Same Game Parlays (SGPs) are one of the fastest-growing bet types in US sports betting. They are also one of the hardest products for bettors to price correctly, because the legs come from the same game environment and sportsbook pricing models are not transparent.
Our SGP Correlation Calculator helps you estimate whether a Same Game Parlay looks expensive or roughly fair by adjusting the independent-leg probability for positive or negative correlation. It is best used as a screening tool for identifying overpriced or potentially fair SGPs — not as a full bookmaker-model replacement.
SGP Correlation Calculator
Same Game ParlayWhat this tool does and does not do:
- Best for: 2-leg NBA SGP screening.
- Usable for 3 legs as a rough directional check.
- Beyond 3 legs: treat results as approximate only — a single average r across many pairs is a very coarse approximation.
- You supply: fair probabilities per leg and a rough correlation estimate.
- Output: directional pricing check, not exact bookmaker replication.
How to Use the SGP Correlation Calculator
- Add your legs: Enter a description, your estimated fair probability (%), and the individual book odds for each leg. Minimum 2 legs, add more with “+ Add Leg”.
- Select correlation: Choose the average pairwise correlation between your legs. See the reference table below for common NBA scenarios.
- Enter SGP odds offered: The combined American odds the sportsbook offers for the parlay.
- Read the results:
- Fair Odds vs Book Odds: Side-by-side comparison.
- Leg-by-leg breakdown: Each leg’s probability and individual book odds.
- Combined probability: Independent vs. correlation-adjusted.
- Edge and EV: Whether the SGP has positive expected value at the offered price.
- Book margin: How much the sportsbook is charging on this SGP.
Why Correlation Matters in SGPs
In a standard parlay, the combined probability is the product of individual probabilities — assuming the legs are independent. In an SGP, that assumption often breaks down because the outcomes come from the same game.
Positive correlation means the legs tend to win together. Example: Team Over 110.5 points AND Player Over 24.5 points. If the team scores a lot, the player is more likely to score a lot too. The true probability of both hitting is higher than the independent product.
Negative correlation means one leg winning makes the other less likely. Example: Team Under 210.5 total AND Player Over 30.5 points. High individual scoring makes a low total less likely. The true combined probability is lower than independent.
Sportsbook pricing: Modern SGP pricing engines (Kambi, Banach, proprietary models) do account for correlation to some degree, but their adjustments are opaque and often conservative. They may also apply a “correlation penalty” that overcompensates, particularly on strongly correlated legs. The result is that SGP prices are difficult for bettors to evaluate without a reference model.
Common NBA Correlation Estimates
| Leg Pair | Typical r | Direction |
|---|---|---|
| Team Total Over + Player Points Over (same team) | +0.15 to +0.25 | Positive — high-scoring game helps player stats |
| Team Win + Player Points Over (same team) | +0.05 to +0.15 | Weak positive — winning team plays more normal minutes |
| Player Points Over + Player Assists Over (same player) | +0.10 to +0.20 | Positive — active player does both |
| Team Total Over + Game Total Over | +0.40 to +0.60 | Strong positive — same underlying event |
| Player Points Over + Opposing Player Points Over | +0.05 to +0.10 | Weak positive — fast-paced game helps both |
| Team Win + Game Total Under | -0.05 to -0.15 | Weak negative — blowout wins tend to have higher totals |
| Player Rebounds Over + Player Assists Over (guard) | ~0.00 to +0.05 | Near independent for guards |
These are rough practical ranges based on common NBA game-state relationships and historical prop analysis. Use as starting estimates, not fixed truths — individual players and matchups may differ significantly.
Worked Example
SGP: Celtics Over 112.5 (fair: 52%) + Tatum Over 27.5 Points (fair: 48%). Sportsbook offers +220.
- Independent: 0.52 × 0.48 = 24.96%
- With r = 0.15: ≈ 24.96% + 0.15 × sqrt(0.52 × 0.48 × 0.48 × 0.52) ≈ 28.7%
- Fair odds: 1 / 0.287 = 3.48 decimal = +248
- Book offers: +220 (implied 31.2%)
- Edge: 28.7% – 31.2% = -2.5% — no edge, book is charging ~2.5% margin
Even after adjusting upward for positive correlation, the model-implied fair price (+248) is still longer than what the book offers (+220). This suggests the parlay is not attractive at the listed odds based on these inputs.
Limitations
- Simplified correlation model. The calculator uses a first-order Gaussian copula approximation with a single pairwise r for all legs. Real correlations vary by leg pair.
- Fair probabilities are your inputs. If your probability estimates for individual legs are wrong, the combined analysis is wrong.
- Correlation estimates are approximate. Precise correlations require historical game-log data and vary by player, opponent, and pace.
- SGP margins are typically high. Even with correct correlation adjustment, SGPs offered by sportsbooks generally carry higher effective margins than standard straight bets. Finding +EV SGPs is rare.
Frequently Asked Questions (FAQ)
What is correlation in a Same Game Parlay?
Correlation measures how strongly two outcomes are linked. In an SGP, if Leg A winning makes Leg B more likely to win, the legs are positively correlated. Standard parlays assume independence; SGPs should account for correlation, but most sportsbooks do not fully adjust their pricing.
Why are SGPs so profitable for sportsbooks?
Two reasons: SGPs often carry materially higher effective margins than standard straight bets, and correlation mispricing adds another layer of complexity. For positively correlated legs, the true combined probability is higher than the independent product, meaning the fair odds are shorter. Sportsbooks apply their own correlation adjustments (often opaque), plus a margin. The result is a product that is structurally difficult for bettors to beat.
Can I find +EV Same Game Parlays?
Rarely. The margins are large enough that you need a significant edge on multiple individual legs AND correct correlation estimates to overcome the built-in house advantage. The most likely path to +EV SGPs is combining boosted/promoted legs where the sportsbook is offering artificially improved odds.
Does this work for NFL, MLB, or soccer SGPs?
Yes. The correlation model is sport-agnostic. Adjust the correlation estimates based on the sport — for example, NFL QB passing yards and team total are very highly correlated (r ≈ 0.3-0.5), while soccer goals and corners have lower correlation.