Who Actually Makes Money on Prediction Markets
Most traders lose. A tiny few take almost everything. And on this market, you can actually see which camp you're walking into.
In the previous post, we established that prediction markets offer positive expected value — but only if you play correctly: be a maker, not a taker, and bet “no” on underdogs.
But that raises a harder question. Prediction markets, unlike equity markets that grow alongside the economy, are a pure zero-sum game. Every dollar won is a dollar lost by someone else. So is there actual skill here — or does money just shuffle endlessly between participants?
The answer is clear.
Winner takes almost all
Of 1.4 million Polymarket users, fewer than one in three ever finished in the green. Those 450,678 winners shared $512.8 million in total gains. But even within that group, the math is brutal.
The top 0.1% captured 58.5% of all profits. The top 1% took 84.1%. By the time you reach the top 5%, you’ve accounted for 95.8% of every dollar ever made on the platform.
And that’s just among those who came out ahead. Overall, 70.8% of all participants finish in the red.
A separate study of 2.5 million traders puts that number even higher. Only 15.9% of Polymarket traders finished in the green — 84.1% lost money. Of those who came out ahead: only 2% ever earned more than $1,000 over their entire trading history. Only 0.32% — roughly 8,000 wallets — crossed $10,000. Fewer than 840 addresses ever made more than $100,0001.
This isn’t money sloshing around randomly. A tiny group extracts nearly the entire surplus, and a much larger pool of losers funds it.
Is it skill or luck?
Skill. And researchers can prove it.
To separate genuine edge from lucky streaks, the studies use a metric called Excess Hit Rate (EHR). EHR measures one thing: do you consistently buy bets cheaper than they should be? If yes, your EHR is positive. If you overpay, it turns negative.
Here’s what makes this hard: Polymarket prices are remarkably well-calibrated in aggregate — events priced at $0.70 really do occur about 70% of the time. The crowd is not stupid. But “accurate on average” doesn’t mean “accurate on every contract.” Across 70 million trades, the EHR declines sharply and consistently through the profit distribution — from +0.15 for the top 2% of earners to -0.15 for the worst performers. The best traders aren’t luckier. They’re systematically better at identifying underpriced contracts.
The fundamental difference between Polymarket and every other market
When you trade on a conventional exchange, your counterparty is completely invisible. It could be a quant fund running sophisticated arbitrage strategies, an algorithm operating at millisecond latency, or someone who just had a feeling. You have no idea. And that invisibility isn’t accidental — it’s valuable.
This asymmetry is so lucrative that hedge funds actually pay to access retail order flow. The practice is called Payment for Order Flow (PFOF), and it’s the entire reason Robinhood charges zero commissions2. They sell your trades to quant funds who happily pay for the privilege of trading against uninformed money. Robinhood earns hundreds of millions per year from this.
Polymarket runs on a public blockchain. Every transaction is visible. You can look up the wallet on the other side of your trade, check its full history, and see whether you’re sitting across from someone who’s been consistently right for two years — or a chronic loser who’s never turned a profit.
Trading against uninformed money doesn't cost you a thing. It's all just there.
Conclusion
Warren Buffett once wrote to his investors: “If you’ve been playing poker for half an hour and you still don’t know who the patsy is, you’re the patsy.”
On a stock exchange, this saying is almost unfair — you genuinely cannot see the table. You’re playing blind against opponents whose skill level is completely hidden. On Polymarket, you can see everything. The wallets, the history, the position sizes.
That makes the question pretty simple. If you've spotted what looks like easy money on a mispriced contract — and every profitable trader on the platform has quietly taken the opposite side — it's worth asking: are you part of the 84% feeding the top 1%?
This study treats one wallet as one person, which may slightly skew the numbers — in practice that's not always the case. But the overall picture doesn't change.
Your “free” trade isn’t free — you pay with execution quality and behavioral data. PFOF is banned in Europe, restricted in the UK, and has survived in the US only because the industry lobbies hard to keep it.