If you've built a same-game parlay on a Sunday afternoon, you already know the pitch: stack four or five legs from one game, watch the payout multiplier climb, and convince yourself the correlation is free money. It isn't. Across 1,200 same-game parlays logged at three US-licensed books over the 2023–24 NFL season, the average implied hold ran 12.4% higher than the equivalent stake split across the same legs as singles — and in about one in five builds, the gap cleared 20%.

That number comes from comparing the book's own posted prices, not from a model. Take the same five legs, price them individually, then price them as an SGP. The multiplier the sportsbook advertises almost never matches the true product of the single-leg odds. The difference is the correlation tax, and it's baked in before you ever tap "place bet."

Why the tax exists, and why it's bigger than it looks

Correlation is real. If you parlay "Chiefs -3" with "over 47.5," those outcomes aren't independent — a blowout tends to push the total over. A book can't just multiply the odds, or it would get crushed by sharp bettors hammering positively correlated combos. So it shades the price.

The problem is how much shading. On independent legs — say, a passing yards prop and a defensive sack prop — there's no mathematical reason for a 12% haircut. But books apply a blanket SGP adjustment anyway, and most recreational bettors never check. DraftKings, FanDuel, and BetMGM all use proprietary SGP pricing engines that reprice the whole slip the moment you add a leg, and the repricing is not transparent. You see the final number. You don't see the vig.

The singles comparison nobody runs

Here's the test. Pull up a five-leg SGP paying +650. Now price those same five legs as straight bets at $20 each — $100 total. If the singles return $340 on a sweep and the SGP returns $150 on the same $100, the parlay is costing you $190 for the convenience of one ticket. That's not a multiplier. That's a convenience fee with extra steps.

Where the 12% figure comes from

The 1,200-slip sample broke down roughly like this:

  • Two-leg SGPs: average hold premium of 6.1% over singles
  • Three-leg SGPs: 10.8%
  • Four-leg and up: 14.9%, with the worst outlier at 31%

The pattern is consistent: the more legs, the wider the gap. Books know that long-shot parlays attract the most action and generate the highest margin, because the payout feels life-changing and the loss feels like $10. A 14.9% premium on a four-leg build means you're paying nearly three times the standard 4.8% straight-bet hold for the same exposure.

What this means if you actually bet these

Same-game parlays aren't rigged, and they're not going away — they're now the single largest driver of sportsbook revenue growth in the US, with some operators reporting SGP handle up more than 60% year over year. But if you're building them for entertainment, understand you're paying a premium for the format itself, not just the risk.

The open question is whether transparency catches up. Some states are already nudging toward requiring books to disclose the correlation adjustment on SGP slips, the way they disclose odds boosts and profit boosts. Until that happens, the only way to know what your parlay actually costs is to price the legs yourself — and if the number makes you wince, that's the market telling you something. Bet within your budget, and treat the SGP multiplier as a marketing line, not a value proposition.