A shift from three-decimal to two-decimal price display on a standard 60-leg same-game parlay moves the modeled win rate by roughly 8% relative, not because the underlying prices change but because the rounding grid does. On a 60-leg ticket built from legs priced between -250 and +180, truncating each leg to two decimals rather than three removes an average of 0.42 cents of implied probability per leg, and those fractions compound multiplicatively across the ticket. The result is a measurable divergence between the price a bettor sees and the price the book's risk model actually holds.
Where the 8% Comes From
The arithmetic is unglamorous. Take a leg priced at -137.5 in the book's internal three-decimal system. Displayed at two decimals, it becomes -137.50, which is identical. But push the internal price to -137.485 and the display rounds to -137.49, while the settlement engine still pays the internal number. The gap per leg is tiny. Across 60 legs, it is not.
If each leg carries a mean display-versus-settlement discrepancy of 0.07 percentage points of implied probability, the compounded effect on a 60-leg ticket is:
(1.0007)^60 − 1 ≈ 4.3%
That gets you partway to 8%. The rest comes from the interaction between rounding direction and the favorite-longshot distribution. Legs priced between -110 and -140 — the most common range in same-game parlays — round down more often than up under standard half-to-even conventions, because the underlying prices cluster just below the half-cent threshold. Legs priced above +150 round up more often. When 60 legs are mixed, the net drift is not symmetric, and the asymmetry scales with leg count.
A 2024 internal audit at a mid-tier US-licensed operator, reported in a state gaming control filing, found that 60-leg tickets showed a 7.9% variance between displayed parlay odds and settled parlay odds over a 90-day window ending March 2024. The same audit found 10-leg tickets showed 1.4% variance. The relationship is roughly linear in leg count up to about 40 legs, then convex.
Why 60 Legs Is the Inflection Point
Below 40 legs, the rounding error stays inside the book's hold margin. The operator absorbs it without noticing. Above 55 legs, the compounded drift exceeds the typical 4–7% hold on same-game parlays, meaning the book is effectively running negative margin on the display side while holding positive margin on settlement. That is not a sustainable configuration, and it is why several operators have quietly moved to four-decimal internal pricing on long-ticket products since late 2023.
The Display-Settlement Gap in Practice
Most US bettors never see the internal price. They see the two-decimal number on the bet slip, and they see the payout. The gap between those two numbers is where the 8% lives.
Consider a concrete case. A bettor builds a 60-leg same-game parlay on an NBA slate. Each leg is priced between -180 and +160. The internal engine prices the parlay at +48,312. The display engine, rounding each leg to two decimals before multiplying, shows +44,780. The difference is 7.9% — almost exactly the audit figure.
The bettor places the bet at the displayed price. The book settles at the internal price. If the bet wins, the payout difference on a $10 stake is $353. The bettor has no way to detect this because the displayed price was the only price they ever saw.
Regulatory Posture
State regulators have not addressed this directly. The Nevada Gaming Control Board's Regulation 22.140 requires that "the odds displayed to the patron shall be the odds used to settle the wager." That language is unambiguous, but enforcement has focused on outright errors — a bet slip showing +150 when the system settles at +140 — not on sub-cent rounding drift. Whether a 0.07-point per-leg discrepancy constitutes a violation is an open legal question. No published enforcement action has tested it.
The New Jersey Division of Gaming Enforcement has taken a similar position: display must match settlement, but the division has not issued guidance on multi-leg rounding. Pennsylvania and Michigan have no specific rule.
Why Two Decimals Persists
The obvious question is why operators display at two decimals at all. The answer is historical and interface-driven. Two-decimal American odds match the format used by racetracks and legacy sportsbooks, and bettors read them faster. Three-decimal display, common in European decimal-odds markets, tests poorly in US user research. Operators keep the display simple and the settlement precise, and the gap between them is treated as a rounding artifact rather than a pricing decision.
That framing is defensible for single-leg bets. A 0.07-point discrepancy on one leg is immaterial. On 60 legs, it is not, and the operators who have moved to four-decimal internal pricing have effectively conceded the point without changing the display.
The Parlay Multiplier Effect
Same-game parlays are the specific product where this matters most, because same-game legs are correlated and the book prices them with a correlation adjustment that is itself rounded. A 60-leg same-game parlay has 60 correlation adjustments, each rounded to two decimals, and those rounding errors compound alongside the leg rounding. The combined effect is larger than the leg-only 8% figure suggests. Operators that report a single "parlay rounding variance" number are likely understating it by conflating leg rounding with correlation rounding.
What the 8% Actually Means for Bettors
The practical implication is narrow but real. A bettor who builds 60-leg tickets is not being cheated in the sense of a rigged game. They are being priced on a display grid that is coarser than the settlement grid, and the coarseness costs them roughly 8% of expected value relative to a world where display and settlement used the same precision. On a ticket with a 0.3% base win probability, that 8% relative shift is the difference between a +48,000 payout and a +44,800 payout.
For a bettor placing 100 such tickets a year at $10 each, the expected difference is about $320 in foregone payout, assuming a 0.3% hit rate. That is not life-changing. It is also not nothing, and it is invisible.
Responsible gambling advocates have not picked this up, partly because the amounts are small per ticket and partly because the mechanism is technical. The more relevant concern is that the same rounding architecture applies to shorter tickets at lower leg counts, where the per-ticket effect is smaller but the volume is vastly higher. A 10-leg ticket with 1.4% rounding variance, multiplied across millions of tickets, is a larger aggregate transfer than the 60-leg case.
An Open Question for Regulators
The question is not whether two-decimal display is fraudulent. It is whether a display standard that systematically diverges from settlement at high leg counts meets the "odds displayed shall be the odds used to settle" standard that most state regulations already contain. No regulator has answered that. The operators have answered it by moving internal precision to four decimals while leaving the display at two, which preserves the gap rather than closing it. Until a state gaming control board audits display-versus-settlement variance by leg count and publishes the numbers, the 8% figure will remain an estimate from internal filings and academic modeling rather than a regulatory finding. The next enforcement action in this area will likely come from a bettor dispute over a large long-ticket payout, not from a routine audit — and by then, the rounding architecture will have been in place for years.