Implied probability explained: how to turn any odds into a probability
Implied probability is the chance of an outcome that a price suggests. Decimal odds of 2.50 imply 1 ÷ 2.50 = 40%; American odds of -150 imply 60%. Turning prices into probabilities is the first step of every betting decision, because it lets you compare a price with your own estimate on the same scale. Below are the formulas for every odds format, how to remove the bookmaker’s margin, and how to use the result.
The formulas for each odds format
| Format | Formula | Example | Implied probability |
|---|---|---|---|
| Decimal | 1 ÷ odds | 1.80 | 55.6% |
| American (negative) | odds ÷ (odds + 100), minus sign dropped | -150 | 60.0% |
| American (positive) | 100 ÷ (odds + 100) | +200 | 33.3% |
| Fractional a/b | b ÷ (a + b) | 5/2 | 28.6% |
All three formats describe the same thing. Decimal 2.00, American +100 and fractional 1/1 all imply 50%. If you use one format, convert the others into it with the odds calculator rather than doing it in your head.
Why the probabilities add up to more than 100%
Add up the implied probabilities of every outcome in a market and you will get more than 100%. The extra is the bookmaker’s margin, called the overround, vig or juice.
- A point spread priced -110 / -110: 52.4% + 52.4% = 104.8%.
- A soccer 1X2 market priced 2.10 / 3.40 / 3.60: 47.6% + 29.4% + 27.8% = 104.8%.
That extra 4.8% is why a coin-flip bet at -110 needs to win 52.4% of the time just to break even, not 50%.
Removing the vig: the fair (no-vig) probability
To see what the market really thinks, strip the margin out. The simplest way is to divide each outcome’s implied probability by the market total:
| Outcome | Odds | Implied | ÷ 104.8% | Fair probability | Fair odds |
|---|---|---|---|---|---|
| Home | 2.10 | 47.6% | → | 45.4% | 2.20 |
| Draw | 3.40 | 29.4% | → | 28.1% | 3.56 |
| Away | 3.60 | 27.8% | → | 26.5% | 3.77 |
The fair probabilities add up to 100%, and the fair odds are what the prices would be with no margin at all. This proportional method is the one most calculators use. Other methods spread the margin differently — markets tend to load a little more of it on longshots — but for comparing prices the simple method is close enough.
Sharp, low-margin markets give the best no-vig numbers, because there is less margin to remove and the price has been tested by more money. A 105% market tells you more than a 115% one.
How to use implied probability
1. Compare it with your own estimate. If you think a team wins 55% of the time and the fair probability is 45%, you disagree with the market by 10 points. That is only worth acting on if you have a reason the market has missed — and most public news is already in the price.
2. Measure line movement in probability, not odds. A move from 2.30 to 2.05 looks small in odds but is 43.5% → 48.8%, a 5.3-point jump. Probability makes moves of different sizes comparable; see what are dropping odds and what is line movement.
3. Check parlays honestly. Probabilities multiply. Two legs at 70% each land together 49% of the time; three legs at 60% land 21.6% of the time.
4. Track your own hit rate against it. If your picks around 1.50 (66.7%) win 60% of the time, you are behind the market in that range. A bet tracker shows this after a few weeks.
A worked example
A match is priced 1.95 / 1.95 on the total, over and under. Each side implies 51.3%, the market adds up to 102.6%, and the no-vig probability is 50% each side. If the over then shortens to 1.80 and the under drifts to 2.05, the implied numbers are 55.6% and 48.8% (104.3%) and the no-vig over probability is 53.2%. The market has moved 3.2 points toward the over — a number you can compare with any other move, in any sport.
Does implied probability come true? 1,921 matches checked
Formulas are only useful if the number they produce means something. So we took every soccer match SharpPick recorded between September 1 and October 13, 2026 that has an official result — 1,921 matches across 65 leagues — and compared the closing no-vig probability of each outcome with what happened. The closing price is the last one recorded before kick-off at a low-margin bookmaker; the average overround on these markets was 104.2%. Each match contributes three rows (home, draw, away), 5,763 outcomes in all.
