Pep Guardiola, the vanguard of possession soccer, had an innocuous quote turned into an amusing headline when he said Manchester City passed too much against Hull.  What the article misses is that the specific quote about passing less is not a contradiction of the possession philosophy.  As Thierry Henry articulates in this video, Guardiola utilizes possession to advance the team to the final third, where advantageous attacking opportunities mitigate the need for more passing.

The major upsets of 2016 – Leicester’s Premier League title and Portugal’s Euro championship – signaled a blow for possession soccer.  Rather than react away from a possession-based philosophy, though, many top European teams spent this season re-enforcing the benefits of pressing and direct play within the possession framework.  With opponents sitting deeper and deeper, there is an elevated need within this structure for players who can create goal-scoring opportunities.  The trade-off calculation for managers is to field players who assume the right amount of risk.  When should players attempt a riskier pass, one that might lead to a shot for a teammate but could increase the chances of killing the attack?


With this framework, let’s take a closer look at Key Passes in MLS (passes that lead to a shot on goal.)  All Key Passing stats here are Key Passes per 90 minutes.  Key Passes has some limitations, primarily that it does not account for the quality of shot created.  However, as a rate statistic it correlates well with xA in MLS. (Pearson correlation of 0.91 for KPp96 by xAp96 from American Soccer Analysis.)  Key Passes is also a more widely available stat, enabling the analysis below.

First, here are passes per 90 by Key Passes per 90.  Forwards and attacking mids are in blue, midfielders are in green, and defenders are in yellow.  Positional data is coded to the first position listed in Who Scored (the source for all data below); I’m happy to make any updates as requested.  All players have at least 700 minutes of playing time in the 2016 MLS season, with the circle size indicating the total minutes played.  Connecting lines appear where a player met the 700 min threshold with two different teams.


(Click here for an interactive version)

Sacha Kljestan topped the league with 3.6 Key Passes per 90 minutes played, and attempted just under 55 passes per 90.  He’s joined by Matias Perez Garcia during his time with San Jose, and the usual suspects of MLS playmakers – Mauro Diaz, Diego Valeri, Federico Higain, Nicolas Lodeiro, and Javier Morales – surround them.

Next, we can look at passing accuracy to gauge the adventurousness required to generate those Key Passes.


(Click here for an interactive version)

Our upper-right quadrant shifted more toward the middle, with just Higuain and Morales registering pass completion rates above the league average.  Morales appears to be an excellent acquisition by Dallas to cover for Diaz’s injury.  Lloyd Sam sacrificed some passing accuracy when he joined DC, but increased his Key Passes per 90 from 1.8 to 2.9.  Patrick Mullins generated the headlines with his goal-scoring, but Sam’s contribution should not be overlooked.

The circulators  like Wil Trapp, Osvaldo Alonso, and Anibal Godoy appear in the lower-right, with Mohammed Saeid and Darlington Nagbe creating chances at a higher rate.  Forwards mostly populate the left side of the chart, with Didier Drogba a creative but inaccurate passing outlier.


Part of what drives inaccuracy, though, is the types of passes played.  We would hope that our central connectors do a better job of playing a 10-yard pass than wingers who attempt a 35-yard cross.  Below, we have the percent of each player’s key passes that were a short pass (Whoscored defines a long pass as 25 or more yards.)  For the sake of visibility, I’ve omitted a few defenders who had one or two long Key Passes only.


(Click here for an interactive version)

On average, four out of five Key Passes in MLS are short Key Passes.  Style indicators emerge in this chart, with deep-lying midfielders like Steven Gerrard, Andrea Pirlo, and Michael Bradley utilizing long Key Passes more frequently than others.  We can also present this data another way, and have a better view of Key Pass versatility.


(Click here for an interactive version)

Again, Perez Garcia stands out for his ability to generate short and long Key Passes well above an average rate.

Why does this all matter?  MLS is about to enter a period of tremendous player upheaval.  The next wave of expansion, begins (or continues?) with Atlanta and Minnesota next season, and a target of 28 teams in the near future.  With the various convoluted methods of player acquisition still in place, it’s vital that a team identifies not just a decent player, but one who aligns with the intended playing style.  As we’ve learned, even those generating easily identifiable stats like Key Passes often do so in different ways.