When you browse prediction markets like Polymarket or Kalshi, you’re looking at real dollars riding on future events. Switch to Metaculus or Manifold Markets, and you’ll see play money or points. Does the difference matter for accuracy? The answer shapes how we should interpret forecasts and where we should place our trust when making decisions.
The theory: skin in the game vs intrinsic motivation
The classic argument for real-money markets is simple. When forecasters risk actual cash, they think harder and research deeper. Skin in the game filters out casual guesses and rewards people who truly understand an event. This is the prediction market definition that economists have championed since the Iowa Electronic Markets launched in 1988.
Play-money advocates counter with a different story. They argue that intrinsic motivation (reputation, intellectual challenge, community status) can drive accuracy just as well. When you remove financial barriers, you attract a wider pool of talent. Hobbyists, domain experts, and curious minds contribute without needing spare capital or regulatory approval.
Metaculus and the play-money accuracy record
Metaculus has built a remarkable track record using only points and reputation scores. The platform has consistently outperformed traditional polls and expert panels across thousands of questions. Its community forecasters nailed pandemic timelines, climate benchmarks, and technology milestones with striking precision.
The Metaculus tournament accuracy data
In 2024 and 2025, Metaculus tournaments showed Brier scores (a measure of forecast accuracy) competitive with or better than real-money platforms on comparable questions. The platform’s aggregation algorithms and active community created what researchers call collective intelligence forecasting. This success challenges the assumption that money is the only way to extract wisdom from crowds.
Manifold Markets as a hybrid case
Manifold Markets offers an interesting middle ground. Users trade with play money, but top forecasters can convert winnings into charitable donations or small prizes. This hybrid model has grown rapidly since 2023, attracting both serious forecasters and casual participants.
The platform shows that even modest stakes (or indirect ones) can maintain quality. Manifold’s accuracy on political and tech questions rivals platforms with real money, suggesting that community engagement and good design matter as much as financial incentives.
When real money helps and when it hurts
Real money shines in high-stakes, information-rich domains. Financial traders and policy experts gravitate to platforms like Kalshi when they have genuine edge. The capital requirements naturally filter for serious participants who do their homework.
But real money creates problems too. Regulatory friction limits participation in many regions. Wealth disparities mean rich forecasters can overwhelm accurate but cash-poor predictors. And on niche questions with thin liquidity, a single whale can distort prices regardless of accuracy.