I’ve heard that #QuarantineLife can make it difficult to focus on work.

So hey it sure seems like lots of @NBA players like tattoos!

Ever wonder whether tattoos correlate with basketball performance? Seems like a super important question. (1/9)

Lo and behold! It turns out that someone coded every NBA player who had or did not have a visible tattoo in the 2013-2014 season! https://github.com/fivethirtyeight/data/blob/master/nba-tattoos/nba-tattoos-data.csv

So I downloaded the dataset & merged it with season player stats. (2/9)

I regressed players’ field goal%, 3-point %, assist%, and free throw% on a binary indicator for whether the player has tattoos...

...controlling for minutes played, usage, age, height, and weight. (3/9)

As you’d expect, minutes & usage are the strongest indicators of everything. Height & weight correlated with the stats that you’d expect.

But are tattoos markers of excellence? Are illustrated men better or worse basketball players? (4/9)

There was no difference in FG%, but it in 2013-14 players with tattoos were a statistically-significant 2% worse at free throws than their non-tatted brethren. (5/9).

Tattoos didn’t predict 3-point shooting, but we again see a statistically-significant 2% gap in assist percentage between tatted and non-tatted players.

Apparently some guys need tattoos that say: “ball hog.” (6/9).

Notably, the tattooed ball-hog effect declined as player minutes increased. Maybe coaches notice that the illustrated men don’t pass effectively and bench them? (7/9).

In case you're wondering: the correlations hold with or without team fixed effects. I couldn't do much more analytically with just one year's tattoo data (8/9).

In any case, I’m glad to have finally resolved this scientific mystery for the ages.(9/9)

Anyway, how is your #socialdistancing going?

Also, #BringBackTheSonics.