| No-vig probability at close | Outcomes | Price said (with vig) | Price said (no vig) | Actually happened |
|---|---|---|---|---|
| under 20% | 881 | 15.2% | 14.6% | 14.3% |
| 20–30% | 2,173 | 26.5% | 25.5% | 25.2% |
| 30–40% | 1,152 | 35.5% | 34.2% | 33.7% |
| 40–50% | 688 | 46.6% | 44.8% | 43.8% |
| 50–60% | 450 | 57.1% | 54.7% | 54.9% |
| 60–70% | 255 | 67.4% | 64.5% | 73.7% |
| 70–80% | 121 | 78.4% | 74.9% | 71.1% |
| 80% and over | 43 | 88.0% | 84.7% | 88.4% |
| All | 5,763 | 34.7% | 33.3% | 33.3% |
Three things stand out.
- The no-vig number is the right one. Across all 5,763 outcomes the raw implied probability (margin included) averaged 34.7%, the no-vig probability 33.3%, and 33.3% of the outcomes happened. The 1.4-point gap is the margin, and it lands on every side. Compare your own estimate with the no-vig column, never the raw one.
- The bands line up almost everywhere. Under 20% landed at 14.3% against a 14.6% label; 20–30% at 25.2% against 25.5%; 50–60% at 54.9% against 54.7%. Home wins were priced at 44.1% and happened 45.1% of the time, draws 25.5% versus 25.0%, away wins 30.4% versus 29.8%.
- The small bands wobble. The 60–70% band won 73.7% and the 70–80% band only 71.1%, which looks like a contradiction until you see the counts: 255 and 121 outcomes. Six weeks is not enough to pin a band of that size to within a few points, and a bigger sample would almost certainly pull both back toward their labels. Do not read a strategy into a short-sample bump.
Two-way markets: hockey and baseball
Moneylines with no draw behave the same way. Across 422 NHL, KBO and NPB games recorded in the same period (146 NHL games from September 19, 276 KBO and NPB games from September 2), the 40–50% band won 46.7% of the time against a 45.4% label, and the 50–60% band 53.1% against 54.4%. The 30–40% band (144 outcomes) won only 24.3% and the 60–70% band (144 outcomes) won 75.7%, mirror images of each other, which again is what a small sample looks like, not a rule about favorites.
One match from the table
Ipswich Town hosted Fulham on October 10, 2026. The last price before kick-off was 2.85 / 3.50 / 2.56 (home / draw / away):
| Outcome | Odds | Implied (1 ÷ odds) | ÷ 102.7% | No-vig probability |
|---|---|---|---|---|
| Ipswich win | 2.85 | 35.1% | → | 34.2% |
| Draw | 3.50 | 28.6% | → | 27.8% |
| Fulham win | 2.56 | 39.1% | → | 38.0% |
Fulham were the narrow favorite. Ipswich won 2–1. That is not the price being wrong: a 34% outcome is supposed to happen about one time in three, and in the table above the 30–40% band did exactly that, 33.7% of the time. One match tells you nothing about a probability; a few thousand do.
SharpPick converts every price it records into no-vig probability, every 5 minutes, and flags the games where it moves most before kick-off. The picks and their full record are on today’s AI picks and the public record.
FAQ
What does implied probability mean?
It is the chance of an outcome that a price suggests. Decimal odds of 2.50 imply 1 ÷ 2.50 = 40%. It is the market's estimate plus the bookmaker's margin, not a fact about the game.
How do you calculate implied probability from American odds?
For negative odds, divide the odds by the odds plus 100 (ignoring the minus sign): -150 is 150 ÷ 250 = 60%. For positive odds, divide 100 by the odds plus 100: +200 is 100 ÷ 300 = 33.3%.
Why do implied probabilities add up to more than 100%?
Because the bookmaker builds its margin into every price. A -110 / -110 market adds up to 104.8%; the extra 4.8% is the overround, often called the vig or juice.
What is a no-vig price?
The price with the margin removed. Divide each outcome's implied probability by the market total so the outcomes add up to 100%, then convert back to odds. It is the fairest single number for what the market thinks.
How accurate is implied probability?
At a low-margin bookmaker, very. Across 1,921 soccer matches we recorded from September 1 to October 13, 2026, the closing no-vig probabilities averaged 33.3% per outcome and 33.3% of those outcomes happened. Every 10-point band from under 20% to over 80% landed within about 2 points of its label except the 60–70% band, which overshot.