What Is a Good Win Rate on Polymarket?

There is no good win rate in isolation, because win rate is set mostly by the prices a wallet buys at. A wallet that only buys favorites at an average entry of 0.85 should win about 85% of the time with no edge at all. The number that carries information is the gap between how often entries resolved true and what the entry prices implied, read with a bootstrap 95% interval against zero. The free Polymarket wallet analyzer computes exactly that on any public 0x address, no signup.

Win rate is a price-preference statistic

Every Polymarket entry price implies a probability: paying 0.85 implies 85%, paying 0.15 implies 15%. If the prices a wallet pays are accurate, its win rate will simply converge to its average entry price, whatever its skill. A 90% win rate from a favorites-only wallet and a 20% win rate from a longshot wallet can describe records of identical quality. That is why “what is a good win rate” has no honest one-number answer, and why any leaderboard sorted by raw win rate is mostly a leaderboard of price preference.

The number to read instead: edge against price

The readable question is whether a wallet's entries resolved true more often than the prices it paid implied. That gap is the realized entry edge, and on its own it is still just a point estimate; the verdict follows the bootstrap 95% interval around it. 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 win rate says. Below 30 resolved positions the interval is too wide for any verdict, and the honest output is “too few resolved trades to read”. How that read works end to end is covered at is a Polymarket trader skilled or lucky.

A high win rate can lose money

Payoffs are asymmetric. A share bought at 0.90 earns 10 cents when right and loses 90 cents when wrong, so nine wins out of ten at that price only breaks even. A wallet can post a winning-majority record and a negative net result at the same time, and many favorites-heavy records do. Win rate, profitability, and skill are three different questions; the free check reports the pieces separately (calibration with the win rate read next to stated probability, sizing, concentration, and the edge with its interval) instead of collapsing them into one flattering number.

What the published record shows

We do not publish a cohort win-rate table; we publish the harder read. 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. The lone uncorrected positive interval belongs to a wallet that is net-negative on PnL in the published cohort artifact. The full per-wallet table and method are at the top-50 skill scan. The honest takeaway is not that these wallets have no skill; it is that at these sample sizes most records cannot be separated from chance either way, and a win rate alone tells you even less.

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