31 May 2012

Where's your varsity? Strong/weak USOC lineups

Philadelphia fans were treated to a relatively boring game;
other MLS fans were not so lucky.
MLS teams struggled in the third round of the U.S. Open Cup this week, with only eight of 16 teams advancing despite facing teams from lower division of American soccer. The most common assumption is that this occurs because MLS teams, to be blunt, don’t care about the tournament. In that vein, the most talked about man of this round of the tournament, Cal FC manager Eric Wynalda, proclaimed via Twitter: “the MLS teams that played a starting lineup- won. Those who didn't, lost.” This supports the common thinking: MLS teams don’t care about the tournament and use reserve-laden lineups with less experience, less talent, and less cohesion than the normal squad and less motivation than the opponent.

From there, I see two main reasons as to why MLS reserves struggle to beat minor-league starters:
  1. MLS teams’ depth sometimes fails to match up with minor-league clubs because, if a club spends big on its starters, the remaining players may have been selected as much on dollar value as on talent.
  2. More likely, the reserves are still more talented and often carry play, but they lack the decisive technical quality to take advantage of their dominance. You can see this in the number of games where MLS teams, even with reserves, post heavy shot and possession advantages against minor-league opposition but fail to advance.
Wynalda made his statement one night before his amateur club stunned the Portland Timbers – fielding a virtually full-strength lineup – on Wednesday night, but it still got me thinking. So I decided to take some data and see how accurate the statement was this year. My results are after the jump ...
I took the starting lineup and three substitutes used by the 16 MLS clubs on Tuesday and Wednesday nights and measured them in four categories:
  1. Total minutes played, 2012 MLS season
  2. Percentage of team minutes played, 2012 MLS season (important because MLS teams’ minutes played range from 1,350 (DC, FCD) to 966 (HOU)
  3. Median MLS minutes played in the 2012 MLS season by the 14 players who saw action
  4. Median MLS minutes played in the 2012 MLS season by the 11 starters
The results were interesting, although not as overwhelming as you might have thought.

% of Team’s MLS Minutes
TM% MLS Mins.Result
To show you how I calculated the numbers in the table at right, consider New York. The 14 players used by the Red Bulls against Charleston have played a combined 8,506 MLS minutes on the season, 66.1% of the total available to the team, which is 12,870 minutes. This weights substitutes, such as New York's Jonathan Borrajo and Ryan Maduro (who have not played an MLS minute this year) evenly with starters.
Based on these percentages, there is a noticeable dip in performance by teams with lineups that have played less than 42 percent of their league minutes this season. All five MLS teams falling below that 42-percent threshold (Chicago, Columbus, New England, LA Galaxy, Houston) were eliminated from the tournament. Three of those teams - Chicago, Columbus, and Houston - started three players each who had not played an MLS minute this year.
Above that 42-percent mark, MLS teams were 8-3, a .727 winning percentage that would be much more normal for the first Open Cup round involving MLS teams. There did not, however, seem to be a clear difference in success rate past that threshold. The three losses all came from the seven highest percentages. The only clubs to lose while playing lineups above the 42-percent threshold were Portland (66.6%), Real Salt Lake (53.9%), and Dallas (52.1%). What’s more, both Portland and Dallas had significantly more shots than their opponents (Portland 37-8, Dallas 28-7) but failed to take advantage of those opportunities.

Median Mins. by Starters
TMMed. Mins.Result
The table at right sorts each team by the number of MLS minutes played in 2012 by the middle man - the player with the sixth-most and sixth-least minutes - in the starting lineup. They ranged from New York's Jan Gunnar Solli (799) to New England's Darrius Barnes (48). I used median in an attempt to eliminate outliers, although most wound up being players who do not start regularly.
The correlation between results and a team’s starting lineup was a little less pronounced, but the chart still provides some insight as to how seriously teams take the tournament.
The results were less clear, with the closest thing to a trend being a 6-4 record for the top 10 and a 2-4 record for the bottom six. Those two wins in the bottom six, however, highlight the importance of substitutes. San Jose (median minutes played 72) and Kansas City (mmp 66) both brought big guns off the bench, with Kansas City (Sapong, Zusi, Besler) trying to hold onto a lead against the very dangerous Orlando City and San Jose (Simon Dawkins, Steve Beitashour, Rafael Baca) coming from behind only after that trio entered the game. Interestingly, San Jose jumps from 16th to 5th in median minutes when its three substitutes are considered; Kansas City remains one of the lowest clubs, albeit with a much higher number.
Not surprisingly, the teams that put the most quality into their outing, only to suffer a home loss, continued to be Portland (744) and Real Salt Lake (472).

Many teams have already had to call on their goalkeeping depth in league play during the season, but seven of the 16 MLS teams still used goalkeepers on Tuesday who have not played in MLS this year: Bobby Shuttleworth (NE), Matt Lampson (CLB), Steward Céus (COL), Tyler Deric (HOU), Jon Kempin (KC), Tim Melia (CHV), David Bingham (SJ). Only two out-and-out starting goalkeepers – Ryan Meara (NY) and Troy Perkins (POR) – saw action in the round, although several others have been starting as their team’s healthiest option.

There were, of course, any number of things I could not take into account here. Several teams were missing a number of starters due to international duty, although many big-name players historically sit out early-round Open Cup games anyway. Some teams were prevented from using regular starters by the little-known five-foreigner limit imposed on professional teams in Open Cup play. The minutes-played statistic does not account for injury, suspension, or form, and I also failed to take into account home-field advantage and the form and/or quality of the opponent (although based on Orlando and Rochester losing and Dayton winning, it wouldn't be significant this year).

I don’t think I’m breaking new ground here, but I think the numbers highlight the danger of naming a lineup with minimal experience – that is, players that have accounted for less than 40 percent of a team’s minutes on the season – to face any opponent. It will be interesting to see how strong lineups are in Round 4, both when MLS teams take on minor-league opposition and when they face each other.


  1. Love the analysis.

    May be tough to figure this into such an analysis, but some of the teams have had players getting minutes that may not be usual starters due to injury.

    Seattle is a great example - only 4 of last night's players would be considered regular starters, Parke, Hurtado, Alonso, Johnson.

    Meredith was in for Gspurning
    Burch for Gonzalez
    Scott for Johansson
    Cato for Rosales
    Ochoa for Montero
    Rose for Evans
    Caskey for Fernandez

    Maybe the Sounders are an anomaly because they rotate their squad so much more..

  2. Another thing to factor in is MLS was 0-4 in matches I watched over the two days. :)