Study: only ~3% of Polymarket accounts drove most price discovery in 2023-2025, suggesting market accuracy comes from an informed minority, not crowd wisdom
Researchers show market accuracy comes from a tiny group of informed traders, not broad participation. … What to know:
Context & Ripple Effects
The finding adds a concentration lens to Polymarket’s recent growth narrative: a separate analysis found that 0.1% of accounts captured 67% of profits, particularly among high-frequency traders. It suggests the traders moving prices and those extracting returns may be similarly narrow cohorts.
That concentration matters because related coverage has also raised questions about wash trading and suspicious profits. If price discovery is driven by relatively few accounts, market-quality claims depend more heavily on the behavior and information advantages of those accounts than on headline participation or volume.
First-order effects
- Polymarket’s apparent forecasting accuracy becomes attributable primarily to a small set of informed accounts, rather than to broad user participation.
- Casual traders face a market in which the most consequential repricing may occur before their activity meaningfully affects odds.
Second-order effects
- Measures of platform health based on account counts or trading volume become less informative; analysts and users will place greater weight on participant quality, concentration, and trading integrity.
- The overlap of concentrated price discovery with prior wash-trading and suspicious-profit concerns raises the stakes for surveillance of influential accounts and for clearer market-quality disclosures.
Third-order effects
- Prediction markets may increasingly be evaluated as information markets operated by specialist liquidity and research traders, not as pure aggregations of public sentiment.
- If this pattern persists across platforms, credibility will hinge on whether concentrated expertise can be distinguished from privileged information or manipulative activity.
The trend: Prediction-market platformization is shifting the debate from whether crowds can forecast well to how a small, identifiable set of traders produces—and potentially distorts—market signals.