How We Research
We have not burned any of these fire pits. This is what we do instead, why it is defensible, and — the part most methodology pages leave out — exactly what it cannot tell you.
Last updated
The claim we refuse to make
Every serious competitor in this category pairs “best” with “tested”. One publisher has bought and tested 29 fire pits over five years with infrared readings and a published weighted rubric. Another researched 125 and tested 15 across 33 hours with three named testers. That work is real and it is valuable.
We have not done it, and we are not going to say we have. Claiming hands-on testing we did not do is fabrication, it is an FTC deception, and it is the single easiest claim in this category to falsify — nobody can produce the unit. So the number of fire pits Ring & Ember claims to have tested is zero, on every page, in writing.
What follows is the method we use instead. It is weaker than a test lab on some questions and genuinely stronger on others, and the point of this page is to be precise about which is which.
The Specification Method
Every ranking on this site is derived from published, verifiable inputs. A reader with the same sources can re-derive the ranking, disagree with the weighting, and check every number. That is the property a test result does not have.
The arithmetic we show
This is the part that fills a genuine gap. Of the competitors we examined, none publishes burn hours per propane tank by BTU rating, and it is a two-line calculation from constants anyone can look up. So we publish it, with the inputs and the formula visible:
- Propane runtime. A 20 lb cylinder holds about 4.7 gallons at 4.24 lb per gallon, and propane’s higher heating value is 91,420 BTU per gallon per the US Department of Energy. That is roughly 431,000 BTU per tank; divide by the burner rating. The full table.
- Chamber size against standard firewood. Firewood is conventionally split to 16 inches. Checking each pit’s published inner diameter against that tells you whether you will be cutting wood before every burn. The comparison.
- Dimensions against the recreational-fire threshold. The fire code defines a recreational fire as a fuel area of 3 ft or less across. A pit’s published interior tells you which side of that line it sits on. Why it matters.
Why this is defensible
- It is reproducible. Every input has a public source and every source is linked. Disagree with the weighting and you can redo the ranking yourself.
- It does not expire. A tested-in-2024 verdict decays as products change. Published specifications plus live prices do not.
- It fills a real gap. The competitors with test labs are not publishing running-cost arithmetic or per-model metallurgy. Those are the numbers buyers actually need and they require no units.
- It is honest about its edges. Which is the next section.
What this method cannot tell you
Every product page carries a “what we could not assess” section for exactly this reason. In general, a specification sheet cannot answer:
- How much smoke a pit actually produces. It depends on the wood, its moisture, the wind and the load. No published figure covers your yard, and every percentage circulating in this category lacks a stated method.
- How warm you will feel. Distance, wind and ambient temperature dominate, and none of them is a product attribute.
- Real burn time on a load of wood. Nobody publishes it and it varies with the fuel.
- Build quality you can feel. Weld consistency, panel fit, how a lid seats. Those need hands.
- Long-run durability. We can tell you the alloy and the gauge, which predict it well. We cannot tell you what four winters did to a specific unit.
- Customer service. A warranty document is not a warranty experience.
If those are what you need, a hands-on review is the right source and we would rather say so than pretend otherwise.
Things that are never inputs
- Commission rate. Ranking is not affected by what pays. Where a cheaper option is the better buy, it is the pick.
- Unsourced claims from listings. “Heavy-duty steel” is not a specification and does not earn a product a place.
- Star ratings and review counts. They render only when the live data layer supplies them, and they are not part of the ranking.
- Our own impressions. We do not have any. That is the point.
When we get it wrong
A method built on published figures fails when a figure is wrong or a listing is out of date, and both happen. If you find one, tell us — a source makes it a five-minute fix. The editorial policy covers how corrections are handled and how the pages are reviewed.