Skip to content
MarketAtlas
Odds and probabilitiesBeginner

How Prediction Market Odds and Probabilities Work

A prediction market price is a probability you can trade. This guide explains how to turn a price into an implied probability, what the spread, volume and liquidity tell you about how far to trust it, and why even a busy market can be wrong.

By MarketAtlas EditorialPublished 3 min read

Key takeaways

  • A Yes price of $0.65 corresponds to a 65% market-implied probability, before fees.
  • The No price is roughly $1 minus the Yes price, so No at $0.35 implies a 35% chance the event does not happen.
  • The bid is the best price someone will pay; the ask is the best price someone will sell at. The gap is the spread.
  • A wide spread or low liquidity means the headline price is less reliable.
  • Market probabilities are estimates. Research on betting markets shows they can be systematically biased, and any single market can simply be wrong.

Implied probability

In a binary market, a winning contract pays a fixed amount. On Polymarket, for example, each winning share pays $1.00 at resolution, and the platform describes prices as the market’s current probability of the event. That makes the price easy to translate:

  • A Yes contract at $0.65 implies a 65% chance the event happens.
  • A Yes contract at $0.10 implies a 10% chance.

This is the market-implied probability. It is not anyone’s official forecast. It is the price at which buyers and sellers are currently willing to trade.

Yes and No contract prices

If you buy a Yes contract at $0.65 and the event happens, you receive $1, so you gain $0.35 before fees. If it does not happen, you lose the $0.65 you paid.

The No side mirrors this. With Yes at $0.65, No trades near $0.35, which implies a 35% chance the event does not happen. Buying No at $0.35 returns $1 if the event does not happen.

Fees change the maths

Both Kalshi and Polymarket publish fee information, and fees can make a trade that looks fairly priced slightly worse in practice. Kalshi, for instance, says it charges a transaction fee and that some markets carry different fees from others. Check the current fee pages linked in the sources before trading.

Bid and ask spreads

Exchange-style markets show two prices for each side:

  • The bid is the highest price someone is currently willing to pay.
  • The ask is the lowest price someone is currently willing to sell at.

The difference is the spread. If the Yes bid is $0.63 and the ask is $0.67, the spread is 4 cents. You would buy at $0.67 or sell at $0.63, not at a neat $0.65. The midpoint, $0.65, is a reasonable rough estimate of the implied probability.

A narrow spread usually means plenty of active traders. A wide spread is a warning that the displayed probability is soft.

Trading volume

Volume measures how much has traded. More volume generally means more people have tested the price with real money.

Be careful comparing platforms. Kalshi reports volume as a number of contracts, while Polymarket reports it in dollars, so the raw figures are not directly comparable.

Liquidity

Liquidity is how easily you can trade without moving the price. A liquid market has plenty of orders close to the current price. In an illiquid market, a single large order can push the price several points, and you may not be able to exit at a fair price.

Signs of good liquidity:

  • A narrow spread
  • Meaningful size available near the best bid and ask
  • Steady trading activity, not one burst

Why prices change

Prices move when traders’ views change or when the balance of buyers and sellers shifts. Common causes:

  • New information: a poll, a data release, an injury report or a court ruling.
  • Time passing: as the deadline nears with no change, uncertainty shrinks.
  • Large orders: one big trader can move a thin market on their own.
  • Attention: a market in the news can draw in new traders with different views.

Why market-implied probabilities can be wrong

Markets are often useful, but they are not always right.

  • Thin trading. A price set by a handful of trades says little.
  • Biased participants. Traders who share the same view can push prices away from reality.
  • The favourite–long shot bias. Long-standing research on horse-race betting has found that long shots tend to be overbet and favourites underbet, so betting odds overstate the chances of unlikely outcomes (Snowberg and Wolfers, 2010). That evidence comes from betting markets; how strongly the effect shows up on a given prediction market will vary, so treat very low prices with some caution.
  • Rule differences. A price reflects that market’s specific resolution rules, which may not match the question you care about.
  • Fees and capital costs. Fees, and tying money up until a distant resolution date, can keep prices some distance from the probability you would otherwise expect.

Where to go next

Interactive explainer

From contract price to probability

Move the slider to see how an illustrative Yes price maps to a market-implied probability.

Implied probability
65%
Illustrative Yes price
$0.65
Illustrative No price
$0.35

An illustrative Yes contract priced at $0.65 corresponds to a 65% market-implied probability before fees and other considerations. If the event happens, it pays $1.00; if not, it pays nothing.

Educational illustration only. This is not live market data or trading advice. Real prices are affected by trading fees, bid and ask spreads and liquidity, and a market-implied probability is an estimate that can be wrong.

Frequently asked questions

Is a 90% market a sure thing?

No. Even if a market is perfectly calibrated, events priced at 90% should fail to happen about one time in ten. Treat the price as an estimate, not a promise.

Why do the Yes and No prices not add up to exactly $1?

Because traders quote separate buy and sell prices, and the gap between them (the spread) means the best prices on each side rarely sum to exactly $1. Fees can widen the difference further.

Which price should I treat as the probability?

The midpoint between the best bid and the best ask is a common rough estimate. The last traded price can be stale if the market is quiet, so check when it last traded and how wide the spread is before relying on any single figure.

Sources

  1. What is Polymarket (opens in a new tab)

    Polymarket Docs · Documentation · Accessed

  2. Get Markets (API reference) (opens in a new tab)

    Kalshi API documentation · Documentation · Accessed

  3. Fees (opens in a new tab)

    Kalshi Help Center · Company · Accessed

  4. Fees (opens in a new tab)

    Polymarket Docs · Documentation · Accessed

  5. Explaining the Favorite–Long Shot Bias: Is it Risk-Love or Misperceptions? (Journal of Political Economy, 118(4), 723–746) (opens in a new tab)

    University of Chicago Press — Snowberg & Wolfers, 2010 · Primary data · Accessed

Written by

MarketAtlas Editorial

Site byline

Articles under this byline are researched and written by the small team that operates MarketAtlas. Every piece is checked against the primary sources listed at the end of the article and approved by a person before it is published. AI tools may help with drafting, but they never publish anything.

Editorial review
Reviewed by Blake on
Standards
Read our methodology and editorial standards.

AI tools helped prepare an early draft of this article. Every fact was checked against the sources listed above, and a human editor is responsible for the final text.

Platform reviews

Keep learning

  • Prediction market basicsBeginner

    What Are Prediction Markets? A Beginner's Guide

    Prediction markets let people buy and sell contracts tied to whether a future event happens. Because a winning contract pays a fixed amount, its price can be read as the market’s current estimate of how likely the event is. This guide covers how that works, how markets are settled and where the numbers can mislead you.

    4 min read