MLS Teams Quaking In The Wake Of The Departed?

Who stands to lose as the luminaries walk away?

Who stands to lose as the luminaries walk away?

Steve Fenn with analytical look at MLS in the wake of The Departed

Which teammates did David Beckham, Landon Donovan, Roger Espinoza, Kei Kamara, and Freddy Montero interact with most often in 2012?

Much been made of MLS departures (or, in Donovan’s case, his walkabout in Cambodia) this offseason, but most of these articles have been conjecture about marketing, merchandising, etc. Let’s see what a quantitative on-field analysis tells us. (Editor’s note: Thanks Steve. Long needed.)

Opta passing matrices specifically list how many passes these players sent to or received from teammates in 2012. Here’s a visualization of that data for the Galaxy:


Interesting, but a little overwhelming. More revealing with a narrower focus.

First, this study only includes outfield players on the club’s 2013 roster who played over 800 total minutes in 2012. It is fairly irrelevant at this point that Beckham didn’t exchange many passes with Edson Buddle last year.

In all of the following visualizations, passes from “The Departed” (Beckham, Donovan, Espinoza, Kamara, Montero) are on the left, and passes to them on the right. Also, teammates are ranked very roughly by position, and all keepers (including those no longer with their teams) are lumped together at the top.

Passes are listed per 90 minutes those 2 players shared the pitch. The width of each bar is driven by the minutes-shared total

LA Galaxy:

Becks and Landy impact...

Becks and Landy impact…

Unsurprisingly, Beckham sent many passes to nearly everyone he shared the field with. Some may have overlooked how often he provided on outlet for LA’s defenders and goalkeepers (though not the Houston Dynamo it’s a good bet who were laid to waste by Beckham being the point guard on the counter in the MLS Cup something awful).

Someone else will need to show for the ball when Omar Gonzalez looks to transition the Galaxy into the attack. Note that.

Most striking, perhaps, is Robbie Keane’s reliance on Donovan and Beckham. Overall Keane received 1180 passes in 2012, and 28% of them were from Beckham and Donovan, even though all three only shared the field at the same time for 1397 of Keane’s 2519 minutes, roughly 55%.

This is as good a place as any to mention that these figures do not include crosses. However, even with a crosser of the ball as skilled as Beckham, none of his teammates received more than 11 in total via his right peg all season (Keane got on the end of that many). A likely issue for the club, but it would be very hard do specify which player will miss this Beckham skill most.

Sporting Kansas City:

To Wigan and Norwich, respectfully, with love...

To Wigan and Norwich, respectfully, with love…

Bobby Convey, Seth Sinovic, and Chance Myers are interesting for their varied interactions with SKC departures. Per 90, Kei Kamara partnered with Myers on 14.4 passes, but only 5.6 with Roger Espinoza.

On the flipside, Convey and Sinovic exchanged the ball with Espinoza 14.9 and 13.9 per 90, respectively, but only 3.5 and 5.0 in an average match with Kamara.

Both Espinoza and Kamara were very active with Graham Zusi, and accounted for 295 of the passes he received and 251 of those he completed over the season. That’s 23.3% & 23.2% of his season totals while sharing only 1960 of his 2788 minutes with both of them.

Seattle Sounders:

Fast Freddy....

Fast Freddy….

Fredy Montero’s data is striking for its peaks and valleys. He combined often with Osvaldo Alonso (7.9 passes to/from per 90), Alex Caskey (8.2), Brad Evans (9.5), Mauro Rosales (9.8), Eddie Johnson (10.4), and David Estrada (10.7).

Meanwhile, passes involved Montero and Jhon Kennedy Hurtado or Patrick Ianni less than once per 90. Montero didn’t pass to Ianni once in 1111 minutes together. Some might say that’s a good thing.

Many of these figures are largely driven by relative positioning in the Sounders 2012 tactics, but they do point out which players are more likely to be effected by Montero’s absence in 2013.


In all of these cases, nominal replacements have either been acquired (Feilhaber, Bieler) or seem inevitable (Martins, Lampard, Donovan himself). No two players are the same though, and it should be interesting to see how Zusi, for example, combines with Feilhaber and Bieler in comparison to his constant partnerships with Espinoza and Kamara.

This data does have its limits, though. There is no weighting for length, location, or quality of these passes. Beckham’s passes to some teammates may have been mostly short and routine, while other players more often got on the end of a jaw-droppingly beautiful 50 yard dime.

Also worth noting that these numbers are rooted in various factors such as relative positioning, tactics, chemistry, etc. Accurately teasing specific causes and effects out of them can be quite difficult.

Even so, passes per 90 minutes are a clear record of how often these MLS departures interacted with their former teammates. It should be interesting to see how the coaches and teammates left behind adjust to their absences and replacements.

21 responses to this post.

  1. I’d be happy to answer questions any readers have about the data here and my analysis of it. Analyzing passing data can be very tricky, and I’d love to here alternate interpretations, especially from fans of LA, SKC, or the Sounders.


  2. What is there to say for the players position in general? For Kamara, he’s rarely one to kick it back to a defender to clear out and, as seen by the lower Besler/Collin passes as opposed to Espinoza having been a midfielder and more inclined to pass it back. Granted, that doesn’t say much about Myers/Sinovic. That difference could literally come down to the side of the field that Kamara was on vs Espinoza playing in the middle.


    • I’d agree that field position is a big driver, but when we find relationships that defy that trend, I think it would be a disservice to just assume that side of the field is the reason.
      Espinoza/Sinovic and Kamara/Myers are absolutely the most interesting pairings to me, especially in comparison to each other.
      Does anyone who follows SKC closely have a theory that would go beyond which side of the field they play on?


