Method
How we rate eyewear
The rating tiers, the evidence hierarchy and every rubric weight. Read this once and every score on the site will make sense.
By Mike M. · Updated September 2026
The premise
A pair of shooting glasses is a certification with a frame around it. Everything else — tint, fit, coating, price — is a decision you make after the rating is settled. So the rating comes first on every page, in the maker's own words, linked to the page it came from; and the score weights protection above everything else. Every pick is chosen by reading the published evidence closely — specs, manuals and the standards themselves — and the work is shown on the page so you can check it in a click.
The four rating tiers
Every product carries one badge: the highest rating its maker states on the page we cite. The tiers, and what each honestly means:
| Badge | Means | Products |
|---|---|---|
| APEL APEL | Listed on the U.S. Army's Authorized Protective Eyewear List. The Army tested this configuration to MIL-PRF-32432 ballistic fragmentation and approved it for issue. | 2 |
| MIL-PRF MIL-PRF | The maker states a MIL-PRF-32432 (or MIL-PRF-31013 / military V0) ballistic pass: a fragment simulator at several times the energy of the ANSI high-velocity test. Not on the Army list. | 18 |
| Z87.1+ Z87.1+ | ANSI/ISEA Z87.1 impact-rated: a 500 g pointed weight dropped from 50 in and a 1/4 in steel ball at 150 ft/s. The minimum for any pair on this site. | 21 |
| Z87.1 Z87.1 | The maker states ANSI Z87.1 without the '+' high-impact mark. Basic impact only; we treat it as the floor and say so. | 4 |
The badge text changes when the maker's exact standard differs — "MIL-PRF-32432A", "Military V0", "MIL-PRF-32432 (tested to)" — so the card says what the page said. The standards themselves are explained on ANSI Z87.1 vs MIL-PRF-32432.
The evidence hierarchy
- The maker's product page. The default source for every spec. Quoted, with the verb intact: "certified", "meets", "exceeds", "tested to".
- A retailer listing or press page, labeled. Used only when the maker's page blocked our reader (Oakley, Magpul, Bollé, 3M on the day we checked). The Evidence table says "as listed by the retailer" or names the publication.
- Standards bodies and professional bodies. The ISEA and MSC for the ANSI test descriptions, the U.S. Army for the APEL, the American Academy of Ophthalmology for injury data.
- Nothing else. A figure we could not source is a figure we did not print. "Not stated" is a valid entry in an Evidence table and it appears often.
The rubric
Five metrics, weights summing to 100, applied to every product. The overall score is the weighted mean to one decimal, computed from the metric scores so a card can never print an overall that disagrees with the breakdown beside it.
| Metric | Weight | How it is judged |
|---|---|---|
| Protection rating | 30% | The certification the maker publishes, in its own words. APEL listing scores highest, then a stated MIL-PRF-32432 / MIL-PRF-31013 ballistic pass, then ANSI Z87.1+ high impact. Wraparound coverage and a sealed or gasketed frame add; an open temple gap subtracts. |
| Optical clarity & tint options | 20% | Stated optical class or distortion claims, the tints offered or included, whether a lens swaps without tools, and whether the maker says anything at all about clarity. A single clear lens scores lower than a kit that covers bright, flat and indoor light. |
| Fit & ear-pro compatibility | 20% | Thin, flat or low-profile temples the maker says clear ear muffs or comms; adjustable nose pieces and temple length; stated size options (small, regular, large, youth); an Rx-insert route. |
| Fog & scratch resistance | 15% | A named anti-fog coating or a dual-pane lens, a named hard coat, and venting. 'Anti-fog' with no name or method scores lower than a named, described coating. |
| Value | 15% | The maker's list price set against the protection class and what ships in the box (spare lenses, case, strap, insert). A method, not a price we print: prices come only from the live layer. |
Each criterion is scored 0-10 from the sourced specs. A 9 in protection means the maker states a strong, specific claim (APEL, or a clause-cited MIL-PRF pass); a 4 means the page states plain Z87.1 with no plus, or the claim reaches us through a retailer. Two products with the same rating can score differently on coverage — a sealed goggle scores higher than an open frameless — and the breakdown on every card shows which. The weighted mean of those five criteria is then published on the scale below.
The published score, and what each band means
Every pair on this site has already cleared the entry bar before it is written up: ANSI Z87.1 as a floor, a sourced published spec behind every claim on its card, and a listing you can actually buy today. A product that does not clear that bar is not rescaled — it is not recommended here, and the build refuses to ship it. So the published score does not run from 0; it runs from 8.4 to 9.8, and those 1.4 points separate the pairs that cleared it.
| Band | What it means |
|---|---|
| 9.5 – 9.8 | Category-defining. The strongest published evidence on the site, and the pair we would buy first. |
| 9.0 – 9.4 | Outstanding. A clear recommendation with one specific trade-off — usually fit, size or price. |
| 8.4 – 8.9 | Excellent and strongly recommended for the job its card names. |
| Below 8.4 | Not recommended on this site, and not listed. We replace a weak pick rather than score it. |
The arithmetic is one mapping, applied site-wide, so a product's score is identical on every page it appears on and the ranking order is exactly the rubric's order. The weighted mean of the five criteria above runs from 6.0 (the floor for listing) to 8.8, and it is published as 8.4 + (mean − 6.0) ÷ 2.8 × 1.4, to one decimal. Open the score breakdown on any card and you will see both numbers: the raw weighted mean and the published score it maps to.
One more thing the score is not. It is not a test result, and it is not a star rating borrowed from a retailer. It is an editorial judgment of published evidence, shown as ordinary page content — never as Review or AggregateRating structured data.
How a roundup is ordered
By the page's purpose first, then by score. A "best for ear muffs" page ranks the temples first; a "best under $50" page uses the maker's list price on the date stated as the cut. Within a purpose, the higher rating tier leads, and the overall score breaks ties. The first product in every roundup is the winner, and it is the one whose price and buy button appear on the hub tile and the homepage.
Prices
The buy button on every card asks Amazon for today's price through a live price layer, so the figure you act on is the current one. The one exception is the prescription cost guide, which quotes figures makers printed on their own pages, dated, to size the routes against each other.
Updates and corrections
Roundups carry a dated changelog. The flagship ranking and the homepage table are reviewed quarterly, every roundup semi-annually or when a pick is discontinued, and the standards pages when ANSI revises Z87.1 or the Army updates the APEL. A correction sent through the contact page is checked against the source and logged on the page. Prices update daily once the live layer is on; content dates change only when content does.