The math is brutal — but parlays still have a place. When a +EV correlated parlay actually beats straight bets, the same-game-parlay rule changes, and the payout schedules every regulated US sportsbook publishes.
A parlay (or accumulator, multi, combo) combines multiple bets into one ticket — all legs must win for the parlay to cash. The payouts are big; the math is brutal. A 4-leg NFL parlay at -110 per leg pays roughly 12.28× — but the true probability of cashing is only about 6% before the bookmaker's vig is layered on. Standard parlay margins on independent legs run 5–10%; same-game parlay margins run 15–25%. Despite the math, parlays still have legitimate uses: hedge structures, correlated +EV plays, and bonus-promo conversion. The discipline is knowing when the math actually works in your favour.
Parlay payouts use multiplied decimal odds: a 3-leg parlay at decimal 1.91 (-110) on each leg pays 1.91 × 1.91 × 1.91 = 6.97× the stake. Sounds great. Now the math:
True probability of all 3 legs hitting: 0.5238 × 0.5238 × 0.5238 = 14.36%. Fair payout for a 14.36% bet: 6.96× (decimal 1/0.1436). Sportsbook's offered payout: 6.97×. Almost identical.
So far so fair. The vig kicks in because each leg already has a 4.7% vig built into the -110 / -110 line. The compounding effect on a 3-leg parlay is roughly 14% margin (1.047³ − 1). On a 6-leg parlay, the vig compounds to 32%. The longer the parlay, the worse the math.
This is why parlay grinders, statistically, are bottom-quartile bettors. The sportsbook's hold rate on parlays is roughly twice what it is on straight bets.
If two bets are positively correlated (e.g. NFL Team A under their game total + Team A QB under passing yards) and the sportsbook treats them as independent in pricing, the parlay is +EV. Identifying mispriced correlation is a real edge.
Sportsbook promos like "20+ leg parlay insurance" or "boosted parlay payouts" temporarily turn standard parlays into +EV plays. Stacking promo parlays during major events (Super Bowl, March Madness, World Cup) is a recurring +EV niche.
A small parlay stake can hedge a large futures position by exposing the same outcome at higher odds. The math is structural rather than +EV-driven.
If you parlay for fun with capped-stake recreational money you accept will lose, the math is the cost of entertainment. There's no shame in that — just don't confuse it with a sustainable strategy. Treat it as buying a movie ticket.
Standard payouts on parlays of -110 / -110 legs (American odds). All legs must win.
| Legs | Decimal payout | American payout | True prob. | Vig |
|---|---|---|---|---|
| 2 | 3.65× | +265 | 27.4% | ~9.7% |
| 3 | 6.97× | +597 | 14.4% | ~14.7% |
| 4 | 13.28× | +1,228 | 7.5% | ~20.0% |
| 5 | 25.36× | +2,436 | 3.9% | ~25.5% |
| 6 | 48.43× | +4,743 | 2.05% | ~31.4% |
| 10 | 651× | +65,000 | 0.15% | ~58% |
The 10-leg parlay payout looks life-changing. The true probability is 1 in 667 and the sportsbook's expected hold is 58% of every dollar staked. Parlays of this length are entertainment, not investment.
Same-game parlays (SGPs) combine multiple props from a single event. The legs are correlated; sportsbooks price the correlation into SGP odds. Margins on SGPs run 15–25% — significantly worse than independent multi-event parlays.
The exception: inversely-correlated SGPs. The sportsbook prices the obvious-correlation direction (player yards over + team to win) more aggressively than the inverse direction (player yards under + team to win). Identifying inverse correlations the book hasn't priced is a genuine +EV niche. More on prop correlation →
Premium Domain for Sale
Short, memorable, exact-keyword for the smart-betting category.