Prediction markets vs sportsbooks

This is the structural difference and everything else follows from it. A sportsbook takes the opposite side of your position itself. It profits when its customers collectively lose, manages its exposure by moving prices, and can decline, limit or close your action if the arrangement stops working in its favour.

A prediction market matches you against another participant. Somebody had to be willing to sell you the contract at the price you paid, which means the price is the point where two opposing views met rather than a figure a firm decided to offer. The venue takes a fee for running the matching and holds no position in the outcome.

That distinction is why an exchange has no reason to restrict a consistently profitable participant, and why a book often does. It is also why an exchange can show you a price nobody is prepared to trade at — the wide-spread problem — while a book always has a price, because it is the one making it.

Both venues charge for the service, and in both cases the charge is visible in the fact that the two sides of a market add to more than certainty. A book quoting decimal odds of 1.80 and 2.10 on the two sides is implying 1 ÷ 1.80 = 55.6% and 1 ÷ 2.10 = 47.6%. Those sum to 103.2%, and the 3.2 points of excess is the book’s margin.

The arithmetic to recover the probabilities is the same one an exchange requires: divide each by the sum. 55.6 ÷ 103.2 = 53.8% and 47.6 ÷ 103.2 = 46.2%, which now add to 100%. An exchange market quoting YES 64¢ and NO 40¢ needs 64 ÷ 104 = 61.5% by exactly the same logic. The formula does not care which kind of venue produced the numbers.

What differs is who receives the excess and what determines its size. A book sets its margin as a commercial decision and keeps it. On an exchange the spread is captured by whichever participant was standing at the front of the queue, and its width is set by competition between them — so it can be far narrower on a heavily traded market, and far wider on a neglected one. Neither arrangement is automatically cheaper. A liquid exchange market is usually cheaper than a book on the same event; an illiquid one frequently is not.

A book’s price is not purely a forecast. It is a forecast adjusted for the book’s exposure and for what its customers are doing: if money piles onto one side, the price moves to attract the other side, whether or not any new information arrived. Reading it as a clean probability estimate means reading a number that has commercial risk management mixed into it.

An exchange price is closer to a forecast because there is no book to balance, but it is not pure either. It reflects who happens to be trading, how much capital they have committed, and what it costs them to hold the position until settlement. On a thin market it can be one participant’s opinion. It is a better probability estimate than a book’s price on average, not by construction.

The honest summary is that the exchange price is a measurable estimate with a measurable distortion, and the book’s price is an estimate with a distortion you cannot see the size of. That is a real advantage, and it is a smaller one than the framing usually suggests.

A prediction market position trades until it settles, so you can sell it at whatever it is worth at any point. A view that turns out to be right early can be realised early; a view that turns out wrong can be closed at a partial loss rather than run to zero. The price you get is set by the book, not by the venue’s goodwill.

A fixed-odds bet is generally held to the event. Where a cash-out is offered it is offered at the book’s price, which includes the book’s margin a second time — so a settled-looking position often cashes out for meaningfully less than its true worth.

You can also be the one quoting. Posting a limit order on an exchange means other participants trade against your price, and if it is filled you have captured part of the spread rather than paid it. There is no equivalent at a sportsbook, where the price is the house’s to make.

The risk of loss is identical in kind. You can lose everything you committed on either, and being correct about the world is not sufficient — paying too much for a correct view still loses money. Neither instrument is a savings product and neither becomes safe by being regulated.

The legal treatment, on the other hand, is not the same and is not stable. Event contracts are regulated as financial instruments in some jurisdictions and as wagering in others, and the classification is being actively argued about in several. Nothing here is legal advice; check your own position before assuming which set of rules applies to you.