A fresh batch of on-chain data is putting a hard number on something many crypto traders quietly suspect, which is that most people aren’t making money on prediction markets as it seems. According to a new analysis of millions of wallets, the overwhelming majority of users on prediction platform Polymarket are operating at a loss, and only a tiny fraction have managed to turn consistent profits.
The findings, published in early April, come at a time when prediction markets are enjoying renewed attention, particularly around elections, sports, and geopolitical events. But behind the hype and viral screenshots of winning bets, the data paints a more sobering picture, one that may challenge the popular narrative that “anyone” can profit from forecasting the future.
The losses is not by a small margin
The report, compiled by independent on-chain analyst Andrey Sergeenkov, examined roughly 2.5 million wallet addresses interacting with the platform. The headline number reads 84.1 percent of traders were found to be in the red, meaning they lost more money than they earned over their trading history.
Even more striking, the pool of meaningful winners appears extremely small. Only about 2 percent of traders have ever earned more than $1,000, while a microscopic 0.033 percent, roughly 840 wallets, managed to surpass $100,000 in total profits.
Those numbers may sound harsh, but they also mirror patterns seen across other speculative markets, from retail stock trading to sports betting. A lot of participants jump in with confidence, ride a few early wins, and then gradually give those gains back. The difference here is that blockchain data leaves a permanent trail, making it easier to quantify the outcome.
The study also explored what it takes to earn a steady income from prediction trading, say, around $5,000 per month. The odds of achieving that consistently were found to be less than 1 percent in any given month, and the chances shrink further over time.
In practical terms, it suggests that turning prediction markets into a reliable livelihood is not impossible, but it’s statistically rare.
Why profits are so concentrated among a few traders
One explanation offered in the research and echoed by academics studying prediction markets is that the playing field isn’t as level as it might appear. While anyone can open the app and place a bet on an event, not everyone has the same tools or speed.
Sophisticated participants often rely on automated strategies, arbitrage models, or real-time data feeds that react faster than manual traders. In markets where prices shift within seconds of new information, those advantages compound quickly. Over time, they create a familiar dynamic whereby a small group captures most of the profits, while the majority absorbs the losses.
There’s also a behavioral angle to consider. Many users treat prediction markets like a mix of trading and entertainment. They might bet on elections, sports outcomes, or trending news stories without applying strict risk management. That approach can work occasionally, but it rarely holds up over dozens or hundreds of trades.
Interestingly, the data suggests that profitability is often short-lived. Among traders who did manage to earn more than $5,000 per month on average, more than half remained active for only a single month before disappearing. Only about 2.6 percent stayed consistently profitable for more than a year. That pattern hints at volatility rather than sustained skill.
Rapid growth may be making matter worse
The timing of the report is worth noting. Prediction markets have grown dramatically over the past two years, fueled by election cycles, sports partnerships, and social media buzz. Trading volumes have surged, and new users continue to sign up in large numbers.
But growth can have unintended side effects. When fresh participants flood into a market, they often arrive with less experience and fewer strategies, making them more vulnerable to losses. Meanwhile, seasoned traders, particularly those running automated systems, benefit from the additional liquidity.
In other words, popularity doesn’t necessarily translate into profitability. It’s also important to remember what prediction markets actually are. Platforms like Polymarket allow users to trade shares representing the likelihood of future events, from political outcomes to weather patterns, using cryptocurrency collateral.
That structure can make the experience feel analytical, even scientific. Yet at its core, the system remains a zero-sum environment: for one trader to win, another must lose.
None of this means prediction markets are inherently flawed. They can provide useful signals about public sentiment and probabilities, and in some cases, they’ve proven surprisingly accurate at forecasting real-world events. Still, accuracy as a forecasting tool doesn’t guarantee profitability for participants, a distinction that’s easy to overlook during a winning streak.