How does a parlay bet work?

A parlay bundles two or more individual bets (called legs) into a single wager. Every leg must win for the parlay to pay out. The odds of each leg multiply together, producing a combined payout that grows exponentially with each addition. If even one leg loses, the entire parlay loses and you forfeit your stake. Most sportsbooks allow parlays with up to 10-15 legs.

What happens if one leg of my parlay loses?

The entire parlay loses. There is no partial credit — one missed leg zeroes out the wager. Some sportsbooks offer parlay insurance promotions that refund your stake (usually as a free bet) if exactly one leg misses, but the standard rule is all-or-nothing. This binary structure is what drives both the high payouts and the low hit rate.

Are parlays a good bet?

Generally, no. Parlays are negative expected value because the vig from each individual leg compounds multiplicatively. A two-leg parlay with standard -110 juice carries roughly twice the vig of a single bet. The exception is correlated parlays — legs within the same game whose outcomes are linked — where the true combined probability exceeds what independent odds imply.

How do you calculate parlay payouts?

Convert each leg to decimal odds, then multiply them together. That product is the parlay's decimal odds. Multiply by your stake to get the total return, then subtract your stake for profit. Example: two legs at -110 each convert to 1.909 decimal. Multiply: 1.909 x 1.909 = 3.645. A $100 bet returns $364.50 total, or $264.50 profit.

What is a same-game parlay?

A same-game parlay (SGP) combines multiple selections from a single game — such as a moneyline pick, a point total, and a player prop. Sportsbooks price SGPs using correlation models rather than simple multiplication because outcomes within the same game are statistically linked. SGPs can occasionally offer better value than cross-game parlays if the book underestimates correlation.

What Is a Parlay Bet? How Parlays Work and When to Use Them

A parlay is a single wager that links two or more bets together. All selections — called legs — must win for the parlay to pay out. The appeal is straightforward: combined odds produce payouts far larger than any individual bet. The tradeoff is equally straightforward: the probability of hitting every leg drops fast, and the sportsbook’s margin compounds with each one.

Key Takeaways

A parlay multiplies the odds of each leg together — all must win or the entire bet loses
The vig compounds on every leg, making most parlays significantly -EV
A standard two-leg parlay at -110/-110 carries about 4.5% total vig; a four-leg parlay exceeds 9%
Correlated same-game parlays are the one scenario where parlays can offer neutral or positive EV
The same probability math applies to prediction market contracts — combine uncorrelated positions and your edge erodes the same way

How Does a Parlay Payout Work?

Parlay payouts are calculated by multiplying the decimal odds of each leg. The product determines the combined odds of the entire wager.

Step 1: Convert each leg to decimal odds.

-110 American = 1.909 decimal
+150 American = 2.500 decimal
-200 American = 1.500 decimal

Step 2: Multiply decimal odds together.

Two-leg example: 1.909 x 2.500 = 4.773
Three-leg example: 1.909 x 2.500 x 1.500 = 7.159

Step 3: Multiply by stake, subtract stake for profit.

$100 at 4.773 = $477.30 total return, $377.30 profit

Use the parlay calculator to compute exact payouts for any combination of legs.

How Do Parlay Payouts and Probability Scale With Legs?

The table below shows how adding legs affects payout, true probability, and the house edge. Each leg assumes a standard -110 line with a true probability of 50%.

Legs	Decimal Odds	Payout on $100	True Probability	Implied Probability	House Edge
1	1.909	$190.91	50.0%	52.4%	4.5%
2	3.645	$364.46	25.0%	27.5%	9.1%
3	6.962	$696.15	12.5%	14.4%	13.2%
4	13.293	$1,329.27	6.25%	7.5%	17.0%
5	25.382	$2,538.16	3.13%	3.9%	20.5%
6	48.454	$4,845.39	1.56%	2.1%	23.8%

The house edge column tells the real story. A single -110 bet costs you about 4.5 cents per dollar in expectation. By the time you reach a six-leg parlay, you are handing back nearly a quarter of your expected return before the games even start. This is the mathematical cost of parlays — each additional leg multiplies the vig.

Why Are Parlays Generally Negative EV?

The expected value of a parlay degrades because the book’s margin applies to every leg independently. Even if each individual bet is only slightly -EV, the compounding effect is significant.

Consider a two-leg parlay where each leg has a true 50% chance. The true probability of hitting both is 25%. At fair odds, that pays 4.00 (decimal). But at -110 each, the parlay pays only 3.645. The difference — the gap between 4.00 and 3.645 — is pure sportsbook margin.

The math works the same way in reverse. If you hold two prediction market contracts and both must resolve “Yes” for you to profit, your combined probability is the product of both individual probabilities. No one is adding vig, but the probability still compounds downward. The Odds Reference dashboard tracks contract prices across prediction markets — you can verify the implied probabilities yourself.

When Can Parlays Actually Offer Value?

The one exception to the parlay-as-bad-bet rule is correlated parlays — legs whose outcomes are statistically linked.

Example: You parlay the Bills moneyline with the “over” on the game total. If the Bills win, it probably means they scored a lot. These outcomes are positively correlated, meaning the combined probability is higher than what you get by multiplying independent probabilities. If the sportsbook prices them as if they were independent, you capture an edge.

Same-game parlays (SGPs) are where this matters most. Common correlations include:

Moneyline + Over: Teams that win tend to score, pushing totals higher
QB passing yards + team total: High team scoring usually means high passing volume
Running back rush yards + team moneyline: Teams that win often run the ball to close out games

Books have gotten better at modeling these correlations, so the edges are thinner than they used to be. But miscalculations still happen, especially on prop-heavy SGPs where the correlation matrix is complex.

How Should You Evaluate a Parlay Before Placing It?

The process mirrors how you evaluate any bet:

Estimate probability for each leg — Your true probability estimate is the starting point. Convert the odds to implied probability and compare.
Multiply true probabilities — This gives the combined true probability of the parlay hitting.
Compare to parlay odds — If the combined true probability exceeds the implied probability of the parlay odds, the bet is +EV. If it does not, the bet is -EV.
Account for correlation — If legs are within the same game, adjust upward for positive correlation or downward for negative correlation.

The implied probability of each leg is the number to beat. If you cannot articulate why your estimate exceeds it for every single leg, you are making a recreational bet — not a calculated one.

How Do Prediction Markets Compare to Parlays?

Prediction markets and parlays rely on identical probability math. Combining two 50-cent contracts into a parlay-style position (both must resolve Yes) gives you a 25% combined probability at fair pricing. The difference is that prediction markets typically carry 1-3% in spread versus 4-10% in sportsbook vig per leg.

This means that if you must combine bets, doing so on prediction markets is mathematically cheaper. Our data across platforms shows effective spreads on liquid contracts running 1-2% on Polymarket and Kalshi, compared to 4.5%+ per leg at standard sportsbook juice.

The data is clear on this: parlays are a high-variance, negative-expectation product for most bettors. The exceptions exist — correlated SGPs, mispriced lines, promotional offers — but they require rigorous analysis, not gut feel. If you want to track line movement and identify pricing inefficiencies across sportsbooks and prediction markets, the Odds Reference dashboard provides real-time data on thousands of markets.