Module 1 · Lesson 2
Why markets can predict
A prediction market price can forecast the future more accurately than the experts — because it aggregates what many independent people privately know into a single number. This lesson builds that mechanism up, tests it against the evidence, and finds its limits.
Key takeaways
- A prediction market price aggregates what many independent people know into a single number — often more accurately than any one expert
- The mechanism is simple: people who know something are willing to bet on it, and their money moves the price toward the truth
- This isn’t magic or crowd-worship. It works under specific conditions — diverse participants, independent information, real incentives — and it fails when those conditions break
- Prediction markets have beaten polls, pundits, and expert panels in real forecasting contests. They’ve also been wrong, and understanding when is as important as understanding why
The claim we need to earn
In Lesson 1, you saw that a contract trading at $0.65 means the market believes there’s a 65% chance the event happens. We treated that as a definition. But it hides a much larger claim — one that sounds almost too good to be true when you say it plainly:
A crowd of strangers, trading for their own profit, can forecast the future more accurately than the experts whose job it is to forecast it.
That’s the claim this lesson has to earn. Not assert — earn. Because if it’s true, it’s the entire reason prediction markets matter. And if it’s only sometimes true, then knowing the difference is the most valuable thing you’ll learn in this course.
Let’s build it up from the mechanism, then test it against the evidence, then find its limits.
Why a price knows more than you do
Start with a single trader. She’s watched every USA game this season, read the injury reports, and built a rough model in her head. She thinks the USA has a 70% chance to win tonight. The market is pricing it at $0.65 — implying 65%. To her, that contract is underpriced. She buys.
Her buying nudges the price up. Not much — she’s one person. But she’s not the only one who knows something.
Somewhere else, a former team physio sees the same $0.65 and knows something the public doesn’t: the USA’s best defender has been quietly playing through an injury. He thinks the real number is 55%. To him, the contract is overpriced. He sells.
Neither of them knows what the other knows. She has form and history; he has private medical insight. But the price absorbs both of their views at once. Her buying pushes it up; his selling pushes it down. Where it settles reflects the combined weight of everything each participant knows — even though no single participant knows all of it.
This is the core idea, and it’s worth stating precisely: a market price is a weighted average of the private information held by everyone trading, with each person’s influence proportional to how much money they’re willing to risk.
That last part matters. In a poll, everyone counts equally — the expert and the guesser get one vote each. In a market, the person who’s confident puts more money down, and moves the price more. The market automatically weights opinions by conviction, and conviction — when people are betting their own money — tends to correlate with actually knowing something.
The idea is older than the technology
This isn’t a new insight that arrived with Kalshi or Polymarket. It’s one of the oldest ideas in economics, and the history is the argument — so it’s worth seeing where it comes from.
In 1945, the economist Friedrich Hayek published an essay called The Use of Knowledge in Society. His question was deceptively simple: how does an economy coordinate itself when the knowledge it needs is scattered across millions of people, none of whom sees the whole picture? A factory manager in one city knows his supply costs; a shopkeeper in another knows her customers’ demand. No central planner could ever gather all of it.
Hayek’s answer was that prices do the gathering. When tin becomes scarce, its price rises, and everyone who uses tin adjusts — without needing to know why it became scarce. The price carries the information. It’s a signal that compresses the private knowledge of thousands of people into a single number that anyone can read.
A prediction market is Hayek’s insight turned into a purpose-built instrument. Instead of letting price discovery happen as a side effect of buying and selling actual goods, a prediction market creates a contract whose only purpose is to have a price — a price that means “the probability of this event.” It’s information aggregation, distilled.
Then, in 1988, some economists at the University of Iowa decided to test whether it actually worked. They built the Iowa Electronic Markets — a small, real-money market where traders bought and sold contracts on the outcome of the US presidential election. The stakes were tiny, capped at a few hundred dollars per trader. The question was whether this little market could beat the polls.
It could. Across the 1988 election and many since, the Iowa markets have tended to be closer to the final result than the major polls — and not just on election day, but for months beforehand. A few hundred people risking lunch money were, collectively, sharper forecasters than professional pollsters surveying thousands.
Why? Because a poll asks “who will you vote for?” — a question about the respondent. A market asks “who will win?” — a question about the world. When you’re betting money on the winner, you don’t answer with your preference; you answer with your best estimate of everyone else’s behavior, including things the polls miss. The incentive changes the question, and the better question gets a better answer.
The condition that makes it work
By now the mechanism should feel intuitive: independent people, private information, real money, one price. But notice the hidden ingredient in every example — independence.
The USA fan and the team physio knew different things. Their errors pointed in different directions. When you average many independent estimates, the errors tend to cancel — one person’s too-high guess offsets another’s too-low guess — and what’s left is the signal they had in common: the truth, or something close to it.
