MLS 2016 Squad Data Confirms: Yup, There’s a Border Wall.

#SoccerNAFTA Initial Data Exploration — The Quick and The Dirty

Gio Dos Santos: 20 percent of the MLS player pool eligible to play for Mexico.

First, thanks to those who read the first #SoccerNAFTA manifesto! Lots of reads and one gold star makes this data scientist a happy boy. I wanted to make one thing clear: when I’m saying #TearDownThisWall, I’m not targeting the fan base culture like certain badly written articles by the New York Times Magazine. I’m not targeting the fact that many gringo fans, myself included, don’t speak Spanish.

I’m explicitly targeting the available player markets being pursued by MLS franchises and MLS itself. I hypothesize that MLS doesn’t target Mexican players, like, at all. Maybe if they did, they could attract more Liga MX fans, and American soccer culture would have a chance to integrate accordingly.

That said, I’ve compiled an initial data study. Because we data scientists love clean plain-text data, I’m putting my trust in footballsquads.co.uk. FIFA Nationality data is available for most clubs in all leagues from 2008 onward, which, well… I can go back further but this is just super convenient for now.

Data is super complicated, and we shouldn’t take too much stock in alarmist socialist narratives. However, my initial perusals of the 2016 MLS player pool (before I even factor in salary) is screaming at me: “This league just doesn't want to hire Mexican players. You need to #WokeUPbro.”

> MLS player pool count by FIFA Nationality
               Country   Nat     n               Pct
1                  USA   USA   269            49.09%
2               Canada   CAN    28             5.11%
3            Argentina   ARG    25             4.56%
4              England   ENG    19             3.47%
5             Colombia   COL    15             2.74%
6               Brazil   BRA    14             2.55%
7              Jamaica   JAM    14             2.55%
8               France   FRA    12             2.19%
9           Costa Rica   CRC    10             1.82%
10               Ghana   GHA     9             1.64%
# ...
16 Republic of Ireland   IRL     5             0.91%
17              Mexico   MEX     5             0.91%
18              Panama   PAN     5             0.91%
# ... with 47 more rows

Not even one percent (N = 548) of all MLS players are eligible to play for Mexico. It’s five players which, well, I guess, since there’s just five:

> MLS %>% filter(season == 2016, Nat == "MEX")
                  Name Nat               team Pos Number
1       Jesse González MEX          FC Dallas   G      1
2       Miguel Aguilar MEX        D.C. United   M     17
3   Giovani dos Santos MEX Los Angeles Galaxy   F     10
4        Raúl Mendiola MEX Los Angeles Galaxy   F     40
5 Eduardo Fern<e1>ndez MEX     Real Salt Lake   G      1

Yes, I could use some help debugging the encoding. And yes, this doesn’t include possible Mexican nationals who are eligible to play for other countries. I was going to include this in a much longer piece, but after just putting in a few lines of code, res ipsa loquitur.

That is, “the data speaks for itself.” #TEARDOWNTHISWALL, Mr. Garber!

+gs

Fifty-Fifties (Stray Observations):

• The data averages out to ~13–14 players per MLS team eligible to play for Team USA, or just over half of each MLS team roster. Obviously, MLS teams take advantage of the green card rule and push through international prospects as permanent residents (Ozzie Alonso, Stefan Frei, Dom Dwyer, etc.), but this just shows how widespread that particular tactic is.
• I’m happy to share this code with anyone who’s interested in looking up the whole dataset. Please feel free to respond, tweet, or message me. Thanks again!