Finite-N Self-Suppression in a Coherence-Weighted Decision Field: Effective Support, Mean–RMS Separation, and Pairwise Ranking Perturbation

This preprint presents a finite-N characterization of a weighted phase-coherent decision mechanism. For normalized nonnegative support weights p_j and independent uniformly distributed phases, the work derives the exact second-moment relation E[r²] = Σ_j p_j², motivating the effective support N_eff = 1/Σ_j p_j² and the resulting RMS scaling 1/√N_eff. The study further separates mean and RMS resultant magnitude, derives the expected candidatewise phase correction, and obtains an exact expression for the variance of centered pairwise ranking perturbations. Monte Carlo experiments across N = 4 to 512, including equal-weight and unequal-weight support models, agree closely with the analytical predictions. The release includes the manuscript, research data, rendered figures, mathematical derivations, evidence records, reproducibility results, and validation materials required to inspect and reproduce the reported results. The work is presented as an EPCF-specific finite-size characterization. The generic N^-1/2 scaling of incoherent random resultants is treated as established prior art; no novelty claim is made for that underlying classical scaling law. Version 1.0.0. DOI: 10.5281/zenodo.23126253