The Lab · Foundations · updated 2026-07-25
The vig — what betting actually costs
The bookmaker's fee is invisible, automatic, and the reason most bettors lose. Here's how to see it and what it does to a season.
Nobody hands you a receipt at a sportsbook. There's no line item that says "fee." And yet every bet you place is taxed — the cost is hidden inside the price itself, and it's large enough to decide whether a season of betting wins or loses.
Where the fee hides
A perfectly fair coin flip should pay +100 on either side: risk 100, win 100. Sportsbooks post that flip at -110 / -110: risk 110 to win 100, both ways.
The implied probability of -110 is 110 / 210 = 52.4%. Both sides together: 104.8%. Real probabilities add to 100% — the extra 4.8 points is the vig (vigorish, juice). It's the book selling both sides of a coin flip for more than the coin is worth.
What it does to your win rate
At -110, breaking even requires winning 52.4% of your bets. Not 50%. That 2.4-point gap sounds small. It isn't:
| Your win rate | Result at -110, flat 1-unit bets |
|---|---|
| 50.0% | lose ~4.5 units per 100 bets |
| 52.4% | break even |
| 55.0% | win ~5 units per 100 bets |
A bettor who genuinely picks winners at 50% — which is what most people manage against the spread — doesn't tread water. They bleed roughly 4.5% of everything they wager, forever. Over a 500-bet season at $50 a bet, that's about $1,100 gone to the fee alone, without a single unlucky streak. The vig doesn't beat you loudly. It grinds.
The fee varies — and that matters
Moneylines don't hold the tidy -110/-110 shape, but the overround is always there. Sum the implied probabilities of both sides and see how far past 100% it lands: a 103.5% game costs less to play than a 105% game. Two habits follow:
- Count the overround before caring about a game at all.
- Shop the price. The same bet posted at -105 instead of -110 cuts the fee roughly in half. Line shopping across books is the only edge that's guaranteed — it's the same bet at a better price. It's why our board quotes the best line across ten books rather than any single book's number.
Prediction-market contracts (priced in cents) carry the same idea in different clothes: a 56¢ contract on a team the books price at 58% implied is the cheaper venue for the identical opinion.
What this doesn't tell you
Beating the vig is necessary, not sufficient. Paying less fee doesn't make a bad bet good — it makes every bet slightly less bad, and a genuinely good bet slightly better. The order of operations matters: first understand what the market believes (the de-vigged probability), then ask whether any price on offer beats that belief by more than the fee. That second question is the hard one — most days, for most games, the honest answer is no. What the vig math gives you is the discipline to know exactly how high the bar is before you put money over it.
Model output is informational and entertainment content, not betting or financial advice. If you bet, bet what you can afford to lose.