 So my unique angle on game theory is what's called behavioral game theory. We're interested in a cognitive hierarchy in which not everybody who plays a game, who chooses a strategy, who's bargaining or competing or helping, completely figures out the structure. Sometimes they just have a snap judgment or a heuristic that they pick. Other people think that other people have that heuristic and they're going to then figure out what to do if someone else chooses randomly, for example, or just does what's most popular. And other people, those people are called level one. Level two people think that they're playing level one and level zero. And so you mathematically build up a hierarchy of more and more steps of understanding about kind of human nature. And this turns out to be, in many cases, a better prediction of what a group of people will actually do than a traditional analysis in which everyone is completely rational picking what's best for them and thinking that everyone else picks what's best for them and so on. So the kind of data we use to test these theories span everything from neural firing rates up to groups of people making complicated decisions in the economy. We have some recent research on movie reviews, about 20% of the time, nowadays in the United States, maybe less in Europe. Movie studios releasing a movie, typically on a Friday at the end of the week, don't actually even show it to film critics in time for the critics to write a review that can be in the newspaper. And we think the reason is that the movies usually are below average, as judged by anybody, not just critics, but average fans, too. We have data on all those things. And the studios think, well, maybe if the critics say it's bad, then people won't come. But if the critics don't even review it, some people might not realize that that's actually bad news. That's associated with the movie really being pretty mediocre. This is called a cold opening. So we find that if you have a mathematical model of how some consumers don't realize that a movie that's not reviewed is not reviewed for a good reason, which is that it's a bad movie, then you can actually explain the types of movies that are not reviewed and also the box office gross, the economic value of how many people go to see the movie and what happens along the line. So it's an example that uses the theories that are very carefully crafted, simple laboratory situation to incorporate one or two steps of strategic thinking instead of lots more. And then it explains what happens in all the movies released in the US for the last 10 years. So anything you would want to do with game theory, you can do even better with behavioral game theory or cognitive hierarchy. So that includes making judgments or regulatory ideas about what is best for consumers. So one lesson from the behavioral game theory approach is that not everybody is strategically savvy or sophisticated all the time. So it means that companies or people may be able to find gullible, sort of naive consumers and lead them to make mistakes. It doesn't mean you necessarily would want to completely regulate what's going on in consumers' lives, but it gives you a mathematical tool for saying when consumers may be likely to say make a mistake in judging whether the content of an advertisement is really truthful. Another thing you might want to do is to design new sets of rules and institutions, we call them in economics, institutions or systems, which are kind of robust to the typical mistakes people will make. So for example, if you introduce a new change of tax or you introduce a new way of buying and selling something, you'd like to be sure that even people who haven't completely figured out how it works won't undermine the system. They either won't make big mistakes themselves or they won't create profit opportunities that get in the way of what you really want the system to do. So far the ideas have mostly been developed in academia. I think there's be some applications in things like understanding bargaining a little bit better, understanding what for regulators are interested after the financial crisis, for example, and if banks and mortgage companies are kind of creating products and describing them in a way that makes it particularly confusing for consumers, maybe even confusing for investors, even fairly sophisticated ones because the mortgage back securities are so opaque and complex, understanding the limits on reasoning and the limits of people's normal ability to guess what other people are kind of telling them or signaling, we may have a more robust kind of user-friendly approach to regulatory design and also for firms thinking about what is it I can offer to people that they'll understand.