How we pick, and what we refuse to claim

What we do not do

We do not test lights. Beam & Bezel owns none of the products on this site, has never handled any of them, and has no laboratory, no integrating sphere, no light meter and no measurement apparatus of any kind. Nobody here has carried, used or examined a single one of these lights.

That sentence is in the first hundred words on purpose. Every competitor in this category leads with hands-on testing, and several of them are telling the truth. We are not, so we say so before anything else.

What we do instead

We compile published ANSI/PLATO FL 1 figures from manufacturer documents, do the arithmetic that connects them, and link every source with the date we read it. That is a smaller claim than testing and it has one advantage: you can check all of it.

1. Find the manufacturer’s own specification page. That is the preferred source and it is what we cite wherever we could reach one.
2. Where we could not, cite the brand’s own retail listing and label it as such. Several manufacturers’ sites would not serve us a product page from this build. In those cases the spec table says the figure came from the brand’s own listing on Amazon.com rather than from a spec sheet, so you can weigh it accordingly.
3. Mark anything unpublished as unpublished. A field reading “not published” is a finding, not a gap we forgot to fill. We do not substitute a figure from another review site, a forum post or a video.
4. Compute the derived values. Runtime, beam distance, cost to run and lumens per dollar are functions over the published inputs, so they cannot drift out of sync with them.
5. Flag conflicts rather than smoothing them. When a maker’s own site and its own listing disagree, we print both numbers and say which we used.

The arithmetic, shown

Every derived number on this site comes from one of the following. Each is stated so you can redo it with a calculator and the sources we link.

How picks are ranked

Our weighting is published so you can predict our conclusions from our stated priorities. If a ranking surprises you given these weights, we have made a mistake and we would like to hear about it.

1. Specification completeness (heaviest). Does the maker publish enough for a buyer to verify the claim? A light with a full per-mode ANSI table outranks a brighter light with a lumen figure and nothing else. This is the single biggest driver of our rankings and it is the one most likely to differ from other sites.
2. Internal consistency. Where candela and beam distance are both published, do they reconcile through the 0.25 lux definition? Agreement is strong evidence of real measurement.
3. Fitness for the specific use case. The right beam shape, the right energy, the right mounting. A page about running headlamps ranks on weight and sustained output, not on peak lumens.
4. Serviceability. A standard replaceable cell beats a sealed one, because it decides whether the light is still a light in five years.
5. Durability, where published. ANSI impact rating in meters, IP rating per IEC 60529.
6. Peak output (lightest). It is the number on the box and it is the least connected to whether the light does the job.

The FL 1 caveat we have to state honestly

ANSI/PLATO FL 1 specifies how flashlight performance is measured. It does not require independent testing, it certifies nobody, and there is no body auditing published figures. Every FL 1 number on this site, and on every other site, is self-reported by the manufacturer.

We cannot fix that. What we can do is check internal consistency and prefer makers who publish enough to be checked, which is why specification completeness is the heaviest term in our ranking. Where our arithmetic disagrees with a manufacturer’s claim, we flag the conflict rather than repeating the claim. The FL 1 standard explained covers this in detail.

Prices

Prices render only from a live data layer with the date they were checked, and the layer expires figures older than 48 hours. When it has nothing current, the button reads “Check price”. We never print a stale, remembered or estimated number, and no price is ever typed into body copy.

At launch the price layer is empty, because this build environment has no credentials for the retail price API. Every buy button therefore reads “Check price on Amazon”. That is the honest state and it is exactly what the system is designed to do when it has nothing to show.

What would make us change a pick

• A manufacturer publishing a figure that contradicts one we used.
• A reader showing our arithmetic is wrong. We correct within 48 hours and log it.
• A product being discontinued or persistently out of stock. Replaced within 7 days.
• A new product whose published figures beat an existing pick on our stated weighting.
• A maker beginning to publish specifications they previously withheld, which can move a light up several places under our first ranking criterion.

Corrections

Material corrections get a dated note on the affected page, stating what was wrong and what it should have been. We do not quietly edit a number and move on. If you find something wrong, tell us — info@beamandbezel.com — and we will fix it and say so.

What we will never do

• Claim to have tested, owned, carried or handled any product.
• Claim a test lab, an integrating sphere, a light meter or any measurement apparatus.
• Publish a “we tested N products” figure, in copy, in a stat band or in schema.
• Invent or imply a credential, certification, license or professional qualification.
• Invent an award, a rating, a review count or a testimonial.
• Publish Review or AggregateRating structured data. There are no reviews to aggregate and the on-page reasoning is editorial judgment, not a measured rating.
• Publish a price that did not come from the live price layer.
• Claim tenure or years in the industry that the domain’s age would contradict.
• Present a manufacturer’s claim as an independent finding.
• Use a stock photograph in a way that implies it is our own product photography. Every image on this site is licensed stock, captioned as such.

Who writes this

Scooter M. — an enthusiast who's genuinely into this — I read the manuals, compile the published specs, and do the math. No lab coat. More on the about page, and our editorial policy covers independence, corrections and how often we re-check everything.