The Truth About Parlay Math
Parlays are marketed as the shortcut from $10 to $1,000. The math behind them is less romantic: odds multiply, but so does the bookmaker's cut. This guide shows exactly how compounded vig turns a 5% per-leg edge into a 30% house edge on an eight-leg ticket, and identifies the handful of situations where parlays are actually mathematically defensible.
Quick Calculator
Every sportsbook's marketing team loves parlays. The reason is simple: parlays carry the highest margin of any bet type on the board. A single spread at -110 costs the bettor about 4.55% in vig; a four-leg parlay of -110 spreads costs about 17% — almost four times as much for no additional edge in the picks. Yet the payout chart makes parlays look irresistible: $10 becomes $148 on four legs hit, $1,318 on seven legs. The marketing is honest about the payout and silent about the compounded margin.
The math is straightforward. Parlay decimal odds are the product of each leg's decimal odds. That works cleanly in the bettor's favor only if each leg's decimal price is the fair (no-vig) price. Since bookmakers use post-vig prices, each leg contributes its own 4-5% margin to the product. Multiply four of those together and you have destroyed any plausible edge most bettors could bring to the ticket.
1. The Parlay Math
# Payout (decimal) for an n-leg parlay parlay_decimal = O_1 * O_2 * ... * O_n # True probability if legs independent p_true = p_1 * p_2 * ... * p_n # Fair (no-vig) parlay decimal fair_decimal = 1 / p_true # Expected value per $1 stake EV = p_true * (parlay_decimal - 1) - (1 - p_true) # Compound vig (identical per-leg vig v for n legs) compound_vig ~ 1 - (1 - v)^n # Example: 4 legs, each -110 (vig ~4.55% per leg) compound_vig = 1 - (0.9545)^4 = 1 - 0.8304 = 0.1696 (16.96%)
2. Worked Example — 4 Legs at -110
Four NFL spread picks, each at -110 (decimal 1.909, implied 52.38%). Assume each leg is a pure coinflip — true probability 50% — which is the unspoken average assumption recreational bettors make when they pick spreads at random.
Quoted parlay payout parlay_decimal = 1.909^4 = 13.29 True parlay probability (independent coinflips) p_true = 0.5^4 = 0.0625 (6.25%) Fair no-vig parlay decimal fair_decimal = 1 / 0.0625 = 16.00 EV per $1 stake EV = 0.0625 * (13.29 - 1) - 0.9375 = 0.0625 * 12.29 - 0.9375 = 0.7681 - 0.9375 = -0.1694 (-16.94% ROI) The bookmaker holds about 17 cents of every expected dollar on a four-leg equally-matched parlay. The fair payout for the probability involved would be 16.00 decimal, not 13.29.
3. Compound Vig by Leg Count
| Legs | Single-Leg Vig 4.55% | Single-Leg Vig 5.5% | Single-Leg Vig 7% | Typical Offer vs Fair |
|---|---|---|---|---|
| 1 | 4.55% | 5.50% | 7.00% | ~ same |
| 2 | 8.89% | 10.70% | 13.51% | 13.29 vs fair 14.58 |
| 3 | 13.04% | 15.60% | 19.56% | 26.50 vs fair 30.50 |
| 4 | 16.96% | 20.20% | 25.19% | 50.50 vs fair 64.00 |
| 5 | 20.71% | 24.56% | 30.43% | 96.30 vs fair 134.10 |
| 6 | 24.31% | 28.75% | 35.30% | 183.80 vs fair 281.20 |
| 7 | 27.73% | 32.72% | 39.83% | 351.00 vs fair 589.00 |
| 8 | 31.00% | 36.45% | 44.04% | 670.00 vs fair 1236.00 |
| 10 | 37.11% | 43.63% | 51.60% | 2,437 vs fair 5,435 |
By eight legs the bookmaker holds roughly 31% of every expected dollar. This is casino slot-machine territory — actually worse than most regulated casino slots (which run at 4-12% house edge). The asymmetry between the advertised "long-shot dream" and the compounded mathematical reality is why parlays are the single most profitable bet type for every sportsbook on the planet.
4. Same-Game Parlays and Correlation
Same-game parlays (SGPs) layer a second mathematical problem on top. Legs within a single game are often correlated: if Patrick Mahomes throws for 350 yards, Travis Kelce is very likely to have 80+ receiving yards; if a team covers a -7.5 spread, the Over on the total is much more likely. A naive parlay that multiplies raw per-leg prices would wildly mis-price those outcomes in the bettor's favor.
To prevent this, bookmakers run correlated outcomes through proprietary pricing engines that strip out the positive correlation bonus and add a margin on top. The pattern, empirically, is that the book removes 100%+ of the correlation value — turning what should be a good bet into a price worse than even the compounded product would give.
