The Vig Math: Why -110 Is a 52.38% Break-Even, and What to Do About the Number
Every bettor hears the number: 52.38%. Win more than that percentage of your -110 bets and you profit. Win less and you lose.
Most explanations stop there. They give you the break-even rate and move on. They skip the overround, the number the book actually builds into the market. They skip the no-vig conversion. They skip what the vig costs in dollars across a full season. And they skip the margin gap between sharp books and US recreational sportsbooks, which determines whether a 54% win rate ends in profit or loss.
This article shows the full machine. The math behind -110, the 104.76% overround, the no-vig removal formula, and three concrete steps to reduce what you hand the book on every bet.
How -110 generates 52.38%
When you bet -110, you risk $110 to win $100. A winning bet returns $210 total: your $110 stake back plus $100 profit.
The break-even calculation is a fraction of the amount you risk divided by the total returned:
break_even = amount_risked / total_returned
# At -110:
break_even = 110 / 210 = 52.381%
At a true 50/50 proposition, the fair price is +100 on both sides. Win a +100 bet and you get back $200 on a $100 stake. Break-even: 100/200 = 50.00%.
The gap between 50.00% and 52.38% is the cost of the vig. You need to win 2.38 extra bets out of every 100 to overcome the margin before a single pick is evaluated on its merit.
The overround: the number inside the price
A standard NFL spread or game total prices both sides at -110. Each side implies a 52.38% probability. Add them:
# Overround = total implied probability - 100%
overround = 104.76% - 100% = 4.76%
A fair market sums to 100%. The 4.76% above 100% is the margin the book builds in before the event starts. The sportsbook prices $104.76 worth of probability into a $100 event. No matter which side wins, the book retains the overrun.
Different lines carry different vig. The overround scales with the price:
| Line (both sides) | Implied prob. each | Total (overround) | Vig % | Break-even |
|---|---|---|---|---|
| -105 / -105 | 51.22% | 102.44% | 2.44% | 51.22% |
| -108 / -108 | 51.92% | 103.85% | 3.85% | 51.92% |
| -110 / -110 | 52.38% | 104.76% | 4.76% | 52.38% |
| -115 / -115 | 53.49% | 106.98% | 6.98% | 53.49% |
| -120 / -120 | 54.55% | 109.09% | 9.09% | 54.55% |
Every row is a symmetric market where both sides carry equal vig. Player props, alternate lines, and futures often run asymmetrically, with the book charging more on one side or burying extra margin in a lopsided favorite/underdog pair.
No-vig conversion: finding the fair price
The no-vig conversion strips the book's margin out and returns the true implied probability each side carries. It is the market's honest opinion on the game, with no house cut.
fair_prob = implied_prob / total_implied_prob
# At -110 / -110:
fair_A = 52.38% / 104.76% = 50.00%
fair_B = 52.38% / 104.76% = 50.00%
In a symmetric -110/-110 market, the no-vig probability is 50.00% each. A true coin flip, priced with juice. The book earns nothing on the game outcome, only on the spread between what bettors pay and what a fair market would charge.
Now run the same conversion on an asymmetric line. Take -145 / +125:
implied_A = 145 / (145 + 100) = 59.18%
implied_B = 100 / (125 + 100) = 44.44%
total = 59.18% + 44.44% = 103.62%
# No-vig
fair_A = 59.18 / 103.62 = 57.11% # fair odds: -133
fair_B = 44.44 / 103.62 = 42.89% # fair odds: +133
The book offers -145 on Side A. The fair price is -133. That 12-cent gap is the price you pay for access. If your own model says Side A wins 60% of the time, you have a 2.89-point edge over the no-vig probability, worth betting. If your model says 57%, you have no edge after the vig.
Comparing your estimate against the raw implied probability (-145 implies 59.18%) overstates your edge by 2.07 points. That error compounds across a full season of bets.
What the vig costs per season
The overround is an abstraction until you put dollar figures on it. Assume you place 500 bets per year at $110 per game on true 50/50 markets (NFL spreads and totals, NBA spreads).
On a $110 bet at -110 in a true 50/50 event, your expected value is:
Over 500 bets, the expected vig drain is 500 × $5.00 = $2,500. That is the floor of what you lose before a single pick is graded on its merit. If your picks produce no edge, $2,500 is gone. If your picks produce a small positive edge that doesn't clear 52.38%, you lose more slowly but still lose.
The same math at different lines:
| Line | EV per bet | Annual drain (500 bets at that stake) | vs. -110 |
|---|---|---|---|
| -105 | -$2.50 | -$1,250 | saves $1,250 |
| -108 | -$3.81 | -$1,905 | saves $595 |
| -110 | -$5.00 | -$2,500 | baseline |
| -115 | -$7.50 | -$3,750 | costs $1,250 more |
| -120 | -$10.00 | -$5,000 | costs $2,500 more |
These are pure vig costs. The EV per bet and annual drain figures assume a 50/50 underlying event and straight wagers only. Moving from -110 to -105 on the same game saves $1,250 per year across 500 bets. The savings come from the line alone, with no change in your pick quality.
The national hold: the industry-level number
The American Gaming Association tracks what US sportsbooks retain on total handle. That figure is the national hold rate.
In 2024, regulated US sportsbooks processed approximately $148 billion in handle and generated $13.71 billion in gross gaming revenue, per reporting by ESPN from AGA data. National hold: 9.3%.
In 2025, the industry processed $166.94 billion in handle and generated $16.96 billion in revenue. Hold rate: approximately 10.15%.
