Skip to main content
Learn/Markets and odds/How prices form on a prediction market
STRATEGYMarkets and odds

How pricing works on a prediction market

The order book at the conceptual level. Bid and ask. How a price emerges from matched orders. Liquidity, depth, and the implications for execution. The mechanics of price formation on an exchange-style operator.

On a sportsbook, a market maker quotes a price and the bettor accepts or declines it. On a prediction market, no one quotes a price for you. The price is what two participants have agreed to transact at, recorded continuously in an order book that any participant can read. Understanding how prices form on an exchange is the foundational mechanic for everything else in this article and the rest of the module.

The order book at the conceptual level

An order book is a live record of every standing offer to buy and every standing offer to sell a contract. Each row in the book has three pieces of information: the side (bid or ask), the price, and the size. A bid is an offer to buy at a specified price. An ask is an offer to sell at a specified price. The size is how many contracts the participant is willing to transact at that price.

ORDER BOOK SNAPSHOT (conceptual)
  Contract: 'TEAM A wins championship. YES.'

  ASK (sellers)              BID (buyers)
    $0.65 × 200 contracts      $0.62 × 150 contracts
    $0.66 × 500 contracts      $0.61 × 400 contracts
    $0.67 × 1,200 contracts    $0.60 × 800 contracts

  Best ask:  $0.65   (the cheapest seller)
  Best bid:  $0.62   (the most expensive buyer)
  Spread:    $0.03   (3 cents wide on a $1 contract)
  Mid:       $0.635  (midpoint of best bid and best ask)

The best ask is the lowest price a seller is currently willing to accept. The best bid is the highest price a buyer is currently willing to pay. The gap between them is the spread. The midpoint is the natural reference price, the closest thing the book has to a single number that represents the contract.

How a trade clears

A trade clears when an incoming order matches an existing order. Two ways this happens.

  1. A market order. The participant sends an order with no price specified, just a size. The platform matches it against the best available price on the other side. If the participant wants to buy 100 contracts and the best ask is $0.65 with 200 contracts available, the trade clears at $0.65 and the participant pays $65 for 100 contracts. The platform takes its transaction fee on top.
  2. A limit order. The participant sends an order with a price and a size. If the limit price meets or beats the best price on the other side, the trade clears at the matching price. If the limit price does not meet the other side, the order rests in the book as a new bid or ask, waiting for a counterparty to match it.

Price as probability on a binary contract

On a binary contract that pays one dollar if the event resolves YES and zero if the event resolves NO, the price is the implied probability of YES. Sixty-two cents in the order book means the market is currently saying the event has a sixty-two percent chance of resolving YES.

The article on reading prediction market prices as probability estimates covers this in depth, including the conversion math, the cross-reference with sportsbook lines, and the cases where the implied probability is meaningfully different from the academic literature's notion of fair value. The takeaway here is that on a prediction market, price and probability are the same number, expressed differently.

Expected value on a contract that pays at $1

The expected value of buying a binary contract at price p, where the participant believes the true probability of YES is q, is straightforward.

EV OF BUYING AT PRICE p, BELIEF q
  payoff_YES = $1.00
  payoff_NO  = $0.00
  cost       = p (per contract)

  EV = q × ($1.00 - p) + (1 - q) × ($0.00 - p)
  EV = q - p

  Example: bought at p = $0.62, belief q = 0.70
  EV = 0.70 - 0.62 = +$0.08 per contract
  Expected return per dollar of cost:
    +$0.08 / $0.62 ≈ +12.9%

The math is identical for selling. Selling at price p, belief that true probability of YES is q, expected value is (p − q). The transaction fee comes off the EV; on a deep contract the fee is small enough that it does not flip the sign on a positive EV trade, but on a thin contract with a wide spread, the effective entry price is closer to the ask than to the midpoint and the EV needs to clear that gap to be worth taking.

Liquidity and what it means

Liquidity is the depth of standing orders in the book on each side. A liquid book has many bids and asks at prices close to the midpoint, with substantial size at each level. An illiquid book has thin standing orders, often with wide gaps between price levels.

Liquidity matters because it determines how much a participant can transact without moving the price. On a deep book, an order for 1,000 contracts can clear at or close to the midpoint. On a thin book, the same order can sweep through several levels and clear at a materially worse average price than the participant expected.

Two contracts, same midpoint, very different liquidity.
AspectLiquid contractThin contract
Best ask × size$0.65 × 1,200$0.65 × 50
Best bid × size$0.64 × 1,500$0.62 × 30
Spread$0.01 (1 cent)$0.03 (3 cents)
Order for 500 contracts (buy)Clears at ~$0.65 averageSweeps to ~$0.69 average
Effective execution costFee + 0.5 cents slippageFee + 4 cents slippage

Time-to-resolution and how the book deepens

Liquidity is generally a function of how much attention the contract has and how close it is to resolution. Contracts close to resolution tend to have deeper books because participants are actively trading on the latest information. Contracts weeks or months from resolution often have thinner books because the marginal information moves the price less frequently.

A bettor who plans to take a position weeks in advance and exit close to resolution gets the benefit of the book deepening over time. A bettor who plans to take a position close to resolution and exit immediately can usually count on the book being its deepest at that moment. A bettor who plans to take a position weeks in advance and exit weeks before resolution may find that the exit liquidity is worse than the entry liquidity, especially on niche contracts.

The transaction fee

Each prediction market platform charges a transaction fee. The exact fee structure varies by operator and is published on the operator's site. Common structures include a flat percentage on each trade, a tiered structure that scales with volume, and a structure that charges fees only on the winning side of the trade. The participant always knows the fee at the time of the trade because it is explicit, not built into the line.

The fee plus the bid-ask spread is the participant's effective transaction cost. On a deeply liquid contract this is well under one percent of position size. On a thin contract it can exceed the hold a sportsbook charges on the same view. The participant who treats the fee as the whole cost is missing the spread; the participant who treats the spread as the whole cost is missing the fee. Both matter.

Why prediction market prices are often a cleaner reflection of probability

On a deep, liquid contract with active two-sided trading, the price represents the consensus of two participants who have both agreed to transact at it. There is no market maker shading the price away from fair value to extract additional hold. There is no managed-position dynamic where the operator wants to attract specific flow. The price is the price two participants are willing to clear at, full stop.

That structural cleanness is part of why the academic and research communities use prediction market prices as inputs into probability estimates. On the contracts where the cleanness holds (deep books, active trading, no manipulation campaign), the price is a credible probability estimate. On the contracts where the cleanness does not hold (thin books, low attention, attempted manipulation), the price is an unreliable estimate and other inputs are needed.

What this article does not cover

DFS pick'em prices and sweepstakes payouts do not work this way. Pick'em prices are fixed odds set by the operator from a published payout schedule. Sweepstakes mechanics use a dual-currency system that does not have an order book in the exchange sense. Both have their own dedicated articles in this module.