One finding of big data, applied to the realm of dating, that is mentioned every now and again on my other blog is the infamous evidence dating site OKCupid provided for the unreasonable pickiness of women. While men rate women so that approximately a normal distribution emerges, where most women are average and very few are either fugly or super hot. Yet, women, which of course includes all those utterly average ones, consider 80 % of men to be below average. When I learned that the guy who is responsible for the OKCupid blog wrote a book on his findings, including the one I just briefly summarized, I was intrigued. His book is called Dataclysm. Its purpose is to communicate to the layman why big data is useful.
Dataclysm is a collection of a few essays that cover, broadly, social interactions online, and what the data you share, knowingly or not, can tell about you. There is not much of a narrative connection between chapters — it’s all about a bunch of anecdotes, really —, so in the end, the author has to pull some heartstrings by telling you how he looked up some soldier in Vietnam who happened to die within a few days after joining the fight, and how much this emotionally affected him. In a nutshell, the message of the book seems to be that big data makes the life and death of the average schmuck more palpable. While the rank and file who got killed in ‘Nam may only have had their date of death recorded publicly, future generations will be able to do research on any shithead because she posted her entire life on Facebook and Instagram. What a wonderful vision of the future that is!
Writing a popular science book on statistical analysis is quite the undertaking. Yet, that is not what Rudder in the end does. The mathematical and statistical content is quite minimal and falls far short from what you should have picked up already in middle school. As such, the book is quite a disappointment. That being said, I’m clearly not the audience. I kept reading, though, but with a more critical lens. Yet, I quickly found myself sifting through politically correct leftist propaganda that made my hair curl. Here you have a Harvard-educated mathematician who sits at a wonderful trove of “dating big data”. Yet, instead of just reporting what is, he reports what ought to be. We get a spin, and plenty of it.
On p. 49, Rudder states that women get more messages if their ratings show more “variance”, meaning that there is a big spread in the ratings they have received. Yet, instead of that kind of waffling, my first thought was that it could very well be that men feel more confident messaging women who are not conventionally attractive. That would strike me as a very plausible hypothesis one could easily have looked into with the data at hand. Yet, no such insight is provided. Instead, his liberal hamster is spinning like crazy. Look at this:
I can imagine our man browsing her profile, circling his cursor, thinking to himself: I bet she doesn’t meet many guys who think she’s awesome. In fact, I’m actually into herf or her quirks, not in spite of them. This is my diamond in the rough, and so on. To some degree, her very unpopularity is what makes her attractive to him.
Nope. It’s much simpler to assume that the average guy knows in what league he is in. Thus, he goes avoids going after the hottest girls because he does not have much of a chance with them. However, this does not mean that he finds average-looking women anything but average looking. In Rudder’s world, though, there is supposedly no such thing as conventional beauty:
Looking at the phenomenon from the opposite angle—the low-variance side—a relatively attractive woman with consistent scores is someone any guy would consider conventionally pretty. And she therefore might seem to be more popular than she really is. Broad appeal gives the impression that other guys are after her, too, and that makes her incrementally less appealing. Our interested but on-the-fence guy moves on.
No, dude, guys find hot chicks hot. However, the average guy does not go after hot chicks because they know, based on experience, that they can’t get them. I bet that they still fantasize about those women when they bang some ugly dumpling with a solid 2.0/5.0 rating.
What do you think of “being yourself”? That’s a fine piece of advice if you are a woman and don’t want to hurt some guy’s feeling. It’s also a fine piece of advice if you are a liberal dipshit writer who does not want to offend any of his readers. Here’s Christian Rudder’s version of it:
So at the end of it, given that everyone on Earth has some kind of flaw, the
real moral here is: be yourself and be brave about it.
The average woman would jump in attractiveness quite considerably if she only managed to slim down those fucking legs. Likewise, the average guy can improve his looks rather substantially by slimming down. Then work out a bit, get a decent haircut, and put on some nice clothes. Suddenly, that woman you did not dare to approach or message is much more in your reach.
I’m skeptical when I see sloppy reasoning. I have read way too many papers in the social sciences — man, you should see some of the crap the supposedly most esteemed departments of economics put out. I tend to bin anything that does not pass the smell test and don’t bother reading on. I made an exception for Rudder’s book, but I ended up regretting it. Here’s one example of his style of sophomoric thinking (p. 66):
The best messages, the ones that get the highest response rate, are now only 40 to 60 characters long.
