THE RANKING METHOD
Not all evidence deserves the same confidence. I separate what the research supports from what is still exploratory, then rank ingredients by how much more one outperforms another — so you know where to put your money.
The strongest evidence I am comfortable leaning on publicly. Usually direct comparisons with clearer endpoints, better controls, and enough context to explain the result honestly.
This is the core layer I use for the strongest public comparisons.
Real signals, but the evidence is incomplete — smaller samples, weaker controls, or fewer replications.
Useful for direction, not strong enough for confident comparative claims.
Worth watching, not worth overstating. Early-stage or mechanistic research that hasn't been confirmed in rigorous trials.
Interesting direction, weak confidence.
SEPARATE FROM EVIDENCE STRENGTH
This is not a fourth evidence tier. It is a separate flag for cases where the market story and the evidence diverge.
A claim can have decent evidence and still be framed badly in the market. Sometimes the problem is hype. Sometimes it's oversimplification. Sometimes it's a good idea being sold with the wrong promise.
"The research suggests..."
"A study found..."
"There is some evidence that..."
"The research is encouraging, but..."
"Early research suggests..."
"Interesting, but not enough to lean on yet."