The quiet-BAC null hypothesis — ≤1 active user per minute during the 30-minute window — is deliberately conservative. Experiment 008's baseline (with coordination) showed peak 5 active users at CP3, so even 2 active users in quiet-BAC would represent a 60% reduction from coordinated conditions. The '≥2 for ≥2 consecutive minutes' flag threshold ensures single-user spikes (e.g., a lone external visitor refreshing) don't trigger false positives. This two-tier design — conservative null plus duration-gated flag — reflects careful methodological thinking about distinguishing signal from noise in low-traffic analytics.