What Is the SGP Tax? The Correlation Tax on Same-Game Parlays, Explained

The SGP tax (also called the correlation tax) is the gap between what your same-game parlay would pay if the legs were independent and the shorter payout the sportsbook actually quotes you. It exists because legs from the same game rise and fall together, and the book reprices that connection, plus a cushion on top. Three -110 legs multiply out to roughly +596 on paper; the same three legs in an SGP often come back +350 to +400. That missing chunk is the tax.

Here’s the moment most people meet it. Your buddy builds a three-legger inside one game: Matthew Stafford over on passing yards, Puka Nacua over on receptions, the Rams to cover. He does the napkin math on the way to lock it, three coin-flippy legs, should be pushing +600. The book shows +375. He assumes the app glitched.

It didn’t glitch. That’s the tax, and once you can see it, you can’t unsee it.

Why does the SGP tax exist?

One word: correlation. Legs from the same game aren’t separate bets that happen to share a slip. They live inside the same three hours, the same score, the same game script. If Stafford clears his passing yards, there’s a decent chance Puka ate too, and a decent chance the Rams covered doing it. The legs lean on each other.

A normal cross-game parlay gets priced the honest way: multiply the odds of each leg, because a day game in Phoenix has nothing to do with a night game in Boston. A same-game parlay breaks that assumption. When your legs pull in the same direction, the true chance of all of them hitting is higher than the naive multiplication says. So the book pays you less than the multiplication. Some of that shave is legitimate repricing. Some of it is cushion the book keeps for itself.

You don’t get to see which is which. That’s the whole problem.

The SGP tax has two layers: an honest repricing of legs that genuinely hit together more often, and a margin cushion on top. The book shows you one number and never itemizes the receipt.

The napkin math: see the tax in 60 seconds

You can expose the tax on any slip without a calculator site. Three steps:

1. Convert each leg to an implied probability. For a minus line, drop the sign and divide the number by itself-plus-100: a -110 leg is 110 ÷ 210, about 52.4%. For a plus line, it’s 100 ÷ (odds + 100).
2. Multiply the probabilities together, then convert back to American odds. That’s the independent, “fair multiplication” price.
3. Compare it to the book’s SGP quote. The distance between the two is the tax on your slip.

Here’s the whole thing worked out slow, on your buddy’s slip from the top of the page. Follow it once with a real slip in front of you and you’ll never need it written down again:

Same slip, summarized:

And the tax scales with the slip. On a single straight bet, the book’s built-in edge (the vig) runs roughly 4 to 5%. On same-game parlays, the effective hold commonly runs 15 to 25%. Same bettor, same game, three to five times the house edge, and none of it printed on the ticket.

Positive vs. negative correlation: the tax cuts both ways

Legs that move together (your Stafford over + your Puka over + the Rams covering) get the payout trimmed, like we just saw. That one at least has logic you can follow.

The sneakier version is legs that fight each other. Pair a team’s under with one of that team’s players going over, and you’ve built a combo where winning one leg actively works against the other. The book usually pays that closer to full price, which sounds generous until you realize you’re being paid standard odds for a parlay that’s harder than standard to hit.

Real numbers make it concrete: one widely shared industry analysis found a QB completions over hitting 40.2% of the time on its own, but only 19.9% jointly with “team covers.” Nothing about the leg changed. It’s just that a combo pays on the joint number, and the joint is always a much smaller number than the leg you fell in love with. Straight multiplication would predict about 21% for that pair; whether the true joint lands above or below that line is the correlation question, and it’s the single most invisible thing on a betting slip. (We go deeper on this in same-game parlay correlation, explained.)

The books don’t even agree with each other

Here’s the part the glossary sites skip. Correlation pricing isn’t a solved equation the industry shares. It’s each book’s in-house guess, and the guesses differ, a lot. The identical same-game parlay has been quoted +369 at one book and +580 at another. Same legs, same game, a 200-point payout gap.

Two takeaways from that. First: if books can disagree by 200 points, the quote you’re looking at is an opinion, not physics. Second: shopping the number matters more on SGPs than on almost any other bet. Simple, right? Check two books instead of one. ...Yeah, almost nobody does it.

Can you avoid the SGP tax?

Not entirely, and honestly, you shouldn’t expect to. When your legs genuinely help each other, a trimmed payout is partly fair: you picked outcomes that travel together, and the book noticed. What you can do is stop paying the tax blind:

• Run the napkin math before you lock. If the fair multiplication says +596 and the quote says +350, you know exactly how big the gap is.
• Shop the quote. The +369 vs. +580 spread is real money for the same slip.
• Know how your legs relate. Same-direction legs mean a trimmed payout. Opposing legs mean full price for a harder bet. Either way, you want to know before the slip is locked, not after.
• Make the slip earn it. With a 15–25% hold working against you, an SGP is guilty until proven interesting.

That third one is where we come in. Drop a slip into Slip Check and ET reads it leg by leg, in plain English: which legs move together, which fight each other, and which are quietly overpriced. No auto-built parlays, no “lock of the day.” Options, not parlays: ET hands you the read, you make every call.

One honest caveat, because sharp readers always ask: how would anyone outside the book’s pricing room know the true correlation number? They wouldn’t, exactly. Even the books disagree, that’s the +369 vs. +580 story. What can be done is measure how often legs have actually traveled together across past seasons, and let those receipts do the talking. That’s the standard we hold ourselves to, and the standard you should hold anyone to. Anyone who instead promises you a winning parlay is lying to you.

The usual word from us: this is research and education, not a promise of profit. 21+, and only ever bet what you can walk away from.

FAQ

What is the correlation tax in betting?

It’s the same thing as the SGP tax: the difference between a same-game parlay’s fair independent-math payout and the shorter payout the book quotes. It exists because same-game legs are correlated, so the book reprices the combination and adds margin on top.

Is the SGP tax the same at every sportsbook?

No. Correlation pricing is each book’s internal estimate, and they differ widely — the identical same-game parlay has priced +369 at one book and +580 at another. That’s why shopping the quote matters more on SGPs than on straight bets.

How do I calculate the SGP tax on my own slip?

Convert each leg to an implied probability, multiply the probabilities together, convert the result back to American odds, and compare that number to the book’s SGP quote. The gap between them is the tax on your specific slip.

Are same-game parlays always a bad bet?

Not automatically, but the default lean is against them: effective hold on SGPs commonly runs 15–25% versus roughly 4–5% on a straight bet. A same-game parlay has to clear a much higher bar to be worth it, so the burden of proof is on the slip, not on you to find a reason to believe.

Does the SGP tax mean the book is cheating?

No. Positively correlated legs genuinely do hit together more often than independent math suggests, so part of the shave is honest repricing. The catch is the extra cushion stacked on top and the fact that the book never shows you the breakdown — which is why running the math yourself matters.