Same-Game Parlays Explained: Correlation, Risk, and the Real Payout Math
A same-game parlay lets you combine a quarterback's passing yards, a receiver's anytime touchdown, and the game's total points into a single bet from a single matchup — and the app will hand you odds for the combination instantly. Those odds look like they came from multiplying the three individual prices together, the same way a traditional multi-game parlay works. Mostly, they didn't. Because the legs of an SGP all depend on the same game, they're statistically linked to each other in ways that plain multiplication ignores, and pricing that correctly is one of the harder problems modern sportsbooks have had to solve.
In this guide
What makes a parlay a "same-game" parlay
A standard parlay links bets from different, unrelated games — a Bills moneyline, a Lakers total, a Yankees run line — where each outcome has essentially nothing to do with the others. A same-game parlay (SGP) instead combines multiple bets from a single contest: a team total, a player prop, the game's over/under, a first-half result, whatever the book's SGP builder offers. That single shared game is exactly what makes the pricing problem different, because outcomes inside one game routinely move together instead of independently.
Sportsbooks didn't always allow this. For years, combining same-game outcomes into one bet was considered too hard to price safely and was simply not offered. What changed is that books built simulation-based and correlation-aware pricing engines — running the game forward thousands of times internally under different scenarios — that estimate how often combinations of outcomes actually occur together, rather than assuming they're unrelated.
Why independent-odds math breaks down inside one game
The standard way to price a parlay of independent events is to convert each leg to a decimal price and multiply them together. Two -110 bets (decimal 1.91 each) become a 3.65 parlay if you just multiply — implying roughly 27.4% probability of both hitting. That math is correct only if the two events don't influence each other at all.
Inside a single game, that assumption rarely holds. A quarterback throwing for a lot of yards and his team scoring a lot of points aren't independent events — they tend to happen together, since one is often a direct driver of the other. If a book priced that parlay using plain multiplication as though the two legs were unrelated, it would be handing sharp bettors a mathematically guaranteed edge, because the true joint probability of both things happening is higher than the product of their individual probabilities suggests. That gap between naive multiplied odds and true correlated probability is the entire reason SGP pricing needed its own approach.
A worked example: two positively correlated legs
Say a sportsbook offers a quarterback Over 250.5 passing yards at -115 (implied probability ~53.5%) and that same team's Over 47.5 total points at -110 (implied probability ~52.4%) in the same game. Multiplying those two decimal prices (1.87 × 1.91) gives a naive combined decimal of about 3.57, implying roughly 28% probability — treating the two outcomes as if they had nothing to do with each other.
| Pricing method | Implied joint probability | What it assumes |
|---|---|---|
| Naive multiplication of both legs | ~28% | The two outcomes are unrelated |
| Correlation-adjusted book pricing | Meaningfully higher, illustratively ~34-36% | High passing yardage and a high team total tend to occur together |
In reality, games where a quarterback clears 250 yards are disproportionately likely to also be games where the team's total climbs past 47.5, since the yardage is often what produces the points. A book that recognizes this shortens the payout on that combination relative to what plain multiplication would suggest, pricing it closer to the true, higher joint probability instead of the artificially low naive figure. That's the "correlation adjustment," and it's the main reason SGP payouts on obviously connected legs look noticeably smaller than multiplying the individual prices would produce.
Negative correlation: when legs work against each other
Correlation cuts the other way too, and it's easy to miss. Consider parlaying a team's Over on team total points with their lead running back going Under on rushing yards. If that team wins by pulling ahead and passes more to build a big total, the running back plausibly sees fewer rushing attempts late — the two outcomes can actually work against each other rather than reinforcing each other. When legs are negatively correlated like this, the true joint probability of both hitting is lower than plain multiplication implies, meaning the "naive" parlay math actually overstates your chances more than it does with independent legs.
This is also why some sportsbooks restrict which same-game combinations can be parlayed at all, rather than pricing every possible combination. Building an accurate correlation model for every conceivable pairing across every sport is genuinely difficult, and a book that gets it wrong in either direction either bleeds value to sharp bettors or offers a parlay that's a worse bet than it appears — neither of which the book wants, which is part of why SGP menus are often narrower than the full list of props available on a game.
The vig still compounds, correlation or not
Correlation pricing changes how a book estimates the true joint probability of your legs hitting together, but it doesn't remove the book's margin — vig is still baked into every individual leg before correlation is even considered, and that margin compounds across legs the same way it does in a standard multi-game parlay. Add a third or fourth leg to a same-game parlay and the effective hold on the combined bet climbs quickly, even before accounting for how well or poorly the correlation was modeled.
| Legs in the parlay | Effect on effective sportsbook margin |
|---|---|
| 1 leg | Standard single-bet vig (e.g., ~4.5% at -110) |
| 2 legs | Vig compounds across both legs before any correlation math is applied |
| 4+ legs | Effective hold often climbs into the double digits, on top of correlation pricing |
See our guide to how vig actually works for the underlying mechanism — everything in that guide about compounding margin across legs applies to SGPs on top of the correlation adjustment described here, not instead of it.
Common same-game parlay mistakes
- Assuming a big combined payout means a mispriced book. A large number on an SGP slip usually reflects vig compounding across several legs, not an oversight in the book's model — books have gotten considerably better at correlation pricing since SGPs launched broadly.
- Stacking legs that all depend on the same outcome. A QB Over on yards, his team's Over on points, and his top receiver's Over on receiving yards are all pulling from the same underlying scenario — that's not diversification, it's the same bet expressed three different ways, and the book has almost certainly already priced that overlap in.
- Ignoring how few legs it takes to erode value. Even with fair correlation pricing, every added leg still carries its own slice of vig. A 2-leg SGP with well-modeled correlation can be a reasonable bet; a 6-leg SGP rarely is, no matter how connected the outcomes feel.
None of this means SGPs are a bad product — they're genuinely useful for expressing a specific view of how a single game will play out, and modern correlation pricing is far more sophisticated than the early, crudely-priced versions of these bets. It does mean the payout on your slip reflects a real (if imperfect) probability model, not a shortcut. For the mechanics of vig itself, see our guide above, and for how parlay structure compares to a hedged alternative, see our closing line value guide on tracking whether your bets are beating the market over time.
Frequently asked questions
Are same-game parlays always worse value than regular parlays?
Not inherently — it depends on how well the book's correlation model matches reality and how many legs you add. A well-priced 2-leg SGP with genuine correlation can be reasonably fair; the bigger risk is that adding legs compounds vig regardless of correlation quality, the same way it does in any parlay.
How do sportsbooks actually calculate correlation for SGPs?
Most major books use simulation-based models that run a game forward many times under varying conditions to estimate how often combinations of outcomes occur together, rather than a simple formula. The exact methodology is proprietary and varies by operator, which is part of why the same SGP can be priced slightly differently at different books.
Can I build an SGP with legs from different games?
Some sportsbooks now offer an "SGP+" or cross-game parlay product that lets you combine same-game legs from one matchup with legs from a separate game in a single bet. The same-game portion still uses correlation pricing; the legs from the other game are priced independently, the same as a traditional parlay.
Why did my SGP payout change between when I built it and when I placed it?
SGP pricing updates in real time as the underlying individual odds move and as the book's correlation model recalculates, so a payout you saw a few minutes earlier can shift before you confirm the bet — similar to how any live odds can move, just applied across multiple linked legs at once.