No bet type is more popular or more misunderstood than the parlay.
Sportsbooks love parlays. They generate more revenue per dollar wagered than almost any other product. And recreational bettors love parlays too — for the thrill of big payouts from small bets.
Here's what the math actually says.
What Is a Parlay?
A parlay combines two or more bets into one. All legs must win for the parlay to pay out. Miss one and you lose the entire bet.
The appeal: a 5-leg parlay at even odds (-110 per leg) pays roughly 24:1. A $10 bet wins $240.
The catch: all five legs must hit.
The Parlay Math Nobody Shows You
Each -110 leg has a 52.38% implied win probability. For a 5-leg parlay where all legs are independent at exactly -110:
Win probability = 0.5238 × 0.5238 × 0.5238 × 0.5238 × 0.5238 = 3.9%
Only about 4% of 5-leg -110 parlays win. The sportsbook pays 24:1 — which implies a 4% probability is fair. But because each leg has vig embedded, the actual combined vig compounds. The true fair payout for a 3.9% win probability is 24.6:1. The book pays 24:1.
The difference is small on a single parlay. It compounds into significant house edge across millions of bets.
Industry average parlay win rate for recreational bettors: 14%. Oddible users following only Great and Good graded legs: 44%.
That's not magic — it's the difference between random parlay construction and selecting legs with positive expected value.
The Correlated Parlay Exception
Most parlays assume independent outcomes. But some outcomes are correlated.
Example: You parlay the Chiefs -7 WITH the over on total points. These are correlated — if the Chiefs cover big, they likely scored a lot, which helps the over. This isn't a free edge, but it does affect the probability calculation.
Some sportsbooks restrict correlated parlays for exactly this reason. When they're permitted, they can be legitimate — if the legs have genuine individual edge.
When Parlays Make Sense
Pure entertainment, small stakes: there's no sin in a $5 parlay if you understand you're paying for entertainment value, not edge. The cost is the expected value you give up.
Same-game parlays with correlated outcomes: a quarterback throwing yards AND the team winning by more points share correlation. Books have started pricing these separately (SGP odds). Compare the SGP price to independent leg multiplication — if it's priced better, there may be value.
When parlays don't make sense:
- As your main volume strategy (the vig compounds with every added leg)
- As a way to chase a target dollar amount with a small bankroll
- When you'd never bet some of the legs individually
The Brutal Variance Problem
Even +EV parlays have extreme variance. A 5-leg +EV parlay might have a 5% win rate — meaning you'll lose 95 out of 100. The 100-bet sample that includes your 5 wins is profitable in EV terms, but the 95 consecutive losses are psychologically brutal.
Recreational bettors almost universally misidentify variance as "my system stopped working." They haven't hit enough volume to see the edge. They switch to a new approach. The cycle repeats.
Flat single-game bets at -110 are higher variance than prop bets but lower variance than parlays. Understanding where you sit on this spectrum helps you set realistic expectations for your results.
Next: [Sports Betting Tools and Apps — The Complete 2026 Guide →] Previous: [Player Props vs. Game Lines →]
[Check what your parlay's real odds of winning are — free with Oddible →]