Sources & method
How BeerHQ decides what it knows
Factual statements are recorded as claims with sources attached, and the state of each one is visible on the page it appears on.
What “verified” requires here
Two things that are commonly treated as one: there is a citation and somebody read the citation. BeerHQ keeps them separate, because the first is trivially easy and the second is the work.
A claim is only verified when a person has opened the source, found the passage and recorded where it is. A source attached without that is supported — the claim is probably right and nobody has checked. Neither is a defect; presenting the second as the first would be.
| Verified | Traced to a primary or authoritative source and checked against that source directly. |
|---|---|
| Supported | Backed by a credible source, but not independently corroborated against a primary record. |
| Provisional | Widely repeated and probably correct, but BeerHQ has not yet traced it to a source it trusts. |
| Editorial | BeerHQ's own synthesis, abstraction or judgement — for example our sensory scale. Not a claim about the external world. |
| Disputed | Credible sources genuinely disagree. BeerHQ must present the disagreement rather than pick a winner. |
| Held | Withheld from publication — known or suspected to be wrong, unverifiable, or carrying legal or safety risk. |
Where the catalogue stands
244 recorded claims, of which 58 are verified in the sense above. 61 sources, 40 of which have a URL a reader can open.
That last figure is the honest weakness. Much of what is genuinely authoritative about brewing lives in textbooks and institute publications that are neither free nor online, so a large part of the evidence layer cannot be independently checked by a reader — or, for that matter, re-checked by the next editor. It is recorded as a ceiling rather than a queue: retrying a source that is behind a paywall does not make it reachable.
What counts as a source
Not every source can support a verified claim. A trade magazine can establish that a brewery said something; it cannot establish a labelling threshold. The tiers below decide the strongest state a claim resting on them can reach.
| Primary — 38 sources | Standards bodies, regulators, legislation, peer-reviewed research and primary historical archives. |
|---|---|
| Authoritative — 11 sources | Brewing textbooks, university brewing programmes, technical institutes and plant-breeding programmes. |
| Reputable — 10 sources | Established trade press, recognised beer historians, museums and archives. Cannot support a verified claim on its own. |
| Industry — 2 sources | Producer, maltster, hop-merchant and trade-association material. Reliable about their own products; promotional beyond them. Cannot support a verified claim on its own. |
| Community — 0 sources | Community-maintained references and forums. Useful for finding leads; never sufficient for a foundational technical claim. Cannot support a verified claim on its own. |
Things that go stale, and are dated
Some facts are true until they are not, and the ones a reader is most likely to repeat are in that group. Every beer strength records the date it was last checked and whether the figure came from the producer or from consistent trade reporting. Every alcohol labelling threshold records the jurisdiction it applies in and when it was reviewed — see alcohol information.
Treat anything reviewed more than a year ago as unverified rather than as current. That is the same instruction the editors work to.
What BeerHQ does not do
- No health claims about alcohol, in either direction. There is no health-evidence programme, so any such claim would be unsupportable by construction.
- No inference from style to product. A beer’s recipe, hops, yeast or production route are only stated where they came from that product’s own sourcing, never from what its style usually does.
- No filling gaps with confident prose. Where the evidence is genuinely disputed, the page says so rather than picking the more readable version.
Telling BeerHQ it is wrong
hello@hekno.co.uk, or see corrections for what happens to a report and which kinds of wrong get which response.