Election Prediction Markets Explained
A closer look at how political contracts are actually built — nominations, general elections, multi-candidate fields, and why the market's number can diverge from the polling average.
Two different questions, two different markets
"Will this person be the nominee?" and "will this person win the general election?" sound related but are separate contracts with separate timelines. A nomination market resolves earlier and depends heavily on primary-specific dynamics — intraparty support, fundraising, early-state momentum. A general-election market depends on the entire electorate and often keeps trading long after the nomination question has already resolved. It's common for a candidate's nomination odds and general-election odds to move independently, because they're pricing genuinely different risks.
How multi-candidate fields get priced
In a race with several plausible candidates, each one trades as a separate contract rather than everything being squeezed into a single yes/no. A crowded primary field might look something like this at a given moment:
Arbitrage keeps the contracts roughly summing to 100%: if all the individual probabilities added up to noticeably more or less, traders could profit by buying the underpriced side or selling the overpriced one, which pulls the total back toward consistency.
Why swing contracts trade differently
In systems where the overall outcome depends on winning specific sub-contests rather than a simple national popular vote, markets on the closely contested sub-contests tend to be far more active and sensitive to news than markets on the safely decided ones. A single poll or news event in a genuinely contested area can move its contract noticeably, while the same story barely touches a race that isn't competitive.
Why markets and polls can tell different stories
A polling average is a statistical aggregation of stated voter intent, built from a defined methodology applied consistently. A market price is an aggregation of trader judgment, which folds in turnout assumptions, views on momentum, and each trader's own opinion of how reliable current polling is. Neither approach is right by definition — they're answering the question through different mechanisms, and in a fast-moving or closely contested race, it's normal for them to disagree for a stretch before converging or not.
What to keep in mind
- Nomination and general-election markets are not interchangeable — check which question a contract is actually asking before reading its odds as an answer to a different one.
- Probabilities describe uncertainty, not outcomes. A well-priced 75% contract still loses a quarter of the time, by construction.
- Informational only. Nothing here is political or investment advice — it's a description of how these markets are structured and priced.
See how current political contracts are trading on the live dashboard →
Quick answers
What's the difference between a nomination market and a general-election market?
A nomination market asks whether a specific person will become their party's candidate, which resolves earlier and depends on primary dynamics. A general-election market asks who wins the final race, which depends on a completely different set of factors and typically trades separately even when the same person is involved.
How do multi-candidate markets keep their probabilities consistent?
Each candidate's contract trades independently, but arbitrage traders keep the sum of all outcomes close to 100% by buying underpriced candidates and selling overpriced ones whenever the total drifts too far from that mark.
Why might a prediction market and a polling average disagree on the same race?
Polls sample stated intent and aggregate it statistically; markets price in trader judgment about turnout, momentum, and the odds that polls themselves are off. Neither method is infallible, and disagreement between them is common in close or fast-moving races rather than a sign that one side is simply wrong.