# Fair price for 3 correlated legs with joint probability p_joint fair_decimal = 1 / p_joint # Example: KC -7, Over 54, Mahomes 300+ yards # Pretend true joint probability = 9.5% (correlation pushes it above # the naive product 8.2%) naive_product = 1.909 * 1.95 * 2.10 = 7.82 decimal fair_from_joint = 1 / 0.095 = 10.53 decimal # Typical bookmaker-offered same-game parlay price book_offered = 6.00 decimal (much worse than either!) # EV at offered price EV = 0.095 * (6.00 - 1) - 0.905 = 0.4750 - 0.905 = -0.43 (-43% ROI) SGPs routinely clear -20% to -40% EV on standard slates.
5. When Parlays Actually Make Sense
If each leg carries +3% EV as a standalone bet, a three-leg parlay of independent events preserves a compound +EV of roughly (1.03)^3 - 1 ~ +9.3%, minus parlay-specific vig rebuild. This only works if you have a verified edge per leg.
Rare, but real in niche markets. A two-leg parlay that the book priced as independent when the legs are actually correlated at rho ~ +0.15 can be strongly +EV. Found mostly in obscure futures combinations and secondary leagues.
A 20% profit boost on a 4-leg parlay turns a -17% EV ticket into roughly +0% EV. If you already wanted the legs, a boost can flip the math. Always compute exact EV rather than trusting the advertised "worth $X" figure.
If the parlay stake is a small amount set aside for entertainment — not part of the Kelly-sized bankroll — the math is less important than the emotional utility. The key is honest accounting: entertainment budget stays separate from investing budget.
6. Parlay Variance vs Straight-Bet Variance
Even if you somehow solved the EV problem, parlays sell variance. A $100 four-leg parlay at 13.29 decimal has a variance of roughly 180 times the variance of a $100 straight bet at -110. For Kelly-style bankroll management, that translates to wildly larger optimal stake cuts. The same edge staked at 4% of bankroll on a straight bet would be staked at under 0.3% in parlay form — destroying the compounding benefit of Kelly sizing.
The net result: even a parlay with genuine +EV is usually worse for long-term bankroll growth than the equivalent stake spread across the legs individually. Variance is not free; it eats compounding. See our Kelly Criterion guide for the underlying geometric-growth argument.
7. Round-Robin Parlays and Teasers
Round-robin combinations break a large parlay list into every possible smaller parlay — e.g., three teams in a round-robin produces three 2-team parlays. Mathematically this reduces leg-count compounding but increases the number of tickets, which re-accumulates vig from another direction. Net effect is usually slightly worse EV than an equivalent set of straight bets.
Teasers let you move the line by 6 or 6.5 points in exchange for reduced payout. In the NFL, 6-point teasers across key numbers (3 and 7) can be slightly +EV at correct prices; outside those specific "Wong teaser" situations teasers are firmly -EV. If you are not specifically attacking the 2.5-to-8.5 NFL spread window, teasers are not a free lunch.
8. Frequently Asked Questions
Are parlays always a bad bet?
They are negative EV by default because of compounded vig, but they can be positive EV when each leg already carries individual edge, when correlation is priced away incorrectly, or when a promotional boost offsets the compound vig.
How is a parlay payout calculated?
Multiply the decimal odds of each leg. 1.91 * 2.00 * 1.85 = 7.07 decimal on a 3-leg parlay. Multiply the result by your stake to get total return including stake.
Can I hedge a parlay that is one leg away?
Yes, by betting the other side of the last remaining leg in proportion to the parlay payout. This locks in guaranteed profit (reduced from the potential max) and is a common freeroll strategy in tournaments.
Do sharp bettors use parlays?
Rarely. Sharps prefer straight bets because they maximize the bankroll-growth rate per unit of edge. A few sharps use correlated-leg SGPs when they can identify a market where correlation was under-priced, but this is an exception, not a strategy.
Why do sportsbooks push same-game parlays so heavily?
Same-game parlays are the highest-margin product they offer, routinely clearing 15-25% vig. They are ideal for maximizing revenue from recreational customers, which is the SGP's entire reason for existing.
What happens if one leg is a push?
Most books drop the pushed leg and recalculate the parlay with one fewer leg at its original stake. Some older books void the entire parlay. Check the parlay rules at your specific book before staking.
Price your ticket with the parlay calculator, confirm per-leg EV using the EV tool, and measure the book's margin with margin analysis.
Responsible gambling notice. Parlays are among the highest-variance products a sportsbook offers. This article is educational and is not an endorsement to wager. Stake only what you can afford to lose. For support with problem gambling visit BeGambleAware.org or call 1-800-GAMBLER (US). Must be of legal betting age in your jurisdiction.