KEY TAKEAWAYS
CONTENTS
To calculate series odds from single-game odds, you convert the game price into a win probability, then add up the probability of every sequence in which that team reaches the required number of wins first. If a team has a 60% chance of winning any one game, and the games are treated as independent, its chance of winning a best-of-7 series is about 71.02%, which converts to fair odds of roughly -245. That is the core of the method: game probability in, binomial arithmetic in the middle, series probability and fair odds out. It matters because a series price is not a stretched-out game price. A team that wins most individual games gets more chances for its edge to show, so the series favorite is priced noticeably shorter than the single-game favorite. This guide works through the formula, a full best-of-7 example, the best-of-5 version, and the conversion back to American and decimal odds. If you need a refresher on reading a price as a percentage first, start with what implied probability means. For the market itself and how it settles, see how a series bet works. Every number below is a hypothetical example, not a current price.
Why Series Odds Are Not the Game Odds
A series is a race to a fixed number of wins: three in a best of 5, four in a best of 7. The series winner is whoever gets there first, not whoever wins a particular game. That changes the math. One upset can decide a single game, but in a series the stronger team can lose a game or two and still advance, because the weaker side has to keep winning to stay alive.
Think of each game as a coin that lands on the favorite’s side slightly more often than not. Flip it once and the favorite wins 60% of the time. Keep flipping until one side has four heads or four tails, and a small per-game edge compounds into a larger series edge. The longer the format, the more the better team’s advantage is expressed, which is why a 60% game favorite becomes roughly a 71% series favorite in a best of 7.
The effect runs both ways. At exactly 50% per game, the series is also 50%, because neither side has an edge to compound. And for an underdog, a longer series is usually worse news: a 40% team wins a single game 40% of the time but a best of 7 only about 28.98% of the time. This is the same idea behind calculating parlay probability, where you multiply leg probabilities, except a series counts paths to a target number of wins rather than requiring every single result to go one way.
How to Calculate Series Odds With the Binomial Formula
The method has three steps, and each one has a clear job. Get a clean per-game probability, count the winning paths, then add them up. The arithmetic is simple once you see why the paths are counted the way they are.
Step 1: Turn the game price into a probability
Convert the single-game odds into implied probability, then strip out the vig. For example, a no-vig price of -150 implies 150 / (150 + 100) = 60%. Use the vig-free number, not the raw implied probability, because the raw number includes the sportsbook margin and overstates both sides. If the quoted prices are -150 and +130, the raw figures are 60.00% and 43.48%, adding up to 103.48%; normalizing gives the favorite about 57.98%. The full procedure is covered in how to calculate no-vig odds.
Step 2: Count the ways to win the series
In a best of 7, the winning team needs four wins and can lose zero, one, two or three games along the way. The last game played must always be the clinching win, so the losses can only be arranged among the earlier games. If the team needs k wins and loses j games, the number of valid orders is C(k − 1 + j, j), the number of ways to place j losses among the first k − 1 + j games.
Step 3: Add the paths together
Each path with k wins and j losses has probability pk × qj, where p is the per-game win probability and q = 1 − p. Multiply by the number of orders and sum over every possible j:
P(series) = Σ C(k − 1 + j, j) × pk × qj, for j = 0 up to k − 1.
For a best of 7, k = 4 and j runs from 0 to 3. For a best of 5, k = 3 and j runs from 0 to 2. In plain terms, you are adding the chance of a sweep, plus the chance of winning in one extra game, and so on up to the longest possible series. The result is the probability that the team wins the series under this simple model.
Worked Example: 60% per Game in a Best of 7 Series
Suppose your vig-free estimate gives Team A a 60% chance of winning any game against Team B, so p = 0.60 and q = 0.40. Each row in the table below is one way the 7 game series can end with Team A winning, computed with the formula from the previous section.
| Team A wins | Paths | Calculation | Probability |
|---|---|---|---|
| 4-0 | 1 | 1 × 0.604 | 12.96% |
| 4-1 | 4 | 4 × 0.604 × 0.40 | 20.74% |
| 4-2 | 10 | 10 × 0.604 × 0.402 | 20.74% |
| 4-3 | 20 | 20 × 0.604 × 0.403 | 16.59% |
Adding the four rows gives 0.710208, or about 71.02% (the rounded rows show 71.03% only because of rounding). Team B’s series probability is the remainder, about 28.98%. Notice that a sweep is the least likely winning path for Team A, even though it is the strongest team: there is only one order in which it can happen, while a 4-3 win can happen 20 different ways.
The same steps for a best of 5, where Team A needs three wins, give 21.60% for 3-0, 25.92% for 3-1 and 20.74% for 3-2. That sums to 68.26%, a little lower than the best-of-7 figure because a shorter series gives the favorite fewer games to express its edge. A best of 3 at the same 60% gives 64.80%.
Converting Series Probability Back to Fair Odds
Once you have a series probability, convert it into a price so it can be compared with what a sportsbook lists. Fair decimal odds are 1 divided by the probability: 1 / 0.710208 = 1.41. For American odds on a favorite, use −(probability / (1 − probability)) × 100, which gives −(71.02 / 28.98) × 100, or about -245. Team B’s fair price is about +245, or decimal 3.45. For the general method in both directions, see converting probability to betting odds.
