OkCupid Data Scientist Max Shron Talks Data Analysis, Next OkTrends Post

It's been nine months already!
mshron headshot2 OkCupid Data Scientist Max Shron Talks Data Analysis, Next OkTrends Post

Mr. Shron

Last Friday, OkCupid’s data scientist Max Shron gave about 30 students at a Hackdays event a peek at what he does everyday: take terrifying amounts of raw data, make sense of it and draw insightful conclusions based on what he finds.
Students—including three girls!—gathered in a classroom on the third floor of NYU’s Courant Institute of Mathematical Sciences and got ready to nerd out.

Before the meetup, students downloaded a 1.4 GB raw data set—or three million

lines of code—from 311, New York City’s information and services hotline. From there, Mr. Shron spent two hours leading students through a series of commands in the terminal program to turn what was an intimidating and unmanageable amount of unusable statistics into a revealing picture about the number and different types of complaints 311 receives from New Yorkers. 

picture 34 OkCupid Data Scientist Max Shron Talks Data Analysis, Next OkTrends Post

It's starting to make sense...

Betabeat confesses that most of Mr. Shron’s presentation was way over our heads, and several students had to help get Betabeat back on track on more than one occasion. While Mr. Shron’s work is essential to OkCupid’s operations—“someone might want to know… how many lesbians there are in Topeka or something”—the only visible way users are able to see what he does is through the popular OkTrends blog that Mr. Shron helps co-founder Christian Rudder write. “Christian and I sit down and hash out how we can [realize] his vision,” Mr. Shron said. “We try to find what some basic ideas are and then I go ahead and try to reduce it down into something that’s useful for him and then he goes ahead and puts it together and writes something great out of it.” But that hasn’t happened since April when OkTrends published “10 Charts About Sex”— over nine months ago. The OkTrends blog has been inactive for nearly as long as Betabeat’s LiveJournal. We pressed Mr. Shron as to when we could expect something new from OkTrends, but he wasn’t budging. “We’re working on it,” Mr. Shron said. “I can tell you that we’ve got a couple things in the pipeline, hoping to get some stuff out soon. But it’s just been a busy time at OkCupid. I’m just really happy with the success of the blog… Both of our goals is to produce something great that helps people.” That and the data analysis. Mr. Shron said despite the fact that OkTrends is so popular it’s not about the notoriety or exposure for him. He just loves the work. “There’s really not a lot of personal motivation in it. I know the blog is popular, but I think about it as just something I do that people happen to know about.” When Mr. Shron is not working on the blog, he works on longer-term statistical and business projects, answering internal questions to calibrate the website’s functionality and creating and maintaining tools that allow him to answer those questions more easily. Mr. Shron, who used to do freelance data analysis for professors and journalists in Chicago, has been at OkCupid for about a year and a half. “I had a bunch of friends send me the job link for the OkCupid blog and I was like, ‘Oh, sounds like a good match, and they liked me,’” Mr. Shron said. “I’ve had to develop tools for dealing with the quantity of data we’ve got. It’s nice to work on things for long periods of time. Previous things I’d worked on had been maybe a couple weeks at most… it’s nice to draw those things out.” But this hiatus on blog posts has been drawn out enough. Give us something new for our chart hungry brains to snack on.

Follow Ben Weitzenkorn via RSS.

Comments

  1. OKCupid ask you questions about you in 3 parts: “First,
    you’ll give your own answer. Second, your ideal match’s answer. And
    finally, you’ll tell us how important (if at all) that question is to
    you. In this way OkCupid lets you – and only you – control how matches are sorted. The more you answer, the better your matches.”

    That is a GREAT
    MISTAKE because people often report partner preferences that are not
    compatible with their choices in real life [uncovered by Eastwick &
    Finkel (2008); Kurzban & Weeden (2007); Todd, Penke, Fasolo, &
    Lenton (2007)].Latest Research in Theories of Romantic Relationships Development shows: compatibility is all about a high level on personality* similarity* between prospective mates for long term mating with commitment.

    *personality measured with a normative test.

    *similarity: there are different ways to calculate similarity, it depends on how mathematically is defined.