to installs one Bernoulli trial at a time while a live z-test tower tracks statistical significance.
This is the same ASO funnel as a Monte Carlo particle system: individual visitors — not averaged rates — stream out of a ranking gate whose throughput depends on keyword relevance and competition, split across two creative variants, and each one converts or bounces as a genuine random trial at that variant's true conversion probability. Installs pile up into two rising towers while a live two-proportion z-test tracks whether the gap between variants is real or just sampling noise, exactly like watching a live App Store Connect or Google Play Console A/B test run to completion.
In this dynamic model, individual visitors are simulated to track which creative variant leads to more installs with statistical significance.
3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install