Prediction markets are the most accurate forecasting tool most people have never used. This is everything you need to know to read them, understand them, and use them.
A prediction market is a place where people bet real money on the outcome of a future event. Will the Fed cut rates this quarter? Will this election go to the incumbent? Will this company get acquired?
Every bet creates a price. That price is a probability. If a contract trades at 65 cents, the crowd is saying there's a 65% chance the event happens. Simple.
The key word is real money. When people have skin in the game, they think harder. They don't guess. They research. They update. That's what makes prediction markets different from polls, punditry, or Twitter takes.
The core insight: When someone puts $500 on an outcome, they're telling you something more honest than any survey could. Money talks. Everything else is just talking.
Every market shows a single number: the probability, expressed as a percentage. That's it. No spread, no moneyline to decipher.
67% means the crowd thinks a rate cut is more likely than not, but not certain. If you think the odds are wrong, you can bet against them and potentially profit if you're right.
The change is where it gets interesting. A market that moved 8 points in 24 hours isn't random noise. Something changed. New economic data dropped. A Fed official said something. Someone with real information started buying. Watch the moves, not just the price.
Volume tells you how much the crowd cares. A market with $10K bet on it is different from one with $2.4M. High volume means the price has been pressure-tested by a lot of informed traders. Low volume means a few people set a price that hasn't been challenged.
Most prediction markets use a simple binary contract system. A contract pays out $1 if the event happens, $0 if it doesn't. If the contract trades at 67 cents, you're paying 67 cents for the right to collect $1 if the Fed cuts. Your upside: 33 cents. Your downside: 67 cents.
Markets run continuously, like a stock exchange. Prices move as new information arrives. When the market moves 10 points in an hour, someone just decided the odds shifted. Maybe they know something. Maybe they're wrong. That tension is what makes it interesting.
Resolution is straightforward. The market picks a clear, verifiable outcome in advance. Did the Fed cut or not? Did the election go to candidate A or B? When the answer is known, contracts pay out automatically based on the result.
The profit motive matters: If you think a market is pricing something at 30% when the real probability is 50%, you buy. If you're right and the market eventually reprices, you profit. This incentive is what keeps prediction markets honest. No other forecasting tool has this built in.
Two platforms dominate the prediction market landscape right now. They cover similar topics but have different strengths, different user bases, and occasionally disagree on the same question. That disagreement is a signal worth watching.
The largest prediction market by volume. Built on crypto rails (USDC on Polygon), which means anyone in the world can trade. No US bank account required. Deep liquidity on major political, crypto, and macro markets.
A federally regulated exchange based in the US. Uses USD directly, which makes it familiar for American traders. CFTC-regulated, which matters for institutional participants and adds a layer of legitimacy.
When Polymarket and Kalshi disagree on the same event, pay attention. A 10-point gap between the two platforms on the same question means at least one of them is wrong. That's a potential trade.
| Polymarket | Kalshi | |
|---|---|---|
| Who can trade | Anyone (crypto-based) | US residents |
| Currency | USDC (stablecoin) | USD |
| Regulation | Decentralized | CFTC-regulated |
| Strongest markets | Politics, crypto, global events | Economic data, macro |
| Total volume | Larger (billions traded) | Growing fast |
Polls ask people what they believe. Pundits say what sounds interesting. Prediction markets ask people to put money where their mouth is. That difference changes everything.
Skin in the game filters out noise. People who don't actually know something don't bet their own money on it. The crowd that moves a prediction market is self-selected for information and conviction. The crowd that responds to a poll is not.
Markets update instantly. When new information arrives, prices move within minutes. A poll takes weeks to design, conduct, and publish. By the time you read it, it's already stale.
The track record holds up. Studies consistently show that prediction markets outperform expert forecasters, polls, and models on a wide range of questions. The research goes back decades, from the Iowa Electronic Markets in the 1990s through every major election since.
They're not perfect. Thin markets can be manipulated. Liquidity constraints mean prices can be wrong for longer than you'd expect. And sometimes the crowd is just wrong. But on average, over time, the money is smarter than the commentary.
The Prob's read: Don't use prediction markets to bet. Use them to think. A 78% probability of a Fed cut tells you something about how informed capital is positioned. That's more useful than any analyst's 2,000-word report.
No. You can read the prices without placing a single bet. Most of what The Prob covers is the signal embedded in the prices, not a recommendation to trade. Think of it like checking stock prices without owning any stocks. The information is still useful.
It usually means new information hit. Earnings reports, polling data, an unexpected announcement, a regulatory decision. Sometimes it's a large trader repositioning. The question worth asking is: what do they know that I don't? Big moves on high-volume markets are rarely random.
90% is not 100%. The crowd is saying this outcome is very likely. They're also saying there's a 1-in-10 chance they're wrong. Events that resolve against 90% odds happen all the time. Rare doesn't mean impossible. Always read a probability as a probability, not a certainty.
Yes, though it's expensive and temporary. Thin markets (low volume) are more vulnerable. High-volume markets are harder to move without real capital, and any artificial price tends to be corrected quickly by other traders who spot the arbitrage. Manipulation is a real concern on smaller markets. On the major liquid ones, it's much harder to sustain.
Over a single event, not much. Prediction markets are calibrated over thousands of outcomes. When the market says 65%, it means that historically, events priced at 65% have resolved YES about 65% of the time. The difference between 60% and 65% is real, but small enough that you shouldn't bet your house on it. Treat the number as a range, not a decimal-precision forecast.
It means the crowd is genuinely uncertain, and has been for a while. That's not a failure of prediction markets. That's them being honest. A coin-flip outcome really is a coin flip. Watch for what breaks the stalemate. Often a price stuck at 50% will stay there until a single piece of decisive information sends it sharply in one direction.
Kalshi is CFTC-regulated and fully legal for US residents. Polymarket uses a decentralized structure and USDC, though US residents have historically faced restrictions. The regulatory landscape is evolving. Check each platform's terms of service for the most current guidance.
The Prob pulls live market data directly from the Polymarket and Kalshi APIs every hour. Prices, volumes, and probability changes are sourced directly from the exchanges and updated continuously. What you see on this site reflects real market activity, not editorial estimates.
The biggest moves, sharpest crowd signals, and what it all means. In your inbox by 7am ET. Free.