Moneyball: An Explanation for 2019

There were several persistent issues fans and pundits had with Gregg Berhalter and the USMNT in 2019, beyond results. 

Four of the biggest were:

1.  The continued presence on rosters of mid-20s players, from MLS, that were given minutes no matter how they performed for the USMNT or in MLS.  
2. The almost Puritanical need to “Play Out of the Back”, no matter the game state or opposition defense.
3. A passive mid-block defense that allowed opponents easy entry into our third of the field.  
4.  Substitutions that made no sense in competitive games.

There are two theories among fans and analysts who try and analyze the situation, beyond, “Gregg is an idiot who was only hired because he is brothers with the CEO to be.”  Whether there is any truth to that statement is counteracted by the idea that nobody could be that big an idiot. Some will disagree of course.

Both theories start from the idea that Gregg, from the day he was hired, had a set system that he wanted to play.  That he would find players that fit the system, rather than devise a system that fit the players he had.  

The first theory is that Gregg wanted to only rely on data and “player profiles” based on data collected by services like Opta.  This tries to explain the #1 complaint by pointing out that most of our talent in Europe is very young, few have first team minutes in large quantities, so there is no Opta data.  That since Gregg was only selecting players he could get data on, they mostly were mid-twenties players in MLS who had mountains of data available. Players with good amounts of data could be added, like Pulisic, Brooks, Ream.  But players like Richie Ledezma could not.

This theory kind of hits the snag at the exclusion of plenty of players with lots of first team minutes like Antonee Robinson, Duane Holmes, Julian Green, Fabian Johnson, etc.  It also doesn’t address #2, 3, 4 in any way other than, Gregg is “naïve”.

My theory, based on things discussed by media members with insider access, is that 2019 was an experiment in a “Moneyball” approach.  This is a popular sporting concept in the USA and increasingly around the world. It is looking for incremental competitive advantages using statistical analysis to offset competitive disadvantages in talent disparity.  The famous example is how the Oakland A’s used the statistical work of Bill James to build a team on par with the Yankees and Red Sox on 1/50^th the budget.

The idea Gregg, and Earnie Stewart, had and possibly still have, is that the MLS season is very different from the European and South American seasons.  The long winter break allows for a long national team training camp in January (Camp Cupcake). Brazil, Germany, France, etc. do not have this opportunity for a long, like a club pre-season, camp.  That the extra time could be used to teach a complicated, Pep Guardiola-esque, positional offense that would allow the USMNT, with less talent, to compete with the previous mentioned teams; who can’t run sophisticated offenses because the short windows national teams have together.  

Data was used to pick the original January pool.  They were drilled in the new system. The idea was that these players would then be able to carry that knowledge forward and help train additional players.  New players would be brought in a few at a time until they got up to speed then more could be integrated. This is why the rosters throughout the year featured so many players from January camp (Cupcakers).  This is why only the top players in Europe were initially brought in in small numbers. In June, longer camps were held and more non-MLS players were brought in. But when the Gold Cup roster was announced, in the end, it contained 14 Cupcakers out of 23 players.  This roughly 2/3rds ratio of Cupcakers to non-Cupcakers was very consistent throughout 2019.

There is no way to teach a possession based positional read and react offense (The System) without playing it.  This is why all year, until the final Canada game, the USA played out of the back almost exclusively. Even in the Friendly against Mexico, with Mexico probably figuring out this entire plan and that the USA was not going to go over the top, against 7 and 8 men presses, they played in to the pressure.  

But how does this explain why the USMNT played a passive mid-block defense that made teams like Curacao and Canada look like Spain?  The defense was deployed by Gregg not to win games or even to win the ball. It was designed to create goal kicks for the USA. The more goal kicks, the more opportunities to play out from the back and practice The System.  Since the most limited thing a national team has is minutes of game time, every minute was maximized as a training exercise.

Down by a goal in the Gold Cup Final, Gregg substituted on Roldan for Morris, Zardes for Altidore, and Lovitz for Ream.  The last one is really perplexing because Lovitz is not only not a goal threat, but is not even good. This one is discussed a lot because it is so questionable.  But all three are Cupcakers subbing in for non-Cupcakers. Gregg must have felt a better execution of The System would lead to a goal or viewed the Gold Cup Final as another training exercise like all the previous games.

Following the second Mexico loss and the loss in Canada, pressure mounted.  Against Canada in Orlando, some things changed. The roster was still heavily infested with Cupcakers, but only 4 ended up playing.  The USMNT played goal kicks short, but never passed back to the goal keeper like previously. If another short pass was not on, the ball was hit long to a forward.  Contrast that with the Mexico friendly where passing back was used to keep possession at all costs; frequently ending up at the feet of a center back on the goal line or in Steffan’s 6 yard box.   The USMNT did not sit in a passive mid-block, but instead used a form of high press that led to turnovers and transition goals. The subs were mostly defense for offense which made sense with a multi-goal lead.

I believe several flaws in the “moneyball” approach became apparent:

1. The Cupcakers that were relied upon were simply not good enough
2. The reliance on a core Group, doesn’t allow for new or in form players to be integrated quickly or at all.
3. Most of the young talent breaking through last year and in the upcoming years, will be in Europe and will not be able to attend Camp Cupcake.
4. The team did not look comfortable executing The System at any point and that is either because Gregg could not teach it, the players could not execute it, or even with the large amount of Cupcakers the European players couldn’t learn it, or the European players could execute it but the MLS players couldn’t keep up.

We don’t know if it has been abandoned and we don’t know if Gregg thinks it was a failure at all.   We probably will not know until March at the earliest, maybe not until the CONCACAF National League Final or even World Cup Qualifying in August.