The theoretical hold on a straight -110/-110 bet is 4.76%. The national hold running near 10% means the product mix is pulling the average far above straight bets. Same-game parlays run 20 to 35% hold for the book, depending on the number of legs. Every additional leg layers another overround on top of the last.
A bettor who only makes straight spread bets is losing at a 4-5% hold rate. A bettor who mixes in four-leg same-game parlays regularly is losing at rates closer to 25%. The book's blended hold climbs because the mix of products shifts toward higher-margin offerings every year.
The margin gap between sharp and recreational books
Not every sportsbook charges the same vig. The gap between sharp books and US recreational books is wide and has a real impact on long-term results.
Pinnacle, the reference sportsbook used by professional bettors for line-shopping and edge measurement, runs margins of 2 to 3% on major sports markets. On NFL and NBA spreads, lines often sit at -107/-107 or -108/-108 rather than -110/-110. The overround on a -107/-107 market is 3.38%. On a -108/-108 market, it's 3.85%.
US recreational sportsbooks like DraftKings, FanDuel, and BetMGM standard at -110 on most spread and total markets, and push toward -115 or worse on player props, alternate lines, and boosted markets.
Pinnacle's own research quantified the margin impact on 1,000 flat-stake bets at a 54% win rate:
- At 2.5% margin (Pinnacle-range): ROI of +5.3%
- At 8% margin (recreational book range): ROI of -0.1%
The picks are identical. The win rate is identical. The only variable is where the bets are placed. A bettor winning 54% of his wagers, a rate that qualifies as sharp by most definitions, loses money at a high-margin book and profits at a sharp book. That gap is the full cost of the vig over a long sample.
What the academic research shows
The overround formula is the standard tool for estimating bettor loss rates, but it understates the actual drain in many markets.
Tadgh Hegarty and Karl Whelan at University College Dublin published "Calculating the Bookmaker's Margin: Why Bets Lose More on Average Than You Are Warned" (MPRA Paper No. 116924, 2023, later published in Applied Economics, 2025). Their finding: the overround accurately predicts loss rates only when a bookmaker applies the same percentage margin to every outcome on an event. When books charge a higher margin on longshots than on favorites, which the data shows is common, actual bettor losses run higher than the overround formula implies.
In soccer betting markets, they found actual average loss rates were one-fifth higher than the overround calculation predicted. In other words: a 5% overround in a soccer market produces expected losses closer to 6% of handle, not 5%.
For NFL and NBA spread bettors, symmetric -110/-110 pricing means the standard formula is accurate. For bettors mixing in underdogs, futures, and player props, actual losses exceed the headline margin figure.
The same research group, in "Market Structure and Prices in Online Betting Markets: Theory and Evidence," published in Oxford Economic Papers in 2026, documented why the margin gap between Pinnacle and retail books exists. Sharp books operate on thin margins and high volume, accepting all comers. Retail books widen margins and limit or ban customers who show positive expected value. The two business models produce structurally different prices for the same underlying event.
Three things to do with this information
1. Strip the vig before evaluating a bet
Before you decide whether a line has value, run the no-vig conversion. Convert both sides to implied probability, add them to get the total, then divide each side's implied probability by the total. The result is the market's fair estimate, with no house thumb on the scale.
Your model's probability estimate needs to beat the no-vig probability, not the raw implied probability. Comparing against the raw number gives you a false read on your edge, especially on lopsided lines where the book distributes the margin asymmetrically.
2. Shop lines on every bet
A -108 line instead of -110 on a $110 bet saves $0.95 in expected value per wager. Over 500 bets, that's $475. Over 1,000 bets, $950. The savings require no improvement in pick quality. Place the bet at the better number.
The floor for line shopping is checking at least two books before placing any bet. Open accounts at multiple books. The line at DraftKings will not always match the line at FanDuel. A half-point or a couple of cents difference in price is real money over volume.
3. Track your win rate against break-even, not your win/loss record alone
A 55% win rate is not automatically profitable. The profitability depends on the average line you bet.
If you bet -110 lines and win 55%, you beat the 52.38% break-even by 2.62 points, a positive edge worth building on. If you bet a mix of -110 and -140 lines and win 55%, your break-even for the mix is higher than 52.38%, and 55% overall may or may not clear it.
Calculate your weighted average break-even across your bet history. Compare your actual win rate against that number. The gap between the two is your real edge, or your real deficit. The vig does not move. Your job is to beat it with better picks, reduce it with better line access, or both.
No-Vig Calculator
Enter American odds for both sides. Negative numbers for favorites, positive for underdogs.
Side A
Side B
The number that separates serious bettors
Every recreational bettor hears "52.38%" and treats the number as a curiosity. Serious bettors treat the number as a cost of goods. The vig is a fixed fee on every wager, charged before the game starts, regardless of who wins.
The full picture is 104.76%: the overround on a standard -110/-110 market. That number tells you what you are paying to participate. Strip it out with the no-vig formula and you see what the market actually thinks. Do the season-loss math and you see what the vig costs over a year of bets. Shop the line to the sharpest available number and you reduce the cost before a single game is played.
US sportsbooks collected $16.96 billion in gross gaming revenue in 2025. That number grows because most bettors ignore the vig, bet into bad lines, and add parlays on top. The bettors who understand the machine and reduce their exposure to the vig by line shopping and no-vig conversion are the ones who give themselves a real chance at the break-even rate and beyond.