This is the real engine behind “the wisdom of crowds,” the phrase James Surowiecki popularized in his 2004 book of the same name. His classic example wasn’t a market at all: at a county fair in 1906, the statistician Francis Galton watched 787 people guess the weight of an ox. No single guess was exact. But the average of all 787 guesses was 1,197 pounds — and the ox weighed 1,198. The crowd, as a whole, was off by one pound, beating every individual expert in the crowd.
The crowd was right for a specific reason: the guesses were independent. Each person looked at the ox and formed their own estimate without conferring. Their individual errors were scattered, so they canceled. The average distilled out the noise and left the signal.
How much does this ox weigh?
Hold onto that word — independent — because it’s also where everything falls apart.
Where it breaks
If independence is the engine, then anything that destroys independence breaks the machine. And plenty of things do. A prediction market is not a crystal ball, and the same forces that make it smart can make it dumb. Here’s where the mechanism fails.
Thin markets. The wisdom of crowds needs a crowd. A market with five traders isn’t aggregating diverse information — it’s reflecting the opinions of five people, who might all be wrong in the same way. On obscure questions with little money and few participants, prices are noisy and easily pushed around. The magic is real but it scales with participation; a market nobody’s trading tells you almost nothing.
Herding. Galton’s fair worked because nobody could see anyone else’s guess. But real markets are visible — everyone sees the current price. That visibility can destroy the very independence the mechanism depends on. If traders stop forming their own estimates and start copying the price (“it’s at 70%, so it must be likely”), the errors stop canceling and start compounding. The crowd stops thinking and starts following. This is how bubbles form in every market, and prediction markets aren’t immune.
Now let the crowd see itself.
Same 787 people, same ox. The only thing that changes is whether each person can see the guesses made before theirs.
Manipulation. Because a prediction market price is a public signal, someone with an agenda has a reason to distort it. A trader might buy a candidate’s contract not because they think the candidate will win, but to create the impression of momentum — to move the number that headlines will quote. In deep, liquid markets this is expensive and hard to sustain: other traders pounce on the mispricing and push it back. But in thin markets, a single motivated actor can move the price meaningfully. The defense against manipulation is the same as the defense against noise — volume, liquidity, participation.
Long-shot bias. Humans are systematically bad at pricing rare events. Across many prediction markets and betting markets, contracts on very unlikely outcomes tend to trade a little too high — people overpay for the thrill of a big payoff, the way they overpay for lottery tickets. The mirror image happens with near-certainties, which often trade slightly too low. The market is still directionally right — the favorite is still the favorite — but the exact probability can be skewed at the extremes.
None of these break the core claim. They bound it. A prediction market forecasts well when it has enough independent participants with real incentives and enough liquidity to punish distortion. Take those away and it degrades — gracefully at first, then completely.
So how much should you trust a price?
Here’s the honest synthesis, and it’s the thing to carry out of this lesson:
A prediction market price is not the truth. It’s the best available estimate, produced by a specific machine, and that machine works better under some conditions than others. When you look at a price, the useful questions aren’t “is this right?” but:
- How many people are trading this? A liquid market with thousands of participants deserves more trust than a thin one with a handful.
- Do participants have real information, or is everyone guessing? A market on a corporate earnings number, traded by people who follow the company, aggregates real insight. A market on a genuinely random event aggregates nothing but noise.
- Is the price near an extreme? At 50%, the market is usually sharp. At 2% or 98%, treat the exact number with a little skepticism — long-shot bias lives at the edges.
- Does anyone have a reason to distort it? A high-profile political market that headlines will quote is a bigger manipulation target than a quiet market on a sports outcome.
Ask those, and a prediction market becomes what it actually is: not an oracle, but the sharpest forecasting tool we have for the specific class of questions it handles well — and a tool you now know how to read critically, which is worth more than trusting it blindly.
The bottom line
Prediction markets can predict because a price aggregates scattered private information, weighted by conviction, into a single number — an idea Hayek described in 1945, the Iowa Electronic Markets proved against real polls in 1988, and Galton’s ox demonstrated at a county fair decades before either. The engine is the independence of the people trading: their errors cancel, and the signal survives.
But the same engine stalls when independence breaks — in thin markets, when traders herd, when someone manipulates, and at the extreme tails where humans misprice rare events. The price is not the truth. It’s the best estimate a particular machine can produce, and now you know both how the machine works and where it fails.
Next, we get concrete about the machinery itself: how contracts are actually bought, sold, and held — order books, liquidity, and the friction between the price you see and the price you get.
Related lessons
Previous in Module 1:
- Lesson 1: What is a prediction market? — What you trade, how the price works as probability, and what happens when an event resolves.
Next in Module 1:
- Lesson 3: How contracts work — The mechanics of buying, selling, and holding contracts. Order books, liquidity, and the friction between the price you see and the price you get.