Why Same-Game Parlays Are -EV: The SGP Math Explained | DMP Learn
Definition
Why Same-Game Parlays Are Usually -EV in sports betting the math behind SGPs, why sportsbooks love them, and where the hidden cost lives.
Think of it this way
An SGP is like ordering a combo meal where the restaurant sets the price. You can't see the cost of each item — just the total. The restaurant bundles items that naturally go together and charges you a premium for the convenience of the "combo."
Why Same-Game Parlays Are Usually -EV
What You'll Learn
Same-game parlays (SGPs) are the fastest-growing product in sports betting — and one of the worst bets most people can make. This lesson explains the math behind SGPs, why sportsbooks love them, and where the hidden cost lives. You'll learn what "correlation" means in this context, how books price it into the parlay, and why the narrative appeal of SGPs is exactly what makes them dangerous.
What Makes SGPs Different from Regular Parlays
A regular parlay combines bets from different games. The outcomes are genuinely independent — whether the Bills cover has nothing to do with whether the Lakers go over their total. The parlay math is straightforward: multiply the individual probabilities.
A same-game parlay combines bets from the same game, and the outcomes are often correlated. If Patrick Mahomes throws for 320 yards, that makes it more likely that Travis Kelce has a big receiving game. If the game goes to overtime, everyone's stats inflate. If one team blows out the other, the losing QB's passing yards might collapse while the winning RB's rushing yards spike.
Correlation means the legs of your SGP aren't independent — and pricing correlated outcomes is where sportsbooks make their money.
The Correlation Problem
Why independent pricing is wrong for SGPs
If you parlay "Mahomes Over 275.5 passing yards" (-110) with "Kelce Over 65.5 receiving yards" (-110), a standard parlay calculator treats them as independent and gives you roughly +264 odds.
But these outcomes are positively correlated. Mahomes going over on passing yards makes Kelce going over on receiving yards more likely (Kelce gets his yards from Mahomes's passing). The true fair odds, accounting for correlation, might be closer to +220.
The sportsbook knows this. They price the SGP at +200 — below even the correlation-adjusted fair value. The gap between the posted +200 and the fair +220 is the SGP tax.
The "SGP tax" quantified
Sportsbooks don't publish how they calculate SGP correlation adjustments. But by comparing posted SGP prices to what the legs would price at independently (and adjusting for known correlation), researchers have found that SGPs typically carry 15-30% total vig — far more than the 4-5% on a standard single bet.
Some books are more aggressive than others. But across the industry, SGPs are consistently the highest-margin product the sportsbook offers.
Why Sportsbooks Push SGPs So Hard
Walk into any sportsbook app and SGPs are front and center. Promotional banners, "popular SGPs," pre-built parlays, boosted SGP odds. This isn't because books want to help you — it's because SGPs are enormously profitable for the house.
Higher vig per bet. A single -110 prop has ~4.5% vig. A 3-leg SGP might carry 20%+ total vig. The more legs, the more vig compounds.
Narrative appeal drives volume. SGPs tell a story: "Mahomes throws 3 TDs, Kelce goes over, and the Chiefs win by a touchdown." That story feels plausible. It feels like a prediction, not a gamble. This narrative framing makes bettors more willing to place bets they'd never make as individual plays.
Correlation pricing is opaque. Most bettors have no way to check whether the posted SGP odds are fair. With a single Over/Under, you can devig and compare across books. With a 4-leg SGP, the correlation adjustment is a black box.
Boosted SGPs are still usually -EV. When a book "boosts" an SGP from +500 to +600, it sounds like free money. But if the fair value is +900, the boosted +600 is still deeply -EV. The boost makes a bad bet slightly less bad — but still bad.
The Math: A Worked Example
Let's walk through a 3-leg NFL SGP to see where the value disappears.
Legs:
- Josh Allen Over 250.5 passing yards (-112)
- Stefon Diggs Over 75.5 receiving yards (-108)
- Bills win (-150)
Step 1: Individual fair probabilities (devigged)
- Allen Over 250.5: 53.8% (devigged from -112/-108)
- Diggs Over 75.5: 52.4% (devigged from -108/-112)
- Bills win: 58.3% (devigged from -150/+130)
Step 2: Independent parlay probability
If these were independent (they're not):
0.538 × 0.524 × 0.583 = 16.4%
Fair odds (independent) = (1 / 0.164) - 1 = +510
Step 3: Correlation adjustment
These legs are positively correlated:
- Allen passing yards → Diggs receiving yards (Diggs catches Allen's passes)
- Allen passing yards → Bills winning (good QB play = more wins)
- Diggs receiving yards → Bills winning (big receiver game = more wins)
With positive correlation, the combined probability is higher than the independent calculation suggests. Let's estimate the correlated probability at 19.5%.
Fair odds (correlated) = (1 / 0.195) - 1 = +413
Step 4: What the sportsbook posts
The book prices this SGP at +350.
Step 5: The real vig
Fair odds: +413 → fair implied probability: 19.5%
Posted odds: +350 → implied probability: 22.2%
Vig = 22.2% - 19.5% = 2.7 percentage points
As a percentage of fair value: 2.7 / 19.5 = 13.8% vig
Compare that to 4.5% vig on a standard single bet. The SGP vig is 3x higher.
And this is a simple 3-leg SGP. Add a 4th or 5th leg and the vig compounds further. 4-5 leg SGPs commonly carry 25-35% total vig.
The Narrative Trap
Here's why SGPs are so seductive and so dangerous.
When you build an SGP, you're constructing a story about how you think the game will unfold. "The Chiefs will control the game, Mahomes will be efficient, Kelce will be the primary target, and they'll win by a comfortable margin." That narrative feels coherent and probable.
But each additional leg doesn't just add a little risk — it multiplies the improbability. Even if each individual leg has a 55% chance, a 4-leg parlay has:
0.55 × 0.55 × 0.55 × 0.55 = 9.2% hit rate
Nine percent. A bet that hits less than once in ten attempts. And that's assuming each leg has genuine 55% probability — which most casual SGP builders are overstating because the narrative feels so convincing.
The appeal of a big payout ("+800!") obscures the fact that the vig-adjusted probability is even lower than the raw math suggests. The sportsbook is selling you a lottery ticket at inflated prices — and the narrative is the marketing.
When SGPs Can Make Sense (Rarely)
SGPs aren't always terrible. There are narrow circumstances where they can be neutral or even positive EV:
Negatively correlated legs. If two outcomes in your SGP are negatively correlated (one going over makes the other going under more likely), the book's correlation adjustment might undercharge you. Example: "QB1 Over passing yards + RB1 Under rushing yards" in a game projected to be pass-heavy. These outcomes support each other, but some pricing algorithms treat them as more independent than they are.
Significant boosts on low-leg SGPs. A 2-leg SGP that's boosted by 30%+ might occasionally cross into +EV territory. But verify by devigging the individual legs and estimating the correlation.
Exploiting slow book adjustments. If news breaks (key defender out, weather change) that affects the game but the SGP pricing hasn't updated, the entire SGP might be mispriced before the book recalculates.
These are edge cases, not strategies. For the vast majority of SGPs on any given day, the vig is simply too high to overcome.
How to Evaluate an SGP
If you're going to bet SGPs, at least check the math:
- Devig each leg independently — calculate the fair probability of each individual bet
- Assess the correlation direction — are these legs positively or negatively correlated?
- Estimate the combined fair probability — for positively correlated legs, the true probability is higher than the independent product
- Compare to the posted SGP odds — convert the posted odds to implied probability and see how much vig the book is embedding
- Decision: If the vig is under 10%, it might be tolerable for entertainment. If it's above 15%, you're donating money.
How DMP Handles This
DMP uses a Gaussian Copula Monte Carlo simulation (10,000 samples) to model correlation between SGP legs in its Slips product. This means DMP's fair probability estimates for multi-leg entries account for the statistical relationship between legs — not just the independent probabilities.
When DMP recommends a multi-leg Slip, the EV calculation already factors in correlation. This is fundamentally different from most SGP analysis, which ignores correlation entirely and overstates the expected value.
DMP's fully customizable Slips also let you remove legs, swap markets, and see how each change affects the overall EV. This transparency is the opposite of the black-box SGP pricing most sportsbooks use.
Key Takeaways
- SGPs carry 15-30% total vig — 3-6x more than standard single bets.
- Sportsbooks price in correlation adjustments that are invisible to most bettors. The posted SGP odds are not just the individual legs multiplied together.
- Narrative appeal is the trap. A story about how a game will unfold feels compelling but obscures the compounding improbability and inflated vig.
- Most "boosted" SGPs are still -EV. A boost makes a bad bet less bad, not good.
- If you bet SGPs, devig each leg, estimate correlation, and calculate the true vig before deciding. Most of the time, you'll find the cost is too high.
- For consistent profitability, single-bet +EV props beat SGPs overwhelmingly over time.
How DMP uses this
DMP uses Gaussian Copula Monte Carlo simulation to model correlation between legs. When DMP shows a multi-leg Slip as +EV, the correlation cost is already factored in — unlike sportsbook SGPs where the pricing is a black box.
Common mistake
Building SGPs based on a narrative ("the Chiefs will dominate") instead of checking the math. A story that feels 70% likely is often only 15-20% likely when you multiply the actual leg probabilities and add the vig.
After this lesson
You understand why SGPs carry 15-30% vig, how correlation pricing works against you, and why single-bet +EV props are more profitable than parlaying correlated legs.