The slot machine industry is a vast realm of beeps, flashing lights and tall, noisy boxes decorated with images of martini glasses and pirates. Inside, well-dressed men and women push buttons and pull levers while watching cherries and sevens spin on liquid crystal displays. But they aren’t there for fun; they’re here to make money.

No one has uncovered the Platonic ideal of slot games, but certain principles undergird most of them: colors tend toward primary or pastel, franchise tie-ins are de rigueur and game soundtracks are in a major key. Most important, the machines fulfill a fundamental psychological need for many players: They allow them to zone out and escape thought.

A player’s choice of machine is based on several factors, including the paytable and how often a particular symbol appears on the reels. The more often a symbol appears, the higher the payout. But the actual frequency of a symbol on the physical reel is irrelevant to the machine’s random number generator (RNG), which selects stops at random and determines the outcome of each spin.

Manufacturers can change the paytable of a specific machine, but routinely changing entire floors full of slots isn’t feasible in terms of time and money. Instead, a casino may alter its payout percentages, which will likely affect the average amount of money a player loses over a long period. This research will test the generalizability of Lucas and Roehl’s model, provide an empirical study of a vital but under-researched aspect of casino operations and improve the understanding of slot machine performance potential.