Every product gets a deterministic 0–100 score computed from its specifications across weighted sections. The same specs always produce the same score — there are no editorial ratings or hidden adjustments, so comparisons are consistent and reproducible.