Rock typing is a vital step in oil and gas reservoir development to achieve predictions of hydrocarbon reserves, recovery, and underground storage capacity for CO 2 or hydrogen. To address inaccurate initial hydrocarbon-in-place prediction and improper rock property distribution in a reservoir model, a recent rock typing method, pore geometry and structure (PGS), has revealed a more accurate prediction on connate water saturation and better grouping of capillary pressure. However, the current state still needs physical interpretations of the PGS rock typing. We have compiled thousands of experimentally measured hydraulic properties, such as permeability k within 12 orders of magnitude, porosity ϕ up to 0.9, specific surface area S S within 4 orders of magnitude, and pore size R ranges around 3 orders of magnitude. We conduct the first-ever holistic physical interpretations of the PGS rock typing using gathered data combined with analytical theory and the Kozeny−Carman equation. Surprisingly, our physics-inspired data-driven study reveals advanced findings on the PGS rock typing. These include (i) why PGS method prevails over the hydraulic flow unit rock typing, (ii) explanations to distinguish between causality and indirect relationships among hydraulic properties, rock type number, and electrical resistivity, (iii) a proposed novel method: permeability prediction from the resistivity and rock type number relationship, and (iv) a suggestion and criticism on how to avoid a recursive prediction on permeability.