Come on! It is much more plausible that you send shorter messages in an ongoing conversation. Consequently, we are looking at a tautology: short messages are the norm in an active exchange. Consequently, these are the messages that get responses. Again, a bit of legwork could have turned this into a much meatier analysis, for instance if he had done some analysis based on message length and duration between messages.
Later on in his book he comes across as even more naive. He writes,
The citizens of most countries are usually only concerned with one constitution— their own—but Google has assembled all nine hundred such documents drafted since 1787. Combined and quantified, they give emerging nations—five new constitutions are written every year—a better chance at a durable government because they can see what’s worked and what hasn’t in the past. Here, data unlocks a better future because, as Constitutes website points out: in a constitution, “even a single comma can make a huge difference.”
This is so absurd that I can’t believe he wrote that. His statement would make sense if we were dealing with a large number of essentially identical countries. As this is not the case, you can scrap the entire project. Who would think that comma placement will turn a highly unstable society in a stable one? Oh, right, that must be the same people who believe that we can turn countries that have never intellectually left the Dark Ages into functioning democracies within a few decades. To me it seems that Christian Rudder is unable to think critically.
There is a bit of an excursion to graph theory as well, but it is only discussed in extremely superficial terms. Thus, I am not sure whether my own understanding of the example Rudder provides is imperfect, or whether he simply deludes himself. On p. 75, he states that he and his wife “are much less likely than other couples to break up”, which he alleges is due to them having many mutual connections in their respective social graphs. So far, so good. Yet, a few pages later (p. 78), there is some seemingly contradictory data. Some piece of research is cited according to which the strongest marriages have the feature that there is no overlap between social circles. I perked up a little bit, like I tend to do whenever I see some bullshit reasoning unfold in front of my very eyes. Rudder did not disappoint, because he then hamsters like crazy and writes that for a certain subset (!) of his social network, he and his wife are the only connectors to some other subset. Consequently, he twists reality until it suits his belief. However, if you show two social graphs with a big overlap, you can’t then afterwards say that there is no overlap, because there is no overlap in particular subsets of that graph. Here you have it, guys: a Harvard-educated mathematician, hamstering worse that some CEO’s side chick.
As I was reading on, I was hoping to see his hot take on “racism” and other progressive talking points as well. Again, he did not disappoint. He even states that he has a vision for “how most of us know the world should be”. Consequently, he is upset because there is not enough inter-racial dating on OKCupid:
As we’ve seen, the pattern is so woven-in that relatively recent additions to our society, Asians and Latinos, have adopted it, too.
For this liberal snowflake, it’s a problem that the least number of OKCupid users wants to date black women, but black men aren’t doing so hot either. He then tries to shame Asians and Latinos for likewise “adopting” societal racism, or whatever the term for that is. Nope, it can’t just be that Asian women and White men are the most attractive. The work of progressives isn’t done until we are all brown, right? You are just not going to make significant numbers of people want to have sex with a man or woman they are not attracted to, but feel free to shout, “raycis!” anyway.
He then drags a piece of research out from under the carpet according to which even blind people prefer their own sex. We read that, “blind people’s attitudes on race reflect a lifetime of cultural absorption, as opposed to any visual reality.” You can’t make this shit up. That Rudder is a complete cultural Marxist has been made clear at that point already, but in order to eradicate any doubt, he later on states that, “dating is just the frontier of a vast cultural mass that will take decades to rearrange.” Decades? There never was significant inter-racial dating, and there never will be.
About halfway through Dataclysm, I dropped all attempt of reading this book to learn anything from it. Instead, I viewed it as a source of entertainment. One tidbit is his statement, referring to women exposing themselves to the world on Instagram, “The blogs, created by women, are truly the epitome of the male gaze — and I say this as a person reflexively skeptical of the language of the academic left.” It probably takes a Harvard education to be a total lefty, yet believe you are not (it’s just how things have to be!), and also that you are “reflexively skeptical” of leftist ideology while you are soaking in it from head to toe. Rudder drops the worst dog whistle of the left when he casually remarks, when discussing politics out of nowhere, that, “whiteness felt most attacked”. Another highlight is his comparison of some guy’s action that “this very well might’ve been some twenty-first-century Nixon working his ratfucking magic.” He really likes Obama, by the way.
