Good Quant, Bad Quant

What is your estimate of the amount of quant analysis or data analysis or data science that does neither descriptive nor inferential statistics? My guess is 99.44%. Sports stats being the most irritating. They're just numbers, fur cryin out loud. Which is not to say that decisions can't be bolstered by mere numbers. It happens a lot. One recent day in the life of The Failing New York Times brought forth some good quant and some bad quant. And no statistics to be found.

First, we get a story about a cranky investor, on the heels of the tenth anniversary of Lehman bullet to the back of the head.

Mr. Rasmussen, founding partner of Verdad Capital in Boston, has written an article and two reports that make a case against investments in private equity, venture capital and private real estate, and he has piles of data to back up his argument.

If you do read the article, there's no analysis beyond stating the obvious: buy small and cheap and sell out expensive. Nice work if you can get it. And no p-values to argue about.

Second, is an article Serena's meltdown, if it were such, and whether her (and others) claim that girls are treated more harshly by officials than boys. Bad quant in this one.