  3. […] MLS Teams Quaking In The Wake Of The Departed? – from The Shin Guardian: Interesting, but a little overwhelming. […]


  4. […] At the Shin Guardian, Steve Fenn does an analytical breakdown of what the departure of some big name players from LA Galaxy, Seattle Sounders, and Sporting Kansas City might mean. […]


  5. Posted by bob on 2013/03/06 at 1:46 PM

    steve, stats are a wonderful thing and for me, it is interesting to see how teams move the ball around and side by side, you can see how the movement works LA vs KC. teams should make good use of this information. of course, there are new players now.

    there is the thing about stats though, that they can be misinterpreted, sort of like the two goal lead senario. its like if you try and say a two goal lead is not the most dangerous, when in fact it very well may be. i dont know. maybe it is just a philosophical way of looking at things.

    stats are powerful, that is for sure.


    • You’re right, stats are only as good as their interpretations and applications (as well as the accuracy with which they’re recorded). That’s why you’ll find hedges and caveats within most of my writings on such subjects.

      Not sure I follow your 2-goal example, though. “Most dangerous” would by definition mean that a team is more likely to lose a 2-goal lead than a 1-goal lead. This is on its face preposterous, and it doesn’t hold up to scrutiny, either. Per Soccer Statistically, the results when 35th minute margin is:
      1 goal lead: win: 78%, draw: 16%, loss: 6%
      2 goal lead: win: 96%, draw: 2%, loss: 2%
      The 1 time out of 50 that the 2-goal team loses doesn’t disprove that 2-goal leads are hugely safer than 1-goal leads.


      • Posted by bob on 2013/03/06 at 2:53 PM

        i get that, but i hope you get that when barcelona is playing getaffe in the nou camp and they are two nil in the 35th minutes, barca im talking, we view this stat differently then if getaffe were up two nil in the nou camp after 35 minutes.

        steve, dont you find yourself looking at and seeing how the goal scoring plays out for all the games out there. i know i do

        i would think also that watching aggrigate scoring like in the champions league is a good example of why a two goal lead is dangerous. and teams do take a two goal lead, give it up, then still win the game., which places the stats in favor of not thinking a two goal lead is what it is, which is the most dangerous lead in soccer. and the pundents say it for a reason, which maybe should not be marginalized by stats.


        • Going off of “looking at and seeing how the goal scoring plays out” can skew your memory of events.
          When you see a team come back from a 2-goal deficit, it sticks in your mind far more than games that are always within 1. That’s why the overall stats tend to disagree with general perception and memory.


          • Posted by bob on 2013/03/09 at 11:38 AM

            celtic with a two goal lead at ross county today, and i dont want to spoil the final score. but ranger fans everywhere must be cheering.


      • Posted by bob on 2013/03/06 at 3:40 PM

        and to me, the most telling stat you present is that when a team has a one goal lead, they dont lose the game 94 out of 100 times. when a team has a two goal lead, they dont lose 98 out of 100 times.


        • Yeah, the 1-goal lead is only 3 times as dangerous if you ignore ties. However, dropping 2 points can be quite damaging, too, and the 1-goal lead is 8x more dangerous in that respect.
          I’m not saying that a 2-goal lead is inherently safe, just that claiming it is more dangerous than a 1-goal lead is absurd on its face and in practice.


          • Posted by bob on 2013/03/09 at 7:15 AM

            steve, maybe this is a case where logic and statistics are at an impasse. nobody is saying a two goal lead is more dangerous than a tie score, just that it is the most dangerous lead, and that is still based on the situation.

            in practice, good teams are aware of the traps and pitfalls of a two goal lead, just as they are aware of the importance of maintaining a one goal lead.

            everyone should put some stock into your stats, but it would also be wise to spread around the investments, it would probably be absurd not to


            • “nobody is saying a two goal lead is more dangerous than a tie score, just that it is the most dangerous lead,”
              Maybe you’re defining “most” in a strange way. If you’re saying that a 2-goal lead is “the most dangerous,” it logically follows that it is more dangerous than any other lead.
              Do you honestly think that, despite all evidence, a team whose greatest lead is two will lose or tie more often than one whose greatest lead is 1?

            • Posted by bob on 2013/03/11 at 10:42 PM

              steve, is it possible that you are wrong, and your stats are wrong? is that even a possibility for you? i think the stats are wrong, and i think you are wrong and i think all the british pundents who use the phrase, they are right. i mean, you gotta come back from a one goal lead in order to come back from two down and goals are a premium in this sport.

            • Posted by bob on 2013/03/11 at 10:51 PM

              steve, do you think that when a team takes a two goal lead in the 90th minute, we call that the most dangerous lead, or do you think this lead is more dangerous than a one goal lead?

            • Posted by bob on 2013/03/12 at 5:38 AM

              steve, do you think teams defend as hard when they have a four goal lead, as they do when they have a one goal lead?

  6. Posted by bob on 2013/03/06 at 2:56 PM

    steve, in the monte hall example, you know you are suppose to switch doors when monte shows you a loser door and there are only two doors left, right?


    • Not seeing how the Let’s make a deal paradox applies here.
      What are you saying is the metaphorical goat and the metaphorical car?


      • Posted by bob on 2013/03/09 at 11:34 AM

        it has me wondering, what is more likely of the three, a three goal lead staying three goals, or going to four, or going down to two? and which do you think is second and which is third?


  7. Anyone else see that L.A. acquired Chandler Hoffman from Philly for a conditional pick in the 2014 Superdraft? They have amassed a ton of potential at forward; Rugg, Villareal, McBean, Zardes, and now Hoffman. Bruce is killing it.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: