Profit boosts and same-game parlay insurance get the marketing budget. Score-boost free bets quietly cost more. Across the four legal U.S. books that still push them — DraftKings, FanDuel, BetMGM, and Caesars — a score-boost token applied to a same-game parlay returns roughly 8% less expected value than the same stake bet straight, once you account for the token's expiry and the correlation tax baked into SGP pricing. The headline number looks like free money. The math says otherwise.
What a score boost actually is
A score boost pays a free bet when a named player hits a threshold — 25+ points, two goals, 6+ strikeouts. It is not a profit boost. It does not multiply your return. It hands you a separate free-bet token if the condition lands, usually $10 to $25, with a 7-day clock and a minimum-odds floor around -200.
The catch sits in the qualifying leg. Books price player-prop markets with a house hold of 5.5% to 7.5%, versus 4.5% to 5% on main lines. You are buying a lottery ticket on the most juiced market on the board to earn a token worth a fraction of the stake.
Where the 8% goes
Take a $50 same-game parlay at +450 with a score-boost leg attached. The book's fair price on that combination, stripped of correlation markup, is closer to +490. That gap — the SGP correlation tax — runs 4% to 9% depending on how tightly the legs are related. Stack a 6% correlation tax on top of a boost token whose expected value is about 65% of face (standard for free bets, since you keep only winnings, not stake), and the effective return drops below the straight-bet equivalent.
Run the same $50 on a two-leg SGP without the boost and the correlation tax is smaller, because the book has fewer dependencies to price. The boost token pushes you toward three- and four-leg builds, where the tax compounds.
The expiry problem
A 7-day expiry on a $25 token forces action. Bettors who redeem inside 48 hours take worse prices than those who wait for a line they like. The behavioral cost is real even when the nominal EV is neutral.
What the books don't disclose
None of the four operators publish the correlation adjustment applied to SGP legs. You see the final price, not the components. That makes the 8% figure an estimate — derived from comparing SGP prices to the same legs priced individually and manually de-vigged. DraftKings' own help pages note that SGP pricing "may differ" from single-bet pricing without quantifying it.
State regulators have started asking. New Jersey's Division of Gaming Enforcement requested SGP pricing methodology from operators in early 2024, though no public findings have been released. Until that changes, bettors are pricing a product with a hidden margin.
The practical read
If a score boost lands on a leg you already wanted, take it as a single. The free-bet token is worth more attached to a straight bet at -110 than buried in a four-leg SGP at +600. If the boost only fires inside an SGP builder — and on some apps it does — treat the token as worth about 60 cents on the dollar and size accordingly.
The open question is whether books will ever show the correlation tax line by line. If they did, the score-boost pitch would lose most of its shine. Set a deposit limit before you test any of this; the tokens are designed to keep you betting past the point where the math works.