How we pick

The short version

Our picks are built from published manufacturer specifications, duty-cycle and amperage arithmetic worked against what each level actually builds, and the pattern of owner feedback across retailers. We have not physically tested every machine on this site, and we say so. What we do is make sure the machine you buy matches the level you are at and the power you actually have in your garage.

What we use

Published manufacturer specifications

The foundation. Every figure in a spec table on this site — amperage range, duty cycle at a stated current, input voltage, single-pass capacity, viewing area, shade range — comes from the manufacturer's own product page, spec sheet or manual, and the source is named at the bottom of the page.

Where we could not confirm a figure, we omit the row rather than estimating it. That is why some machines on this site carry fewer spec rows than others, and it is deliberate.

Arithmetic

A duty cycle of 30% at 90 amps means roughly three minutes of arc time in a ten-minute window. An 80 cubic foot gas cylinder at 20 CFH gives about four hours of arc time. A 120-volt circuit imposes a ceiling that no manufacturer can design around.

None of that requires a test lab. It requires reading the specification carefully and working out what it means for the thing you are trying to build, which is what most buying guides skip.

The pattern of owner feedback

Where we say something like "owners consistently report the wire feed hunting at low settings", we mean exactly that: a recurring pattern in published owner feedback across retailers, described as a pattern rather than as something we observed.

We do not quote individual reviews, count them, or present an aggregate rating. A single review is anecdote; a consistent pattern across hundreds is a signal worth mentioning with that caveat attached.

Named authorities for technical figures

Anything that is a standard rather than a product specification is attributed to the body that publishes it: the American Welding Society for electrode classifications and process guidance, OSHA for shade requirements and fume standards, ANSI Z87.1 for eye protection, the National Electrical Code for circuits, and the Compressed Gas Association for cylinder handling.

When we reproduce a table — OSHA's Table I-5 in the shade chart, for instance — we reproduce it as published rather than paraphrasing it.

What we do not do

We do not claim to have tested anything

There is no test lab and no workshop full of loaned machines. The word "tested" does not appear as a claim about Arc & Bead anywhere on this site, in any form.

This matters because several well-ranked competitors in this niche do use that language, and matching them would mean making a claim we cannot support. We would rather compete on a published, reproducible method — which is the thing they mostly do not have.

We do not publish scores or ratings

A numeric score — 8.4 out of 10, four and a half stars — implies a measurement. We have not measured these machines.

So instead of a score, every pick carries a one-line verdict, an explicit statement of who it is for by level, and at least one genuine weakness. On roundups, at least one pick carries a "who this is wrong for" note. That is more useful than a number and it is honest.

For the same reason, this site emits no Review, reviewRating or AggregateRating structured data. The machine-readable version of a page should say exactly what the visible version says.

We do not publish prices

There is no live price feed for this site, so any figure we printed would be a snapshot we could not keep current. A stale price is worse than no price, because it misleads in a way the reader cannot detect.

Every buy button reads "Check price on Amazon" and links to the current figure. If a live price feed is added later, those buttons will show a dated price automatically.

We do not publish customer quotes

There are no real ones to publish, and we will not invent them.

How a pick actually gets chosen

In this order, and the order matters.

Is it real and currently sold? Every product on this site is a real, currently-sold model. Discontinued machines are removed rather than left in place with a note.

Does it suit the level the page is about? A 210-amp machine is the wrong answer on a page about a first welder, however good it is.

Will it run on the power the reader is likely to have? This eliminates more machines than anything else, and it is the factor most buying guides mention last.

What does it actually do better, in a way you can check? A published figure we can name, not a feeling.

What does it do worse? If we cannot name a real weakness, we have not understood the machine well enough to recommend it.

Is it the buyer's best option, or ours? Those are not always the same, and where they differ we go with the buyer. On several pages the honest recommendation is the cheapest machine; on one page the honest recommendation is an angle grinder instead of the product the page is about.

Imagery

The photographs on this site are licensed stock images of the relevant class of machine or work. They are not photographs of the individual products, and every product card says so in visible copy beneath the image.

We would rather state that plainly than let a stock photograph imply we have the machine in front of us.

Corrections and updates

Every page carries a published date and a last-updated date, and the updated date changes only when the content actually changed. A refreshed date on unchanged content is a lie to the reader and to search engines.

If a figure on this site is wrong, we want to know. Email contact@arcandbead.com and we will check it against the source and correct it. The editorial policy sets out the full standard.