Using Math to Move a Courtship Forward
SCOTT SIMON, host:
Going to try and change the mood a bit now. Love is a many-splendored thing, but it should come with a diamond ring? Researchers at the University College in London have come with a formula they believe predicts the success of a budding romance. And here to help us through this formula is our Math Guy, Keith Devlin, at Stanford University.
Keith, thanks very much for being with us.
KEITH DEVLIN reporting:
Hi, Scott. Nice to be here.
SIMON: You know, we get self-help books in the mail all the time, how-to books--you know, how-to-find-your-true-love tests, you know, answer 60 questions and it'll tell you whether or not the relationship is right for you. Why are mathematicians stooping to this kind of thing?
DEVLIN: Well, as I've said on air many times before, you can apply mathematics to anything, and why leave love out of the picture? And indeed, that's what these guys at University College, London, did. They took some standard mathematics, mathematics of game theory, which was invented in the 1940s and originally applied to military conflicts, and they applied it to courtship. And they said: What's the optimal strategy a man should adopt when it comes to giving a present to the object of his fancy, to the woman he fancies? They look at three kinds of presents. They look at cheap presents or valuable presents, like the diamond ring that you mentioned in the introduction, or the intermediary, which they call extravagant gifts, which are presents which are expensive to buy but have no lasting value, like taking the lady to the opera or to an expensive restaurant. Which one is most likely to achieve success in getting the lady to accept your advances?
SIMON: You're asking me?
DEVLIN: Well, I'm going to tell you the answer because they came up with the answer, and that's what...
DEVLIN: ...they give in this paper that they've just published. And the conclusion is to cut to the chase, the optimal strategy is to buy the lady an extravagant gift, something that costs a lot of money but has no lasting value, like a meal or a visit to the opera. You should save the diamond ring until later in the courtship.
SIMON: But, I mean, is this statistically supportable or what's their theory on this?
DEVLIN: Certainly, when you take this technique and you apply it to other kinds of creatures, like insects and so forth, it turns out to be remarkably accurate. You can get very, very good predictions of how these creatures will behave. Now...
SIMON: What do you mean, apply it to insects? I mean, what, insects take each other out for expensive meals at four-star restaurants or what?
DEVLIN: Well, indeed, they do. In fact, one of my favorite insect examples is the hanging fly.
DEVLIN: The male hanging fly will give the object of his affection a gift--it's usually a dead insect, which I guess, if you're a female hanging fly, is a pretty attractive thing to receive. He'll give her this gift. After mating takes place, assuming the female accepts his advances, the male then takes the gift back and uses it with the next female he fancies.
SIMON: An expensive dinner, theater tickets, a trip to Paris, a better idea? That's a better idea statistically than a diamond ring?
DEVLIN: Indeed. When you're making that first approach to try and persuade the lady to accept your advances, it's what's called a Nash equilibrium. That's named after John Nash, the famous mathematician who won the Nobel Price, played by Russell Crowe in the movie "A Beautiful Mind."
DEVLIN: And what Nash developed was an understanding of what the optimal outcome is when you have one of these conflicts or pseudo-conflicts of situations where two or more parties are trying to maximize an outcome.
SIMON: You or I, or maybe I should just limit it to me--I could buy theater tickets and expensive dinners for Elizabeth Hurley all day, and she'd still want nothing to do with me.
DEVLIN: Right. Right. And if you look at the paper that these two mathematicians...
SIMON: You didn't have to agree that quickly, Keith, OK? I was hoping you'd think that over a little more.
DEVLIN: (Laughing) But if you look at the paper that these two mathematicians in England have written, among the parameters that they've put into the model is whether each person finds the other one attractive in the first place. So their model does account for the fact that Elizabeth Hurley may well say no to you, Scott.
SIMON: Yeah, may well. I think that's safe to say. But I haven't tried the dead fly trick with her. So...
DEVLIN: Give it a try. It works for the hanging fly.
SIMON: Yeah, what have I got to lose?
Now there's another study, too, isn't there, that predicts the success of a relationship over a three-year time span?
DEVLIN: Oh, right. I think you're talking about the work of John Gottman and James Murray up in Seattle, Washington.
SIMON: Of course.
DEVLIN: Yeah, they've come up with some remarkable results. They can take a couple into their facility up in Seattle, Washington; they will film them, they will videotape them with a multicamera setup in a 15-minute conversation. They will then analyze that 15-minute conversation and reduce it to a whole string of numbers with different kind of behavioral activities. And they've been able to predict with 90 percent accuracy whether that couple will stay together or divorce. And the accuracy is in terms of a period up to 15 years.
SIMON: Could it be as simple as during the 15-minute conversation one party looks at the other says, `I can't stand you'? I mean...
SIMON: ...you don't need to be Albert Einstein for that, Keith.
DEVLIN: They look at, I think, at least a couple of dozen different--maybe even 30 or 40 or more different parameters, things like whether one person rolls the eyes when talking to either one, how much eye contact there is between them, whether they tend to agree more or disagree more. So the mathematics picks up some things that people wouldn't necessarily pick up on just by looking at their behavior.
SIMON: Have these folks looked at any videotape of Tom Cruise and Katie Holmes?
DEVLIN: (Laughs) I have no idea whether they've done that or not. But without looking at the videotape, my prediction is that it won't last. That's not based on mathematics; that's just based on looking at the magazines in the supermarket check-out line and seeing what happens three years down the line--or three months down the line.
SIMON: Well, a big dead fly to you, Keith.
DEVLIN: Thanks very much, Scott, and a big dead fly to you.
SIMON: Keith Devlin is executive director at the Center for the Study of Language and Information at Stanford University. His most recent book is "The Math Instinct: Why You're a Mathematical Genius (Along With Lobsters, Birds, Cats, and Dogs)."