Intermediate Strategy

NFL Parlay Strategy

The truth about parlays and when they actually make sense

The Hard Truth About Parlays

Sportsbooks Love Parlay Bettors

Parlays are one of the most profitable products for sportsbooks. The house edge on parlays typically runs 10-30% depending on legs and the payout table, compared to about 4.5% on straight bets. This is because parlay odds don't reflect true independent probabilities.

The Math Problem

For a 2-leg parlay with -110 odds on each leg:

Fair Odds

Each leg at -110 = 52.4% win rate

True parlay odds: 0.524 × 0.524 = 27.5%

Fair payout: +264

Actual Odds

Most books pay around +260

Some pay as low as +250

You're getting shortchanged

This gap widens significantly with more legs.

Parlay Odds & Payouts

Legs Typical Payout Win Probability* House Edge
2 +260 27.5% 10.0%
3 +600 14.4% 12.5%
4 +1100 7.5% 25.0%
5 +2200 3.9% 28.1%
6 +4500 2.0% 28.1%
*Win probability assumes 52.4% per leg (the -110 break-even); house edge is exact math with fair-coin (50%) legs against the listed payout

When Parlays Can Work

Despite the math, there are specific situations where parlays can have positive expected value:

1. Correlated Parlays

When outcomes are positively correlated, the combined probability is higher than the product of individual probabilities.

Examples of Positive Correlation:
Strong Correlations
  • Team ML + Game Over
  • Team spread + QB passing yards Over
  • Big underdog + Team total Over (garbage time)
  • Heavy favorite + RB rushing yards Over
Moderate Correlations
  • Team win + star player TD
  • Game Under + favorite cover
  • Low total + defensive/special teams TD
Caution: Many sportsbooks now price correlations into same-game parlays, reducing or eliminating this edge.
2. Parlaying +EV Bets

If each individual leg has positive expected value, the parlay maintains positive EV (though variance increases).

Example:

If you have two bets each with 55% true win probability at -110:

  • Straight bet EV: +4.5% per bet
  • Parlay: 0.55 × 0.55 = 30.25% win rate
  • At +264 payout, EV is still positive

The catch: You need to actually be able to identify +EV bets consistently, which is very difficult.

3. Cross-Book Parlays

Building "synthetic parlays" by taking the best line at each book:

Example:
  • Book A has Chiefs -3 at -105 (best spread price)
  • Book B has Bills ML at +155 (best ML price)
  • Book C offers parlay calculator with true odds

By shopping, you reduce the house edge on each leg, making the combined bet closer to fair value.

Same Game Parlays (SGPs)

SGPs combine multiple bets from the same game. Sportsbooks love them because:

Why Books Love SGPs
  • Higher hold percentage
  • Correlations favor the house
  • More legs = more edge
  • Entertainment value drives volume
When SGPs Can Work
  • Strong positive correlation
  • Props that books misprice
  • 2-3 legs maximum
  • High-confidence game thesis
SGP Strategy:

If you use SGPs, build them around a core thesis. Example: "The Lions will dominate and run the ball" → Lions spread + D'Andre Swift rushing Over + Game script-dependent props.

Bankroll Management for Parlays

Key Rule: Size parlay bets much smaller than straight bets due to higher variance.
Suggested Sizing:
Bet Type % of Bankroll Reasoning
Straight bet 1-3% Standard unit
2-leg parlay 0.5-1% Higher variance
3-leg parlay 0.25-0.5% Much higher variance
4+ leg parlay 0.1-0.25% Entertainment only

The Bottom Line

Do This
  • Focus 90%+ on straight bets
  • Only parlay correlated outcomes
  • Keep parlays to 2-3 legs
  • Size parlays much smaller
  • Shop for best parlay odds
Avoid This
  • Long-shot parlays for "fun"
  • Parlaying unrelated games
  • Chasing losses with parlays
  • Overconfident "lock" parlays
  • Daily/weekly parlay habits

We measured the correlation excuse

“My legs are correlated” is the standard defense of the cross-game parlay. For legs from different games it is checkable, and for the most popular same-slate combination — a favorite covering and a game going over — it is checkable even within one game. From the bundled nflverse log, across 6,947 completed games with both a spread and total (pushes excluded):

EventRate
Favorite covers48.7%
Game goes over49.5%
Both happen (measured)24.2%
Both happen if independent (48.7% × 49.5%)24.1%

Data: nflverse games.csv bundled with this site (6,947 games with spread and total decisions, 1999–2025), computed by the author.

The ratio of measured to independent is 1.004 — effectively zero correlation, in the very pairing bettors most often describe as connected. Cover/over correlation within a single game rounds to nothing, and legs from different games are independent by construction. The practical rule: unless your legs share a causal engine inside one game (a quarterback's passing yards and his top receiver's receiving yards), assume independence, which means the parlay math on this page applies at full force and the house edge compounds exactly as shown.

The compounding ladder, exact

Modern online books usually price parlays multiplicatively from the leg odds, which sounds fair — and quietly compounds their margin. Exact math with all legs at −110 (decimal 1.909, true leg probability 50% for a fairly-priced coin-flip market):

LegsMultiplicative payoutTrue chance (0.5ⁿ)Expected return per $1House keeps
11.90950.0%$0.9554.5%
23.64425.0%$0.9118.9%
36.95812.5%$0.87013.0%
413.2866.25%$0.83017.0%
525.3633.125%$0.79320.7%

Exact formula math: payout = 1.909ⁿ; expected return = 0.5ⁿ × payout; house keep = 1 − expected return. Books using fixed odds tables (+260 for 2 legs, +600 for 3) keep slightly more.

Each added leg multiplies your expected return by 0.955. That is the entire story of parlays in one number: the same 4.5% toll, paid once per leg, on a road where you only get paid if every tollbooth waves you through. The flip side is honest too — if every leg is genuinely +EV, parlaying multiplies the edge as well as the variance. The catch is the “genuinely.”

Honest limitations

  • Our correlation test covers spread–total pairs only. Player-prop correlations (QB yards with WR yards) are real but cannot be measured from the bundled game log, which has no player data — and books already price SGP legs jointly for exactly that reason.
  • The 50% leg assumption is the friendly case. Historically favorites covered 48.7% and overs hit 49.4%, so naive legs are a shade worse than coin flips after the vig.
  • Payout schedules differ by book. Check whether yours prices multiplicatively or from a fixed table before trusting any parlay EV calculation.

Keep reading: teasers for the one multi-leg format with a measurable historical case, and the EV calculator to price single legs first.

Nothing here is betting advice, and no number on this page predicts any single game. Sports betting is legal only in some jurisdictions and only for adults (21+ in most U.S. states). If betting stops being entertainment, call or text 1-800-GAMBLER. Read our full disclaimer.

Quick Math

At -110 odds per leg:

  • 2 legs: 27.5% win rate
  • 3 legs: 14.4% win rate
  • 4 legs: 7.5% win rate
  • 5 legs: 3.9% win rate
  • 6 legs: 2.0% win rate
Correlated Bets
  • Team spread + Team total Over
  • Favorite ML + Game Over
  • Underdog + QB pass yards Over
  • Favorite + RB rush yards Over