Performance audit
Am I good, and am I safe, answered for any Polymarket wallet from public data. No account, no API key, nothing to install.
Scoring a forecast requires knowing what someone believed before the outcome, which is why every other tool asks traders to log a probability against each position and then wait dozens of trades to learn anything. That step is unnecessary here. Every fill on Polymarket settles on Polygon with a price and a timestamp attached, so the record already exists: the price paid is the probability revealed.
What this measures
Positions are rebuilt from fills rather than read from the current book. That matters: the positions endpoint is a snapshot, and on one wallet it showed 7 holdings against 294 fills across 32 markets. Auditing the snapshot scored two resolutions and called the trader unmeasurable; rebuilding from fills found 27 and a Brier of 0.076.
One caveat is worth stating rather than burying. An entry price is not a stated forecast, buying at 0.57 means the trader thought the true probability was at least 0.57, not exactly 0.57. So this measures whether entry prices were well calibrated, which is a slightly harsher question than whether stated beliefs were. It is also impossible to fake, which a self-reported number is not.
Polymarket only. Kalshi is a regulated exchange with a private database, its per-account endpoints require the account holder's own credentials, so there is no equivalent public record to read. Method →