A/B inference simulator
Simulate an experiment, then compare Bayesian posteriors. This is what decision-making under uncertainty looks like.
Controls
Observed A
150 / 2000
90% CI: 0.066 – 0.085
Observed B
165 / 2000
90% CI: 0.073 – 0.093
Decision signal
P(B > A) = 81.4%
With a uniform prior. Increase traffic to shrink uncertainty.