There has been a long-standing debate about whether or not players can detect differences in theoretical hold between slot machines. Academics have used statistical arguments to show that hold differences are very difficult for players to detect due to the high volatility of most slot games. However, casino operators remain largely unconvinced and are hesitant to increase slot holds, fearing a backlash from their players. Are these operators unnecessarily leaving money on the table? This talk explores this fascinating topic, using detailed casino simulations to demonstrate that operators of locals markets are right to be concerned. Although the academics have generally done their math correctly, they have missed an obvious but important fact – that players in locals casinos often have friends who also play slots, that these friends have other friends, and that friends talk to each other. I will show simulations demonstrating how information about slot hold propagates through a social network of players, allowing the community as a whole to detect changes in slot hold, often with surprising speed. I will also propose a strategy showing how to increase hold selectively and with vigilant monitoring, allowing operators to improve the casino’s bottom line with relative safety.

Learning objectives:

- Discover why slot hold is difficult for players to detect, and why this important topic has been so controversial.

- Explore simulations showing how a community of players can discover changes in slot hold that would be imperceptible to any individual player.

- Learn how to increase slot revenue by increasing hold on machines with specific characteristics and monitoring for signs that the community might be discovering the increase.