/ anti-portfolio

Where we were wrong

Most analysis tools never tell you when their score was wrong. We do. Every founder who opts in publishes their actual outcome — shipped, killed, succeeded, failed — alongside the engine’s original prediction. The misses are the proof we’re honest.

Methodology: Pseudonyms are user-chosen at publish time; idea text is never public. Engine scores are immutable from analysis time (verifiable via /receipt/[id]). Outcomes are self-reported and random-spot-audited via founder follow-up. A “miss” is defined as engine score < 60 + shipped-succeeded, score ≥ 75 + shipped-failed, or killed-incorrectly.