Every AU bookmaker price includes a margin. Strip it out and you get the fair odds — the bookmaker's actual probability estimate. Here's how to do it properly.
NFL is the second-most-bet sport on the planet. AU bookmakers price it as an afterthought — which is where the edge lives.
The Super Bowl is a 500-market firehose. About 30 markets are genuinely playable. The rest are an entertainment tax.
De-vigging odds is the technique of removing the bookmaker's margin from a price to reveal the fair probability — what the bookmaker actually thinks the likelihood of the outcome is, stripped of the built-in vig. It's one of the most important calculations in advantage betting, because it converts a raw bookmaker price into something you can meaningfully compare to your own true- probability estimate.
This guide covers the three main de-vig methods (proportional, Shin, power), when to use each, how to do the calculation manually, and how AU advantage bettors use de-vigged prices to find positive expected value bets.
Every decimal price at an AU bookmaker includes a margin (the vig). The implied probability derived directly from the raw price is systematically higher than the bookmaker's real probability estimate, because part of the “probability” in the price is actually margin.
Example: Sportsbet has Collingwood at $1.91 and Essendon at $1.91 in an AFL H2H. Raw implied probabilities are 52.36% each. But these sum to 104.72%, meaning there's 4.72% of margin baked in. The bookmaker's actual probability estimate for both teams is closer to 50% — the raw implied probabilities are each inflated by roughly 2.36 percentage points.
If you compared your true probability estimate of Collingwood at 50% against the raw implied probability of 52.36%, you'd incorrectly conclude that Collingwood is overpriced and Essendon is the value bet. That's wrong — both bets are roughly fair against your estimate. De-vigging removes this distortion.
The simplest and most widely used method. It assumes the vig is distributed proportionally across outcomes.
Formula:
Fair probability = implied probability / sum of all implied probabilities
AFL H2H at Sportsbet: Collingwood $1.91 / Essendon $1.91.
A-League match: Home $2.30 / Draw $3.40 / Away $3.20.
The Shin method, developed by Hyun Song Shin in 1993, addresses a known issue with proportional de-vigging: the vig isn't actually distributed proportionally. Bookmakers tend to over-shorten favourites and under-shorten underdogs because they're accounting for informed bettors concentrated on the favourite side.
The Shin method is most useful on two-way markets with a heavy favourite (implied probability above roughly 65%). For balanced markets, the Shin method and proportional method give nearly identical results.
The mathematical derivation is complex (involves solving a quadratic), but the calculator above handles it for two-way markets. For hand calculation in a two-way market with raw probabilities p₁ and p₂ summing to B:
z = (p₁² + p₂² − 2·p₁·p₂·B) / B² ... plus additional algebra
If that looks painful, use the calculator. The practical effect of the Shin method vs proportional is to slightly increase the de-vigged probability of underdogs and slightly decrease favourites — typically by 1-3 percentage points on heavily lopsided markets.
The power method works by finding an exponent k such that raising every implied probability to the power k produces a sum of exactly 100%. It's particularly useful on three-way and multi-way markets where proportional de-vigging can be systematically biased.
Formula: Find k such that Σ pᵢ^k = 1. Then each fair probability is pᵢ^k (normalized to sum to 1 after the small numerical error).
Hand calculation requires iterative search. The calculator above uses binary search to find k for any number of outcomes. For two-way markets the power method and proportional method are functionally identical.
Practical recommendations for AU bettors:
For the vast majority of bets the practical differences between methods are small enough that proportional de-vigging is “good enough.” If you're building a sophisticated model, use Shin for two-way favourites and power for everything else.
The main use cases:
Calculating EV. Your expected value calculation needs a true probability estimate. The de-vigged probability from the sharpest bookmaker (typically Bet365 in AU) is a reasonable baseline estimate. Compare your own true probability estimate (or another bookmaker's de-vigged estimate) to determine whether a given price is +EV.
Building market consensus probability. The Krok Odds +EV Finder de-vigs every AU bookmaker's price on a market, then takes the median across all de-vigged probabilities to produce a market-consensus true probability estimate. Market consensus across 12+ AU books is more robust than any single bookmaker's estimate.
Finding outlier prices. If most AU bookmakers have a market at roughly 52% implied probability (after de-vigging) and one bookmaker has it at 48%, the 48% bookmaker is offering value. Their de-vigged price is below market consensus, meaning their raw price is above consensus — that's a +EV bet against the market.
Verifying arbitrage. Calculating the book percentage of the best-available prices across AU bookmakers directly tells you if there's an arbitrage. Book percentage below 100% is an arb. De-vigging isn't technically required for arb detection, but the framework carries over.
A few honest caveats:
De-vigged probabilities aren't the true probabilities. They're the bookmaker's estimate of the true probabilities. If the bookmaker's estimate is systematically wrong (which happens on soft markets like player props), de-vigging doesn't magically fix it. You still need accurate probability information from somewhere.
Single-bookmaker de-vigging is less reliable than multi-bookmaker de-vigging. Any individual bookmaker's de-vigged estimate can be biased. The median of de-vigged estimates across 140+ bookmakers is substantially more reliable than any single book's number.
Vig distribution assumptions matter on lopsided markets. The proportional method assumes vig is distributed proportionally. For two-way markets with a heavy favourite this assumption breaks down, and the Shin method is more accurate. Use the right tool for the market.
“No-vig” describes a price or probability with the bookmaker's margin stripped out. A no-vig price is the fair odds the bookmaker thinks the market is worth before adding margin. No-vig probabilities sum to exactly 100% across all outcomes.
Same thing. “Fair odds” and “no-vig odds” both describe prices with the bookmaker's margin removed. “True odds” is sometimes used the same way, though it technically refers to odds reflecting actual underlying probability rather than the bookmaker's estimate of it.
De-vigging accurately removes bookmaker margin from a given set of prices. Whether the resulting fair probabilities accurately reflect reality depends on the bookmaker's pricing accuracy. On main markets at sharp AU bookmakers (Bet365, Sportsbet), de-vigged probabilities are generally within 1-2 percentage points of actual outcome frequencies over large samples. On soft markets they can be much less accurate.
Yes, using the same methods as main markets. Over/under totals at a bookmaker have two sides (over and under); de-vig them like any two-way market. Player props with yes/no structure de-vig identically.
Betfair Exchange prices have minimal vig built into the odds because the exchange matches backers with layers at market prices. The 6.5% commission isn't part of the odds — it's charged on net winnings separately. So raw Betfair back prices don't need traditional de-vigging, but you should account for the commission when comparing Betfair prices to corporate bookmaker prices.
Daniel writes about the maths underneath advantage betting — expected value, Kelly sizing, closing line value, bankroll theory. Translates the theoretical side into practical decisions AU punters can actually apply.
The Melbourne Cup is the only day of the year when half of Australia bets. The other 364 days are when sharp punters earn from the half that does not.
