Per Capita Player Production in American Men’s Soccer: A Sort-Of MLS Season Preview
With the good news that the MLS season will indeed start this week, I’d like to offer my esoteric version of a season preview: where exactly does the current crop of elite American players come from? Who are the boys (Luis Gil at 16 is—as far as I know—the youngest player in the league) and men (Pat Onstad at 42 is—as far as I can tell—the oldest) we’ll watch this season?
My questions are mostly born out of an amateur interest in cultural geography and a general curiosity about youth development; knowing where players comes from provides an important indicator of how the game works in different places. Ironically, the globalization and commercialization of the modern professional game often obscures the importance of place—while pro teams symbolically represent cities and regions, very few of their players actually come from those locales. It seems to me, for example, that knowing Real Salt Lake won last year’s MLS Cup tells me much less about soccer in Utah than the fact that the state, with a population of 2.5 million, has produced only one current MLS player (Justin Braun of Chivas USA grew up in Salt Lake City).
I was also provoked by an interview published here on Pitch Invasion a few months ago between Peter Wilt and ‘St. Louis based soccer executive Jeff Cooper.’ Wilt asked “Is the ‘St. Louis as a soccer hotbed’ notion a myth associated with the history of the sport’s support there or is St. Louis truly still ahead of the rest of the Midwest, and nation, in soccer interest and development?” and Cooper replied: “Per capita, St. Louis still produces more elite level players than any market.” That, as we say in the social science business, is an empirical question.
Of course, if you try to parse Cooper’s statement it gets tricky: what exactly qualifies as an “elite level player” and how do you define a “market” (and, I might add, has hyper-capitalism advanced to the point that we now live in “markets” rather than “cities”)? But I gave it a stab—and by my calculations Cooper is not quite right. Though, in fairness, he’s also not far off.
So in the spirit of Soccernomics (which tried to statistically analyze soccer success internationally through cross-national comparisons), The Best Eleven (which has had some excellent maps of where MLS players come from), Kenn Tomasch (who has done some interesting comparisons of things such as the percentage of American born players in MLS vs the old NASL), and others I’m probably not aware of, I present and interpret a poor-man’s geographic analysis of the current MLS (and US Men’s National Team—USMNT) players based on where they spent their formative years.
What counts as “formative years?” Honestly, that is something we could debate. I went with where players spent their high school years. Yes, I know most elite players don’t really bother with high school anymore. But it seems to me as though those years (while focused on club soccer) are still key developmentally since they are when most players have to finally decide if they are going to commit to the game or move on to real life.
That also means that I included MLS players who are not necessarily US nationals, but who spent their adolescence in a US city. So, for an example from down the street, I counted both Danny Mwanga and Alex Nimo as Oregonians—even though Mwanga emigrated from Congo at age 14 and Nimo’s family emigrated from Liberia by way of Ghana at age 9. I was honestly surprised by how many ‘international’ players in MLS actually spent significant portions of their youth in the US (other prominent examples include Toronto FC’s ‘Swiss’ goalkeeper Stefan Frei, who went to high school in the San Francisco Bay Area, Philadelphia’s ‘Brazilian’ midfielder Stefani Miglioranzi, who spent his adolescence in New York, and New England’s ‘Grenadian’ midfielder Shalrie Joseph, who spent his teen years in Brooklyn). In fact, one of the main inferences I took away from looking at everything was that some degree of immigrant presence is probably a significant part of success in American soccer—both for immigrants and for non-immigrants challenged to compete.
Otherwise, my main method involved way too much (and, sadly, also way too little) time on Wikipedia and MLSnet. I also learned that I was not the first to indulge this strange curiosity, as there is a whole Wikipedia page devoted to the home states of MLS, USL, Yanks Abroad, etc.—but however lovingly it must have been put together, that particular page now seems several years out of date. So I mostly used the more current general list of MLS players and USMNT players currently in the World Cup picture, and cross-checked for the best information I could find about where they spent adolescence. I ended up with a spreadsheet of 249 MLS / USMNT ‘American’ players. I’m sure I got some of the specifics wrong (if I ever had the time and resources to do a more rigorous analysis, I’d jump at the opportunity), but the beauty of statistics is that with large enough numbers the patterns can still be meaningful. And after putting it all together I’m confident the general picture is about right.
Looking at a state by state comparison, the most obvious conclusion is that California dominates American soccer. I count 58 players produced by the Golden State, and no other state is even close (Texas is second with 17, Florida is third with 15, and Illinois fourth with 13). Further, most of those Californians are actually from the greater LA or greater San Diego area (41 by my count)—Cal South as they say in US Youth Soccer.
Of course, California is also the most populous US state (with Texas second, Florida fourth, and Illinois fifth)—so it may actually be more informative to look at players produced by population. From that angle California’s 58 players from 36.5 million people is still not bad, but other ratios compare favorably:
- Colorado: 9 players, 4.7 million people.
- Oregon: 7 players, 3.7 million people.
- Hawaii: 2 players (Brian Ching and Zach Scott), 1.3 million people.
- Maryland: 8 players, 5.6 million people.
- Missouri: 8 players, 5.8 million people.
- Arizona: 8 players, 6.2 million people.
Then there are the states that don’t seem to be doing so well. In terms of large states, the most glaring absentee is New York: with around 20 million people the state has produced only 11 players. Of course, there are more players from the New Jersey suburbs that could count as the New York metropolitan area—but even calculated by metropolitan area (which I look at below) New York does not do as well as it should based on population. Why not? My guesses would be a combination of bad soccer weather, lots of other stuff for a young athlete to do, and maybe a lack of field space in the more populated areas? But those are just guesses.
