Online Dating Stats Reveal A 'Dataclysm' Of Telling Trends OkCupid co-founder Christian Rudder knows a lot about his site's users. He explains how he uses mass data to explore behavior in his new book Dataclysm: Who We Are (When We Think No One's Looking).
NPR logo

Online Dating Stats Reveal A 'Dataclysm' Of Telling Trends

  • Download
  • <iframe src="https://www.npr.org/player/embed/345884282/346391349" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Online Dating Stats Reveal A 'Dataclysm' Of Telling Trends

Online Dating Stats Reveal A 'Dataclysm' Of Telling Trends

  • Download
  • <iframe src="https://www.npr.org/player/embed/345884282/346391349" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

ARUN RATH, HOST:

It's ALL THING CONSIDERED from NPR West. I Arun Rath. Thanks to a bunch of extremely popular online dating sites there's a growing mountain of data about how we hook up, date and generally judge people we don't know. Christian Rudder is the co-founder of the dating site OkCupid. In his new book, "Dataclysm" he looks at the information OkCupid and other sites gather from their users and tries to reveal the gap between what people say and they do. Christian welcome to the program.

CHRISTIAN RUDDER: Thank you. I'm very happy to be here.

RATH: There are a lot of very interesting charts and graphs in this book. So, I want to start with one that shows the average length of messages people post on OkCupid. You show that messages now are about just a third as long as they were back in 2005. What happened? Well, I mean smartphones happened basically and OkCupid developed a App and people started typing on their phones, which just kind of in a physical sense is harder but also it's just more chatty, I guess. You don't have a keyboard, it's very hard to go back and edit all this kind of stuff. So, it's a lot more concise.

RATH: One of the nicer surprises in this book for me was that shorter messages don't necessarily mean that the writing is bad or getting worse. Can you talk about how Twitter compares with Shakespeare?

(Laughter)

RUDDER: Yes, of course. I looked at a large sample of tweets and looked at the average word length in tweets and of course, you know, the first blush would be, you know, with all the abbreviations like, U, for Y-O-U, and thx, you know T-H -X, - that there would be shorter words on Twitter and I found that just wasn't the case. They were longer, not only in Shakespearean English also in kind of long form journalism. It's much shorter obviously because there's only 140 characters but it is just as robust in a certain sense, that it's the kind of an unfettered writing environment.

RATH: You spend a lot of time in the book on race. Some really fascinating stuff because your tracking what people are actually doing and saying as opposed to what they might think or profess in other ways. What were the biggest surprises that you found?

RUDDER: All the data on race that I have is from dating, but on these sites black users especially, there's a bias against them. Every kind of way you can measure their success on a site. How people rate them, how often they reply to their messages, how many messages they get, it's all reduced and so I obviously race is always a topic in this country but especially now with Ferguson. It's such a emotional issue and it's rare that you can find data that speaks of how one person of whatever race treats another person of race and an aggregate and kind of measurable kind of way. And so online data is very good at that and specifically dating data because it's all just strangers mixing with each other. And the whole premise of a dating site is to judge people. And so you really are able to tease that out free of any sort of off-line social constraints. You know, your friends don't know what you do a dating site, like they do know what you do on Facebook. So, that's why I relied so heavily dating data for that chapter.

RATH: You found that, you know, people who might say that, no, I'm racist, but in the aggregate the numbers are showing something different. They're revealing clear racial biases.

RUDDER: Oh yeah of course. When you ask anybody directly, I mean, you know, even if it's just a computer asking the question, you know, people are like, oh yeah, interracial marriage is great. I don't care what race my match is, you know all that stuff you'd expect a kind of decent forward thinking person to say. But when you actually are observing passively what they're doing, when they don't think you're keeping track, you see a totally different story.

RATH: Well, this kind of brings us to the political culture. What do you find there - is the country as polarized as it looks on cable news?

(Laughter)

RUDDER: Well, in the book I do look at speech patterns and political speech in particular and found that certainly if you look at divisive speech that is increasing and that does correlate to kind of more gerrymandered political consciousness in the country.

RATH: So, what does that mean? Like on your dating site are Republicans likely to date Democrats or does that not happen?

RUDDER: (Laughter) Oh. Well, on OkCupid, I mean we have found that the thing that matters the most is whether you care about politics one way or the other. If you're ambivalent you will have a hard time relating to I assume someone who's passionate whether it's red or blue.

RATH: With all this data from a dating site you write about attractiveness. When you break down the data, what have you learned about what men and women think about attractiveness when they're dating?

RUDDER: Well, age is a huge variable in online dating. Women generally like a guy to be the same age as them - up until the guy's hit about 40. Then when you flip it around, when you look at how men perceive women, it is pretty much just straight ticket vote for 20, which is the lowest age that I looked at in my dataset. So, you know, even 45 year old guys will rate 20-year-old women the best. It's very pervasive opinion among - at least the people on OkCupid.

RATH: So, the data confirmed men are in fact creepy.

(Laughter)

RUDDER: Yeah or dreamers. Either one, how ever you want to look at it. I mean what happens is - this is just measuring people's opinions not what they actually go out and do, what you see when you actually look at what people do, you see the realisms set end. So, what these 40-year-old guys who, yeah, you know, that they - 20-year-old girls are the 21-year-old girls. The people they actually have the courage to actually is a lot older. It's 30, 35-year-old women.

RATH: So, Christian if I asked you to make a value judgment, because you clearly love diving into all this time, and we'll leave aside government intrusion, but just collecting all this data, tons and tons of info from various companies, is getting all this information about ourselves good or bad?

RUDDER: Oh, I mean I definitely think it's good. I think to speak to your sort of complicate concern about privacy, for sure, I mean, you know all of this data, everything in the book and generally anything that you read online about people's behavior on sites is all aggregated and anonymous. Nobody's looking at your personal account, but when we put all this stuff together you're able to look at people in a way the people have never been able to look at people before. You know, like you have millions and millions of people living their lives through an interface that records what they're doing as they live. And, you know, look there's no way OkCupid, Facebook, Twitter, these sites even added altogether can stand in for the entirety of the human condition. People do all kinds of thing that don't do online. But they're doing fewer and fewer things that they don't do online and it's the beginning of I think a revolution in how social science and behavioral science are done.

RATH: Christian Rudder is a co-founder of the online dating site OkCupid. His new book is, "Dataclysm: Who We Are When We Think No One Is Watching." Christian Rudder, thank you.

RUDDER: Oh, my pleasure.

Copyright © 2014 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.