Methodology: listing-stated features, review blending and exact listings
Massage chair marketing runs ahead of massage chair documentation. Our method deliberately runs behind it: a chair is scored on what its exact Amazon listing states, a device on how completely its listing is documented, and both on the customer review record Amazon publishes. These pages are research summaries of listing data — not physical tests of any chair.
The chair spec-fit formula
Where the exact Amazon listing title or feature bullets state them, our documented spec-fit score for a full-body massage chair is 100 × [0.35 × mechanism (4D stated = 1, 3D stated = 0.7, otherwise 0.35) + 0.25 × (zero-gravity recline stated ? 1 : 0) + 0.20 × roller track (SL-track stated = 1, L-track = 0.7, S-track = 0.4, none stated = 0) + 0.20 × (a heat function stated ? 1 : 0)], rounded to a whole number. A term the listing does not state contributes its minimum — nothing is ever inferred from photos, brand reputation or similar models. This index reflects how the seller documents the listing; it is not measured motion quality, not comfort, and not a customer review.
Other categories
Massage guns, foot and leg massagers, seat and back massagers and accessories use a documentation completeness score: 100 × [0.5 × min(documented feature bullets / 8, 1) + 0.3 × min(product images / 5, 1) + 0.2 × (brand documented ? 1 : 0)], rounded to a whole number. Every input is read from the stored Amazon listing record, so the same listing always produces the same score. It measures how completely a listing is documented — not performance, quality or customer satisfaction.
Blending in the public review record
Where Amazon publicly displays customer reviews for a listing, the raw editorial score is 100 × [0.45 × (Amazon average / 5) × min(review count / 50, 1) + 0.35 × min(log10(1 + review count) / log10(1 + 5000), 1) + 0.20 × documentation score / 100], rounded to a whole number. The confidence damper keeps a three-review chair from outranking one with a thousand reviews, and the logarithmic term rewards genuinely large review volumes. Listings with no captured public review data use the documentation score alone.
The displayed scale
Raw documentation scores are normalized across our full catalog to a 5.1–9.9 editorial display scale; ranking order always follows the raw documented score. The displayed value is an editorial index of listing documentation — not a physical test result and not a customer review.
Every listing stands alone
An Emperor in a second color, a "Pro" refresh and a bundle with an ottoman are separate listings with separate scores. No specification, review count or price ever transfers between them. Price appears only from a stored buy-box snapshot no older than 24 hours; the under-$2,500 comparison is gated on that snapshot, and a chair without a current snapshot simply sits out.
What we never publish
No comfort verdicts, no motion-quality claims, no massage technique effectiveness statements, and no invented specifications. Amazon product images render only within their supplied expiry, never beyond 24 hours after retrieval. Refreshing source data is a deliberate maintenance operation.
Corrections
Send the page URL, the exact listing (ASIN) and what the listing actually states to [email protected]. Our affiliate disclosure details the Amazon relationship.