Books love parlays. They love them because, on average, parlays make the book a lot more money per dollar wagered than singles do. The standard 3-leg parlay carries about 12% hold; SGPs and SGP+ products often run 15-25%. That's a significant tax to pay every time you click.
So when, if ever, are parlays worth betting? Surprisingly often — but only when the legs satisfy specific conditions.
The math of parlays
A parlay's payout is the product of all its legs' decimal odds. Three legs at 1.91 each = 6.97 decimal (about +597 American). The implied probability of all three legs hitting is the product of the leg probabilities — at 52.4% per leg, that's 14.4%.
A 12% hold means the book is offering you ~12% less than fair value. Fair would pay roughly 7.95×; the book pays 6.97×.
When a parlay is +EV
The math of independent +EV legs:
- Each leg has positive EV individually. If you wouldn't bet leg A, leg B and leg C as singles, parlaying them doesn't fix the math.
- The legs are statistically independent. Two unrelated games. A coin-flip in NYC and a coin-flip in LA. Combining them parlay-style preserves the EV math.
- The book hasn't applied a "parlay tax". Some books quietly worsen pricing on parlay tickets. Compare the implied probability of the combined parlay to the product of single-bet implieds.
If all three conditions hold, parlaying maintains your edge while compounding it through multiplication. The variance is much higher than singles — but the EV is real.
When SGPs are actually worth betting
Same Game Parlays look terrible on paper because of correlation pricing — but they're not always traps.
Good SGP candidates:
- Heavy favorite + their star + game total. If the Chiefs are -10, Mahomes 2+ TDs and over 51.5 are strongly correlated. The book prices that, but sometimes underprices it.
- Game-script-driven props. If Team A wins big, Team A's RBs run more, Team A's defense gets sacks, Team B's QB throws more. Stack legs that all assume the same game flow.
- Books that don't fully correlate-price. Smaller US books and offshore books often have less sophisticated SGP engines than DraftKings/FanDuel.
Bad SGP candidates:
- Random unrelated props. "Team A wins + Team B's QB over 250 yards" — no narrative coherence, you're paying full hold.
- Anything in the SGP+ "boosted parlay" feature. The boost is the book's way of telling you they've priced it well and want your action.
- "Parlay of the day" promotions. Almost universally negative-EV.
The correlation principle, explained
Most bettors don't understand why SGP pricing works the way it does. Here's the intuition:
Independent legs multiply: probability of A and B = P(A) × P(B). But if A and B are positively correlated — both more likely if a certain underlying event occurs — the joint probability is higher than the product. Books "fix" this by adjusting the combined price downward.
If they correlate-price perfectly, you're stuck paying the same hold as the underlying market. If they over-correlate (overprice the correlation), the parlay is +EV. If they under-correlate (underprice it), the parlay is -EV.
Identifying the spots where books under-correlate is where the SGP edge lives.
Practical rules
- Cap parlays at 3 legs. Anything 4+ is mostly entertainment.
- Round robins beat single big parlays for capital efficiency. A 4-team round robin in 3-team parlays gives you 4 tickets, surviving 1 loss.
- Use the parlay calculator to verify the implied probability of the combined ticket vs. the product of single-leg implieds. If they match, the book is correlate-pricing. If the combined is a worse implied, the book is taking extra hold.
- Never bet a parlay you wouldn't bet the legs as singles. The "if I just hit one ticket" mindset is how recreational bettors lose.
Want more? Browse all our strategy guides, or jump straight to the calculators.