Implied Probability: What the Odds Are Really Telling You
Implied probability is the break-even win rate hidden in every betting price. How to calculate it, strip out the vig, and test it against your own estimate.
By Lines & Limits Editorial6 min read
Implied probability is the win rate a price is built around — the exact percentage at which a bet breaks even and returns nothing. A −110 bet has an implied probability of 52.38%. Win more often than that and you profit; win less and you lose.
It is the single most useful conversion in betting, because it turns an arbitrary-looking number like +265 into something you can actually reason about: “this happens a bit more than a quarter of the time.”
How to calculate implied probability
The formula depends only on the format of the price. Full derivations of the odds formats themselves are in our guide to reading betting odds.
| Format | Formula | Example | Result |
|---|---|---|---|
| Negative American | abs(odds) ÷ (abs(odds) + 100) |
−110 | 110 ÷ 210 = 52.38% |
| Positive American | 100 ÷ (odds + 100) |
+150 | 100 ÷ 250 = 40.00% |
| Decimal | 1 ÷ decimal |
1.91 | 1 ÷ 1.91 = 52.36% |
| Fractional | denominator ÷ (numerator + denominator) |
10/11 | 11 ÷ 21 = 52.38% |
Decimal is the fastest to work with because the conversion is a single division. If you can only remember one formula, remember 1 ÷ decimal.
Break-even win rates at standard prices
This table is the practical version of implied probability. The right-hand column is the record you need over 100 bets to be ahead.
| Price | Implied probability | Break-even record per 100 bets |
|---|---|---|
| −200 | 66.67% | 67–33 |
| −150 | 60.00% | 60–40 |
| −130 | 56.52% | 57–43 |
| −120 | 54.55% | 55–45 |
| −115 | 53.49% | 54–46 |
| −110 | 52.38% | 53–47 |
| −105 | 51.22% | 52–48 |
| +100 | 50.00% | 50–50 |
| +110 | 47.62% | 48–52 |
| +120 | 45.45% | 46–54 |
| +150 | 40.00% | 40–60 |
| +200 | 33.33% | 34–66 |
At −110 with $100 bets, a 52–48 record loses (52 × $90.91) − (48 × $100) = $4,727.27 − $4,800 = −$72.73. A 53–47 record makes (53 × $90.91) − (47 × $100) = $4,818.18 − $4,700 = +$118.18. One win per hundred separates a losing bettor from a winning one at standard pricing, which is a fair summary of how thin this activity is.
Why raw implied probability is always wrong
Convert both sides of any market and add them up. They will exceed 100%.
A standard total priced −110 / −110:
- Over:
110 ÷ 210 = 52.3810% - Under:
110 ÷ 210 = 52.3810% - Sum:
104.7619%
Two outcomes that cannot both happen have a combined implied probability of 104.76%. The 4.76% surplus is the overround — the sportsbook’s margin expressed as excess probability. It is the reason raw implied probability overstates every single outcome on the board. The full mechanics of that margin are in our explainer on what the vig is.
So before implied probability means anything, you have to take the fee back out.
The two-step no-vig calculation
No-vig probability (also called fair or devigged probability) is implied probability rescaled so the market sums to exactly 100%.
Step one: convert every side to implied probability. Step two: divide each one by the sum of all of them.
That is it. The second step is just normalization.
Worked example: a −110 / −110 total
- Over implied:
110 ÷ 210 = 52.3810% - Under implied:
110 ÷ 210 = 52.3810% - Sum:
104.7619% - Over no-vig:
52.3810 ÷ 104.7619 = 50.0000% - Under no-vig:
52.3810 ÷ 104.7619 = 50.0000%
A −110 / −110 market is the book saying “this is a coin flip, and the fee is 4.76% of a unit.” The symmetry makes it easy, but the method is identical for lopsided markets.
Worked example: a −150 / +130 moneyline
- Favorite implied:
150 ÷ 250 = 60.0000% - Underdog implied:
100 ÷ 230 = 43.4783% - Sum:
103.4783% - Favorite no-vig:
60.0000 ÷ 103.4783 = 57.9832% - Underdog no-vig:
43.4783 ÷ 103.4783 = 42.0168%
Convert those back to prices and the fair line is exactly −138 on the favorite and +138 on the underdog. Check it: 1 ÷ 0.579832 = 1.7246 decimal, and −100 ÷ 0.7246 = −138. On the other side, 1 ÷ 0.420168 = 2.38 decimal, and (2.38 − 1) × 100 = +138.
