Does Polymarket skill persist?

By the Cent Signals editorial desk. Reviewed June 17, 2026.

A trader posts a strong month on Polymarket. Does that tell us anything about how the next month goes? The honest answer, grounded in the public record, is: not much. The most thorough study of the platform so far finds that performance persists only modestly from one period to the next, and that high turnover among participants makes any single good month a weak signal of durable skill. This guide explains what persistence means, what the research found, and how to read a profit ranking in that light. Like everything on this site, it is description, not direction.

What persistence means

Persistence is whether a trader who does well in one period tends to do well in the next. It is the statistical test that separates a repeatable edge from a lucky run. Strong persistence, where good periods reliably follow good periods, is evidence of durable skill: something about how a trader operates carries forward. Weak persistence, where a strong period is about as likely to be followed by an ordinary one, is consistent with luck, or with a population of participants that keeps changing so fast that no single account has the chance to compound an advantage.

The finding: modest, and concentrated

In Who Wins and Who Loses in Prediction Markets? Evidence from Polymarket (Akey, Gregoire, Harvie, and Martineau, 2026), the authors examined roughly 67 billion dollars of trading volume. Month-to-month performance turned out to be only modestly persistent, and what persistence there was concentrated among traders who worked primarily through limit orders, the liquidity providers who post quotes rather than take them. That result held under risk-adjusted measures such as the modified Sharpe and Sortino ratios, which discount returns that come with heavy swings. For the broad population of order-takers, the link between one month and the next was faint.

Churn complicates the picture

A large part of why persistence is hard to pin down is that most accounts do not stay long. The study found about 44 percent of users stopped trading within a month and up to 66 percent within six months, including roughly 55 percent of the best performers. Because so many accounts enter and exit, the active population is continually reshaped. That means apparent persistence may partly reflect who happens to keep trading, a selection effect, rather than a repeatable edge held by the same people over time. When the cohort turns over this quickly, the group that looks skilled this month is not the same group next month.

Why a hot streak is weak evidence

Put the pieces together and a single strong month reads as a noisy signal. Most participants are short-lived, and gains concentrate in a small group, so one good period is easily produced by chance within a large and shifting field. The study is explicit about the caution it implies: short-run results should not be read as proof of skill. It also notes a structural limit on what can be concluded, since the data lacks the long time series and the benchmark model that traditional-market research leans on to judge whether returns reflect skill or simply exposure. Without those tools, a hot streak stays an observation, not a verdict.

How to read our rankings in that light

Cent Signals ranks wallets by realized profit as an observation of public data, not a forecast. A high position on the board records what a wallet has already done; it carries no claim about what comes next, and past results say nothing about future outcomes. The research above is the reason we keep that framing strict: persistence is weak, turnover is high, and a profit figure is a fact about the past rather than a guide to the future. For the wider question of who ends up ahead on the platform, see who actually wins on Polymarket, and to read the underlying figures see traders by realized profit. The numbers describe; they do not predict.

Frequently asked questions

Does a good Polymarket month predict the next one?

Only weakly. The largest study of Polymarket to date found month-to-month performance was modestly persistent at best, and what persistence existed concentrated among a narrow group of traders who worked mainly through limit orders. For most accounts a strong month is a noisy signal that says little about the month that follows.

Is Polymarket success skill or luck?

The evidence points to a mix, with luck and selection accounting for much of what a single good month appears to show. Some durable performance is visible among liquidity providers under risk-adjusted measures, which is consistent with skill, but the short trading lives of most participants make it hard to separate repeatable edge from chance across the wider population.

Why do so many top traders stop?

Churn on Polymarket is high across the board. The study found roughly 44 percent of users stopped within a month and up to 66 percent within six months, including about 55 percent of the best performers. Strong results in one period therefore do not reliably carry into the next, partly because many of the people who produced them simply leave.

Does Cent Signals predict which traders will profit?

No. Cent Signals ranks wallets by realized profit as an observation of public on-chain data. It is a record of what has already happened, not a forecast. Past results say nothing about future outcomes, and nothing on the site is a recommendation to follow or copy any account.