Tool signal brief
The Evidence-Gap Penalty I Use for Running-Shoe Claims
Soft, responsive and stable are not comparable claims when one listing shows measurements and the other shows only marketing copy. Penalize the missing proof.
One listing calls a running shoe “responsive.” Another calls its midsole “soft and stable.” Those phrases cannot be compared until I know what, if anything, supports them.
I give each claim an evidence-gap penalty. The penalty is not a verdict that the shoe is bad. It stops unsupported language from winning a comparison.
Score the support, not the adjective
A visible midsole shape, outsole coverage, pair weight, flex view or model-specific review can support part of a claim. None proves the whole ride. I write what the evidence actually establishes and leave the rest open.
I subtract confidence when the photos hide the side profile, when every statement comes from seller copy, or when a stability claim conflicts with a very narrow-looking base. A contradiction deserves more weight than another repeated adjective.
I use the comparison view to keep price, evidence and uncertainty on the same card. Two shoes may finish tied because they answer different use cases.
Then I return to the running-shoe category and ask a better question: which candidate has enough proof for the job I named? A listing with modest claims and readable construction can beat a louder one. Missing evidence should remain visible, not be smoothed into a score.