Gambling is a topic that receives attention on these pages from time to time. I am currently working on a paper dealing with Internet gambling, which is a growing phenomenon. PartyGaming PLC, a Gibraltar-based gaming enterprise focusing on Internet operations, did an IPO last fall. Its market capitalization has recently been higher than that of British Airways - which is a pretty big company itself.
There are many aspects of this form of amusement and/or addiction (for at least some people) that merit attention. One of them is the matter of payback ratios. Given that all Internet gambling operations are located abroad, U.S. citizens who want to gamble (and many do, despite possible legal barriers to participation) have to trust a foreign enterprise enough to send money to gamble.
In the U.S., casino patrons may trust government regulation to ensure the fairness of the games. On the net, private ordering has prevailed. In effect, the international accounting firm of PriceWaterhouseCoopers has lent its trust to these sites (for a fee, I'm sure, and I don't fault them for that in the least) by providing an attest function regarding payout rates for various Internet games.
My limited research indicates that the payout rates advertised for slot machines (or more accurately its electronic equivalent on the web) are anywhere from 2 to 8 points higher than those reflected in traditional casinos in Missouri, where payout data is monitored and reported to the public by the state gaming commission. This means that an Internet player can waste more time (oops - is that a value judgment slipping in?) before he/she loses her money.
The laws of probability are hard to evade. Even a payout ratio of 98 percent, which is approximated at some Internet sites, means you lose, on average, 2 percent of all you wager. So if you bet $1 every 6 seconds, that means over the course of an hour you bet $600, and on average you lose $12. Over time, that adds up. You can lose more, or you can lose less. But if the payout ratio is only 90% (or less- as at one Missouri casino), that means more like $60/hour.
It seems to me that we could use more price transparency on slot machines in traditional locales. People should have access, as in Missouri, to reports about the routine payout levels. This would show you, if you were rational, where to go to lose your money more quickly, or more slowly, as you preferred. Of course, particular machines could be better or worse, but at least you would have some idea of the overall opportunities.
But I think rationality does not hold a monopoly on this form of behavior, which is based on the optimistic outlook of feeling lucky today. (And I love that Mary Chapin Carpenter song, even though I personally don't organize my life on the principle of blind luck.)