Every score on this site is computed from public data, editorial review and (over time) verified owner submissions. The formulas, weights, sources and refresh cadence are published below. If a score is wrong, dispute it — we publish every resolution.
Most car-buying advice is opinion. Scores published without a methodology behind them are decoration — they sell ad inventory but they don't help a buyer. Our scores are designed to be auditable. Every ScoreCalculation in our database stores its full input snapshot and the methodology version it was calculated under. If our weights change, you can see what the score was before the change and why it moved.
Severe complaint count increased by 1 (21 -> 22).
NHTSA complaint count increased by 8 (1646 -> 1654).
NHTSA complaint count increased by 8 (688 -> 696).
NHTSA complaint count increased by 6 (1260 -> 1266).
NHTSA complaint count increased by 5 (1641 -> 1646).
RPI = 0.35·complaint_density + 0.30·recall_severity_weighted + 0.20·expensive_system_concentration + 0.15·brand_reliability_proxy. complaint_density is normalized against the peer set (same vehicle_type) with peer median = 50. Recall severity uses minor=1, moderate=3, critical=6.
| Input | Weight |
|---|---|
| complaint_density | 0.35 |
| brand_reliability_proxy | 0.15 |
| recall_severity_weighted | 0.3 |
| expensive_system_concentration | 0.2 |
ISS = 0.50·vehicle_type_baseline + 0.30·repair_complexity_proxy (= RPI) + 0.20·ev_hybrid_surcharge.
| Input | Weight |
|---|---|
| ev_hybrid_surcharge | 0.2 |
| vehicle_type_baseline | 0.5 |
| repair_complexity_proxy | 0.3 |
ORS = 0.50·complaint_density + 0.30·severe_complaint_share (injuries+deaths/total) + 0.20·user_submission_signal (would_buy_again rate when N>=25). HIGH confidence requires user data.
| Input | Weight |
|---|---|
| complaint_density | 0.5 |
| severe_complaint_share | 0.3 |
| user_submission_signal | 0.2 |
RSS = 0.50·brand_reliability + 0.35·vehicle_type_demand_premium + 0.15·platform_freshness. Higher = better. Pricing-data integration (Black Book / MMR) deferred to month 7-9 per spec.
| Input | Weight |
|---|---|
| brand_reliability | 0.5 |
| platform_freshness | 0.15 |
| vehicle_type_demand_premium | 0.35 |
The True Monthly Cost output is a transparent calculator, not a single number. Six components are estimated separately, each with its own assumptions visible in the result. Insurance varies 3–5× by personal record and carrier; the calculator is calibrated to a state-aware national average and is explicitly labeled "estimate, not a quote."
TMC = Loan + Insurance + Fuel + Maintenance + Repair Reserve + Depreciation
Manufacturers, owners, journalists and readers can dispute any score, year-verdict or recall summary on this site. Every dispute is reviewed by a human editor. Outcomes — upheld, amended, retracted, or rejected — are published in a public log. This is what protects both the brand and our readers from invisible editorial drift.
File a methodology dispute →Every score carries a confidence label visible on every Car File page: