Introduction to Correlation & Why It Matters for SGPs
It is the Divisional Playoffs in January 2026. You're scrolling your sportsbook app and you see a Same Game Parlay (SGP) in the Bears-Rams matchup:
- Caleb Williams Over 1.5 Passing TDs (+109)
- Colston Loveland Over 0.5 Receiving TDs (+190)
- SGP payout: +305
Your brain immediately tells a story: if Williams throws multiple touchdowns, Loveland is more likely to score. The payout looks attractive. Then the right questions kick in:
- How much more likely is Loveland to score when Williams is having a big passing day?
- Is +305 fair, or is the book charging extra for the story?
- How can you price this SGP yourself before you click "Place Bet"?
Key Insight
In SGPs, the sportsbook is selling you dependence. If you can measure it and price it, you can decide when the "book story tax" is too high.
What You'll Learn in This Chapter
By the end of this chapter, you will be able to:
- Explain covariance and correlation in plain English and why SGPs depend on them
- Calculate correlation from game logs in Excel and interpret the result
- Price 2-leg SGPs using conditional probability when you have paired history
- Price 2-leg SGPs using a simple r-adjustment model when you do not
- Back out the sportsbook's implied r from an SGP price and compare it to your own assumption
- Extend the logic to 3-leg SGPs using the chain rule and a practical r-based shortcut
- Apply these methods across NFL, NBA, and NHL examples
Tip
The goal is not perfect modeling. It is consistent, explainable pricing that lets you spot when the market's dependence assumptions are wrong.
When Props Move Together (and Why Books Care)
Props in the same game share the same game environment: pace, play-calling, efficiency, injuries, and game script. That shared environment creates dependence.
Two Common Patterns
| Pattern | Description | Example |
|---|---|---|
| Positive dependence | Two outcomes tend to happen together | QB 2+ pass TDs and top pass-catcher TD |
| Negative dependence | One outcome makes the other less likely | QB high pass attempts and WR under receptions |
Sportsbooks do not price an SGP by multiplying the legs and calling it a day. They price in dependence because recreational bettors love the story and will overpay for it.
Key Insight
If the legs share the same story, the book will usually shorten the payout relative to independence.
The Independence Assumption (And Why It's Wrong)
If two events were completely independent, you could simply multiply their probabilities:
Independence Baseline
P(A ∩ B) = P(A) × P(B)=A1*B1For our Williams-Loveland example:
- P(Williams 2+ TDs) ≈ 47.8%
- P(Loveland 1+ TD) ≈ 34.5%
- Independence: 47.8% × 34.5% = 16.5% (fair odds ≈ +506)
But the market is offering +305, which implies a 24.7% probability—much higher than independence would suggest.
Why? Because the book knows these legs are positively correlated. When Williams has a big passing day, Loveland is more likely to score. The book has priced in that dependence.
The Sportsbook's Business Model
Here's the uncomfortable truth: sportsbooks know you love SGPs that tell a compelling story. They price them accordingly.
How Books Think About SGPs
- Calculate the independence baseline (what the parlay would pay if legs were uncorrelated)
- Estimate the correlation between legs
- Adjust the payout down based on positive correlation (or up for negative correlation)
- Add their margin on top
The "story tax" is the premium you pay for betting on narratives that feel obvious.
Warning
Compelling stories are exactly what sportsbooks want you to bet. They price these aggressively because they know recreational bettors love them.
Quick Self-Assessment
Before we dive into the math, test your intuition on these SGP scenarios:
Scenario A: QB + WR Same Team
- Patrick Mahomes Over 2.5 Passing TDs
- Travis Kelce Over 0.5 Receiving TDs
What type of correlation would you expect?
Scenario B: Opposing RBs
- Chiefs RB Over 75.5 Rushing Yards
- Raiders RB Over 75.5 Rushing Yards
What type of correlation would you expect?
Scenario C: Same Player Stats
- LeBron James Over 25.5 Points
- LeBron James Over 7.5 Rebounds
What type of correlation would you expect?
📝 Exercise
For Scenario A (Mahomes TDs + Kelce TD), what correlation would you expect?
The Big Picture
Understanding correlation in SGPs comes down to one simple question:
Are you getting paid fairly for the actual probability, or is the book charging you a story tax?
In the next lesson, we'll learn exactly how to measure correlation using covariance and the correlation coefficient (r)—the mathematical tools that let you answer this question with precision.
Note
Coming Up Next: We'll define covariance and correlation mathematically, learn the Excel formulas to calculate them, and understand how to interpret the results for prop betting.