Catalog utility profiles
Thermal market, physical market, historical market, and evidence exception are named executable profiles. Passing a catalog profile does not by itself create a product route or an indexable page.
R-value
Reported values display with their attribution. ASTM-qualified filtering, comparison, and charting require value-level attribution, product-or-value evidence scope, a resolvable evidence source, and explicit manufacturer ASTM wording. Brand, page, FAQ, or methodology context remains context only. Manufacturer estimates are visibly labeled and excluded from numeric filtering, ranking, charting, and calculations. OutdoorSpecs does not convert R-value to temperature, stack values, or derive a universal efficiency score.
Weight and variants
Canonical numeric weight functions require a verified canonical reference-variant assertion. Model-wide and canonical-variant facts may pair only when the model-wide fact genuinely applies and the canonical assertion resolves to the product’s reference variant. A non-current canonical variant cannot provide current-market measurements just because another variant remains current. Noncanonical observations never enter canonical derived output.
How market comparisons are calculated
Typical is the median: the middle value after the current pads that qualify for the calculation are ordered from lowest to highest. Weight uses fields whose filter and current-market capabilities allow the operation. R-value additionally requires value-level ASTM qualification. At least five products must qualify. The calculation retains its build date, exact comparison set, and above/below counts; a smaller set produces no statement.
Plain-language calculation contract
The page leads with the product value and its buying implication, then shows Typical and Products compared. A tap- and keyboard-accessible disclosure explains the median and the exact qualification set.
- Can an ordinary buyer understand it in roughly five seconds?
- Can its primary label avoid statistical jargon?
- Can the calculation be explained accurately in one or two sentences?
- Does it answer a real buying question?
- Would removing it make the buying comparison less useful?
If question 4 or question 5 is no, the metric must not be displayed merely because it can be calculated.
Future calculations that combine two dimensions may use formal statistical methods internally, but the public result must state the buying meaning—for example, “3 current pads are both lighter and have a higher qualifying R-value.” Internal analytical labels never become the public label.
Price
A current numeric reference price requires verified reference-price basis, canonical variant scope, a non-future evidence date, and evidence no older than 30 calendar days at build time. The older applicable source/assertion date is used conservatively.
Current market
Only records explicitly marked current or current unavailable enter the default current comparison set. Variant-level lifecycle is applied to canonical measurements. Delisted, legacy/orphaned, probable-successor, and explicitly discontinued records remain historical and are excluded from default current counts.
Successor evidence
A probable-successor relationship is a candidate relationship, not confirmed equivalence or a recommendation. Specifications are never inherited between records.
Static aging
Every artifact exposes its UTC build time, market-data as-of date, and expiry. A live deployment must be regenerated at least every 14 days.
R-values explained · ASTM F3340 explained · Data states explained