Same-Game Parlays: The Math

A three-leg parlay of standard -110 bets carries roughly triple the hold of a single wager. A same-game parlay carries more. This is the fastest-growing market in American betting, and the growth is not an accident.

Same-game parlays did not exist in American sportsbooks a decade ago. Today they are heavily promoted, prominently placed, and a substantial share of operator revenue. Understanding why requires understanding the arithmetic, which almost no one placing them has done.

This post is not an argument that you should never place one. It is an argument that you should know what you are buying. The math is not complicated, it is simply never displayed.

01Start With the Ordinary Parlay

Before same-game, understand the standard version, because the same-game problem is this problem plus another layer.

A parlay multiplies the odds of independent legs. Three bets at -110 each have decimal odds of 1.909. Multiplied: 1.909 cubed equals 6.96, which is about +596.

Now compute what it should pay. Each -110 leg, devigged, is a genuine coin flip at 50 percent. Three independent coin flips hit 12.5 percent of the time. Fair odds on a 12.5 percent event are 8.00 decimal, or +700.

You are offered 6.96 on something worth 8.00. The hold is 1 minus 6.96 divided by 8.00, or 13 percent.

The progression, all using -110 legs:

The mechanism is that margin compounds multiplicatively rather than adding. Each leg's vig gets multiplied by the next leg's vig. Nothing dishonest is happening. It is simply what multiplication does.

02What Makes a Same-Game Parlay Different

The ordinary parlay math assumes independence. Legs from different games generally are independent: a Lakers result tells you nothing about a Yankees result.

Legs from the same game are not independent. They are correlated, sometimes strongly, and correlation breaks the multiplication.

Positive correlation. A quarterback going over his passing yards total and his top receiver going over his receiving yards total. If one happens, the other is substantially more likely. Multiplying the individual probabilities badly understates the combined chance.

Negative correlation. A team winning outright and the opposing running back going over a rushing total. Teams that fall behind abandon the run. If one happens, the other is less likely.

Before books offered same-game parlays, they simply prohibited correlated combinations, because a bettor stacking positively correlated legs at independent pricing would have a genuine and obvious edge.

03How Books Price Correlation

Modern same-game parlay engines model the joint probability of the selected legs rather than multiplying them. Conceptually, the book simulates the game many times and counts how often all of your legs hit together.

The effect on pricing runs in both directions.

Positively correlated legs pay less than naive multiplication implies. Stacking a quarterback's passing yards with his receiver's receiving yards produces a price noticeably shorter than multiplying the two individual prices.

Negatively correlated legs pay more. Combining a team moneyline with the opposing running back going over produces a longer price than multiplication suggests, because the book knows those outcomes fight each other.

The pricing is doing something genuinely sophisticated. The issue is not that the correlation model is wrong. It is what happens to the margin on top of it.

04Why the Hold Is Higher

Same-game parlays typically carry a larger hold than standard parlays with the same number of legs, and there are structural reasons.

• The base markets are softer. Player props carry wider margins than sides and totals to begin with, and most same-game legs are props.
• Correlation modeling carries uncertainty, and books price uncertainty conservatively, in their favor.
• The product is not comparable. You can shop a spread across five books in twenty seconds. You cannot meaningfully compare a five-leg same-game parlay, because no other book offers that exact combination. Line shopping, the highest-return habit from Post 77, does not function here at all.
• Demand is inelastic. Customers building these are not price-sensitive, and operators price accordingly.

That third point deserves emphasis, because it is the quiet one. Every other market in this series can be shopped. Same-game parlays are the one product where you are structurally unable to check whether the price is fair.

05Where Correlation Might Be Underpriced

In fairness, the honest case for same-game parlays.

Correlation models are built on historical data and general relationships. They handle common combinations well, because those combinations occur constantly and there is abundant data. They handle unusual combinations more crudely.

