Book of the Week: Dataclysm

20 Jul 2016

dataclysm This week I read Dataclysm by Christian Rudder, because the OkCupid blog, OkTrends, is probably the best example of content marketing with big data. You may wonder why would you need to read the book, if you’ve already read the blog posts. Rudder redid the analysis with new data to double check his previous work and took data from sources other than OkCupid to talk about what brings us together, what pulls us apart and what makes us who we are. You may hear individual anecdotes of racism, but when you aggregated lots of data together, insights rise above the noise. You can actually say something about people. He covers linguistics, employment, beauty, sexuality, race and also the Greater Internet Fuckwad Theory. You get a lot more than just data about dating from the book.

Tonight, some thirty thousand couples will have their first date because of OkCupid.

Men vs Women

Women are inclined to regret the sex they had, and men the sex they didn’t. —Harper

Men are less biased than women when rating attractiveness. The ratings following a symmetric beta distribution, while women say that only 1 in 6 guys are above average. This means that men have a more realistic or statically unbiased expectation of attractiveness even after being bombarded with unrealistic media depictions of photoshopped beauty. Women being more picky about their mates probably has evolutionary origins.

In the mathematical sense, a man’s age and his sexual aims are independent variables: the former changes while the latter never does. —Wooderson’s Law

As women age, they prefer slightly older guys until 30. Then after they like them slightly younger. Men reach their sex appeal limit at 40. As men age, their age preference in women don’t change. Women are over hill at age 21 and women over 35 don’t exist. WEIRD

This is the first time I heard the acronym. It means most studies and reported papers use research subjects that are plentiful and convenient. This means getting WEIRD students from college campuses. Some of the analyze done by Rudder is WEIRD since that just makes things easier. This is fine, but if you’re doing drug development, you may find that some heart drugs work better on African Americans or have different effects on women. Social Graph Analysis

High assimilated couples function—the two people together—as the bond between many otherwise unconnected cliques. They are the special glue in a given spread of dots, and furthermore, they’re a glue like epoxy: it takes both ingredients to make the thing hold together.

You can predict how likely a couple is going to stick together by analyzing their social graph. You’re likely to say together if the only path from your friends to your significant other’s friends are through the both of you. If you want to stick together, find someone from a different social circle. The reason why you two are together must be strong. Strong like epoxy. Race Even though a machine learning expert called me racist, I can take solace that the data shows that everyone else is racist too. Blacks are rated lower by non-black users. Asian men also. People don’t like Black men, Black women and Asian men. Women prefer men of their, but also also express a preference for white men. Blind people are also racist since they are conditioned by the same society. If you send a resume with a black sounding name, you’re going to get less replies. The dice is loaded, but you need to roll the dice a lot to find out that people are racist. Social desirability bias forces people to answer things in a way that make them look good, but in private they are biased. Beauty

Online, you can always get what you want. But what you need, that’s a much harder thing to find.

Success and beauty correlated for both sexes. Beautiful people have easier lives, because they seem more intelligent, more competent, so they get better jobs. Beautiful people have more Facebook friends. For men that relationship is linear, but for women it is exponential. The problem is that women are treated like they are on OkCupid even though they are interviewing for a job. Female employees are always viewed through same (sexualized) lens despite there being no romantic intent. A man’s looks has no effect on job prospects. Since women are hired based on looks, you are guaranteed statically poor performance from women versus men.

An there the conversation takes a turn. But not too pretty, right? Yeah, we wouldn’t want that. We both sit back, and the conversation moves on to something else. This is what it comes down to: I can’t imagine anyone wishing limits on a son.

Pretty people live in a different world than me. They get treated better by other people. They don’t have to develop the survival skills that us ugly people have to obtain to get through life. I want my daughter to be pretty, but not pretty enough where she becomes stupid. OkCupid launched an app called Crazy Blind Date. You didn’t get to see the person before the date.

In short, people people appear to be heavily preselecting online for something that, once they sit down in person, doesn’t seem important to them.

Turns out that looks didn’t affect people’s satisfaction from the dates. We heavily filter based on beauty, but after you get to know a person, that matters less. But beauty is apparent, so that is what we react to. Once you get to know an ugly person, you might found out that they are really nice. They have to be, you can’t go through life being ugly inside and out. Data More than dating, this book is about data and data journalism. One prime example is Nate Silver. He writes popular mainstream articles that capture the attention of the populous based on data. Google knows everything about you. Data from Google, Facebook and Nate Silver show that 1 in 20 men are gay. You can look at percentage of gay porn searches, social graph and polls. In areas where there are less than 5% reported gay men, the difference is made up of people in the closet. They are out there, hiding. You can also figure out that 6-7% of people are looking for casual sex. Well, people except for straight women who are statistical outliers at 0.8 %. This indicates a cultural taboo for straight women. Sucks to be a straight women. This means there are 5% of straight women who won’t admit they are after casual sex. The median lifetime sexual partners is 4-5. The top 2% of gay men have 28% of the gay sex. You can learn so much from the data. Data is used to give you targeted ads that are relevant to you. These ads are what allows awesome services like Facebook and Google to be free. You pay for it with your data. Data is everywhere and you need a data scientist to harness it.

Stuff like height, political views, photos, essays, all of it is right there, easily sortable, easily searchable. It’s there to help people make judgments and fulfill their desires, and as fascinating as those judgments and desires may be to pick apart, there’s a side of it that I think does love a disservice. People make choices from the information we provide because they can, not because they necessarily should.

Funny how love works. Resources

Purchase Dataclysm on Amazon.com or check it our from your local library.