Election betting odds and how to read them as prices

Election betting odds are somebody's willingness to pay for a political result. A sportsbook posts a number and will take the other side under its margin. A prediction market lets people trade a share that pays if the result happens, and the last price is whatever the buyers and sellers just agreed. Both get called odds. They are produced differently, and the percent you read off them still is not a ballot count.

The inputs are public and private. Polls, fundraising, economic numbers, and scandals move the conversation. So do hedges, partisans who want a side regardless of price, and thin books where a small order shifts the quote. A campaign contract can be a research object and a billboard at the same time. Treat the billboard as a price.

This page will not quote a live candidate. Any cent figure below is a teaching input, labeled as one, so the conversion can be checked without pretending this guide knows tonight's market. If you want the live number, open the venue and read it there. Then run the arithmetic yourself.

Adults 18+ only. Political contracts can go to zero. This is not financial, legal, or tax advice, and it is not a voting guide. The cluster home is Prediction market guides. How the share itself trades is how Polymarket works. If the stake is hurting, use responsible gambling.

On a market that pays $1 for a correct Yes share and $0 otherwise, the price in cents is already the implied percent. Example 1: a Yes share at 40 cents. Divide the price by the payout. 40 / 100 = 0.40, which is 40%. One hundred shares cost 100 × $0.40 = $40. If that side resolves Yes, the payout is $100 and the profit is $60 before fees. If it resolves No, the loss is $40. The 40% described the invoice. It did not survey voters.

The same conversion is the one in how Polymarket works: a 63 cent share is a 63% price. Nothing about elections makes the translation more official. A political market can be thinner than a liquid game, so the 40 cents may be a last trade between a much lower bid and a much higher offer. Write down the price you can actually buy and the price you can actually sell. The midpoint is a summary, not a fill.

Fees move the percent you need. If it costs more than 40 cents all-in to hold a share that pays $1, the implied price is higher than the headline cent. Read the schedule on that venue. This page will not supply a universal political-market fee.

Two contracts on "the election" can be different bets. One pays on the call on election night. One pays on certification. One pays if a party wins the popular vote, another if it wins the electoral vote. Convert the price only after you know which sentence you bought. A 40% price on the wrong sentence is a precise number about the wrong event.

Books often post elections in American odds. Implied probability is the general converter. The two formulas you need are short.

Positive odds +A imply 100 / (A + 100). Example, still teaching and still not a candidate: +150. 100 / (150 + 100) = 100 / 250 = 0.40, or 40%. The same 40% as the 40 cent share, by design, so you can see the formats meet. A $100 stake at +150 returns $250 if it wins, which is the $100 stake plus $150 profit. You are being paid like an event the book treats as 40% for pricing purposes, before you ask whether the book's hold makes the true break-even stricter.

Negative odds −A imply A / (A + 100). Teaching number: −200. 200 / (200 + 100) = 200 / 300 = 0.667, about 66.7%. You risk $200 to win $100. The price talks like a two-thirds event. Favorite prices in elections often look like this, and the hold means a pair of candidates can both be priced so their implied percents sum to more than 100%. That extra is the book's margin, not extra probability hiding in the electorate.

Check the row against the ticket in your hand. Then stop. A converted percent is the start of a comparison with your own estimate, not the end of an argument.

Researchers have studied prediction markets for years, including the Iowa Electronic Markets and the survey by Justin Wolfers and Eric Zitzewitz in the Journal of Economic Perspectives (2004). The careful summary is that markets can aggregate information well in some settings, that prices often move when news is real, and that they are not magic. They have tracked competitive races in ways polls also tracked, and they have been confidently wrong. A market is a running price. Accuracy is something you score afterward, on a set of resolved events, not something a single quote carries on its face.

Beware a precise accuracy percent in a headline. "Markets are 80% accurate" usually hides a choice about which markets, which lead time, and what "accurate" means. A favorite priced at 80% who wins is not the same test as a set of 50% races called within a point. This page will not invent that percent. Read the paper if you want the historical tables, and expect the tables to disagree with a slogan.

Political markets add problems the lab does not. They can be thin. A motivated buyer can push a small contract. Rules about who may trade, and from which country, change the population whose beliefs are in the price. Resolution can lag the public call, so a contract that looks "settled" on television is still open. Favorite-longshot bias, known from other betting markets, can make long shots too expensive and favorites a bit short. None of that yields a correction factor you should subtract from tonight's quote without doing the work.

Use the price as one input next to polls, not as a replacement that ends the need to read. When they diverge, you have a question, not a trade you are obligated to make.

Pick the contract sentence first. Office, date, vote type, and what counts as a winner. Then write the price in one format. Convert it with the table above. Then write a second percent that is yours, from whatever evidence you actually looked at, and label it as yours. The gap between the price and your percent is the only thing a trade could be about. If you do not have a second percent, you are not disagreeing with the market. You are cheering it.

Example 2 stays fictional. Suppose your own estimate for a teaching outcome is 50%, and the share is 40 cents, a 40% price. You are 10 percentage points apart. On 100 shares the cost is $40 and the Yes payout is $100. You are saying the contract is cheap relative to your 50% figure. Expected payout under your estimate is 0.50 × $100 = $50, against a $40 cost, or $10 of expected profit before fees. If your estimate is wrong and the true chance is the market's 40%, expected payout is $40 and the edge is zero before fees. The $10 exists only inside your estimate. Fees can erase it. So can a resolution rule that does not match the outcome you estimated.

That is the whole discipline. Convert. Separate your number from their number. Subtract costs. Respect the sentence being settled. Do not let a televised percent skip those steps. And do not update your estimate by copying the price and then claiming you found a gap. That gap is zero by construction.

Prediction markets and US law is the separate question of whether you may trade a given contract at all. A clean conversion is not permission.

They cannot tell you how to vote. A price is not a moral argument and not a forecast you owe allegiance to. They cannot tell you that a candidate is "safe" because the implied percent is high. High prices lose. They cannot be averaged across venues that settle on different rules and then called a consensus. They cannot be turned into a betting system by this page, because this page has no live prices and no slate.

They also go stale. A quote from this morning is not a quote after a debate, a court ruling, or a dropout. If you are going to use a number, use the number on the screen at the time you act, and keep a copy of the contract text with it.

The table is a brake. Election screens are built to feel like news. The brake is how you keep a news feeling from becoming a stake you did not price.

Election betting odds move when beliefs about an election move. A hashed player-versus-player pot does not. On Jackpot, players fund one pot and a committed seed selects a ticket. There is no candidate, no electoral vote, and no American odds. The result is a draw you can recompute, not a civic outcome.

Fairness runs that recomputation after the seed is published. It will not convert +150 into a percent, and an election market will not publish a Jackpot seed. PVPspinArena does not list political contracts. Jackpot, Coinflip, and Roulette are the games, funded with USDC or ETH on Base. This guide invents no accuracy rate for them and no fee percent.

Checklist when you see a political price:

• Ignore any candidate quote that came from this page, because there is not one.
• Name the exact settlement rule.
• Convert the share or the American odds with the formulas above.
• Write your own percent on a separate line.
• Subtract fees before you call the gap an edge.
• Stop if the stake is there to prove a point rather than to match a price you can lose.

Adults 18+ only. Not financial, legal, or tax advice.