How to Remove Vig from Betting Odds: WC26 True Probability

When France is priced at +450 to win WC26, the naive read is "18.2% chance." It is not. That number is inflated by the vig — the sportsbook's margin — baked into every price on the board. Learning how to remove vig from betting odds is the single highest-leverage skill in prediction-market trading, because the de-vigged number is your prior: the clean probability you carry over to Kalshi and Polymarket to decide whether their contract is cheap or rich.

With WC26 kicking off June 11, the outright board is set: France +450, Spain +500, England +600, Brazil +700, Argentina +850. Add up the raw implied probabilities of all 48 teams and you will not get 100% — you will get something like 125–135%. That excess is the vig, and how you strip it changes your fair number by enough to flip a trade from +EV to −EV.

What the vig actually is

Convert American odds to implied probability and the formula for a positive line is 100 / (odds + 100). So +450 implies 100 / 550 = 18.2%, and +500 implies 16.7%. Do this for every team and sum the column.

In a fair, marginless market that sum is exactly 100%. In a real sportsbook it is more — say 130%. That 30 percentage points of overround is the vig (also called juice or the overround). It is how the book guarantees profit: it sells you $1.30 of "probability" for every $1.00 that can actually happen.

The implied probabilities you read straight off the screen are therefore systematically too high. Every one of them. To recover an honest estimate you must scale the whole column back down to 100% — and how you scale is where the methods diverge.

Method 1: Multiplicative devig (the default)

The simplest and most common approach: divide each raw implied probability by the book sum (the overround). If France is 18.2% raw and the book totals 130%, the fair number is 18.2 / 1.30 = 14.0%.

This is multiplicative (or proportional) devigging. It assumes the vig is spread proportionally — every outcome's true probability is its raw probability times the same constant factor (here, 1/1.30). It is fast, it always returns a valid distribution summing to 100%, and for two-way and balanced markets it is perfectly serviceable.

Plug the WC26 outright board into the calculator below. Enter each team's price in cents (the implied %), watch the overround appear, and read the de-vigged fair column.

The catch: multiplicative devig assumes the book applies its margin evenly across favourites and longshots. It does not. And that assumption is exactly where naive traders bleed.

Method 2: Additive devig

Additive (or "Shin-lite") devigging subtracts an equal slice of the overround from each outcome rather than scaling proportionally. If the book is 30 points over a 6-way market, you subtract 5 points from each: France 18% → 13%, a longshot 2% → could even go negative, which is its fatal flaw.

Because subtracting a flat amount can push small probabilities below zero, pure additive devig is rarely used for full outright boards. It can be reasonable for balanced two- or three-way markets — a single match's home/draw/away, where the three numbers are close — but it falls apart the moment you have a 40% favourite and a 1% longshot in the same market. For World Cup outrights, skip it.

Method 3: Power devig (the sharp's choice)

The most accurate method for lopsided markets like a 48-team outright is the power method. Instead of dividing every raw probability by a constant, you raise each to a power k and choose k so the column sums to exactly 100%.

The trick: with k slightly above 1, favourites shrink less and longshots shrink more than under multiplicative devig. That is the correct direction, because of a real phenomenon called the favourite-longshot bias.

The favourite-longshot bias that ruins naive devigging

Here is the empirical truth that decades of market data confirm: bettors systematically overbet longshots and underbet favourites. People love a +5000 ticket; they find a −200 favourite boring. Books know this and price accordingly, padding the longshots more heavily than the favourites.

The consequence for devigging is brutal. If you strip vig multiplicatively — applying the same shrink to everyone — you under-correct the longshots and over-correct the favourites. Your fair number for USA at +4000 comes out too high (because the book already padded it and you did not remove the extra pad), and your fair number for France comes out a touch too low.

This matters because longshots are exactly where Kalshi and Polymarket prices diverge most. If your naive multiplicative prior says USA-to-win is 2.4% but a power devig says 1.9%, and Polymarket is offering the contract at 2¢, multiplicative says "fair," power says "rich — fade it." Same data, opposite trade.

The bars show it cleanly: the two methods nearly agree in the middle of the board and split hardest at the extremes — favourites nudged up, longshots cut down.

A worked example: turning +450 into a clean prior

Walk it through end to end.

1. Convert. France +450 → 100 / 550 = 18.2% raw. (Use the converter below to do this for any team in any format.)
2. Sum the board. All 48 teams' raw implied probabilities total ~130% — that is your overround.
3. Multiplicative pass. 18.2 / 1.30 = 14.0% fair.
4. Power adjustment. Because France is a heavy favourite and the bias under-shrinks it, the power method nudges it back up to roughly 14.6%.
5. Compare. If Polymarket's France contract trades at 18¢ (18%), your power-devigged prior of ~14.6% says the contract is rich — you would lean toward NO, not YES.

That single half-point difference between 14.0% and 14.6% is not academic. Multiply it across a season of bets and it is the gap between a profitable book and a losing one — and on a single +450 favourite it can move your edge calculation by enough to change the trade.

How to actually use your de-vigged number

A clean prior is only useful if you deploy it with discipline.

• Devig the sharpest book, not the softest. Pinnacle and other low-margin, high-limit books carry the most information. De-vig their line and treat it as the market consensus, then go hunting for venues that disagree.
• Use power (or at least multiplicative) for outrights; reserve additive for tight two-ways. Match a method to the shape of the market.
• Carry the prior to Kalshi and Polymarket. Your fair % is the input to every edge and EV check. If the contract trades below your fair number you have value; above it, you are the sucker. The honest prior is also the foundation of any cross-venue arbitrage between Kalshi, Polymarket and sportsbooks.
• Re-devig when lines move. The overround drifts as money comes in. A prior built off Monday's board is stale by Friday — recompute before WC26's group stage scrambles every number.

Master this and the entire outright board stops being a list of odds and becomes a list of testable claims. France at 14.6%, Spain at 15.1%, USA at 1.9% — now you can compare each to Kalshi and Polymarket and find the handful that are genuinely mispriced before kickoff on June 11.