Since I have a background in marketing, I’ve received a minimum of training in advertising via Facebook, LinkedIn, Twitter, and Instagram. Because of this experience, I was already aware of the somewhat creepy predictive capabilities of these social networks even before I read Christian Rudder’s new book, Cataclysm: Who We Are (When We Think No One’s Looking). For example, I already knew that the all powerful “Like” button, a now universal micro-currency for the internet social scene, allows networks to predict with an incredible degree of accuracy whether or not you are gay (85% accurate), whether you are African American or Caucasian (95% accurate), Republican or Democrat (85% accurate), and even whether or not your parents got divorced before you turned 21 (60% accurate). By monitoring your commentary, your clicks, and what posts you linger over and for how long, Facebook knows if you have children, how many, and how old they are. All of this, mind you, without you telling them directly.
And it gets even worse. There was once a time when the embarrassing (and often idiotic) things people said as teenagers evaporated into the past, sometimes for good, only to be remembered when mom digs up an old diary or yearbook from some dusty boxes in the attic. But those days are gone. The young (and old) of today will create a digital footprint that will persist for decades, ready to come back and haunt them at any time. It’s all just a Google search away, for those who know how to look. That thing you tweeted when your girlfriend cheated on you the other day?–Well, if you ever run for public office, you can expect to have to explain that one to the public. Thanks to “Big Data,” we each leave a trail of breadcrumbs throughout our digital lives, and sometimes that trail doesn’t lead anywhere good.
Christian Rudder is the founder of the online dating site, OkCupid, and is therefore an expert in mining data. The nature of his work has also made him very good at quantifying human interactions, at least as much that can possibly be done. Now there’s obviously much to be said against reducing the most subtle and subjective of human phenomena–love, for instance–to a collection of charts compiled from rating systems and photo “likes”–but what Rudder demonstrates in this book is that, regardless of whether or not the meanings we draw from Big Data are complete pictures of reality, they can nonetheless tell us something valid.
For one thing, readers of this book will come to realize that we may not know ourselves as well as we think we do. Men who try online dating, for example, tend to specify that they are looking for women slightly younger than them–say, within 2-3 years of their own age. Data mined from the actual behaviors of these same men, on the other hand, says that the women they actually message are much younger than that. And when it comes to positive reactions, snap judgments, in other words, men of every age will prefer (at least on the internet) the women ages 20-22, with positive reactions dropping off drastically after that, regardless of the age of the man being observed. To sum up the phenomenon, women in the online dating world are “over the hill” at age 21.
Women, of course, tend to have responses that more accurately reflect their stated desires, which is to say that tend to response positively to men about their own age. This leads to a situation in which men and women past 21 are like “two ships passing in the night,” both looking for someone but at cross-purposes. It isn’t all bad, though. Rudder also presents data from blind-date experiments he’s conducted and explains that, if you surprise two people by sitting them down to chat for a while, more often than not they enjoy the experience and all of those things they thought were important tend to be forgotten.
We are superficial beings, it seems, but Rudder uses his experiences to show how we might get around our superficiality. He explores the hatred and “incompatibility” of someone who identifies as a Republican vs. another individual who identifies as a Democrat. If you start things off with that, then the two won’t get anywhere. But taking a different angle, and asking the right questions, such as: “Are you passionate about political issues?” it is possible to bring two such people together as civic-minded individuals who have a common passion, even if their conclusions are at odds.
In another chapter, he finds a relationship between the configuration of a married couple’s Facebook network and the strength of their marriage. What’s more, in this case he has developed tools to bring the reader along on the experiment. Simply go here and enter a pair of Facebook credentials. You “assimilation” will be charted for you, showing to what degree you and your spouse’s “social circles” overlap. Statistically speaking, the more overlap the better, and the stronger the relationship, for reasons which you’ll have to read about in order to understand.
Now this may all seem of little interest to anyone who isn’t trying to find that special someone via the internet, and who doesn’t care about novel relationship tests, but Rudder is only using his peculiar background as a point of departure. He develops each of these insights in the general direction of human nature of social development, highlighting the positives and the negatives of what he’s seen. The positives he highlights in Part I of the book, titled “What Brings Us Together”; the negatives, such as those ugly outbursts of collective rage and closet racism that we’ve all seen at one point or another, are covered in Part 2, “What Pulls Us Apart”; finally, in Part 3, “What Makes Us Who We Are,” he tries to pull everything together.
Numbers can’t tell us everything, and most of the statistics we hear on a daily basis are designed to either scare us, sell us something, or both. Rudder admits as much from the get-go, and so this book is not about a data-fanatic thinking he has finally gotten enough information to interpret the human experience in quantitative form. He is just a mathematician who loves his job, and who wants to take the reader on a ride through the numbers. Data points may be nothing more than so many dots on a pointillist painting, but acknowledged as such, and rightly interpreted, they have plenty of insights to offer.