1. People don’t actually know what you do. Sometimes you’re not so sure either, tbh.
2. 85% of your time is spent cleaning data. And it is SO MUCH FUN. #reallynotreally
3. You start putting priors on everything you do. Because if a prior distribution works in Bayesian statistics, it should work everywhere, amiright?
"Who wants to prior-game before the party tonight? I'm thinking non-informative."
4. Chances are high that you get bored really quickly with any one topic. So basically you have ADD.
5. Your idols include Hilary Mason and Nate Silver.
Also Taylor Swift because she'd prob be your bestie IRL.
6. When someone drops a statistics joke, you’re the first one to get it.
7. People expect you to know exactly how much to tip at big group dinners since you’re the math geek. But, like, you work with probability distributions. Not real numbers.
“The mean tip should be $5.03, plus or minus $0.76.”
8. That time when you realized you would never learn everything about computer science and statistics really sucked.
9. But then five minutes later you were like, “Oh shooooot. That means I’ll never stop learning!”
10. You like to brainstorm applications of data science to everything. Everythiiiing.
Sees excruciatingly long line at Whole Foods.
"Well I think I can use travelling salesman here."
11. You get into heated discussions about artificial intelligence.
"What if we're all just part of a giant experiment that's being run by another species??"
12. You know that “big data” and “deep learning” are highly controversial topics. How big is big? How deep is deep? We may never know…
13. You want the fastest, best, most robust solution, so you automate, optimize, and parallelize WHENEVER POSSIBLE.
*Cue "Stronger" by Kanye*
14. Only your data science friends truly understand the joy that comes from creating a great logistic classifier, or from visualizing publicly-available data sets in your spare time.
15. At your core, you’re really just a kick-ass story-teller.
Also you preface all stories with a list of assumptions about the data sets used.