How we grade evidence

Proof, Promise & Risk — and why we split them

We refuse to hide behind a single vague 'hype' number that conflates how proven something is with how much upside it has. So we show three honest scores, then let you see the trade-off:

• PROOF — how strong the actual evidence is (this is the S–F grade, shown as a 0–5 bar). High Proof = real human data.
• PROMISE — how effective it could be IF the claims pan out (the upside). A compound can be high Promise but low Proof — exciting but still unproven. That's a feature worth chasing, not a crime.
• RISK — a composite of known side effects, unknowns, market/quality risk, and irreversibility.

RISK-TO-REWARD % is computed, not hand-set: reward (Proof + Promise) weighed against Risk. It reads as a score, not a scolding. Tirzepatide (Proof 5, Promise 5, Risk 2) lands ~83% — strong. BPC-157 (Proof 2, Promise 4, Risk 3) lands ~67% — a genuinely promising bet the human trials haven't caught up with yet. High Promise on thin Proof isn't a mark of shame; it's exactly where the interesting frontier lives.

Evidence Grade (S–F) — the Proof scale

The grade reflects the quality of evidence for the effect people actually want — not the best-case reading. A compound can have fascinating rat data and still be a D, because animal results routinely fail to translate to humans.

• S — Gold standard: robust human trials AND regulatory approval.
• A — Robust human randomised trials.
• B — Decent human data, some gaps.
• C — Limited or early human data.
• D — Animal-only, with biological plausibility.
• E — Anecdotal or purely theoretical.
• F — Hype with little or no support.

The regulatory badge (the category-error killer)

This badge exists because the single biggest mistake in peptide culture is treating an approved medicine and a grey-market experiment as the same kind of thing. They aren't, and we refuse to blur them.

• Approved medicine — has real regulatory approval and human data (somewhere, for some use).
• In human trials — genuine human trial data, not yet approved.
• Cosmetic / topical — modest real evidence for topical or cosmetic use.
• Animal data only — interesting in animals or cells; not shown in humans.
• Grey-market — unapproved, sold 'research only', human evidence thin or absent.

If we ever score something in a way you can't square with the sources we cite, that's a bug — tell us. Faking certainty is the one thing we're not allowed to do.