84% of Polymarket wallets lose money. What the winners do differently

Prediction markets look easy: you have an opinion, so you buy YES or NO. The numbers are less friendly. An on-chain analysis of about 2.5 million Polymarket wallets, published in April 2026, found that 84% of them were in the red. Only 2% had made more than $1,000, and just 0.033% (about 840 addresses) had made more than $100,000.

One caveat: the study counts wallets. One trader can run many of them, and plenty belong to people who tried once and left. Still, the direction is clear: the default outcome is a small loss.

Half the profits go to a tiny group

A separate study by Solidus Labs looked at Polymarket's political markets between December 2025 and February 2026. It found that 0.55% of profitable maker wallets captured half of all maker gains, and 0.26% of winning taker wallets took nearly the same share. A research paper from IMDEA Networks estimated that arbitrage alone extracted about $40 million in a year, and that the biggest earners looked like bots.

So who wins? Judging by these studies, mostly automated strategies and liquidity providers (makers), plus a small number of experienced discretionary traders. Here is how one of the best known of them describes his approach.

1. Let big payoffs cover small losses

Domer, a former professional poker player, put it bluntly in a 2024 interview: “I make a lot of bets, and lose a lot of bets.” And: “I think I might be losing slightly more bets than I am winning, but winning more money than I am losing.” The edge in his approach is the research behind each bet. A low price alone is no edge: in a 2026 study of 588 million Polymarket trades, buys below 10¢ lost about 19 cents per dollar on average, although the result flips when related contracts are grouped by event.

2. Size by confidence

From the same interview: “If you find an edge but you're not sure if it's any good, bet a little bit. If you find what you think is a big edge, bet a lot.” And his warning to beginners: “A lot of people bet way too much when they don't have an edge.”

3. Scan hundreds of markets, trade a few

In a 2026 interview Domer described his routine as “going through hundreds of markets” and added that “nine times out of ten, I see the price, and it seems fine.” The money is in the tenth market, where the price is wrong. He also keeps “a core group of people that I trust and talk to”.

4. Build your own probability estimate

Winning traders rarely accept the market price as the answer. Some commission their own research; others build models. A weather trader might compare forecast models with market prices and trade only the large gaps, after checking which weather station the rules use. An edge starts with a number you worked out yourself.

5. Be the maker

Makers pay no fees and receive rebates. Posting limit orders inside the spread on markets you understand is one of the few structural advantages an ordinary trader has. The catch: better-informed traders will fill your orders exactly when you're wrong, so pull them before major news.

Five questions before every trade

• Can I write my edge in one sentence?
• What probability do I estimate, and how far is it from the price?
• Have I read the rules to the end?
• Is the position small enough that being wrong doesn't hurt?
• Am I taking liquidity (and paying a fee) when I could be providing it?