A parlay makes the prize grow faster than it makes the price fair.
A $100 wager on one selection at -110 offers a fairly modest $90.91 profit. Combine three similar selections, however, and the potential return suddenly looks substantial: about $696 in total, including the original stake. That larger number is the appeal of the parlay—it turns several ordinary bets into one eye-catching payout.
But every leg arrives with its own slightly reduced price. If each event were a true 50–50 proposition, fair decimal odds would be 2.00 per leg, making a three-leg return $800. At typical -110 pricing, each leg contributes only 1.91 to the multiplication, cutting the return to roughly $696. The gap is not merely three separate fees added together; the reduced prices multiply through the entire ticket. More selections raise the advertised prize, but they also give the bookmaker’s margin another chance to compound—and require every prediction to be correct before anything is paid.
What the -110 price actually contains
- Implied probability
Odds converted into a break-even rate. At -110, the calculation is 110 ÷ (110 + 100) = 52.38%.
- Vig or overround
The amount by which all implied probabilities exceed 100%. Two sides priced at -110 total 104.76%, so the market has a 4.76-percentage-point overround.
- Expected bettor loss
The average loss under an assumed true probability—not simply the overround. If each side is genuinely 50%, risking $110 to win $100 loses an average $5, or 4.55% of the stake.
- Parlay price
The combined price found by multiplying the decimal odds for every leg. Because each sportsbook price already includes margin, that margin carries into the combined payout.
A standard -110 selection has decimal odds of about 1.909. Three such legs therefore return roughly 1.909 × 1.909 × 1.909 = 6.96 times the stake, including the original stake.
If those independent legs are each true 50–50 outcomes, fair decimal odds would be 2.00 × 2.00 × 2.00 = 8.00. This is the key mechanism behind how parlay bets combine odds and payouts: prices multiply, so the shortfall from fair value compounds as legs are added.
The 4.76% overround should not be multiplied directly by the number of legs. Overround describes the probability total quoted by the market; expected loss depends on the true chance of each outcome and the exact payout offered.
Why the margin multiplies
For independent legs, a parlay’s decimal price is found by multiplying the decimal price of each selection. The fair price follows the same structure, but it uses true probabilities rather than the bookmaker’s posted odds:
- Offered decimal odds: leg price × leg price
- Fair decimal odds: 1 ÷ (probability of leg 1 × probability of leg 2)
Consider two propositions that are genuine 50–50 coin flips. Each has fair decimal odds of 2.00, but the bookmaker lists both at -110, equivalent to about 1.909 decimal. Multiplying the posted prices gives 1.909 × 1.909 = 3.64, while the fair price is 2.00 × 2.00 = 4.00.
That gap is easier to understand through payout efficiency. One -110 leg pays 1.909 ÷ 2.00, or about 95.45% of its fair return. With two legs, those efficiencies multiply: 95.45% × 95.45% = 91.12%. The parlay therefore retains only about 91.1% of the fair payout value.
A $100 stake wins only when both coin flips land correctly, which occurs 25% of the time. At 3.64 decimal odds, the expected returned amount is approximately $91.12: 25% × $364.46. Subtracting the $100 stake produces an expected loss of about $8.88, or 8.9%.
This method is useful when adjusting payout calculations for the bookmaker’s margin, but it depends on the legs being independent. Correlated outcomes require a joint probability rather than a simple product of individual probabilities.
From two -110 legs to an 8.9% expected loss
- Convert each price
A price of -110 converts to decimal odds of 1 + 100/110 = 1.909.
- Multiply the offered prices
The two-leg parlay pays 1.909² = 3.6446 decimal, or roughly 3.64 times the stake.
- Calculate the fair price
Two independent 50% events both win with probability 0.50² = 0.25. Fair decimal odds are 1/0.25 = 4.00.
- Measure expected value
Multiply the 3.6446 return by the 25% win probability: 0.25 × 3.6446 = 0.9112. Relative to a full stake of 1.00, the expected loss is 8.88%.
Decimal odds include the returned stake, not just profit.
The shrinking value behind bigger payouts
For independent coin-flip events, each leg has a true decimal price of 2.00. A standard -110 price pays only about 1.91, and multiplying that reduced price across a parlay makes the value gap increasingly visible.
| Legs | Offered payout | Fair payout | Theoretical expected loss |
|---|---|---|---|
| 2 | 3.64 | 4.00 | 8.9% |
| 3 | 6.96 | 8.00 | 13.0% |
| 4 | 13.28 | 16.00 | 17.0% |
| 5 | 25.36 | 32.00 | 20.8% |
Decimal payouts include the returned stake. Thus, a $100 five-leg parlay appears capable of returning about $2,536, an eye-catching result compared with the original wager. At fair odds, however, the same combination would return $3,200.
