15 November 2026
A standard point spread priced at -110 on both sides looks like an even proposition, yet each side carries an implied probability of 52.38%. Add the two together and you get 104.76%, which is more than any single event can deliver. That gap is how the sportsbook gets paid, and once you can see it in the numbers, every price on the board starts to read differently.
American odds come in two flavors, and each has its own conversion. Positive odds, such as +120, show up on underdogs. DraftKings Sportsbook states that the implied probability for positive odds is 100 ÷ (odds + 100). Plug in +120 and you get 100 ÷ 220, or 45.45%, which matches the figure DraftKings shows for that price.
Negative odds work the other way around. DraftKings gives the formula as (−odds) ÷ (−odds + 100), meaning you drop the minus sign, then divide that number by itself plus 100. At -140, that is 140 ÷ 240, or 58.33%. Lumina's glossary writes the same rule with X standing for the odds without the sign, X ÷ (X + 100), and puts -110 at 52.38%. For positive odds it gives 100 ÷ (X + 100), and it lists +130 at 43.48%.
The practical point is that an odds number is a price, not a prediction, but you can translate it into a percentage the book is effectively asking you to accept. A bettor looking at +130 is being asked to treat the outcome as roughly a 43% proposition. If your own read is that the team wins more often than that, the price has some value for you. If your read is lower, the number is telling you to pass.
If the books priced both sides of a market at their true chances, the two implied probabilities would add up to exactly 100%. They do not. Lumina defines vig, short for vigorish, as the sportsbook's margin, calculated by converting each side to implied probability, adding the results, and subtracting 100%. Whatever is left over is what the book keeps for taking the action.
The -110 / -110 market is the clearest example. Bankroll Guardian's glossary says such a market sums to 104.76% implied probability, which corresponds to 4.76 points of overround. Put another way, the book has priced in nearly five points of probability that do not exist. A bettor is paying for them whichever side is chosen.
There is a second way to express the same cost, and the two get confused. Bankroll Guardian says those 4.76 points of overround at -110 against -110 equal a 4.55% vig when stated as a share of everything staked. The first number is the amount by which the market overshoots 100%. The second is the book's cut of the money wagered. When you see a vig figure quoted somewhere, check which of the two it is before comparing it with anything else.
The same arithmetic works on any two-way market, not just a spread. Take a line with one side at +130 and the other at -150. The first converts to 43.48%, the second to 60.00% (150 ÷ 250), and together they reach 103.48%. That market carries a thinner margin than the standard -110 pairing, even though the prices look lopsided. A big gap between the two numbers tells you which side the market favors; only the sum tells you what the book is charging.
American odds are the standard on US boards, but many tools, odds-comparison charts and international markets use decimal odds, and decimals make the probability math simpler. Ultragambler states that American odds convert to decimal with 1 + (A ÷ 100) for positive odds and 1 + (100 ÷ |A|) for negative odds. It also states that implied probability from decimal odds is 1 ÷ decimal odds. That second step is a single division, with no branching between plus and minus prices.
The table below runs the four prices used above through those formulas. The implied probabilities are the figures given by DraftKings and Lumina, and the decimal prices are worked out from Ultragambler's conversion rules.
| American odds | Decimal odds | Implied probability |
|---|---|---|
| +120 | 2.200 | 45.45% |
| +130 | 2.300 | 43.48% |
| -110 | 1.909 | 52.38% |
| -140 | 1.714 | 58.33% |
Reading across a row is a good check on your own work. Take -140: 1 + (100 ÷ 140) is 1.714, and 1 ÷ 1.714 lands on 58.33%, the same number the American-odds formula produced. If the two routes disagree, one of the steps has gone wrong. Decimal odds also show at a glance what a winning unit stake returns, including the stake itself, which is why they are handy when you are comparing a price across several screens.
For the fractional-style way of thinking, the useful habit is to stop treating plus and minus as separate worlds. Every price, whatever its format, reduces to one number between 0 and 100%. Once everything is a percentage, an underdog at +130 and a favorite at -140 sit on the same scale and can be added, compared and set against your own estimate.
The -110 price shows up on point spreads and totals so often that most US bettors treat it as the default. Ultragambler states that at -110 a bettor must win 52.38% of bets just to break even. That is the same figure as the implied probability, and it is no coincidence. The break-even rate for any price is its implied probability.
So a bettor who picks sides at a perfectly coin-flip 50% hits will lose money over time at -110, and not because of bad luck. The 2.38 points between 50% and 52.38% are the vig showing up in a win rate. Over a long run of bets, that gap is the amount a bettor has to overcome through skill, information or shopping for better prices before reaching a profit.
Consider what that means for a typical weekend of play. Someone going 5-5 on ten spread bets has hit 50% and is behind, because break-even at this price sits just above 5.2 wins per ten. Someone going 11-9 over twenty bets has won 55%, which clears the hurdle. The difference between a break-even bettor and a losing one can be a single result in either direction, which is why short streaks tell you very little and why tracking the price you took matters as much as tracking wins and losses.
The three ideas lock together into one routine you can run on any market before placing a bet. First, convert each side's odds to implied probability, using the plus or minus formula above or by turning the price into decimal and dividing it into 1. Second, add the two sides and subtract 100 to see the margin. Third, compare the implied probability of the side you like to your own honest estimate of its chances.
The margin step also shows you something less obvious. If you want a rough sense of what the market thinks without the book's cut, you can scale each side down by the total. In the -110 / -110 case, each side's 52.38% divided by 104.76% comes to 50%, which is what you would expect from a market that sees the two outcomes as equally likely. The scaling is a quick approximation, not an official figure, but it keeps you from mistaking a book's padded number for a real forecast.
A few habits follow from this. Compare the same game at more than one book when you can, since a lower combined percentage is a cheaper market for you. Be careful about when you quote a vig figure and which definition it uses, whether the overround of 4.76 points or the 4.55% share of money staked. And remember that your break-even rate depends on the price you actually took, not the one you meant to take, so a point or two of price difference on a single line changes the hurdle on every bet like it.
For the next bet you place, try the exercise by hand once. Write down both prices, convert them, add them up and see how far past 100% the market runs. If your own estimate for the side you like does not clear its implied probability, the price is asking you to pay for something you do not think is there.