A 96.2% RTP slot and a 87.5% RTP slot can carry the same game title, the same symbols, and the same manufacturer logo. The difference is the cabinet. On three-reel mechanical and "mechanical-style" machines — the ones with physical reels, a lever option, and a single payline — operators in several U.S. jurisdictions are permitted to run a lower theoretical return version than the video equivalent sitting twenty feet away on the same floor. The gap is not a rounding error. On some titles it exceeds 8 percentage points.

Why the Same Game Has Two Math Models

A slot's RTP is not a property of the theme. It's a property of the paytable and the reel strips — the physical or virtual symbol distribution on each reel. A manufacturer ships a game with a range of approved configurations, and the operator picks one.

Three-reel steppers have a structural constraint video slots don't: fewer reel positions. A classic three-reel game might have 22 stops per reel, giving 10,648 possible outcomes (22³). A five-reel video slot with 32 stops per reel gives 33.5 million. Fewer outcomes means the paytable has to do more work to hit a target return, and it means the manufacturer can only offer so many distinct configurations. When a jurisdiction or an operator wants a higher hold — the percentage the house keeps — the three-reel version is often where the adjustment lands, because the video version's math is locked to a configuration that's already been certified and widely deployed.

The Numbers Behind the Gap

Take a title that exists in both formats. The video version might run 95.8% RTP. The three-reel mechanical version of the same theme might run 87.5% to 89.0%, depending on the state and the operator's chosen configuration. That's a 6.8 to 8.3 point spread on identical branding.

This isn't hypothetical. Nevada's Regulation 14 and the equivalent rules in states like Louisiana, Mississippi, and West Virginia allow operators to select among manufacturer-approved RTP configurations, and the approved floor varies. Some jurisdictions set a minimum theoretical return as low as 80% for mechanical reel games; video slots in the same state often carry an 85% floor. A player who assumes the floor is the same across cabinet types is wrong by design.

The practical effect: a $1.00 spin on the 87.5% version loses, on average, 12.5 cents per spin in theoretical hold. The 95.8% version loses 4.2 cents. Over 1,000 spins, that's a $125 expected loss versus $42.

What the Paytable Actually Tells You

The RTP isn't printed on the glass in most U.S. casinos. What is printed is the paytable, and the paytable is where the difference hides.

On a three-reel game, look at the top jackpot relative to the coin size and the number of coins required to activate it. A game that pays 1,000 coins on the third coin but only 5 coins on the first two is a max-coin game — playing fewer coins drops your effective return sharply, often by 2 to 4 points on its own. Combine that with a low-RTP configuration and you're looking at a return in the low 80s.

Video versions of the same title rarely have that structure. They typically pay proportionally across bet sizes.

Why This Persists

Three-reel mechanicals are expensive to maintain and appeal to a shrinking but loyal demographic. Operators keep them because the hold is higher and the players who favor them are less likely to comparison-shop RTP. The math is legal, disclosed in the game's PAR sheet, and filed with the state — but the PAR sheet isn't on the cabinet.

The open question is whether state regulators will eventually require RTP disclosure on the machine itself, the way some European jurisdictions do. Until then, the only reliable signal is the cabinet type, and the assumption that a familiar title means a familiar return is exactly the assumption the configuration system is built to exploit.