Bitquery Reports Concentration and Timing Disputes in Polymarket Resolution
Bitquery published an on-chain analysis on October 9, 2026, matching 470 Polymarket disputes to 489 resolved UMA voting questions over September 9 to October 7. The study reports that the nine largest wallets held an average of 56.3 percent of voting weight and agreed on 483 of the 489 questions. Removing those wallets changed none of the sampled results. Of the 470 matched disputes, 243 ended in a too early ruling that returned the question for a later proposal. The most active disputer won 80 of 80 disputes, including 72 timing corrections. Four wallets lost 46 disputes over one football match, posting 500 USDC per dispute; Bitquery reports 23,000 USDC in forfeited bonds but did not trace settlement payouts. These are attributed findings from a limited sample, not a finding of collusion, common wallet ownership, erroneous real-world outcomes or protocol misconduct.
PARADIGM INSIGHTS:
Voting concentration and demonstrated manipulation are different questions. This sample shows why proposal timing, evidence rules, bond incentives and a clear challenge process matter alongside who holds voting weight. It does not establish that concentrated voters changed the outcomes or that a particular institutional model is required. A wallet is not necessarily a person, and a correct vote under market rules is not by itself proof that the underlying real-world answer is correct.
Sign up for Paradigm alerts.
Join the Paradigm Insights list for future case alerts and selected firm updates.