It is also interesting to look at the relatively large states (with over a million people) that don’t seem to produce any players—in these cases I’d guess the issue is more about the absence of ‘soccer culture’ than anything else:
- Alabama (pop. 4.6 million)
- South Carolina (pop. 4.3 million)
- Kentucky (pop. 4.2 million)
- Iowa (pop. 3 million)
- Arkansas (pop. 2.8 million)
- Nevada (pop. 2.5 million *though I probably could have counted Hercules Gomez—since he played youth soccer in Las Vegas)
- New Mexico (pop. 2 million)
- Idaho (pop. 1.5 million)
- Nebraska (pop. 1.8 million)
- West Virginia (pop. 1.8 million)
By Metropolitan Area
To get back to Jeff Cooper’s claim that per capita St. Louis “still produces more elite level players than any market,” I looked at players produced by metropolitan areas—and I tended to be liberal in defining a metropolitan area. So, for example, I grouped Baltimore, Washington DC, and Northern Virginia together even though that may have meant including players developed in very different youth scenes (and, by the way, that greater DC agglomeration has produced 12 current MLS/USMNT players—which is not bad for 8.3 million in population, but not great either).
With that caveat, by my calculations St. Louis is fourth in players produced per capita among US metropolitan areas. Here they are in order:
- The greater Denver area (including Colorado Springs) has produced 9 players with about 3 million in population – a rate of about 1 per every 330,000 people.
- The greater Raleigh / Durham area of North Carolina (which I stretched to include Greenville to put Michael Harrington in the mix) has produced 5 players with only about 1.8 million people – a rate of about 1 per every 370,000 people.
- The neighboring Greensboro / Winston-Salem area is not far behind—4 players from 1.5 million people means about 1 per 380,000 people.
- St. Louis has produced 7 players from about 2.9 million people—about 1 per every 410,000.
- (Actually, Honorable Mention) The Columbus Ohio area has produced 4 players from about 2 million people—about the same rate as the San Diego area (7 players from 3 million people), the Sacramento area (5 players from 2.4 million people), and even greater Los Angeles (34 players from about 17.8 million people) despite a much less amenable climate.
On the other side of the slate, several metropolitan areas struck me as underperforming—these are not exactly in order, but they all seem as though they should produce more players:
- The greater New York area has produced 14 (or so) players, but has about 22 million people (a rate of about 1 per every 1.6 million people).
- The Minneapolis / St. Paul area currently only seems to have one active player two active players who are both non-Minnesota natives (Abdus Ibrahim—who is an Ethiopian immigrant playing for Toronto FC –correction–and Teal Bunbury who is listed as Canadian), despite about 3.5 million in the area.
- By stretching, I could find 2 players from the Detroit area (Michael Holody is from the suburb of Clarkston, but he is a developmental player for Colorado; while Jacob Peterson of Toronto FC is actually from Portage, Michigan—which is only slightly closer to Detroit than Chicago), despite the Detroit metro area having about 5.3 million people.
- I only count 4 players from the Boston area, which has about 7.5 million people (a rate of about 1 per 1.9 million).
- Miami / Fort Lauderdale is not terrible, with 5 players from about 5.5 million people, but considering the weather and the locale I might have expected a bit more (the Tampa area, with about half the population, has also produced 5 players).
As a Canadian side-note, Toronto compares favorably to other North American cities with 9 players out of about 5 million people, but I could only find 1 from Vancouver’s 2.1 million (Pat Onstad playing for Houston).
So what is the moral of the story? If I were to speculate on a Soccernomics-esque formula for the variables that contribute to American player production my very preliminary analysis would include:
- Population size. This is boring and obvious, but it also seems to be the strongest predictor: With more people there are better odds that some will turn out to be good soccer players.
- Climate. It’s probably not controversial to speculate that places such as Arizona (8 players) and Georgia (5 players) do better than places such as Michigan (2 players) and Minnesota (1 2 players) for the simple reason that a developing player can get outside all year round.
- Soccer culture. This is a fuzzy concept, but I suspect in the US more than other parts of the world there are wide regional differences in attitudes toward the game. I’d hypothesize that this is why North Carolina (11 players) does better than South Carolina (0 players), and Oregon (7 players) does better than Alabama (0 players).
- Immigrants. This is admittedly more speculative, but in looking up current MLS/USMNT player backgrounds I was struck by how many are either first or second generation immigrants. Charlie Davies is the only player from New Hampshire, which is not exactly a soccer hotbed, so it seems significant that his father is a Gambian immigrant. Even in southern California, I suspect players such as Sacha Kljestan (whose father emigrated from Bosnia) and Carlos Bocanegra (whose father was born in Mexico, but raised in California) benefitted from exposure to diverse soccer communities.
For me what is most conspicuously absent from the list of variables that seem to matter in American player production is the presence of an MLS team itself. With the possible exception of Columbus, most MLS cities look pretty average in producing youth players compared with non-MLS cities. In fact, in some cases it seems that having good college programs in an area may better correlate with player production than having an MLS team: I suspect that relatively small parts of North Carolina produce more players than relatively large MLS cities such as Boston largely because Wake Forest, UNC, NC State, and Duke are viable options for quality soccer.
That may well change over time as MLS youth academies become better developed—but it also may not since American soccer is a persistently odd duck. So ultimately, in the spirit of a season preview, the lesson here seems to be that in the current American soccer scene Jeff Cooper was right: MLS teams are indeed more representative of “markets” than “cities.”
(Note: there ended up being too many specific locales and names to list each individually—but I now have most of them in my spreadsheet. So if anyone is curious about other specific places, players, and proportions, feel free to leave a comment with any queries and I will try to respond)