Modeling like a model

Wayne Lee
The Startup
Published in
3 min readMay 25, 2020
A true model, RuPaul from RuPaul’s Drag Race Season 10
A true model, RuPaul from RuPaul’s Drag Race Season 10

A student in my office hours asked me “do you think some people are born with it?” I resisted the urge to reply “Maybe it’s Maybelline!” and instead asked her about her profession before coming back to school.

Turns out she was a professional model who came back to college for a career change. She was in my data wrangling course that targeted students who used a calculator before but couldn’t tell you what separated the values in a .csv file (hint: the “c” stands for comma!).

So I asked her if modeling was just about putting on pretty clothes and walking down the aisle? The answer was an obvious no. As a dancer, I knew a walk is much harder than a choreographed combination where you can distract your audience with your technique. Everyone knows how to walk, so what differentiates a walk that captivates the audience vs my stroll from the bedroom to the living room (#covidlife)?

In a walk, where you look, how far your knees bend, do you smile, and who is around you (the audience could be in the dance!) will impact its perception. The amount of detail is nauseating! Now imagine the impact of those choices on each camera pointed at you from every angle, yikes! Who knew all the things we took for granted could have so much nuance?

I argue that the difference is when any movement or stillness is a result of an thoughtful choice rather than a mindless reaction. This might be weird given our assumptions about artists being individuals who have heightened emotional responses but I think true artists are those who can consistently produce art rather than the one-shot wonders.

These choices are just painfully obvious to people when they’re learning programming because of the error messages. If you do not tell the program to only pass the numeric values to the function, it cannot read your mind and it will fail. My students quickly learn to not take automation for granted.

It’s not always obvious when we are making choices though. In data science, when I see junior data scientists not questioning the data, not trying out simpler models, not diagnosing the models, or not knowing the business, you’re just going through the motions like a model-wanna-be. Sometimes the final product is the same between those who made those choices consciously vs not but only one will be able to consistently produce good work.

So I told my student that few people are “born with it” when it comes to programming. Some exceptions may exist, but the more reliable approach to learning is to unlearn what you took for granted, make choices, and keep coming to office hours ;)

It’s almost surely Mabelline!

--

--

Wayne Lee
The Startup

Educator who believes Statistics can save the world!