In the name of all things good and true, New York Times, will you keep fucking idiots like David Brooks out of my wheelhouse? It's just embarrassing, is what it is.
One would have thought that after Brooks went on national TV to explain how Nate Silver's election model was an example of cupidity and intellectual overreach, only for said model to end up not so much correct as spot on to a legitimately terrifying extent, he'd have the good sense to shut his pie-hole regarding the nature of data manipulation.
But no! Not this man! Not this man who's teaching a university class on humility and making his students read his own fucking book on humility (next semester: Silvio Berlusconi will show up to fuck as many co-eds as necessary before they understand the importance of virginity). Once again it's critically important that someone - who always happens to be David Brooks - must stand up and argue that only the arrogant believe the discipline they've dedicated their lives to can do what they say it does. One might think the true arrogance lies in a man with no detailed knowledge of a field lecturing us on what that field's limitations are, but since Brooks is an expert on humility, you'd look pretty stupid trying to point that out to him. It's not like humility is an easy concept to grasp without training, after all; it's not statistical analysis or anything.
I could rebut "What Data Can't Do" point by point, starting with the idea that the human mind is bad at maths because you can't calculate an irrational number in your head - roughly equivalent to arguing the human mind is bad at language because lot's of people can't spell chrysanthemum - and moving on to more serious points, such as his entire paragraph on data haystacks is countered using exactly five words: "Bonferroni correction, you smug prick".
Mainly, though, it's not worth the effort, because Brooks thinking regarding statistics suffers from the same problem as his thinking on everything else; a tireless desire to paint intellectual laziness as a principled position. At least 80% of the Brooks pieces I've read (and I admit that's not a huge sample, life being too short and all) are based on or at least involve the same maddening logical fallacy: "I don't know how you would do this so I don't believe it can be done". That in itself, of course, is just a gussied-up version of "I don't have the knowledge needed to rebut this argument I dislike, so I'll insist you don't have the knowledge either". Brooks course on humility sounds like the worst use of university time since my alma mater let Tony Blair in to get bloodstains all over our new building, but if he wanted to teach a course in using ignorance as a weapon, I'd sign up for it like a shot.
The aforementioned haranguing of Nate Silver (in which Brooks referred to him and statisticians like him as "wizards" living in "silly-land") is just the most obvious example of this particular brand of strategic ignorance, but his latest article comes close. Brooks knows sweet fuck-all about how to construct, test, improve and implement a statistical model (seriously, he doesn't think we have ways of dealing with the lack of a control group?), but that doesn't mean he should have to feel uncomfortable slapping together the thoughts he came up with in the shower and showing it in exchange for money to millions of people, many of whom could have put him right in seconds had he asked.
How fucking humble of him.