Polymarket: bots dominate, 69% lose
- Bloomberg reported on April 29 that Polymarket’s profits are being captured by a tiny set of high-activity accounts while most traders lose money. - The sharpest numbers are ugly: 69% of traders have lost money since 2022, and the top 1% captured roughly three-quarters of profits. - That matters because prediction markets sell “wisdom of crowds,” but the money increasingly looks concentrated in bots, arbs, and market-making specialists.
Prediction markets are supposed to look like crowdsourcing with money on the line. But Polymarket now looks a lot more like a speed game — one where bots and highly active traders do most of the winning, and everyone else mostly funds them. That was the real news in the latest Bloomberg reporting on April 28 and 29, built on wallet-level analysis and a recent academic paper. The headline number is simple: most people lose. (bloomberg.com) ### What actually changed? What changed this week is that the broad suspicion finally got hard numbers attached to it. Bloomberg said more than 100,000 Polymarket accounts have lost at least $1,000 since the start of 2025, nearly twice the number that made that much, after reviewing every wallet active since early 20(bloomberg.com)lost money since 2022. (bloomberg.com) ### Why are bots so good here? Because this is less about “being right” in some grand sense and more about getting the better price first. Bots can monitor dozens or hundreds of markets at once, react instantly to news, and exploit tiny gaps between related contracts. Humans show up later, click worse prices, and pay the spread. On a platform where contracts move in pennies and timing matters, that edge compounds fast. (briefs.co) ### Is this just a few whales? Not exactly. It’s a few whales, a few market makers, and a lot of automation. The paper cited in this week’s coverage said the top 1% of traders captured about three-quarters of profits. Other recent on-chain analyses point in the same direction — profits are extremely concentrated, with a tiny sliver of addresses taking mo(briefs.co)in electronic trading everywhere else. (fa-mag.com) ### Why do regular users lose? Mostly because they trade like fans, not like market makers. One useful detail from the research: lower-performing users trade much more often at extreme prices — below 10 cents or above 90 cents. Basically, they chase certainty, overpay for “obvious” outcomes, and leave little upside even when they’re directionally correct. If you buy a contract at 94 cents and it wins, you were right — but you still only make 6 cents. (beincrypto.com) ### Doesn’t that break the “wisdom of crowds” idea? It weakens the romantic version of it. Prices can still be informative even if profits are concentrated. In fact, one reason bots make money is that they help keep prices aligned across markets. But there’s a catch — if the platform is marketed socially as an easy side hustle, while the actual economics reward spe(beincrypto.com)don’t really understand. (theedgemalaysia.com) ### Why does this matter now? Because Polymarket is no longer some niche crypto toy. It has become one of the best-known prediction platforms in the world, and the whole category is pushing toward mainstream finance and politics. At the same time, the industry is already dealing with questions about manipulation, insider-style advantages, market integrity, and who should regulate these venues. (theedgemalaysia.com)or casual users. (en.wikipedia.org) ### So what’s the bottom line? The new reporting did not reveal a scandal so much as reveal the business model’s real shape. Polymarket looks less like “everyone can monetize their opinions” and more like a thin, fast market where professionals harvest mistakes. That does not make the platform fake. But it does make the pitch much shakier for ordinary users — especially if they think winning the news cycle is the same thing as winning the trade. (bloomberg.com)
