# How the reviews work

## The unit is a lot, not a strain

Every review is attached to a producer, lot identifier, package date when available, and review date. A name can remain familiar while genetics, cultivation, and handling change. The site therefore does not turn one lot report into a claim about every product sold under the same cultivar name.

<p class="claim-label"><strong>Claim strength: Published evidence</strong></p>

Genetic testing has found that commercial cannabis samples sharing a strain name can be genetically inconsistent. The source stays attached to that statement: [Schwabe and McGlaughlin, *Journal of Cannabis Research* (2019)](https://link.springer.com/article/10.1186/s42238-019-0001-1).

## Three visible claim states

The label is written into the text, not communicated by color alone.

| Visible label | Meaning |
|---|---|
| Claim strength: Published evidence | A statement tied to a linked paper, dataset, statute, rule, bulletin, or agency page. |
| Claim strength: Industry tradition | A practice or interpretation common in the trade but not established here as research evidence. |
| Claim strength: Reviewer experience — one person, one lot | A person's report about the identified lot at the recorded dose and method. |

## Separate reports, never a pre-averaged verdict

Each reviewer report is its own `<article class="reviewer-report">` in the HTML. If two reviewers disagree, both reports and their scores remain visible. The spread is part of the result; the site does not replace it with a mean or consensus score.

## Evidence and illustration are different things

Product photographs, label photographs, and linked source documents can establish facts within their limits. Generated illustrations establish none. Every generated illustration carries a visible disclosure beside it.

## Estate and carried brands

An **estate** operation grows what it sells. A **carried brand** is product grown by someone else and sold by the retailer. Producer descriptions state which relationship applies rather than assuming that everything on one shelf came from one grow.

## Medical information boundary

Dosing, timing, symptom, and interaction material is informational, not medical advice. A reviewer's experience remains attributed to that reviewer. General empirical claims link to their source, and summaries should carry those citations.