So the market’s actual opinion is that the favorite wins about 58% of the time. The −150 you were quoted is that opinion plus 2% of extra probability charged as a fee.
Why the no-vig number is the book’s real opinion
A sportsbook that priced a genuine coin flip at −110 / −110 and moved the line whenever money got lopsided is not expressing a view about the game. It is expressing a view about where the market clears. On liquid markets — major-league sides and totals — that clearing price aggregates a very large amount of money from people who are trying hard to be right.
It also explains why betting lines move: the no-vig number is a live estimate, and it updates as information and money arrive. The closing no-vig probability is the market’s final answer.
That is why the no-vig implied probability of a major market is the strongest freely available estimate of an outcome that most bettors will ever have. It is not perfect. It is better than what you produced in twenty minutes.
It is also worth knowing that simple normalization — dividing by the sum — is the standard first approximation, not the only method. It distributes the margin proportionally across both sides, and there is reasonable argument that books load more of their margin onto longshots than onto favorites, which means normalization slightly understates the fair price of heavy favorites and overstates longshots. Alternative devigging methods exist for exactly this reason. For a two-way market at short prices, the proportional method is close enough.
Using implied probability to test your own opinion
The workflow is short:
- Convert both sides of the market to implied probability.
- Normalize to get the no-vig probabilities.
- Write down your own estimate before you look at step 2, so the market does not anchor you.
- Compare.
If your estimate is higher than the no-vig number, the bet has positive expected value given your estimate. Turning that gap into a dollar figure is the job of expected value in betting: at a price of −110 with a true probability of 55%, EV = (0.55 × $90.91) − (0.45 × $100) = $50.00 − $45.00 = +$5.00 per $100 risked, or +5%.
Change the assumption to 50% and the same bet is (0.50 × $90.91) − (0.50 × $100) = −$4.55 per $100 — a 4.55% loss, which is exactly the theoretical hold on a −110 / −110 market.
The entire distance between those two outcomes is the accuracy of one number you made up.
What to do with a gap you think is real
Do not size up on it. Test it. The cheapest test available is whether your bets are beating the number the market settles on — if you are consistently getting a better price than the closing line, your estimates are carrying information; if you are not, they are not. That is the argument behind closing line value, and it is a far more honest scoreboard than short-run profit and loss.
The second cheapest test is to keep making the estimate and never bet it. Record your number, record the closing no-vig number, and compare after a hundred games. Nothing about a spreadsheet gets more accurate because money is on it, and the version of this exercise with no money on it will tell you the same thing for free.
Frequently asked questions
What is implied probability in betting?
Implied probability is the probability of an outcome embedded in its price — specifically, the win rate at which that bet breaks even over the long run. A −110 bet has an implied probability of 52.38%, meaning it must win 52.38% of the time to return exactly what you risk. Anything below that is a losing bet.
How do you calculate implied probability from American odds?
For a negative price, divide the absolute value by itself plus 100: −150 becomes 150 ÷ 250 = 60%. For a positive price, divide 100 by the price plus 100: +200 becomes 100 ÷ 300 = 33.33%. From decimal odds it is simpler still: 1 divided by the decimal price.
What does no-vig probability mean?
No-vig probability is the implied probability after the sportsbook's margin has been removed, so the two sides sum to exactly 100%. You calculate it by converting both prices to implied probability and dividing each by their total. It is the market's estimate of the outcome with the fee stripped out.
What win percentage do you need to break even at -110?
You need to win 52.38% of your bets to break even at −110, which is roughly 53 wins in every 100 bets. At 52 wins and 48 losses on $100 bets you lose $72.73; at 53 and 47 you make $118.18. The margin between profitable and unprofitable is one win per hundred.
Does implied probability tell you who will win?
It tells you what the market believes, priced with a fee on top. Markets are usually well calibrated on major leagues, so implied probability is a good default estimate — but it is a price, not a forecast, and it moves in response to money as well as information.
How do you find value using implied probability?
Convert the price to a no-vig probability, then compare it to your own estimate of the outcome. If you think a team wins 57% of the time and the no-vig price says 52%, that gap is a theoretical edge. The hard part is not the arithmetic — it is having an estimate that is genuinely better than the market's.