This project presents an empirical analysis of 135 disk galaxies from the SPARC dataset to evaluate whether rotation curve residuals contain coherent structural information. Instead of interpreting residuals as noise or evidence of unobserved mass, the study introduces a mass‑ordered structural regime model and tests its predictive power using Ordinary Least Squares regression and Akaike Information Criterion comparison. The results show that structural regime classification significantly improves model performance relative to a standard mass‑only approach, with a ΔAIC of 119.89. This provides statistical evidence that rotation curve behavior reflects underlying structural invariants rather than arbitrary mass tuning. The project includes the full preprint, validation scripts, and processed datasets required for reproducibility.
Lee Holmes (Thu,) studied this question.