How did the MLS do in the 2014 World Cup?

Taking another look at Jürgen Klinsmann’s theory that his USMNT players should push themselves in a top UEFA league rather than the MLS

With Jozy Altidore kicking off his MLS return the way he did after a truly horrible goal drought while playing for Sunderland (only scoring once in 42 appearances) and Jürgen Klinsmann’s USMNT floundering since the 2014 World Cup, it seems as good a time as any to take a closer examination at Klinsmann’s controversial (to U.S. soccer, at least) opinion that the USMNT players are better off playing in the top European leagues than the MLS.

I’m no expert at statistics (as much as I wish I was in this world of increasingly advanced sports metrics), but I know enough about the basics to be able to run some tests that can move discussion beyond mere conjecture (which is really all that Klinsmann’s theory is, since there’s no way to observe 2 parallel universes where the only difference is that Landon Donovan plays for Everton year-round in one and the Los Angeles Galaxy year-round in the other). Thus, I won’t be engaging Klinsmann’s theory directly: instead of trying to calculate whether a player would’ve done better if they’d played in a top European league and not the MLS (and vice versa), I want to analyze how the MLS did in the 2014 World Cup, especially in relation to the rest of the leagues that were represented.

To start with, I compiled a table of all the 2014 World Cup USMNT players, including their name, league they belonged to at the start of the World Cup, rating for each game they appeared in during the tournament, and resulting average rating. I then did the same for each of the 11 MLS players representing other countries.

10 of the 23 USA players belonged to the MLS, as well as 4 from Honduras, 3 from Costa Rica, and 1 from Australia, Brazil, Ecuador, and Iran respectively. That total of 21 (which would have been 22 if Álvaro Saborío hadn’t suffered a foot fracture right before the tournament, causing him to miss all of Costa Rica’s games) marked the highest ever amount of MLS players appearing in a World Cup. But how did they fare?

As you can see, the USMNT’s MLS players averaged a worse rating than their non-MLS counterparts by a difference of 0.37, enough to separate the non-MLS players into “Good” territory (7.0–7.9), as opposed to the MLS being stuck in the “Average” range (6.0–6.9).


Even after adjusting the averages to only include players who appeared in all 4 of the USMNT’s games (in an attempt to measure a more similar caliber of players, along with equal sample sizes), the MLS still winds up inferior, still Average to the non-MLS Good, although the difference is cut to 0.16. However, neither of these 2 differences in means are statistically significant at the 95% level.

There’s a problem, of course, with this definition of MLS and non-MLS. Saying a player represented only the league they were with at the start of the World Cup not only ignores the other leagues they may have played in since the last World Cup, all of which impacted their journey towards making their national team, but it also weights each of these leagues represented as equal, when in fact each player has often appeared in a different amount of games in each league. Thus, I attempted to compile a more accurate record for each USMNT player, using ESPN FC to track each league they appeared in since the end of the 2010 World Cup up until the beginning of the 2014 World Cup, adding up all the appearances they made in each league.

1 player is missing from this chart, as Julian Green didn’t make any professional appearances during this time span (although he, John Brooks, and Timothy Chandler all made appearances before their professional debuts in the Regionalliga for their respective clubs’ reserve teams). DeAndre Yedlin also played amateur soccer, first at the youth level in the Seattle Sounders FC youth academy, then for Akron in the NCAA while also making appearances with the Seattle Sounders FC U-23 in the Premier Development League, all before he finally made his professional debut.

So now how do we define whether a player is MLS or not? Instead of coming up with an arbitrary requirement for appearances or picking the league the player has appeared in most, I’m settling with a clear-cut “did he play in the MLS at any point or not?” This way, the thinking is that if a player is in the MLS, he’s tainted with the inferior development he’ll experience by playing at a lower competition level than his European counterparts (according to Klinsmann’s logic). As a result, only Geoff Cameron switches allegiances, from non-MLS to MLS, thanks to the final years of his tenure with the Houston Dynamo, with whom he made his name before carving out a regular spot for Stoke City.

Predictably, 1 player switching sides doesn’t change the average rating for either that much. Non-MLS players still have a higher average rating than MLS players and that non-MLS average rating is still Good while the MLS’s is still Average, but the difference between the 2 is now 0.26 when accounting for all players and unchanged at 0.16 when only including those who played all 4 games (which leaves Cameron out, since he was benched for the game against Germany). Once again, the 2 differences in means remain insignificant at the 95% level.

In other words, MLS players on the 2014 World Cup USMNT did play worse than their non-MLS counterparts no matter how I spin it, but not significantly so. There are countless possible explanations for why this is true (the most initially apparent seems to be Tim Howard’s perfect 10.00 rating against Belgium having the effect of an extreme outlier), including Klinsmann’s theory that these MLS players would be better if they played in a league with better competition.

At this point in the 2018 World Cup cycle, though, we have a lot more questions than proven explanations, such as:

Is shining for Toronto FC better for Jozy’s confidence

than struggling with Sunderland?

Is Brek Shea better off trying to match the flair of Kaká while picking up free kick pointers from the 2007 Ballon d’Or winner at Orlando City SC


rather than struggling to find any minutes for Stoke City?

Is it more beneficial for DeAndre Yedlin’s development to train at Tottenham while playing for their U21 team than starting for perennial MLS contender Seattle Sounders FC?

Should Freddy Adu stop barnstorming around the world on teams he’ll barely play for and instead return home to the U.S., where everyone has realized he won’t be the next Pele but wouldn’t mind him playing at his 2011 Gold Cup level of heroics?

Alright, maybe those are all rhetorical.

Or maybe not. Klinsmann’s theory that competition level is the most important factor for his USMNT players developmental success and not, say, playing time, still lingers. In fact, my next post will seek to determine which of those factors has a higher correlation with such success. Until then, observe all those parallel universes with your mind’s eye.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.