Why prediction market odds can add up to more than 100%

You open a market on an upcoming MMA fight and price buying both sides. The cheapest Oliveira share costs 34¢. The cheapest Holloway share costs 94¢. One of each costs $1.28, for a pair that pays exactly $1.00 because exactly one fighter can win.

Those prices are asks: the lowest price a seller will take on each side. They are what a buyer pays, not consensus probabilities, and nothing forces them to sum to 100%. They are also not the prices Polymarket displays. The displayed price is the midpoint between the best bid and the best ask, or the last traded price when the spread is wider than 10¢.

The excess over $1.00 is called the overround, also vigorish or the book margin. It appears in prediction markets, in sportsbook odds, and in any market where each outcome is priced on its own. Its size tells you what it costs to trade and how much participants disagree.

On Polymarket every outcome is a token with its own order book, its own bids and asks, and its own participants. The lowest ask is the cheapest price at which you can buy that outcome right now. It is what you pay, not what the market page shows.

In a single two-outcome market the two books are linked. A bid for Yes at 60¢ matches a bid for No at 40¢, so every bid on one side works as an offer on the other at $1.00 minus its price, and the two asks always add up to $1.00 plus the spread. Take the fight instead as 2 Yes/No markets, one per fighter, the way Polymarket lists an event with several possible outcomes. Now each fighter's Yes shares trade on a book of their own.

Holloway at 94¢ means the cheapest Holloway Yes share costs 94¢. Oliveira at 34¢ means the cheapest Oliveira Yes share costs 34¢. Those are offers from different sellers on different books.

In a liquid market the two lowest asks stay near $1.00 combined, because arbitrageurs profit from any deviation. In a thin market a handful of participants set each book, may not be watching the other one, and the two asks drift apart.

The bid is the highest price someone will pay. The ask is the lowest price someone will sell at. The gap is the spread. Polymarket displays the midpoint of the two while the spread is 10¢ or less, and the last traded price once the spread is wider. Neither is the price you pay to buy.

Say Holloway's highest bid is 82¢ and his lowest ask is 94¢. The spread is 12¢, enormous by liquid-market standards. Buy now and you pay 94¢; sell now and you get 82¢. The spread is over 10¢, so the market page shows Holloway's last traded price rather than a midpoint. Whatever that number is, a buyer pays 94¢.

The midpoint is a better, still imperfect, estimate. For Holloway it is (82¢ + 94¢) / 2 = 88¢. For Oliveira, with a bid of 4¢ and an ask of 34¢, it is 19¢. The midpoints sum to $1.07: still above 100% and much closer. What remains is the cost of providing liquidity in a thin market.

The overround a buyer faces is half of each spread, added together, plus however far the midpoints sit above $1.00. In the fight that is 6¢ from Holloway's 12¢ spread, 15¢ from Oliveira's 30¢ spread, and 7¢ from midpoints that sum to $1.07: 28¢ in all. With a 3¢ spread on each of 2 outcomes and midpoints that sum to $1.00, the overround is 3¢. A 10¢ spread on each side adds 10¢ before the midpoints move at all. The calculator above shows it for any pair of spreads.

3 things widen a spread. Low volume means fewer participants posting competitive orders. Uncertainty makes market makers quote wider to cover the risk of being wrong. And a market with no professional market makers has nobody posting tight bids and asks on both sides.

The MMA market likely has all 3: low daily volume, few market makers, and a handful of participants each pricing in a wide margin for safety. Nobody thinks both fighters have a strong chance.

A highly liquid market such as a US presidential election might have asks of 52¢ and 49¢: $1.01 to buy both, from spreads of roughly half a cent per side. Displayed prices there are close to true implied probabilities.

A moderately liquid market, such as a popular NBA game, might have asks of 62¢ and 45¢: $1.07, from a 7¢ spread. The true implied probabilities are closer to 58% and 42%.

An illiquid niche market, such as an undercard fight, a local election or a long-tail sports prop, can have asks of 80¢ and 80¢: $1.60. The few participants posting orders demand wide margins and nobody has stepped in to tighten them. The true probabilities might be 55% and 45%.

If you can buy every outcome for less than $1.00 combined, you lock in a profit, because exactly one pays $1.00. At 128% there is no such trade on the buy side: you would pay $1.28 for something worth $1.00.

The reverse trade needs the bids to add up to more than $1.00. A full set, one share of every outcome, costs $1.00 to create on Polymarket, so when the bids sum higher you can create a set and sell each share at its bid for more than it cost. Here the bids are 82¢ and 4¢, 86¢ combined, so that trade fails too. Between 86¢ and $1.28 there is no free money, only a wide market.

Market makers are the natural arbitrageurs. Quoting both sides of each outcome earns them the spread and pulls combined prices toward 100%. In a thin market it is risky: as the only liquidity provider you can lose more on one fill against an informed trader than you earned from weeks of spread capture. Many markets stay wide because fixing them does not pay enough.

When odds sum well above 100%, normalize them: divide each price by the total. In the 128% example Holloway's ask is 94 / 128 = 73.4% and Oliveira's is 34 / 128 = 26.6%. The odds calculator at 0xinsider.com/tools/odds-calculator does this for any set of 2 to 64 outcomes and reports the overround it removed.

Better, use the midpoint of each book's bid and ask instead of the ask alone. That removes the liquidity premium in the best offer: the fight's midpoints of 88¢ and 19¢ normalize to 82.2% and 17.8%. Polymarket's displayed price is already the midpoint while the spread is 10¢ or less. Above that it shows the last trade, so read the bid and ask off the book yourself. When you read a large trade on 0xinsider.com/trades, check the spread in its market first: the same size means less in a thin market with wide spreads than in a deep, tight one.

Before any position in a high-overround market, check the spread on both sides. At 10¢ or more you pay a significant premium to get in. A 5% edge in a market with a 15% overround likely loses money on every trade.

Buy at the ask in a market with a 20% overround and you pay 10% more than the fair midpoint. To break even, the true probability has to be at least the price you paid, 10% above the midpoint. The displayed price hides that cost, which is how beginners lose money in thin markets without seeing where it went.

Prefer liquid markets for this reason. At a 1-2% overround the entry price sits close to fair value and a small informational edge turns into profit. At 15% or more the edge has to be proportionally larger, so stay out unless your view is strong.

The overround is one of five checks on a single price. The guide at 0xinsider.com/learn/how-to-tell-if-a-polymarket-price-is-fair runs through the other four: the price you pay, the depth of the book, how the price moved, and who holds each side.