Dataclysm completely jumps the shark when Rudder writes, “In fact, there’s a growing consensus among psychologists that men and women are fundamentally very similar”. Right aferwards, he shows top words used by men and women on Twitter. The top terms of the males were “good bro” and “ps4”, i.e. the PlayStation 4 video game console. On the side of the empowered women, it’s “my nails done” and “my sissy”. Oh, but of course, both sexes are “fundamentally very similar”! The Orwellian double-think on display is truly staggering.
We also learn that is wife is not very attractive, and insecure about it, too, because both of them want their daughter to “not be too pretty.” Somehow, this is then spun into a problem of the sexes because no parent supposedly would want any limitations for their son. Well, there are uneducated stupid poor people out there who don’t want their kids to get a proper education because this would mean that they are “better than them.” Some lack brains, others looks, but the poverty mindset with the resulting crab mentality can apparently affect people of all social strata, even one-percenter couples who would feel uncomfortable if the daughter turned out prettier than the wife.
While Dataclysm is only a few years old — it came out in 2014 — it feels quite dated at times. When discussing some large-scale data analytics at Facebook, we learn:
This stuff was computed from three years of data collected from people who joined Facebook after decades of being on Earth without it. What will be possible when someone’s been using these services since she was a child?
That won’t happen. Teenage engagement on Facebook is down, so the NSA wet dream of people creating their own dossiers for them is not going to happen. The same is true with “Google Flu”, which supposedly tracks flu epidemics. That did turn out about as well as IBM’s Watson, i.e. not so very well. That product has been shuttered in the meantime.
I think Dataclysm could have been a decent book, had it been shrunk down and stripped of its insufferable leftist bias. Also, the language is incredibly juvenile. I mentioned “ratfucking magic” above, but Rudder can do a lot worse, like in the following:
So social scientists are very cagey with data sets; actually, more than yeti, they treat them like big bags of weed—possessive, slightly paranoid, always curious who else is holding and how dank that shit is.
Oh my! Sometimes, Dataclysm just makes you cringe.
Don’t buy Dataclysm. It’s shallow, incredibly biased, and full of sloppy reasoning. The less you know about statistics and the poorer your reasoning skills are developed, the more you will be impressed by it, though. This would say a lot more about you than that book, though.
Did you enjoy this article? Excellent! If you want to support what I am doing, then please consider buying my excellent books, the latest of which is Meditation Without Bullshit or donating to the upkeep of this site. If you want tailored advice, I am available for one-on-one consultation sessions.
5 thoughts on “Review: Dataclysm – Who We Are When We Think No One is Looking by Christian Rudder”
“Some piece of research is cited according to which the strongest marriages have the feature that there is no overlap between social circles.”
Do you know what he means by “overlap between social circles”?
I’ll tell you why I pick this up. My social circle (and man, it’s a small one) has zero overlap with the one of my girlfriend. But that’s more interest-driven. Her friends (she has no close male friends in her circle) consist of girls that I’d never hang out with. We just don’t have the same topics. And she would’t hang out with my boys, because we talk about stuff that doesn’t interest her. BUT: I would definitely say that both circles share some common ground when it comes to values.
So how does he even get to observe “overlap”, when it’s never really tested unless you force people to discuss? I find such statements strange to be honest. There seems to be no basis for them.
It’s a lot simpler than that. In this context, two social circles not having any overlap means that there is nobody who is part of both social circles.
Well his argument is shit then.
I remember when this book came out…didn’t read it…thx for summing it up!
I recall that when the book came out, there was a blog post (I can’t remmeber if the OKCupid founder wrote it) but the gist was looking a racial preferences of women. Pretty much showed that white guys fared better, then blacks, then asians (in terms of women preference).
For guys, the preference was white chicks but asian women were also highly ranked.
Freakonomics touched on racial preferences between the genders and I recall it bore out the same trend. What I drew from Freakonomics is basically ‘people say one thing in a survey, but act differently in reality’.
I wonder, Aaron, if you heard of the book “Billion wicked thoughts” (link: https://bit.ly/2E7fP4w) ?
Apparently, it’s a stats and data analysis about human sexual preferences based on historic web searches. Sounds promising but I wonder if it’s all hokum.
I’ll check out that book.