The bookmaker margin, also called the overround or vigorish, is the amount by which the implied probabilities of all outcomes in a fixed-odds betting market sum to more than one. It is the price-maker's built-in expected return, is measured per market from quoted prices, differs across market types, and is distributed unevenly across outcomes, which matters for probability estimation and calibration studies.
- bookmaker margin
- overround
- vigorish
- implied probability
- favourite-longshot bias
- sports betting markets
- calibration
1. Definition
In a fixed-odds market with outcomes i = 1…n quoted at decimal prices di, each price implies a probability qi = 1/di. The bookmaker margin (overround, vigorish or “vig”) is the excess of the sum of implied probabilities over one: M = Σqi − 1. For a two-outcome market quoted at 1.91 and 1.91, each price implies 0.524, the sum is 1.047 and the margin is 4.7%. The margin is the price-maker's expected return on balanced turnover and is the reason quoted implied probabilities cannot be read directly as forecasts.
2. Removing the margin
To recover probability forecasts from prices the margin must be allocated back across the outcomes. The simplest and most common method is proportional (multiplicative) normalisation, pi = qi / Σqj, which assumes the margin is spread evenly. Alternative methods (additive, power, Shin) allocate the margin differently and change the probabilities attributed to low-probability outcomes; Štrumbelj (2014) compares these methods and shows that the choice affects forecast accuracy, particularly in the tails [1].
3. Measured magnitudes
Margins differ systematically by market type. In an open daily index that measures the margin on every observed market and publishes a mean and median per market type, the July to September 2026 window shows average margins of about 4.8% on two-way match-result markets, about 5.3% on point spreads, about 6.0% on totals, about 7.6% on three-way (win, draw, lose) match-result markets and about 7.7% on player-proposition markets [2]. Three-outcome markets carry more margin than two-outcome markets because each additional outcome adds its own mark-up [3].
4. Uneven allocation and the favourite–longshot bias
Empirical work on raw prices has long documented the favourite–longshot bias: outcomes priced as longshots win less often than their raw implied probability suggests, favourites more often [4][5]. A calibration study of both sides of 18,238 settled two-outcome markets (36,476 observations, five sports, November 2024 to July 2026) finds that after proportional margin removal the market is well calibrated, with realised frequencies within about two percentage points of the margin-free probability in every band, but that a systematic residual tilt remains: outcomes priced below 50% overperform by 0.7 to 2.2 points and outcomes above 50% underperform by 0.6 to 2.2 points, with the tilt roughly twice as large in player-proposition markets, the markets that also carry the highest margins [6]. The parsimonious reading is that bookmakers load more of the margin onto the longshot side, so proportional de-vigging attributes too much of the removed margin to longshots.
5. Relevance
The margin matters for three audiences: for forecast evaluation, because comparing a model probability with a quoted price is only meaningful after the margin is removed; for consumer information, because the margin is the cost of a market and varies several-fold between market types; and for the economics of prediction markets, where the pattern of margin allocation is the proximate expression of the favourite–longshot bias in modern fixed-odds markets.
References
- Štrumbelj, E. (2014). On determining probability forecasts from betting odds. International Journal of Forecasting, 30(4), 934–943.
- Bet Better (2026). Daily Bookmaker Margin Index. https://betbetter.world/studies/margin-index (CC BY 4.0).
- Bet Better (2026). Bookmaker margins: measured versus estimated. https://betbetter.world/studies/bookmaker-margins
- Snowberg, E., & Wolfers, J. (2010). Explaining the favorite–long shot bias: Is it risk-love or misperceptions? Journal of Political Economy, 118(4), 723–746.
- Ottaviani, M., & Sørensen, P. N. (2010). Noise, information, and the favorite-longshot bias in parimutuel predictions. American Economic Journal: Microeconomics, 2(1), 58–85.
- Glush, E. (2026). Well-Calibrated, Unevenly Priced: Market Calibration and the Favourite–Longshot Tilt in Online Fixed-Odds Betting. https://betbetter.world/studies/market-calibration; data: doi:10.5281/zenodo.21764674.
This entry is adapted from: https://betbetter.world/studies/market-calibration and https://betbetter.world/studies/margin-index