MLS Club Data Review: Progressions & Regressions

Ferreira: No regresión

Ferreira: No regresión

Steve Fenn with a continuing series looking into & behind the data as the MLS season progresses or regresses for some. (Twitter) (Blog)

In any sport it is difficult to tell which teams are truly better than others, in a prescribed span, especially two months into a season. It may make for great fan fodder—chest-beating and trolling—but that same fan who puffed himself out might be scurrying away a few months later when his team has taken an injury to a short Colombian play maker with brittle ankles. Nevermind.

Guaging soccer can be more difficult in soccer, where scoring occurrences are few and a random whistle or bounce can easily be a two-point, seemingly unearned, difference to a match’s outcome. Thank you MLS for your schedule randomization too: When the Seattle Sounders have only played 6 matches while the New York Red Bulls are already through 10, it gets more challenging in MLS.

Building off of last month’s attempt to quantify early season accomplishments it shows the writer was spot-on: The biggest MLS outliers in March results have already started to regress, but in tandem a few clubs are separating themselves– good or bad–from the high-parity center of the league.

March ended with data showing a few clubs whose offenses and/or defenses were operating at a highly-unsustainable level. This month instead of comparing 2013 results to historical ones, we look at March to April.


… Club crests sized by games played….

(Click here for interactive version where you can bring in previous season, sort by club, etc.)

Like Tebowmania, Chivas’ offensive madness has come crashing down under the weight of a larger sample size. Chelis chest-beating so to speak should be in regression.

The shots on goal in Goatland have stayed consistent, but their goals are down from 2.0 per game to 1.5 for the season to date. Still higher than expected, but back to Earth. Likewise, Houston has fallen off and Chicago has improved by half a goal per game.

Interestingly, shots on target have been even more volatile month-to-month than goals. The averages of Los Angeles, Dallas, Houston, and Columbus have all fallen by at least a shot on goal, while those of Philadelphia and Sporting Kansas City have improved by the same degree.

San Jose came close, getting on frame 3.89 times per game so far, versus 3.0 through the first month.

Defensively, the Fire bounced back across goals and shots on target after their horrid opening month. Portland reduced their goals significantly though Donovan Ricketts was challenged just as often. Meanwhile, the DC United defense looks as broken down as Kobe Bryant up against Russell Westbrook post-Achilles, with their goals allowed climbing from 1.0 to 1.6, despite their shots on frame per match dropping by 1½.


Click here for interactive version sortable by conference and with a quantitative power ranking.

Using the same methology from last month, borrowed & adapted from Infostrada’s Simon Gleave, illustrates the overall parity of MLS. (See that last month’s article for full explanation.)

For results through April, a slight adjustment was made to xGDPG to correct for the skew given extreme 2012 performers. Rather than make schedule appear easier or harder simply because clubs don’t play themselves, their full 2013 xGDPG was subtracted from the cumulative xGDPG through April.

The majority of clubs are huddled together around the middle of this graph, performing very close to expectations and showing that thus far none of them seem to be particularly stronger than the others.

Houston, LA, Dallas, Seattle, & San Jose have had the easiest schedules so far, with LA & FCD overacheiving, the Sounders & Quakes disappointing, & the Dynamo hitting expectations almost exactly.

"Uh, Steve, this data doesn't look so good..."

“Uh, Steve, this data doesn’t look so good…”

DC United, Chivas USA, New England, and Portland have had the easiest road to this point, and while DC managed to perform even worse than expected, the Goats and Timbers outdid themselves, and the Revs more or less toed the line.

One of the best applications of this data could be comparing those clubs whose records appear very similar on the surface.

Houston is currently sitting on 14 point, while Portland has 13, but when you see that Portland has done so through tougher fixtures, the Timbers achievement is much more impressive. Especially when you consider that schedule strength will more or less even out in the end.

There’s certainly a lot of season left, but if you think wait and see is still the order of the day, consider last season.

At this time in 2012, the 4 clubs with the most PPG, Sporting KC, San Jose, Seattle, and RSL were still top four season’s end. In fact, points per game on May 1, 2012 correlated to full season PPG at R²=0.6415. Goal difference per game was a slightly better predictor, with an R² of 0.677.

For the non-mathletes on PEDs who’ve gotten this far, that means that even though the MLS season isn’t a-third over, the results thus far are likely very meaningful to the final 2013 table.

Someone else will have to go on a tear to keep LA, Montreal, or Dallas from taking the Supporters Shield, with the Timbers and Crew seeming to be the most likely usurpers.

Meanwhile if DC United and Chicago don’t figure things out very quickly they will get stuck slogging through a long, hopeless summer and fall.

13 responses to this post.

  1. Posted by dth on 2013/05/02 at 8:46 AM

    Bring back Boskovic.


