A deep analysis of on-chain data and high-frequency trading has revealed systematic signs of manipulation on the Polymarket prediction market platform. This concerns five-minute contracts on Bitcoin price movement, launched in February 2026. The study, covering over 60 million records, showed that a group of 821 wallets (just 0.34% of 243,000 active addresses) managed to extract $8.22 million in profit through short-term impact on the spot market.
Mechanics of the Alleged Scheme
The Polymarket five-minute contract pays $1 if the BTC/USD price according to the Chainlink oracle at the time of settlement is not lower than the initial value. A key element of the scheme is directional spot trades in the final seconds before the oracle fix. The trader first buys a contract for an increase or decrease, then, 10 seconds before the deadline, executes a large spot market trade, temporarily shifting the price in the desired direction.
Data confirms: after the launch of five-minute markets, the flow of directional orders in the last 10 seconds of the window increased by 50% relative to the background. At the same time, the price movement began to reverse within the next 10 seconds — a classic sign of temporary impact, rather than a reaction to fundamental news. In cycles with a near 50/50 probability of outcome, the surge in activity was 3.9 times stronger than normal, and the average volume of directional spot trades reached $1.7 million compared to $68,000 in regular cycles.
Concentration During Low Liquidity Periods
56% of suspicious episodes occurred during nighttime hours UTC, and another 44% on weekends. This is no coincidence: in periods of low liquidity, less capital is required to shift the price. Analysis showed that in 34.2% of allegedly manipulated cycles where Polymarket estimated the favorite's probability at 90–100%, directional movement against it led to the underdog's victory. Without anomalous activity, such a scenario occurred in only 1% of cases.
Who Gained and Who Lost
The group of 821 wallets, classified as likely manipulators, earned $8.22 million in anomalous cycles and only $90,000 in others. Retail and unclassified participants lost $7.61 million — 93% of the manipulators' profit. In the overall loss distribution, retail accounted for 70.5%, the manipulators themselves in unfavorable cycles for 20.2%, and market makers for 9.3%.
It is important to note: the study does not prove intent, and wallet addresses may belong to the same individuals. Additionally, a direct link between spot accounts and Polymarket accounts has not been established.
Platform Response and Alternative Measures
Polymarket has already announced plans to move some markets to a mechanism that considers prices over a longer period rather than a single value. In 15-minute contracts, the surge in activity was significantly weaker, confirming that increasing the time window reduces the effectiveness of short-term spot manipulations. Other protective measures include calculation based on average price, random settlement timing, position limits, and additional costs for operations before closing.
My Expert Opinion: This case is a classic example of an "oracle attack" vulnerability under low liquidity conditions. Polymarket urgently needs to implement mechanisms to prevent temporary price distortion, otherwise trust in the platform will be undermined. Retail traders should avoid contracts with short settlement windows, especially during periods of low market activity.