Is a Polymarket Trader Skilled or Lucky?

You cannot tell from the P&L. The honest read of a record is the realized entry edge (how much more often the wallet's entries resolved true than the price it paid implied) with its bootstrap 95% interval against zero. If the interval clears zero on the positive side, the record is consistent with an edge. If it includes zero, the record cannot be told apart from chance, whatever the dollar figure says. The free Polymarket wallet analyzer computes this on any public 0x address; first wallet free, no signup, more wallets free with an account.

The concrete test, in four parts

1. Realized entry edge vs the price paid. Every entry price implies a probability (paying 0.62 implies 62%). The edge is the gap between how often the entries actually resolved true and what the prices implied, averaged over the resolved record.
2. The 95% interval vs zero. A point estimate alone is a screenshot. The check reports a bootstrap 95% interval around the edge, and the verdict follows the interval: clears zero on the positive side means consistent with an edge on this record; includes zero means not separable from chance.
3. The readability floor: 30 resolved positions. Below 30 resolved positions the interval is so wide that no verdict is honest, positive or negative. The check says “too few resolved trades to read” and stops there.
4. The independence floor and the concentration note. The interval is an event-clustered bootstrap, so below 20 independent resolved events the read stays insufficient, whatever the point estimate. And when a single event is at least 60% of the net result, the verdict carries the concentrated style note: the per-position edge can be clean while the dollar outcome rode on one outcome, so the caveat travels with the read.

Why most records read “cannot tell”

Binary outcomes are noisy, and at retail sample sizes the interval around a realized edge is wide. On 2026-06-09 we ran our skill read on our own published top-50 cohort; 35 of 50 had enough resolved positions to test; exactly one interval cleared zero on the positive side, roughly what chance predicts across 35 tests at a 2.5% threshold (about 0.9 false positives expected), and it did not survive a multiple-comparisons correction. That is our own board, scanned with our own method, and the honest summary is that a readable edge is rare. Who actually profits, read more broadly, is covered at who wins and who loses in prediction markets.

The badge says one honest thing

Wallet owners can claim a public badge computed from the same read. It renders exactly one honest state: the interval clears zero on the positive side (uncorrected, with a style note saying how the edge was earned), the edge is not separable from chance, or too few resolved positions or independent events to read. “Luck” is a legitimate outcome, not an error state; a verdict can never disagree with its own interval, and the sufficiency floors cap any high point estimate.

Where Edge Score fits, and where it does not

The analyzer also reports an Edge Score percentile against the frozen reference cohort. Edge Score is a descriptive in-sample composite of past behavior; its only out-of-sample forward test held a +0.11 rank correlation (95% CI 0.05 to 0.18) against future PnL and did not clear the pre-set +0.30 bar. So treat the score as a description of how a record looks next to its peers, and treat the realized edge with its interval as the skill-or-luck read. The two answer different questions, and only the second one carries its own uncertainty.

Diagnostics on public on-chain records, not investment advice; a past read is not a forecast.