Poll, Bet, or Signal? What Prediction Markets Really Are

A prediction market looks like a poll, trades like a bet, and behaves like a signal. Only one of those is what it actually is.

Published 2026-07-17 · Market & data intelligence · Educational, not advice.

Three words people use interchangeably

When someone points at a prediction market and says the odds of an event are running around a certain level, they usually reach for one of three words to describe what they are looking at. They call it a poll, a bet, or a signal. Those words are not synonyms, and the difference is the whole game.

Get the category right and the number in front of you starts to make sense. Get it wrong and you will either trust it too much or dismiss it too fast.

Why it is not a poll

A poll asks a sample of people what they think and reports the average. Everyone counts equally. The retiree with a strong hunch and the analyst who has read every filing get one vote each. A poll captures opinion, and it captures it at a single moment, usually days before you read about it.

A prediction market does not ask what you think. It asks what you are willing to risk money on. That is a different question, and it filters the crowd in a useful way. People who are confident and informed tend to stake more. People who are guessing tend to stay small or stay out. The result is a weighted average of conviction, not a flat average of opinion.

Polls also stop. A prediction market keeps moving every second the market is open, absorbing news as it lands rather than the last time a pollster called.

The tell

If a number can only change when someone runs a new survey, it is a poll. If it can move on a headline at two in the morning, it is not.

Why it is not just a bet

A bet is a private wager between two sides, usually on an outcome, often for the thrill. It is settled once and forgotten. The number attached to a bet is a price two people agreed on, and it tells you almost nothing about what the wider world believes.

A prediction market is built from bets, but it is not one bet. It is thousands of them, continuously matched, with the current price visible to everyone. That price is public information the moment it forms. One person betting on rain tomorrow is a bet. Ten thousand people trading contracts on rain, with the price settling near a level, is something else.

The mechanics matter here. On venues like Kalshi and Polymarket, contracts typically settle at one dollar if the event happens and zero if it does not. So a contract trading at, say, sixty cents is not a vibe. It is the market's implied probability of about sixty percent, expressed in money. That is why the price reads as a percentage. The dollar is the ruler.

What it actually is: a signal

A signal is information produced by many people acting on their own knowledge, aggregated into one moving number. That is the honest description of a prediction market. It is a live estimate of probability, priced by people who lose money when they are wrong and make money when they are right.

That incentive is the engine. A poll has no cost for a lazy answer. A market punishes lazy answers automatically. If a price is clearly too high, someone can profit by pushing it down, and that act of profit-seeking corrects the number. The signal cleans itself.

It is not magic, and it is not prophecy. A signal can be wrong. Thin markets with little money trading can wobble. Markets can move together in a herd. Surprises still surprise everyone. Treat the price as the best current estimate available, not as a promise about the future.

How to read a signal well

• Read the level and the move. A price sitting at seventy is one thing. A price that jumped from forty to seventy this morning is telling you news just broke.
• Check the depth. A lot of money trading means the signal is sturdier. A quiet market means take it lightly.
• Compare sources. When the market and the polls disagree, that gap is itself information worth sitting with.

Why the distinction earns its keep

Call it a poll and you will treat a live, self-correcting number as a stale snapshot. Call it a bet and you will treat a crowd's best estimate as one person's gamble. Call it a signal and you finally have the right tool in your hand: a probability, priced in real time, that you can weigh against everything else you know.

This is also why watching prices across venues matters. The same question can trade at slightly different levels on different platforms, and the spread between them is its own quiet tell. Delta Arc's Prediction Markets product pulls the live board from Kalshi and Polymarket into one place, so you can read the signal without hopping between tabs.

Next in The Line: how a contract price converts into a probability, and the small trap that makes people misread sixty cents as a coin flip when it is not.