data science, metrics, Uncategorized

Easy Metrics Versus Real-Life Calculus

I’ve always hated those annoying people who have “500+” LinkedIn connections. Obviously that meant I had to become one of them. So, last year, as a half-joke, I made that a yearly goal.

I didn’t make my goal, but that’s okay because if you make your goals 100% of the time, you’re not setting ambitious enough goals. Right? Continuing to use marketing clichés, I achieved 90% of my goal and increased my network size by 60%. Let’s make a chart or something!

I’ve always thought of numbers as something you use to convince people — a part of the rhetorical toolkit, like pathos. So I could use my 60% growth figure to defend my claim that I’d achieved my actual goal, which was to expand my network. And as a content strategist, I am very into success metrics. I get it. They’re reassuring and tangible.

I’m less reassured when I see people earnestly measuring their end-of-year worth by, for example, number of books read or miles run or lines of code written. This may seem cruel, but I totally want to send those people a small existential crisis, so they have to grapple with the potential meaninglessness of the figures they covet. The dirty secret of the quantified self, I suspect, is that it’s just a numbers fetish trying to fill the gaping void of potential existential meaninglessness.

Don’t get me wrong. There are obviously times when hard numbers help. I’ve been counting calories, which is tedious but effective, and seeing results on the scale. The numbers here provide a concrete sense of what’s happening, whereas otherwise I’d be guessing. And it’s also rhetorically useful to have numbers and records, if a doctor wants act like an ass.

But it’s a slippery slope. You start by taking comfort in the assuredness of straightforward numbers. Then that comfort sneakily seeps into all numbers, rendering otherwise good people incapable of finding meaning, or even reality, in things that aren’t quantifiable.

That said, the rhetorical value of numbers isn’t going away anytime soon. Lately I’ve started to pick up the calculus class I started last year. I signed up for it because so many code people were insistent that math was integral to code and I wanted to know what the fuss was about. So far, it’s pretty much the opposite: math seems exactly what I’ve already learned from doing programming. The good news is that I’ve nobly repressed my desire to scream, “These numbers are meaningless!” at every turn.

Because real life calculus, as I definite it, is never that simple. For example, the calculus I have to do when I’m thinking about seeing a movie. I have to calculate where the closest theater is, how the reviews are and which actors/writers/directors were involved, how much I trust reviews and how much I have to correct for my own system of not seeing rapey/boringly sexist/male-centered “art”, how much energy I have that day, what else I could do with my time, and then factor in the simple boolean of would I really, truly rather just sit on my couch and watch a Parks and Rec rerun, which has a known value of enjoyment.

Or a real-life situation that happened recently, in which I was super unhappy with my haircut and color, and I had to calculate exactly how unhappy I was versus the fact that it was the holiday season and correcting it would take a lot of time I didn’t particularly have.

I’m deliberately choosing first-world examples here. There are, of course, way less privileged examples, such as any time anyone has to make a major medical decision. Or the fact that people living in poverty have neither money nor time, and therefor suffer from decision fatigue at every turn.

My point is, mathematical calculus can’t help you, the individual, in any of these situations. I guess stats might help, if you had the wherewithal to transfer all your personal, abstract variables into a numeric scale, which would be yet more work. It’s not actually going to remedy decision fatigue if I have to make my movie calculations into a set of variables representing convenience, quality, get-off-the-couch-worthiness and the like. I’ve also seen people do math that (I think) shows how complex a given calculation is, but again, I fail to see how that’s going to help you when you have a problem.

And that’s where other people’s math steps in to impose itself on you. It happens every time a credit score is run on you, or Facebook’s spy network decides what ad to show you. And there are people who’d love to make an app that helps you make these decisions, I’m sure. And some of them may even be decent people with a genuine sense of the greater good. Lately I’ve been reading Cathy O’Neil’s Weapons of Math Destruction, and thinking about Vivienne Ming’s use of math to calculate the “tax” that non-straight, non-white, non-men pay for being in the word, as well as her algorithm that tries to challenge the meritocracy. It’s people like this who make me think that okay, maybe numbers and big data can be useful, in a moral sense.

So, that is the challenge of 2017: try to have an epiphany where I’m like, “OMG yes, I now see why people love numbers so much!!” Well, that and continuing to deliver small existential crises where needed. As a user researcher who cares about qualitative data, I need to make sure I can speak the language that the quant dogmatists do, both to see their side and to challenge it where needed.

Oh, and that stupid 500+ thing. I mean, I’m just so close now, it would be silly to stop.

UPDATE 1/20/2017: I just reached my goal. So, pretty darned close to what I’d planned, no?