Parlay Math: Why Multi-Leg Bets Cost More Than You Think

How Parlays Work

A parlay is a single wager that ties two or more individual selections together. All legs must win for the parlay to pay out. If any leg loses or pushes (in most rulebooks), the entire bet loses.

The appeal is straightforward: parlay payouts grow exponentially as you add legs, because each leg's odds multiply against the previous total. A bettor who can pick five winners from five games stands to collect a return that no straight bet could match.

The mechanics are simple. Convert each American odds line to its decimal equivalent, multiply those decimals together, and convert back to American odds. Here is the formula for converting American odds to decimal:

• For a favorite (negative American odds): Decimal = 1 + (100 / |American odds|)
• For an underdog (positive American odds): Decimal = 1 + (American odds / 100)

Example: A 3-leg parlay, all at -110

Each -110 line converts to a decimal of 1 + (100/110) = 1.9091.

Multiply three legs together: 1.9091 x 1.9091 x 1.9091 = 6.958.

Convert back to American odds: (6.958 - 1) x 100 = +596 (rounded).

So a $100 bet on this parlay returns $696 total ($596 profit). That sounds impressive. The problem is what the same bet would pay at fair odds, which is the subject of the next section.

For a full breakdown of how American, decimal, and fractional odds interact, see Sports Betting Odds Explained.

The Compounding Vig Problem

Every -110 line embeds a vigorish (vig or juice) of approximately 4.5% on a single bet. When you combine multiple -110 lines into a parlay, that vig does not simply add up. It compounds.

To understand why, you need to separate the "true" fair-odds probability from the implied probability the book is charging you.

A -110 line implies a probability of: 110 / (110 + 100) = 52.38%.

But in a 50/50 event (like a point spread), the true probability is exactly 50%. The gap between 52.38% and 50% is the vig the book collects on that leg.

When you chain multiple legs together, each with their own 52.38% implied probability instead of the true 50%, the compounding effect accelerates the gap between what you receive and what fair value would pay.

Two-leg parlay at -110 / -110

Fair combined probability: 0.50 x 0.50 = 25.00%. Fair decimal odds: 4.000. Fair American odds: +300.

Book's implied combined probability: 0.5238 x 0.5238 = 27.44%. Book decimal odds: 3.6418. Book American odds: approximately +264.

The book pays +264. Fair value is +300. You are giving up 36 points of juice on a single two-leg parlay.

Five-leg parlay at -110 each leg

Fair combined probability: 0.50^5 = 3.125%. Fair decimal odds: 32.00. Fair American odds: +3100.

Book's implied combined probability: 0.5238^5 = 3.992%. Book decimal odds: 25.05. Book American odds: approximately +2405.

The difference between +3100 and +2405 is enormous. On a $100 wager, you are foregoing $695 in expected value compared to a theoretical no-vig market. The more legs you stack, the wider this gap becomes.

True Odds vs Sportsbook Odds

The table below shows the divergence between fair payout and typical sportsbook payout as parlay legs increase. All calculations assume each leg is priced at -110 on a 50/50 outcome (standard point spreads or totals).

The "Juice Lost" column is the expected monetary cost of the vig on a $100 wager, not an accounting of any single bet outcome. It represents the long-run disadvantage per $100 staked on that parlay size.

Parlay House Edge by Number of Legs

Another way to frame this is through effective house edge. On a single -110 straight bet, the house edge is approximately 4.55%. As you stack legs, the effective edge against you increases dramatically.

The effective house edge is calculated as: 1 - (Book payout decimal / Fair payout decimal). The "Comparable Casino Game" column is included for perspective only. By the time you reach a 5-leg parlay, you are operating in territory closer to keno than to sports betting.

Crypto sportsbooks sometimes offer reduced-juice lines (e.g., -108 or -105), which helps somewhat. But the compounding effect means even a small per-leg vig creates a substantial cumulative disadvantage on multi-leg parlays. For a look at which platforms offer the best lines, see Best Crypto Sportsbooks and the full comparison at /best/sports-betting.

Same-Game Parlays

Same-game parlays (SGPs) allow bettors to combine outcomes from a single game into one parlay. For example: Team A moneyline + Team A first-half spread + Player X anytime touchdown scorer.

SGPs have become one of the fastest-growing product categories at sportsbooks, and for good reason. From the book's perspective, they are an exceptionally profitable offering.

Here is the core problem with SGPs: the legs within a single game are correlated. If Team A wins by a lot, they are more likely to cover the spread. A team winning is correlated with their passing game working, which is correlated with quarterback yardage props. These outcomes are not independent.

When you parlay independent events, you multiply their probabilities together. When you parlay correlated events, the math changes significantly because the outcomes are related.

The sportsbook's standard parlay engine prices legs as if they are independent. On an SGP, the book then applies an internal correlation adjustment (a markup) to its favor to account for the risk it is absorbing from the correlated outcomes. This markup is not transparent; it is embedded in the offered payout.

