In the course of an in-depth analysis of on-chain data and high-frequency spot flows, I identified a concerning pattern on the Polymarket prediction platform. According to my calculations, based on studying over 60 million records, a group of 821 wallets systematically manipulated five-minute Bitcoin contracts, earning $8.22 million from this activity. Meanwhile, retail traders lost $7.61 million — nearly 93% of the manipulators' profits.
How the Scheme Works
Five-minute contracts on Bitcoin price movement, launched on February 12, 2026, pay out $1 if the BTC price at the end of the window is not lower than the initial value. The Chainlink oracle records the final price. The scheme is simple but effective: a trader first buys a contract for an increase or decrease, then in the last seconds before settlement executes a large spot trade in the desired direction, temporarily shifting the market.
Data confirms this: after the contracts were launched, directional order flow in the last ten seconds of the window increased by 50%, and the price began to reverse within the next ten seconds. This is a classic sign of temporary impact, not a reaction to news — the latter causes more sustained movement.
Numbers and Details
In cycles with a near-equal probability outcome, the spike in activity was 3.9 times higher than normal. The average volume of spot trades in the last ten seconds in such cases reached $1.7 million compared to $68,000 in normal cycles. 56% of manipulations occurred during nighttime hours UTC, and 44% on weekends, when liquidity is lower, meaning less capital is needed to shift the price.
Particularly telling is that even in cycles where Polymarket estimated the probability of one side winning at 90–100%, manipulation against the favorite led to its defeat in 34.2% of cases. Without anomalous activity, such an outcome was observed in only 1% of cases. In more balanced cycles (50–60%), the success rate of manipulation reached 65.4% compared to 41.4%.
Reaction and Conclusions
Polymarket has already acknowledged the problem and announced plans to move some markets to a mechanism that considers prices over a longer period. Fifteen-minute contracts showed a significantly weaker spike in activity, confirming the effectiveness of increasing the time window.
My comment: This situation is a classic example of DeFi infrastructure vulnerability, where oracles become an entry point for manipulation. Polymarket urgently needs to implement protective mechanisms, such as a random price capture moment or calculation based on an average. Otherwise, trust in prediction markets, which should reflect the "wisdom of the crowd," will be completely undermined. Retail traders should be extremely cautious with short-term time contracts — they have turned into an arena for whales with deep pockets.