Data Driven

Heartache by the Numbers and OkCupid’s Founder Has Got Yours

Christian Rudder knows who’s been bad or good, and he really knows who’s lying about who they are and what they want. And he has the numbers to prove it.

If you're a single woman over 20, do yourself a favor and stop reading now. For the rest of you, here's a seedy little fact: Men of pretty much any age are mainly attracted to 20-year-olds. It doesn't matter if the man is 20 himself, or 28 or 36 or 49. If we're to believe the search habits of dating-website users, then men, no matter what age, want a woman two years out of high school.

This isn't what men say they want, of course. Read the profile of a 36-year-old man, and he'll say he's looking for a mate between, oh, 26 and 42. And that may be. But when he rates OkCupid's female users in purely visual, ruthless "hot or not" style, as the website asks him to do, the 20 year olds win in a landslide.

Even more pernicious is this: Look at who these men are actually contacting. By a large margin, it's the youngest women they believe they can plausibly nab. Guys in their late ’30s overwhelmingly contact 30-year-old women. But the moment a man turns 40, he immediately switches to 35-year-old women, and keeps plugging away at 35-year-olds till the day he turns 45, when lo and behold, he begins approaching 40-year-olds instead. "You can almost see the gears turning," writes the author—the gulf that separates "what men imagine they're supposed to desire, versus what they actually do."

So begins the fascinating and depressing Dataclysm: Who We Are When We Think No One's Looking, in which OkCupid cofounder and president Christian Rudder exposes us for the frauds we are. Armed with a container ship's worth of internal data from his own website—as well as Google, Facebook, Twitter, etc.—Rudder proves that we're a highly deceptive people, not to mention pervy, racist, and probably gay.

Like a Kinsey Report for the big-data era, there's a gold mine of findings here to titter about. For instance, in the first year of the relationship of an OkCupid couple, the single best predictor of whether they'll stay together is how often they look at each other's profile after they've already started dating—basically, it's an indication of how obsessed they are with each other. Later on, the best predictor becomes how assimilated their Facebook friend networks are. Rudder even pinpoints the two OkCupid profile questions that best predict a couple's longevity if they both answer the same way: Do you like horror movies? And, have you ever traveled to another country alone?

Random, right? Except not at all. If the book has a message it's that virtually every supposedly random blip of human activity can be demystified if we can just access the right data. Let's take the eternal question of how many people are gay. There's absolutely no consensus on this—some researchers say it's 1 percent, others say 15. Instead of asking people about what they do, the book suggests, let's look at their porn searches. Across every U.S. state, from West Virginia to Vermont, the data is remarkably consistent: 5 percent of Internet porn searches are gay.

Fun facts like this abound, often displayed via amusing graphs and infographics. There's the U.S. map that shows where most Craigslist "Missed Connections" take place depending on where you live. For New Yorkers, it's the subway; Portlandians, the bus. But for much of the country, these fleeting moments of wistful longing happen in the everyday-low-prices aisles of Walmart. Rudder makes lists of the most common keywords used in the dating profiles of different racial groups, and creates maps that show which states shower the most and least often. He waxes theoretical about the catalysts behind some of these data (are New Jersey's frequent bathing habits somehow connected to the preening Jersey Shore mentality?). But what he doesn't do much of—and this is refreshing—is delve into the humungous What Does It All Mean? questions, as so many books about big data can't resist. Rudder seems content to play the record keeper and let the philosophers sort out the sigificance.

But it's impossible to read a book like this without conjuring some sort of grand takeaway, so maybe it's this: When it comes to love and sex, we are highly predictable creatures. All our seemingly irrational actions can be easily explained by data. And data, we have. Simply by using websites like Facebook, Google, and OkCupid, we as a society have inadvertently embarked on a global longitudinal study of human behavior. All this data is a mirror held up to ourselves, and in a way, it's rather comforting. It lets us see that our weird and inexplicable urges are actually normal and easily explained. There's no longer a need to make up a story, because now, as Rudder writes, "The numbers are the narrative."