The result: SGPs consistently pay out at a lower rate than standard parlays built from independent events across different games, which are already significantly below fair value. Independent research and tracking from sharp betting communities consistently finds SGP payouts running 3 to 7 percentage points worse than equivalent standard parlays.

If you are building SGPs at Stake or any other sportsbook for the "big payout" potential, be aware that you are accepting some of the worst expected value available in any sports wagering product.

Correlated Parlays

Not all correlated parlays are bad for the bettor. If two outcomes are positively correlated, and the book prices them independently (not adjusting for the correlation), the parlay can carry positive expected value.

The most commonly cited example is the team total over combined with the team's moneyline. If a team is expected to score a lot of points, they are also more likely to win the game outright. These two outcomes are meaningfully correlated. A sharp bettor who identifies a game where a team is significantly undervalued on the moneyline and also likely to put up points might find that combining the moneyline and team total over creates a correlated parlay edge.

The mechanics of why this works: the true combined probability of (win AND score over X) is higher than (P(win) x P(score over X)) because they are not independent. If the book calculates the parlay payout by simply multiplying independent implied probabilities, it is underpricing the correlated parlay from a probability standpoint, and the bettor's expected value improves.

In practice, exploiting this edge is extremely difficult for three reasons:

1. Most major sportsbooks now restrict or limit correlated parlays on the most obvious pairings. A same-team moneyline plus team total over is frequently flagged and either blocked or manually repriced.
2. Finding games where you have a genuine edge on both legs simultaneously is hard. You need to beat the book's line on leg one AND leg two AND correctly assess the correlation. Each additional requirement narrows your opportunity set.
3. Even where the edge exists on paper, limits on correlated parlays at sharp books mean your action size is capped at levels that may not be worth the research time invested.

Correlated parlay hunting is a legitimate area of study for advanced bettors, but it is not a practical strategy for recreational bettors and it does not rehabilitate standard parlays as a general product.

When Parlays Make Sense

There are two narrow scenarios where parlays can be rationally justified.

Scenario 1: You have identified a genuine correlated edge the book has not priced.

As described above, this is the one mathematical case where parlays can carry positive expected value. The conditions are strict: you need a real edge on multiple correlated legs, the book must not have adjusted for the correlation, and you need access to bet the correlated parlay without restrictions. This describes a very small number of bettors in a very small number of situations.

Scenario 2: The entertainment value is worth the cost to you personally.

This is an honest framing, not a mathematical one. If someone wants to stake $20 on a 10-leg parlay because watching ten games become emotionally connected enhances their enjoyment of the weekend, that is a legitimate use of their entertainment budget. The math is terrible, but entertainment spending does not need to optimize for expected value.

The critical point in Scenario 2 is that you know you are paying for entertainment, not making an investment. A $20 parlay with a 22% effective house edge costs roughly $4.40 in expected value on a $20 bet. If that $4.40 is the price of several hours of heightened engagement with sports you would watch anyway, it may be a reasonable leisure expense. It is not a betting strategy.

What parlays cannot rationally be justified as, for most bettors, is a path to long-term profitability. The math does not support it. The vig compounds too aggressively.

Better Alternatives

If your goal is to maximize long-run expected value from sports betting, straight bets with disciplined bankroll management consistently outperform parlays.

Straight bets and line shopping. A single -110 line carries a 4.55% house edge. If you can find the same game at -108 at another book, that drops to approximately 3.70%. If you can identify a line where you have a genuine edge (the book's implied probability is lower than your assessed true probability), you may be operating at positive expected value. None of this is possible when you stack legs and compound the vig.

Unit-based bankroll management. Betting a consistent percentage of your bankroll per wager (commonly 1-5% depending on edge confidence) allows you to survive variance and apply your edge over a large sample size. A parlay strategy, by contrast, concentrates all of that bankroll risk into a single outcome with a compounded vig disadvantage. For a full framework on sizing bets correctly, see Bankroll Management for Sports Betting.

Understanding expected value. Before placing any bet, you should be able to answer: what is my assessed probability of this outcome, and does the offered payout imply a lower probability than my assessment? If you cannot answer that question, you are gambling on intuition rather than edge. The comparison between sports betting EV and casino EV is covered in depth in Crypto Sports Betting vs Casino: EV Compared.

Round-robin parlays as a middle ground. For bettors who want parlay-style exposure across multiple games without putting everything on a single bet, round-robin parlays create multiple smaller parlays from a set of selections. They reduce variance compared to a single large parlay while still capturing some of the higher-payout potential. They do not fix the compounding vig problem; they just spread it across more outcomes.

The most honest summary is this: parlays are fun, they are occasionally profitable in the same way that any negative-EV wager occasionally pays out, and they generate outsized revenue for sportsbooks relative to their handle for a reason. Every additional leg you add moves the effective house edge further from sports betting territory and closer to the house edge of slot machines. Know what you are buying before you buy it.

For a broader look at where sports betting fits in the crypto gambling landscape, the Best Crypto Sportsbooks guide covers platforms, bonuses, and line quality across the major options.