The theoretical opportunities:

• Game-script correlations the model treats generically. A specific matchup where you have a strong view about how the game unfolds, involving legs whose relationship the model prices with a league-average assumption.
• Situational factors outside the model. Weather, a specific defensive scheme, or a rotation change that alters the relationship between two outcomes.
• Unusual combinations that occur rarely enough that the model has thin data.

The honest caveat: these edges must overcome a hold that is often 15 to 25 percent. Finding a correlation the model missed is not enough. You must find one it missed by more than the margin, which is a substantially higher bar than any other market in this series asks you to clear.

06Boosted Same-Game Parlays

This is where the two subjects genuinely meet, and it is the one context where these tickets can make sense.

Books heavily promote same-game parlays with profit boosts and odds boosts, for the obvious reason that the underlying product is highly profitable and a promotion on it costs less than it appears to.

Applying the framework from Post 82:

A 50 percent profit boost on a ticket carrying 20 percent hold does not produce a 50 percent edge. It reduces the effective hold substantially, and depending on the specific numbers it can push the ticket toward neutral or slightly positive.

The practical rule: a boosted same-game parlay can be a reasonable use of a promotion. An unboosted one rarely is. If you are going to place these, place them with a boost attached, and recognize that the boost is doing the work.

07Round Robins and Combination Tickets

A round robin breaks selections into multiple smaller parlays. Six selections might become fifteen two-leg parlays, or twenty three-leg parlays.

The pitch is that you do not need everything to hit. The arithmetic is less friendly: you are placing many parlays, each carrying compounded hold, and the total margin you are paying goes up with the number of tickets rather than down.

A round robin reduces variance relative to a single large parlay. It does not reduce the house edge, and it usually increases total exposure because bettors size each small ticket as though it were a single bet.

08Correlation Examples Worth Knowing

The last row is a reminder that correlation is sport-specific and situational. In basketball, garbage time removes starters, so a blowout can suppress a star's totals. In football, a trailing team throws more, which can inflate passing volume. The direction reverses depending on the sport and the role.

09The Honest Assessment

The case against. Hold routinely runs 15 to 25 percent or higher. You cannot shop the price. The underlying markets are the softest on the board. And overcoming that margin requires an edge far larger than anything else in this series demands.

The case for. They are enjoyable, they turn a single game into sustained interest, and with a boost attached the arithmetic becomes defensible. If entertainment is part of why you bet, this is a legitimate place to spend some of that budget.

The framing that works, and it is the same one we applied to parlays in Post 1: budget them separately from your handicapping bankroll. A recreational allocation, sized as entertainment, tracked separately, and not confused with the operation described in the previous eighty-five posts.

What does not work is placing them as though they were bets, then wondering why a year of solid handicapping produced no profit. A bettor grinding out 3 percent on sides while giving back 20 percent on a weekly same-game parlay is running two businesses, and the second one is much larger than they think.

10Common Same-Game Parlay Mistakes

• Not knowing the hold. The bet slip shows the payout and never the margin.
• Adding legs for the payout. Every leg multiplies the margin along with the return.
• Assuming correlation is free. The book priced it. That is the entire product.
• Placing them unboosted. The promotion is what makes the arithmetic tolerable.
• Treating a round robin as risk reduction. It lowers variance, not the edge, and usually raises total exposure.
• Mixing them into the handicapping bankroll. Two activities, two budgets.
• Chasing a near-miss. A ticket that lost on one leg was not close in any meaningful sense.
• Forgetting they cannot be shopped. Your main protection does not exist here.

11The Bigger Picture

Same-game parlays are the clearest illustration in modern betting of a product designed around customer psychology rather than customer outcomes.

The appeal is real and worth acknowledging honestly. A same-game parlay makes a single game absorbing for three hours. It offers a large payout from a small stake. It rewards knowing things about individual players. Everything about the experience is well designed.

What is also true is that the design includes a margin several times larger than a straight bet, delivered through a product that cannot be price-compared. Both statements are true simultaneously, and a bettor who understands both can make an informed decision about how much of their activity belongs there.

That is really all this post is asking. Not abstinence. Information. The bet slip will tell you what you can win, and this post tells you the part it leaves out.