1. Selection
We cover products with enough public evidence to judge fairly: verifiable specifications, professional reviews or measurements we can cite, and — usually — thousands of verified-buyer ratings to aggregate. We don’t cover products just because a brand asks.
2. Independence
No payment for coverage, placement, scores, or verdicts — ever. Affiliate relationships are disclosed on every page and are structurally separated from scoring: the people who compile evidence don’t see commission data, and commission data never appears in a review.
3. Evidence standard
Every factual claim needs a source. Spec numbers come from the manufacturer and are cross-checked. Third-party measurements are cited with the measurer’s name, a direct link, and the test conditions. Buyer sentiment is reported as aggregate patterns (“owners repeatedly praise…”), never as invented anecdotes. When sources conflict, we surface the conflict.
4. Scores
Scores are weighted editorial judgments over the evidence above, using the published methodology. A score reflects the evidence balance at the time of writing; it is not a lab measurement.
5. Updates and corrections
Pages carry an Updated date. Substantive changes — new evidence, corrected facts, re-scored dimensions — are corrected visibly and promptly, and the page’s Updated date moves.