How we score, and where every number comes from

Every product on Shade & Signal carries a score out of 10 on each of four or five rubric criteria, and one published score on the site's 8.4–9.8 recommendation scale. This page is the whole method: where the facts come from, how they turn into criterion scores, how those become the published score, and how the “best for” awards are decided.

What the scores are

Editorial judgments built from the published record. Manufacturer spec sheets, manuals and help pages supply the ratings; the reviewers who measured a battery, a weight or a frame width — Android Central, Tom's Guide, Wareable, UploadVR — are cited by name wherever we use their numbers. Those facts are then weighed for the buyer comparing models on paper before a considered purchase, which is why battery and fit carry more weight here than headline megapixels. The scores stay out of structured data: there is no Review or rating markup on this site, because a score built from the published record belongs to the reader rather than to a search engine.

Where every fact comes from

1. Manufacturer specs — Meta's newsroom and help pages, Amazon's device page for Echo Frames, Even Realities' spec sheet — for resolution, video modes, rated battery, IP rating, lens and prescription options. Meta's store pages block server-side fetching, so where we could not read them directly we cite a reputable spec table (Android Central, UploadVR) that reproduces them.
2. Measured numbers — battery hours under real use, weight on a scale, frame dimensions in millimeters — from the named review that measured them, with the reviewer's conditions stated. Where the rated and measured numbers differ (battery is the classic case), both appear, labeled.
3. Market, legal and privacy facts from primary sources first: Meta's own capture-LED FAQ, court and regulator announcements, the lawsuit filings as reported by TechCrunch and Fortune.
4. Unverifiable = say so. If a spec is not published — the Echo Frames' weight, for example — the row says “not published” rather than a guess. Where sources disagree (the Gen 2 prescription range is quoted three different ways), we show the disagreement and tell you to confirm at order.

The published score, and the scale it is on

Two numbers do two jobs, and we show both so neither can hide. The criterion scores are absolute: 10 is the best available in that category today, 5 is adequate. The published score is the weighted rubric result mapped onto this site's recommendation scale, which runs from 8.4 to 9.8.

The mapping is one straight line, applied to every product in every category, with no exceptions and no per-product adjustments: a weighted rubric result of 6.6 publishes as 8.4 and a result of 8.4 publishes as 9.8. Because it is a single monotonic line, the ranking order on every page is unchanged by it, a product scores identically everywhere it appears, and you can divide back out to recover the rubric result exactly.

Why map at all? Because the rubric is absolute and no shipping pair of smart glasses is excellent at camera and audio and battery and fit and price. On an absolute scale, a genuinely outstanding pair lands near 8 out of 10 — which a shopper reads as “mediocre”, and that is not what the rubric said about it. The published scale fixes the reading without touching the judgment.

What each published band means

• 9.5 - 9.8 — Category-defining. The pair we would hand a friend with no caveats.
• 9.0 - 9.4 — Outstanding. Beaten only on one axis, and usually on price.
• 8.4 - 8.9 — Excellent and strongly recommended, with a clear best-use.

There is no band below 8.4, because there is nothing below it on the site. A product whose rubric result falls under the floor is not scored down and published anyway — it is replaced with a stronger in-stock alternative, or its slot is retired and the card comes off every page. That rule is enforced in the build: the scoring function throws and the site fails to compile if a product below the floor is ever added. So a score here is not a grade on a curve of everything for sale; it is a ranking among products that already passed the bar.

The rubrics

Each product category has its own weighted rubric, and the recalibration of the published scale in November 2026 did not change a single criterion or weight below. A criterion score of 10 means the best available in the category on that metric today; 5 means adequate; below 5 means a real limitation you should know about before buying. The weights favor what the spec-comparison buyer is actually deciding on.

What each metric means

• Camera & capture — sensor resolution, field of view, video modes and clip limits, stabilization, and low-light performance as reported by reviewers. A centered lens scores above an offset one for first-person video.
• Audio & mics — open-ear speaker loudness and leakage, microphone count and call clarity as reported, wind handling.
• Battery & case — the rated hours, the measured hours under heavy use, how much the case adds, and charging speed. Rated and measured are both shown on the product card.
• Fit & lens options — sizes available, weight, adjustable nose pads or temples, prescription range, Transitions / polarized / clear options, and whether an optician can adjust the frame.
• Everyday value — what you get at the launch price tier relative to the category, plus ecosystem: iOS and Android parity, whether it is sold on Amazon, how often it gets software updates.
• Display usefulness (display glasses only) — what the display actually does day to day: captions, navigation, viewfinder, notifications — and whether it is worth the weight and price it adds.

How “best for” awards are decided

Each roundup ranks the products that fit its question by published score, then the first-column criterion for that question breaks ties and can override: on the big-heads page a wider frame beats a higher score; on the prescription page a wider prescription range wins. The #1 pick of each roundup is the “category winner” shown on the homepage. Every pick — including the winner — carries one “good to know” line: a specific fit, size or price trade-off, because the right pair for most people is still not the right pair for everyone.

When the lineup is re-checked

The lineup was last checked against manufacturer pages and current retail listings on October 16, 2026. Meta Connect 2026 takes place on September 23-24, 2026, and new frames are expected; every roundup and comparison is re-verified immediately after any Meta or EssilorLuxottica launch event and at least quarterly otherwise. When a model is discontinued or replaced, the page says so with the date rather than quietly swapping the pick.

How prices work here

Prices come from Amazon's own product data, carry the date they were checked, and refresh daily — so the number on a button is today's number, and it disappears rather than going stale if a refresh is ever missed. Where the site quotes a launch price — “launched at $379 in September 2025” — it is dated, sourced context about where a model started, not a claim about today.

Corrections

If you find a spec we got wrong, a model we missed, or a measurement that has been superseded, email contact@shadeandsignal.com. Verified corrections are fixed within seven days and dated on the page — see the editorial policy.