World Cup Odds & Implied Probability: Vig and Fair Value
World cup odds and implied probability, explained for traders: convert American, decimal and prediction-market cents into a true probability, then strip the vig to find fair value.
When a sportsbook lists France at +450 to win the World Cup, it is not telling you France has an 18% chance. It is telling you it will pay 18.2% odds — and then quietly charging you a margin on top so that all the prices add up to more than 100%. Learn to read the difference between those two numbers and you have the single most valuable literacy in trading: turning world cup odds into implied probability, and implied probability into fair value.
This is the from-scratch but rigorous version. By the end you will convert American odds, decimal odds and prediction-market cents into a probability in your head, spot the vig baked into any board, and strip it out to find what the market really thinks. Every number here ties to the actual WC26 outright board so you can practice on live markets.
Three ways to quote the same bet
A price is just a probability wearing different clothes. Three dialects dominate WC26 markets.
- American odds. A positive number (+450) is the profit on a $100 stake. A negative number (-150) is the stake needed to win $100. Sportsbooks in the US default to this.
- Decimal odds. The total return per $1 staked, stake included. +450 American equals 5.50 decimal — bet $1, get $5.50 back. Europe and exchanges default to this.
- Prediction-market cents. On Kalshi and Polymarket a contract pays $1.00 if it resolves YES. If it costs 22¢, that 22 is the implied probability: 22%, by definition.
That last one is the trader's cheat code. On a $1 contract, the price in cents equals the implied probability in percent. No conversion needed. 22¢ means the market prices it at 22%. Everything else is just translating American and decimal back into that same cents-equals-percent frame.
Converting American and decimal odds to implied probability
Here are the formulas, then a calculator so you never do them by hand again.
Decimal to probability: implied % = 100 / decimal. So 5.50 decimal gives 100 / 5.50 = 18.2%.
American to probability:
- Positive odds: implied % = 100 / (odds + 100), times 100. For +450: 100 / 550 = 18.2%.
- Negative odds: implied % = (−odds) / (−odds + 100), times 100. For −150: 150 / 250 = 60%.
Run the WC26 board through it and the favourites line up cleanly:
France +450 to 18.2%, Spain +500 to 16.7%, England +600 to 14.3%, Brazil +700 to 12.5%, Argentina +850 to 10.5%. Notice these are raw implied probabilities — they still contain the vig. Plug your own price into the converter and watch all three formats move together:
Translate any price into every format
Implied probability includes the book's margin — devig a full market to get true fair value.
Try +500 (Spain), +600 (England), or type 22 as a probability and read off the matching odds. Once the conversion is automatic, you stop seeing "+450" and start seeing "18.2% before vig" — which is the number that actually matters.
What the vig is, and why every raw probability is too high
Here is the catch that trips up most casual bettors. Add up the raw implied probabilities of every outcome in a market and they sum to more than 100%. That excess is the vig (also called the juice, the margin, or the overround). It is the book's built-in fee.
Take a clean two-way example. A coin flip should be 50/50, decimal 2.00 each side. But a book might price both sides at 1.91 decimal — 52.4% implied each. Add them: 52.4% + 52.4% = 104.8%. That extra 4.8% is the margin. Whichever side you back, you are paying a price that assumes you win slightly less often than fair.
“Raw implied probability always overstates your real chance — by exactly the slice of margin the book has folded into the price.”
On a big outright market the effect compounds across dozens of teams. Sum the raw implied probabilities of all 48 WC26 sides and you might get 115–125%. That 15–25% overround is spread across every team. So France's "18.2%" is not the market's true estimate — it is 18.2% inflated by France's share of the vig. To find what the market actually believes, you have to take the vig back out.
Removing the vig to find fair value
Fair value is the no-vig probability — the market's honest estimate after you subtract its margin. The cleanest method is proportional (multiplicative) devigging: divide each raw implied probability by the total of all raw implied probabilities (the "book sum").
The formula is simple: fair % = raw % / book sum.
Work a tight three-way example — a group winner market with three live contenders priced at 28¢, 40¢ and 38¢:
- Raw implied: 28% + 40% + 38% = 106% book sum (6% vig).
- Brazil fair = 28 / 1.06 = 26.4%
- Field-A fair = 40 / 1.06 = 37.7%
- Field-B fair = 38 / 1.06 = 35.8%
- Check: 26.4 + 37.7 + 35.8 = 100%. The vig is gone.
That France contract at 18.2% raw, inside a 120% book, devigs to roughly 18.2 / 1.20 = 15.2% fair. The market doesn't think France is an 18% shot — it thinks France is about a 15% shot, and the other 3 points were margin. Strip the vig from any market with the calculator below:
Devig a three-way World Cup group winner market
Multiplicative devig. The fair column is what your model has to beat — not the raw price.
Change the prices to your own market and read the fair column. For the longer treatment — including the power method and when to prefer it over proportional devigging — see how to devig World Cup markets.
Comparing fair value to your own model
Devigging gives you the market's fair value. The edge appears when you have an independent estimate of your own and the two disagree. That is the whole point of the exercise: a devigged price is the bar your opinion has to clear.
Say your model — built from Elo, squad strength, draw difficulty, whatever you trust — rates France at 19% to win it all. The market's devigged fair value is 15.2%. France is underpriced by your read. If you can buy France near its raw 18¢–22¢ and you believe 19% is true, you may have value, because your fair (19%) sits above the price you pay. Flip it: if your model said 12%, France would be a sell/avoid even at the "favourite" price.
Visualize the gap across the board — market fair vs your model:
WC26 outright: market fair vs your model
Bars where model exceeds market are candidate buys; where market exceeds model, the price is rich. This is exactly how a desk builds a value board. Turning that gap into a stake and an expected-value number is the next step, covered in expected-value betting for WC26.
Putting it together: read any WC26 board in four steps
Here is the full workflow, start to finish, on any market you pull up:
- Convert. Turn every price into cents-equals-percent. American and decimal go through the formulas; prediction-market cents are already there.
- Sum the book. Add the raw implied probabilities. The amount over 100% is the vig.
- Devig. Divide each raw probability by the book sum to get fair value.
- Compare. Hold fair value against your own model. Where your number is higher than the price you can pay, you have an edge; where it is lower, pass.
Do this on the outright winner board, on group markets via the groups page, and on the favourites and dark horses before they kick off on June 11. The same four steps work for a +450 sportsbook line, a 22¢ Kalshi contract, and a 21¢ Polymarket share — because underneath the costume, they are all just probability.
“Stop reading odds. Start reading probabilities — then read the price the market charges you on top of them.”
FAQ
Frequently asked
How do I convert World Cup odds to implied probability?
What is the vig in World Cup betting odds?
How do I remove the vig to find fair value?
Why does France at +450 not mean an 18% chance?
Are prediction-market cents the same as implied probability?
How do I know if a World Cup bet has value?
Sources (5)
- Polymarket — 2026 FIFA World Cup Winneraccessed 2026-06-06
- Kalshi — Sports event contractsaccessed 2026-06-06
- Pinnacle — Betting resources: margin and oddsaccessed 2026-06-06
- FIFA — 2026 World Cupaccessed 2026-06-06
- Opta / Stats Perform — football analyticsaccessed 2026-06-06