The table below shows how the same per-game edge produces different series probabilities depending on format. Read across a row to see the compounding effect: the further right you go, the more the per-game edge is magnified.
| Per-game win probability | Single game | Best of 3 | Best of 5 | Best of 7 |
|---|---|---|---|---|
| 55% | 55.00% | 57.48% | 59.31% | 60.83% |
| 60% | 60.00% | 64.80% | 68.26% | 71.02% |
| 65% | 65.00% | 71.83% | 76.48% | 80.02% |
| 70% | 70.00% | 78.40% | 83.69% | 87.40% |
Now compare a fair price with a listed one. Suppose a sportsbook offers Team A at -270 and Team B at +220 for the series. Those imply 72.97% and 31.25%, which add up to 104.22%, so the listed series price carries a margin of about 4.22 percentage points. Against your fair -245, the -270 asks you to risk more for the same profit: $100 at -270 earns about $37.04 in profit for a $137.04 payout, while $100 at a fair -245 would earn about $40.82. That gap is what the comparison reveals, but it is only as good as the 60% estimate behind it.
The best-of-5 conversion works the same way: 68.26% converts to about -215, decimal 1.47, with the opponent at about +215, decimal 3.15.
Common Mistakes When Calculating Series Odds
The most common error is plugging a vig-inflated game probability into the formula. The margin then gets magnified along with the edge. In the -150/+130 example, the raw 60% figure gives a best-of-7 estimate of 71.02%, while the no-vig 57.98% gives about 67.02%. Four points of series probability is a large difference when you are judging a price.
A second mistake is assuming the series price should equal the game price. A 60% favorite at about -150 per game is not a -150 series favorite. If you see a series price close to the game price, either the market expects the games to be very different from one another, or something in your inputs is off.
Third, many people confuse “4 wins out of 7” with a fixed seven-game binomial. The series stops when a team reaches four wins, so you count paths that end on a clinching win, not every combination of seven results. Both approaches happen to give the same series-winner probability, but only the path-counting method gives correct figures for each series length, such as 4-2 or 4-3.
Finally, a calculated edge is not a guaranteed result. A 71.02% favorite still loses the series about 29 times in 100 under the model, and the model itself is only an estimate. Treat the output as one input to a price decision, never as a reason to stake more than you can afford to lose.
Home Games, Dependence and Vig: Where the Model Bends
The binomial model assumes every game has the same probability and no game affects the next. Real series break both rules to some degree. The biggest structured difference is home advantage. In a common 2-2-1-1-1 home-and-away pattern, the higher seed hosts games 1, 2, 5 and 7. Suppose Team A wins 65% at home and 55% on the road. Working through every possible sequence gives Team A about a 72.50% chance in the best of 7, compared with 71.02% from a flat 60%. The extra home game is the reason.
Independence is the softer problem. Injuries, rest, schedule changes and tactical adjustments mean that games in a series are related, not separate draws. When outcomes are linked, simple multiplication can misstate the true probability, the same issue described in why correlated parlays break odds multiplication.
In practice, use the model as a baseline to compare against the listed series price, then ask what the market might know that a flat per-game number does not capture.
Related Odds Concepts and Next Steps
Series math sits on top of several basic odds skills. You need to convert a price into a probability, remove the margin, and turn a probability back into a price. If any step felt unclear, revisit implied probability and the no-vig method linked earlier, then return to the worked example. The natural next step is to apply the same thinking to the exact series result market, where each row of the worked table above becomes its own selection with its own price. Comparing those listed prices with the 12.96%, 20.74%, 20.74% and 16.59% figures from the model shows how the sportsbook spreads its margin across the outcome list. The same path-counting logic also helps explain why longer series lengths are not automatically less likely.
Frequently Asked Questions
How many wins do you need to win a best of 7?
You need four wins. The series ends as soon as one team reaches four, so it can last four, five, six or seven games. A best of 5 needs three wins and a best of 3 needs two. The winning team always wins the final game played.
How do you calculate the odds of winning a 7 game series?
Convert the game price to a no-vig win probability, then add the probability of winning 4-0, 4-1, 4-2 and 4-3. Each term is the number of possible orders times p to the fourth power times the loss probability raised to the number of losses. At 60% per game the total is about 71.02%.
What percentage of Game 1 winners win the series?
It depends on the matchup and the sample studied. In a simple model where both teams are 50% per game, the Game 1 winner then needs three wins before the opponent gets four, which happens about 65.63% of the time. Real results vary with team strength and home advantage.
Why is the series favorite priced shorter than the game favorite?
Because a series gives the stronger team several chances for its edge to show. One upset can decide a single game, but the underdog must win most of several games. A 60% game favorite becomes about a 71.02% favorite in a best of 7, which is roughly -245 in fair odds.
How likely is a series to go to Game 7?
Under a simple independent-games model, two evenly matched teams reach Game 7 about 31.25% of the time, since the series must be tied 3-3 after six games. If one team is 60% per game, the chance drops to about 27.65%, because the stronger team often finishes earlier.
Is the binomial series model accurate?
It is a useful baseline, not a forecast. It assumes every game has the same probability and that results are independent. Home advantage, injuries, rest and adjustments all break those assumptions, so treat the result as an approximate fair price to compare against the market.