  2. Please check out the interactive versions. I put a lot of data in here, and I’d love to discuss and debate any alternate interpretations of it.


    • Oops. Just realized that I forgot to click something that opened up the secondary pages in my interactive Tableau visualizations.
      That’s fixed now, so in the G/SOG one you can now see a G/SOG differential graph, and offensive and goalkeeping totals for every club. Also, the ΔPPG vs ΔGDPG viz has extra pages with a quantitative power ranking and expectations for ΔGDPG for each club over every month of the season.


  3. Posted by thedudeandwalter on 2013/05/02 at 12:53 PM

    It seems that the Red Bulls offense is clicking more in April and there defense is slightly better, if I am reading this right (I may not be)


    • In a sense. NY scored a little more and allowed a couple less goals. Their shots on both sides of the ball didn’t change much, though.
      The Red Bulls’ schedule has had pretty much an average difficulty, and they’ve slightly underperformed expectations based on that schedule so far.


  4. Posted by dth on 2013/05/02 at 2:04 PM

    An interesting team to watch will be Colorado, who’ve managed to tread water while dealing with an injury apocalypse.


    • Yeah, it’s important to note that all of my metrics are ignorant of injuries, cards, and suspensions, so this kind of analysis will always underrate a team in a situation like Colorado’s, assuming they can get healthy going forward. Definitely one to watch.


  5. Posted by Crow on 2013/05/02 at 4:59 PM

    I will take Hackworth over Nowak any day but actually believe the Union are UNDERachieving by a bit so far this year. Hackworth has certainly been inconsistent- nailing some games or halves and completely getting outcoached in others. Trying to keep up the proud Philly tradition of Andy Reid and others.

    Its past time to get Kleberson on the field. His quality is evident in the few minutes he’s received. Keon Daniel started off the year strong but has faded. Kleberson and Le Toux should be far ahead of Daniel and Cruz (who makes two stupid decisions/for every good one) and Farfan who just isnt the same player he was a year or two ago. I don’t like Casey at all but Jack Mac has benefited from him so I’m fine for that for the time being.

    I love Okugo but he’s getting exposed more this year- I really wish Soumare was on the roster still and Okugo would go back to the CDM position. Carroll is a great guy but Okugo would be a better option now in that slot.

    I had always been a big MacMath fan but he has not been playing well at all.

    I see Philly sneaking into the playoffs picking up their first win and then getting outclassed by a quality team.


    • I don’t follow the Union nearly as closely as you do, but my ΔGDPG has them underachieving, while the bookies’ ΔPPG has them overachieving. Hard to figure them out so far.
      Overall I’ve got them 6th best in the East, so sneaking into the playoffs is definitely possible. Based on 2012 goal differentials, June and July look to be their easiest months, so I’d say that they need to get it together and go on a tear by then,


      • Posted by Crow on 2013/05/03 at 10:19 PM

        I’ve seen alot of people ‘surprised’ that the Union have been ‘competitive’ so far this year. I guess its mainly because how dysfunctional they were last season under Nowak. They actually are playing alot like they were the first few months under Hackworth last year.

        I really think that the only teams better than the Union on their best day are Houston and SKC and the Red Bulls when they aren’t self-destructing. I guess I’m looking at the potential of the Union, even maybe a full season down the road, but I feel that they should be better than they are right now. Maybe the team will come together by the end of the year, but I think Hackworth is leaving too much talent on the bench. I didn’t even mention Roger Torres.


  6. Posted by Jorge on 2013/05/02 at 6:08 PM

    This is the reason why most football/soccer experts laugh at Americans and call us “idiots.” And now thanks to you..they are right!
    Americans are way too much in love and dependent on graphs, charts, analytics…that’s great for work. However, for this sport…it’s not. To understand and appreciate this sport…you have to embrace the irrational. Sure your tactics, management approaches are there, but ultimately it is between the player and the ball. He has split of seconds to assess and execute. They don’t have the time to study graphs, charts, analytics to make decisions. This is not Apple.


    • To each his own. You’re right that in every individual moment the game is about players’ ability to assess and execute. But as thousands of those moments add up at the club level you’re left with measurable outcomes that can be analyzed.
      For me it’s idiotic to just stare at the raw MLS table at this point and praise NY for being on 14 points despite taking 10 not-terribly-difficult games to get there.
      All I’m offering is a different, deeper view of results so far. If that’s not your cup of tea, OK. For others I hope this article brought some insight into different teams’ accomplishments so far.

      By the way, footie analysis isn’t a solely American thing. Simon Gleave, Mark Taylor, Chris Anderson, and other Europeans write on quantitative analysis of the game regularly, and most EPL clubs have analysts on staff.


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 )

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: