The short answer
The vig (short for "vigorish," also called the juice, the cut, or the hold) is the sportsbook's built-in margin. It is the reason a book can take balanced action on both sides of a game and still expect to keep a slice of every dollar wagered. You never see it as a line item. Instead, it hides inside the odds themselves: the two sides of a market are priced so their implied probabilities add up to more than 100%. That extra slice above 100% is the vig, and it quietly erodes bankrolls one bet at a time.
How odds turn into implied probability
Every odds price is really a probability in disguise. To see the vig, you first convert American odds into an implied probability:
- Negative odds: implied % = odds / (odds + 100). So -110 → 110 / (110 + 100) = 110 / 210 = 52.38%.
- Positive odds: implied % = 100 / (odds + 100). So +150 → 100 / (150 + 100) = 100 / 250 = 40.00%.
These are "raw" implied probabilities. They tell you what price you are being charged, not the true chance of the outcome.
Why -110/-110 is really about 4.5% vig
Take the classic point-spread market: both sides at -110. Each side implies 52.38%. Add them together:
52.38% + 52.38% = 104.76%
A fair, no-margin market would sum to exactly 100%. This one sums to 104.76%. That extra 4.76% is the overround — the raw vig baked into the line. The number you will more often hear quoted, ~4.5%, is the hold percentage: the overround expressed as a share of the total priced-in probability, or 4.76 / 104.76 = 4.55%. That is the book's theoretical margin if action is perfectly balanced.
To find the fair (no-vig) price, remove the margin by dividing each side by the total: 52.38% / 104.76% = 50.0%. So a -110/-110 market is really a 50/50 coin flip that has been dressed up. A true 50/50 outcome should pay +100 (even money), but you are being asked to lay -110. That gap is the tax.
Why it quietly erodes your bankroll
At -110, you risk $110 to win $100. On a genuine coin flip you win half the time. The expected value of one bet is:
EV = (0.50 × +$100) + (0.50 × -$110) = $50 - $55 = -$5 per $110 staked
That is roughly -4.55% per bet — the exact hold. Put differently, you need to win 52.38% of your -110 bets just to break even, not 50%. Win a "respectable" 50% and you slowly bleed. This is why casual bettors who go roughly even on picks still end the season down: the vig is a headwind on every single wager, and it compounds across hundreds of bets.
De-vigging an uneven market
Most markets are not symmetric. Say a moneyline reads Team A -170, Team B +150:
- A: 170 / 270 = 62.96%
- B: 100 / 250 = 40.00%
- Sum = 102.96% (overround 2.96%, hold ~2.87%)
Remove the vig by normalizing to 100%:
- Fair A = 62.96 / 102.96 = 61.15%
- Fair B = 40.00 / 102.96 = 38.85%
A fair price for Team B at 38.85% is about +157 in American odds. So if this book lists B at +150 but another book across town lists B at +160, that second price is actually better than the fair estimate — a potential positive expected value (+EV) bet. This is the whole game: comparing the price you can get against a vig-free estimate of the true probability.
Line shopping is how you fight back
Because every book sets its own margin and its own numbers, the same game can be -105 at one book and -115 at another. Over a season, consistently taking -105 instead of -115 is the difference between a bettor who survives and one who doesn't — it directly shrinks the hold you pay. No single edge matters more for a beginner than line shopping across many books.
Doing all of this by hand — converting odds, summing implied probabilities, de-vigging, then hunting the best price across dozens of books — is tedious and error-prone. This is exactly what EdgeFinder automates: it pulls live odds from 40+ sportsbooks across MLB, NBA, NFL, NHL, college, and 3-way soccer, de-vigs each market to a fair price, and flags where a real book's number beats that fair estimate. You can see the current free list of today's edges without doing a single calculation. Our self-training model (finals-updated Elo, Pythagorean expectation, starting-pitcher and situational terms, anchored to the market) and its publicly logged pregame track record — graded on win rate, ROI, and closing-line value — are open so you can judge the method rather than take our word for it.
A note on CLV and staking
Closing line value (CLV) measures whether you consistently beat the price the market settles at by kickoff. If you took B at +160 and it closed at +145, you got positive CLV — a strong long-run signal that you are finding real edges, not noise. And when you do have an edge, fractional Kelly staking sizes bets to the size of that edge rather than to a hunch. None of this makes any single bet a sure thing — outcomes are probabilistic and variance is real — but understanding the vig is the foundation everything else is built on.
EdgeFinder is an odds-analysis and education tool, not a sportsbook; it accepts no bets. Betting involves risk, legality varies by jurisdiction, and this is not financial advice. 21+. Please bet responsibly.