In the wake of the Rapids trading for Mark-Anthony Kaye for big money, it sure seemed like there was a lot of consternation among FCD’s and MLS’ various social circles that MLS teams were broadly undervaluing such deals for a proven, domestic veteran who have the “know-how” needed to win in MLS. Here are a few exemplary tweets.
As best I can tell, this statistical quirk was first pointed out by our Philadelphian sister site Brotherly Game while commenting on a shift in roster-building under Ernst Tanner. Since taking over, FCD Technical Director Andre Zanotta has brought in just two players from inside the league: Fafa Picault and Eric Alexander. The national writers at The Athletic would call that a “neglectfully low” investment in the domestic player market.
Look, I want it said out front that a player’s experiences prior to his current club are not a good way to judge “MLS savvy”. Jimmy Maurer won two titles in the NASL and has played in MLS for half a decade, but isn’t counted as a veteran by the above. One-club draftees like Graham Zusi, Matt Hedges, and Ryan Hollingshead aren’t either, nor are one-club foreign signings like Boniek Garcia and Nico Lodeiro. If there’s such a thing as MLS experience, such players have it. Additionally, there’s an emerging generation of Academy products with more MLS experience than most journeymen – I’m thinking of guys like Bill Hamid, Russell Canouse, and Scott Caldwell.
Still, I wanted to see how FCD compared to the rest of MLS systematically, so I sorted every player on an MLS roster into one of four categories:
- Domestic Vets – Players with any professional experience as a player signed to a different club in the US or Canada (captures the Maurers of the world vs the above definition);
- Amateurs – Mostly college draftees, but sometimes players signed directly from other amateur setups (eg: Kalil ElMedkhar, Julian Araujo, and Bryce Duke);
- Academy players – Players signed directly from the club’s youth developmental setup;
- Foreigners – players signed without domestic experience from foreign countries (which includes some Americans, like Indiana Vassilev, who jumped overseas as amateurs).
Here’s FCD as an example.
You can see the whole league arrayed below, sorted by their reliance on Domestic Vets. FC Dallas falls on the far right side, using only 7% of its roster on Domestic Vets. While fans seem to position this shortage against the struggling foreign signings (for good reason, as we’ll see below), it’s actually the club’s much higher than average use of amateur and Academy signings that crowds out domestic vets, with FCD spending an almost exactly average number of roster spots on foreigners (even despite Zanotta’s emphasis thereon since he took over).
If you run simple correlations between the percentages of MLS rosters taken up by each player class, you can get an idea of how decision-makers weigh them against one another. While all are negative, since a spot spent on one class takes a possible spot away from the others, some seem to interact more directly than others. For example, the number of Academy players on a roster doesn’t seem to affect the number of foreign players (r = -0.03), suggesting those two classes of players rarely serve the same purposes. On the other hand, adding domestic vets seems to directly reduce the spots for foreign (r=-0.69) and Academy players on the roster. Academy and amateur signings tend to crowd each other out, while foreign signings seem to reduce spots for amateurs.
Why do those scores pop up? I argue it’s a simple function of expected costs (in transfer fees, other acquisition fees, salaries, and scarce roster assets like international spots) and expected ability. The below rubric is approximately how I see it, recognizing there are exceptions to every rule (obviously there are many college players that perform well; foreign players bomb out all the time) but trying to keep an eye on expected ability level, which is informed by average past experience.
Taking those (illustrative) rankings into account, you would probably build your generic MLS roster something like the following. High-salary or high-TAM/transfer-fee veterans have to be at least key substitutes or they’re a misallocation of resources. Maybe you can justify paying Eric Alexander $270k to be the 18th or 19th player on the roster, but it’s a stretch. Conversely, if your club is bringing in a draft pick to be a key starter, odds are something’s gone wrong with your roster build.
So why won’t FC Dallas buy domestic veterans? A few thoughts:
- When Paxton Pomykal, or Jesus Ferreira, or Ricardo Pepi get fat contracts and starting roles, they are taking roster spots that go to domestic vets or foreign signings 80%+ of the time elsewhere in the league. Yes, Pepi may get sold and need to be replaced, but it sure seems like Pomykal and Ferreira may stick around for a while, and there’s a fair portion of FCD’s Academy talent that will stick around too and fill up the upper half of the roster. If Edwin Cerrillo’s in the midfield rotation for the next decade, then FCD will have less need for the Quignons and Alexanders and Ricaurtes of the world. FCD’s Academy success reduces the club’s need for outside domestic talent to support the middle- and upper-thirds of the roster.
- FC Dallas is as good as anyone in the league at finding value in the draft and retaining those players. Hedges, Hollingshead, Nkosi Tafari, and John Nelson have all played significant roles for the team this season. Only San Jose and New England devote more roster space to drafted players. For the majority of the league, those spots are filled by domestic vets or foreign players.
- Competencies are expensive to build and building the “foreign player acquisition” competency is more strategically important than the “domestic vet acquisition” competency. If you want to find undervalued domestic players, that takes scouting resources and data resources and relationships and the scarce time and attention of the front office. It’s the same thing for foreign players, true, but in an under-resourced FCD front office it may not be feasible to do everything (as richer clubs like Seattle and Atlanta can). So why choose to try and get your foreign signings right rather than the domestic signings? Because signing Raul Ruidiaz for $7 million instead of Franco Jara for $3 million a year is WAY more important to the club’s success than signing Mark-Anthony Kaye for $1 million instead of Bryan Acosta for $2+ million or Corey Baird for $500k instead of Jader Obrian for a high six-figure fee. If you want to compete at a high level, you have to find high-level players to be your DPs, and that means going to foreign markets 9/10 times. The nice thing is that when you invest in finding stars, you can use the same infrastructure to find guys like Obrian or Freddy Vargas too.
“I’m sorry, what’s that? Folks, I’m now being told Obrian and Vargas are bad.” That may be true (I think it’s early to tell), but that criticism should be leveled against FCD’s execution of its roster-building strategy rather than the strategy itself. Would it be better if FCD could be good at everything? Yes, but that may not be possible given the team’s current level of investment. This is why I’ve long advocated for FCD to reinvest transfer surpluses into things like adding scouts, bringing FCD out of the stone age of soccer analytics, and funding better NTSC salaries rather than solely into transfer fees and salaries for first-team targets. This stuff may already be on the way.
FC Dallas’ recent history of roster construction under Zanotta has been as checkered as can be. However, I think taking that and dismissing it as the result of poor strategy is naïve. Given FCD’s proficiencies (getting contributors from the draft and the Academy better than almost anyone in the league) and level of player acquisition investment (low), the decision to pass over the domestic market to focus on getting a high rate of return on foreign transfers is an easily defensible one. Next step? Stop not getting a high rate of return on foreign transfers.