As reported in recent articles in Fast Company and the Wall Street Journal, a pair of young entrepreneurs created a website, Ultrinsic.com, that lets students bet money on their future grades. You can bet in two ways--essentially on or against good grades. So you can either bet on getting at least a certain grade or set of grades (say, an A in calculus), or you can buy "insurance" that pays you if your grade or grades fall below a certain point. The odds of the bet and the size of the insurance premium are calculated with an algorithm that uses the student's GPA and the grading history of the course as inputs.

Ultrinsic's founders said that they hope to turn a profit and inspire students to work harder. The jury is still out whether a small monetary incentive will encourage students to work harder or whether it would create a perverse incentive to aim low. But as a business model, the whole setup is deeply flawed. The founders acknowledge that there are kinks in their system but say their information will improve as more bets are placed.

What the entrepreneurs don't seem to realize is that they are up against two big problems that no algorithm can solve: moral hazard and self-selection. Students have fundamentally have better information than statistics compiled from a few numbers. (They know, for example, that they already took calculus in high school and are likely to do better than the average student in the class.) What's more, students have a great deal of control over their grades—so betting on your own grades is not like betting on a horse. And most important, students have control over not making bets that are unfavorable to them: they'd be a fool to bet that they'll ace a class in a subject that's not their strong suit.

For example, you will only get into a bet that your GPA is below 3.0 if you can live with a lower than 3.0 GPA. So no matter what odds the site sets, you can always guarantee a win by working less hard, or not answering all the questions in a final. If we can think of such a simple way to exploit the system in a few minutes, consider what students from MIT, Stanford, and the like can do if given more time to consider the problem. In fact, we'd be willing to bet that some are already working on that.

Copyright Kay-Yut Chen and Marina Krakovsky,http://www.secretsofthemoneylab.com

About the Authors

Kay-Yut Chen
Kay-Yut Chen is a principal scientist at Hewlett-Packard Laboratories. He established behavioral economics research at HP Labs in 1994, a first in any corporation.
Marina Krakovsky

Marina Krakovsky has reported for Scientific American, the Washington Post, and the New York Times Magazine.

You are reading

Secrets of the Moneylab

The Helpful Eye Doctor and the Pricey Glasses

An unexpected visit to the eye doctor shows the limits of reciprocity.

The World Series, the Midterm Elections, and Dancing with the Stars

What these three competitions have in common

Winning the Trust Game (Book Excerpt in Financial Planning Magazine)

Free excerpt from the October issue of Financial Planning magazine