Parlays vs Single Bets: The Math Behind the Sportsbook's Favorite Product

Parlays are marketed as the fun way to turn a small stake into a life-changing payout — string together a few picks, and the odds multiply into something exciting. They're also, by a wide margin, the most profitable single product for sportsbooks. Those two facts are connected. This guide shows exactly why, with the actual math, and the narrow conditions under which a parlay can still make sense.

How parlay odds actually compound

A parlay's payout is the product of each leg's decimal odds, not their sum. Two legs at 1.91 (-110) each don't combine into a 2× multiplier — they combine into 1.91 × 1.91 = 3.65. That part is just multiplication, and it's fair: your true odds of hitting two independent 50/50 propositions really are 1 in 4. The problem is what's hiding inside each leg's price.

The vig compounds too — and that's the part bettors miss

Every leg you add to a parlay isn't just multiplying your odds of winning — it's also multiplying the bookmaker's margin baked into each leg. A single -110 bet carries roughly a 5% built-in edge for the house (a rounded version of the 4.76% overround from our devigging guide). The intuitive assumption is that four legs at 5% each should cost you about 20% in total. The real math is worse than that, because the vig compounds multiplicatively across legs, not additively.

• Naive assumption: 4 legs × 5% vig ≈ 20% combined house edge
• Actual compounding: (1.05)⁴ − 1 ≈ 21.6% combined overround
• At 6 legs the gap widens further: (1.05)⁶ − 1 ≈ 34% — nearly double the naive estimate
• Each extra leg doesn't add a fixed cost, it multiplies the existing one

Single bets vs. parlays in real numbers

If you have zero edge and bet a single game at -110 repeatedly, you lose roughly 4.5–5% of your total action to the house over time — a slow bleed, but a survivable one at reasonable stakes. Bet the same four coin-flip propositions as a 4-leg parlay instead, and you're handing over more than 20% of your expected value before a single leg has even been graded. The stake might look the same on the ticket; the embedded cost is not.

When a parlay can actually be defensible

Two conditions have to hold simultaneously. First, the legs need to be genuinely independent — no shared game, team, or outcome that could correlate them (more on why that matters below). Second, you need a real, de-vigged positive edge on every single leg, not just a hunch. Because the vig compounds, a thin 1–2% edge per leg that would be marginally +EV as a single bet usually isn't enough to overcome a 20%+ combined overround across four legs. The math demands your edge be unusually large and repeated across every leg — which is rare. For a genuine edge, sizing it properly with a fractional Kelly approach on single bets will almost always out-earn stacking the same picks into a parlay.

Same-game parlays: the correlation trap

Same-game parlays (SGPs) make the problem worse, not better. Legs inside one game are rarely independent — a quarterback going over his passing yards line and his top receiver going over his receiving yards line tend to happen together, because both depend on the same underlying event: the team throwing more than expected. Treating those as independent (multiplying their odds as if uncorrelated) would understate the joint probability and hand the bettor an edge — which is exactly why sportsbooks don't price them that way.

None of this means every parlay is automatically a bad bet — it means the bar for making one worthwhile is much higher than most bettors assume, and same-game legs raise that bar again. If you want to see exactly how much vig a specific combination is carrying before you place it, run it through our parlay calculator. And if you're deciding between a multi-leg parlay and a single larger stake on your best individual pick, expected value betting and our track record are the better starting points than the multiplier on the bet slip.