That difference is the central tension in vig effects across multi-team parlays: the possible cash return rises rapidly, but the bettor retains a smaller share of the combination’s fair value. Payout efficiency falls from roughly 91.1% with two legs to 79.2% with five.
The expected-loss column does not mean every five-leg parlay loses 20.8% in practice. Individual outcomes remain all-or-nothing; it describes the average loss per dollar staked over many repetitions under the coin-flip and independence assumptions. Correlated legs, altered parlay prices, promotions, or genuine pricing edges can change the result, but merely adding legs does not dilute the vig—it compounds it.
When an edge survives the parlay
An edge is the difference between an estimated true win probability, p, and the price’s break-even probability. At decimal odds d, break-even probability is 1 ÷ d. A selection estimated at 55% has a five-point edge if its offered odds are 2.00, which require 50% to break even.
The estimate should be compared with a fairer market baseline, not treated as certainty. Understanding how removing vig changes implied probability estimates helps separate the bookmaker’s margin from a bettor’s actual opinion.
Combining the return factors
For independent legs, each selection’s expected gross-return factor is p × d. Multiplying those factors estimates the parlay’s expected return:
Parlay EV factor = (p₁ × d₁) × (p₂ × d₂) × …
A result above 1.00 indicates positive expected value; below 1.00 indicates a negative expectation. Two independent legs each assessed at 55% and priced at 1.91 produce factors of 1.0505. Together, they yield about 1.1036, equivalent to a theoretical 10.36% return on stake.
Real edges can therefore compound. However, adding a weak leg multiplies the entire result by that leg’s factor. If a third 1.91 leg has a true probability of only 47%, its factor is about 0.898, reducing the combined factor to roughly 0.991 and erasing the advantage.
The calculation is only as sound as its inputs. Overconfident probability estimates, hidden correlation, or limits on the best available price can turn an apparent edge into ordinary bookmaker margin.
Common assumptions about parlay edge
It raises estimated EV, not the chance of winning.
More legs still reduce the probability that every selection succeeds.
Its return factor multiplies the whole parlay.
A sufficiently weak leg can erase several small advantages.
It depends on reliable probabilities and an appropriate independence assumption.
Bad estimates or unmodeled correlation distort the result.
Where multiplication stops working
The simple multiplication model assumes two things: each outcome is independent, and the bookmaker builds the parlay by multiplying the displayed decimal odds in full. That works reasonably well for unrelated events, such as bets on games in different leagues.
Correlation breaks the first assumption. In the same football game, for example, a bet on the favorite may be positively related to its quarterback exceeding a passing-yard total. Multiplying both standalone prices would treat the shared game script as if it had no effect, potentially producing an overly generous payout.
Sportsbooks may respond by:
- rejecting combinations they consider too closely related;
- reducing or otherwise adjusting the calculated payout;
- offering a same-game parlay price generated by a proprietary model.
The exact treatment depends on the sportsbook and its rules for correlated betting markets. Negative correlation can also affect pricing, even when a combination remains available.
With same-game parlays, the final quote may not reveal how much margin is attached to each leg or to the correlation adjustment. The earlier multiplication examples therefore serve as a clean baseline—not a universal method for reverse-engineering every offered payout.
For correlated legs, the offered parlay price is the relevant number. Comparing it with a personal joint-probability estimate is more informative than multiplying the displayed leg prices.
Test the value before combining legs
- Record every decimal price
Use the bookmaker’s displayed parlay odds rather than assuming all listed prices multiply without adjustment.
- Estimate each true probability
Assign a realistic probability to every outcome using information available before seeing the potential payout.
- Calculate each leg’s factor
Multiply probability by decimal odds: p × d. A result above 1 suggests positive estimated value; below 1 weakens the ticket.
- Calculate combined EV
For independent legs with fully multiplied prices, multiply all p × d factors, then subtract 1. If the quoted parlay price differs, multiply the combined probability by that price instead.
- Compare and stress-test
Check each leg’s standalone EV (p × d − 1) and compare separate bets with the parlay’s all-or-nothing risk. Lower uncertain probability estimates; any leg that then fails the test should be removed.
Leg count should be determined by the number of defensible edges, not by the size of the advertised return.
- A positive-looking ticket can depend on one fragile probability estimate.
- Separate bets preserve the value of winning selections when another pick loses.
A large payout is not evidence of a good bet. The parlay is justified only when its estimated combined return remains positive after realistic probability adjustments and any bookmaker pricing changes.