What is a prediction market? It’s a platform where people buy and sell contracts based on future events. Think of it as a stock market for forecasts. These markets turn opinions into prices, and those prices often beat expert predictions. The prediction market definition goes beyond simple betting. It captures the collective intelligence of diverse participants who put real money behind their beliefs.
The Surowiecki framework, two decades later
James Surowiecki’s 2004 book “The Wisdom of Crowds” laid the foundation for understanding prediction market basics. He argued that groups often make better decisions than individual experts. Two decades later, platforms like Polymarket and Kalshi prove his thesis daily. These modern prediction markets aggregate thousands of forecasts into a single, constantly updating probability.
Diversity, independence, decentralization, aggregation
Surowiecki identified four conditions for crowd wisdom to work. First, you need diversity of opinion. Second, people must think independently without copying others. Third, individuals should draw on local knowledge. Fourth, you need a mechanism to aggregate all those private judgments into one collective decision. Prediction market mechanics fulfill all four requirements naturally.
How prediction markets work is simple. Traders buy contracts that pay out if an event happens. If you think a candidate will win an election, you buy a “Yes” contract. The current price reflects the crowd’s probability estimate. A contract trading at 65 cents suggests a 65% chance of that outcome.
When crowds outperform experts: the research evidence
Prediction markets vs polls reveals a clear winner. The Iowa Electronic Markets, running since 1988, consistently beat traditional polling. During the 2008 presidential election, prediction markets called the outcome more accurately than most expert forecasters. These markets updated faster as new information emerged, while pundits stuck to outdated narratives.
Famous accurate predictions: 2008 election, COVID timelines
Prediction markets nailed the 2008 election results within percentage points. During COVID, platforms like Metaculus tracked vaccine timelines and case counts more accurately than official estimates. The collective intelligence forecasting model worked because traders had skin in the game. Bad forecasts cost them money, creating a powerful incentive for accuracy.
When crowds fail: bubbles, manipulation, low liquidity
Prediction markets aren’t perfect. Low liquidity creates wild price swings that don’t reflect true probabilities. When only a few traders participate, the wisdom of crowds prediction markets disappears. Manipulation is another risk. Deep-pocketed actors can sometimes move prices temporarily, though this usually creates profit opportunities for informed traders who correct the distortion.
Binary contracts explained show one structural weakness. Yes-or-no questions work well, but complex multi-outcome events can fragment liquidity. Types of prediction markets vary: binary markets vs scalar markets each have strengths. Categorical prediction markets handle multiple outcomes but need enough volume in each contract.