Decomposing the Distortion Factor Φ: Myopic and Retry Branches of DTN Routing Efficiency

We decompose the policy distortion factor Φ into myopic (topology-dependent) and retry (stochastic) branches via the scale-response ln(Φ) = β·E[H]. The myopic branch follows β_myopic = δ₀·E[H]·(1−exp(−p/p₀)) with δ₀ = 0.311, p₀ = 0.042 (R² = 0.68 on 105k configurations, 8 targets). The retry branch contributes only noise (R² = 0.006). The story is one dominant analytic branch plus a weak empirical correction. Validated on both synthetic orbital and real CRAWDAD social traces.