Casino gambling: Hold 'em or fold 'em?

A new study by Puneet Manchanda, associate professor of marketing at the U-M's Ross School of Business, shows that gamblers are less likely to continue betting if they're winning, but more likely to keep on wagering if they're losing—a phenomenon known as the "gambler's fallacy."

"This is an important set of results since it is, to our knowledge, the first study to document the gambler's fallacy among casino gamblers, using individual-level behavioral data," Manchanda said. "It also confirms the results of previous studies using laboratory experiments or direct observation of small samples of consumers."

Manchanda and colleague Sridhar Narayanan of Stanford University studied a representative sample of 2,000 consumers who gambled at a single casino in the southwestern United States during 2004 and 2005. The consumers were members of the casino's free loyalty card program, which monitors their gambling activity and provides incentives, such as coupons and comps. The cards track the amounts bet and won or lost on each play, the comp amount delivered daily to the customer and some demographic information of each consumer.

Using various statistical models, the researchers found that for more than 95 percent of consumers who won on their last play (mostly on slot machines as opposed to table games), the probability of playing again decreased—a result consistent with the gambler's fallacy.

Likewise, the more times consumers lost, the more they kept on betting—as long as the losses were small. Once their losses exceeded a certain amount, however, they stopped playing.

"Consumers behave somewhat rationally on average, cashing out when the going is good and cutting their losses when they are losing a lot," Manchanda said. "They continue to play for small wins, but seem to cash out after their cumulative wins have reached a high level."

The researchers also found, using a well-accepted definition of "addiction" from the economics literature, that 8 percent of the gamblers in the sample showed evidence of being addicted to gambling. Specifically, for these gamblers, the utility gained from playing was higher if they had played and bet more in the past. In other words, these gamblers exhibited positively reinforcing gambling behavior over a sequence of bets.

"While this proportion may look small (in absolute terms), it is consistent with other research that has focused on casino gamblers," Manchanda said. "This also suggests that for a majority of casino gamblers, the failure to find patterns of addiction may be interpreted as support for the view that the role of casinos for these gamblers is to provide entertainment."

Manchanda and Narayanan also examined whether in-casino marketing activity—in the form of comps (complimentary gifts, such as free drinks, meals and hotel rooms, to entice gamblers)—affects gambling behavior. They found that such marketing does indeed influence the decision to play and the amount to bet, especially for consumers who exhibit more addictive behavior.

Provided by University of Michigan