Illuminations: Gold Cup 2011 vs. “The 5-Game Tournament”

2010-2011

2011 – ?

Darius Tahir bifurcates the past two USMNT seasonal campaigns using the Klinsmann hiring as the fulcrum point.

The US takes on Mexico precisely one month from today. The game is a ceremonious marker in time for the US Men’s National Team Program. It was one year ago as Bob Bradley was making phone calls to begin picking up the pieces of a Gold Cup Final gone horribly off the rails that Sunil Gulati swooped in and in a move that many had expected for some time, relieved the Princeton grad of his second term of service to the men’s program.

With Jurgen Klinsmann now having a full term in office, his recently concluded “5-Game Tournament” presents an interesting set of observations to match against Bob Bradley’s Confederation Cup-seeking Gold Cup campaign.

Now our cursory analysis by Darius Tahir

(All statistics shared in this spreadsheet)

Okay, Darius, what you got?

It’s awfully helpful when a comparison makes itself. In this case, with Klinsmann rapidly approaching his first-year anniversary of taking the head job with the USMNT, a natural comparison practically begs itself to be made: Bob Bradley’s Gold Cup run and Klinsmann’s “Five Game Tournament.” So let’s make it, then—what are the similarities, differences, and what might give you confidence (or lack of it) for the future)? And we’ll try to make these comparisons, as often as possible, through cold hard statistics—the OPTA statistics provided by MLS, in this case.

Headline Record

The US with a masterful performance against Jamaica at the Gold Cup…

Five games and six games, as the case may be, is hardly the biggest sample size to draw really firm conclusions from. But it’s probably big enough to make some preliminary ones. Some other caveats must be applied—all of Bradley’s games were competitive; only two of Klinsmann’s were. On the other hand, all of Bradley’s games were at home; two of Klinsmann’s were in the hostile environs of Guatemala City, and, ah, Toronto Canada.

On the other hand, Bradley’s team faced better opponents. Bradley’s team, at the time of playing them, had an Elo rating average of 50; Klinsmann’s, 65.

In terms of results—with no regard to those caveats—Bradley’s look better.

Bradley’s 4-2 record, for 12 points over 6 games, makes 2 points per game. Klinsmann’s record, of 8 points in 5 games for 1.6 points per game, is an appreciable cut worse. Both teams had a goal differential of +3, making Klinsmann’s goal differential per game ever-so-slightly better.

Roster

Did one coach have a better roster to deal with? A heuristic for judging the quality of the roster is to look at the leagues they play in—Bradley’s had 11 players playing for teams in the big 5 leagues (i.e. England, Germany, Spain, Italy, and France) whereas Klinsmann’s featured only 6. It’s probably a mistake to read too much into this heuristic—a big chunk of the difference has to do with players like Jozy Altidore moving to leagues that more appropriately reflect current quality.

That said, Klinsmann did make some curious decisions in filling out his roster—it’s very hard to believe Eric Lichaj is a worse fullback than Edgar Castillo, Jose Torres and Michael Parkhurst.

Injury and other issues blighted availability for each team; Altidore was only able to go a limited amount of time for each iteration of the team—he got injured in the knockout game against Jamaica in the Gold Cup; and, due to AZ’s intransigence, was not allowed to show up until late for the five game tournament. Fabian Johnson missed a game for Klinsmann; as you don’t want to remember, Cherundolo missed most of the Mexico game to injury. Torres missed the Guatemala game. Eddie Gaven refused Bradley’s call-up for the Gold Cup.

Performance, by stats

The US took down the Euro runner-ups….

It’s here that any short-term optimism from Klinsmann’s performance should be drawn. There have been some rumblings that Klinsmann has betrayed the high-flown promises of  a more attractive, better-passing side. Perhaps this is true—beauty is always on the eye of the beholder, and no amount of stats to convince you otherwise can accomplish that task. Nevertheless, we can identify a few possible principles of an attractive side—better passing, percentage-wise; better passing in the final third, percentage-wise; better build-up from defenders, percentage-wise; and more attempts (and more successful attempts) to take on defenders off the dribble.

Klinsmann’s team improved relative to Bradley’s at these metrics. Klinsmann’s team averaged 556 passes attempted and competed roughly 460 of team, resulting in an 82.7% passes completed. By contrast, Bradley’s teams averaged roughly 516 passes per game and completed 80.45% of them. In the final third*, Klinsmann’s and Bradley’s teams averaged a similar number of passes—73 (Klinsmann) or 74 (Bradley)—but Klinsmann’s team was dramatically more proficient, completing 68.5% of passes in the final third versus 61.26% for Bradley. From the backline, Bradley’s team averaged 229 passes per game at a 78.7% success rate; Klinsmann’s team averaged slightly fewer passes from the backline, but completed nearly 4% more of those passes, at 82.6%. Klinsmann’s teams took their opponents on more frequently, at 10.6 times per game versus 7.5 times per game for Bradley. That decision was appropriate, as they were successful 36.2% of the time versus 28.9% under Bradley.

Klinsmann’s team averaged 13.2 shots per game versus 11 in 2011.

There are a lot of styles that lead to success, even offensive success; a direct team can nevertheless average a lot of goals and a technically proficient team can—like Spain—use that proficiency for defensive rather than offensive purposes. Nevertheless, Klinsmann’s shift appears to have lead to more offensive success, averaging 2 goals per game versus 1.5 goals per game. (With the caveats for sample size raised.)

Defensively, it would appear from the overall statistics that Klinsmann has given up ground relative to Bradley; Klinsmann’s teams gave up 1.4 goals per game versus 1 goal per game under Bradley.

Again, many styles can lead to offensive success; looking at some of the opponent’s statistics might nevertheless be instructive. Opponents attempted nearly ten more passes per game against Klinsmann than Bradley (399 vs. 389, roughly in both cases), but were ever-so-slightly less accurate against Klinsmann than Bradley (75.97% vs. 77.31%). In the opponent’s final third, however, Klinsmann’s teams were more stingy, allowing fewer passes (49.4 vs. 53) and allowing a much lower percentage (54.25% vs. 62.89%). Again, this may reflect a change of style on opponents’ part rather than a more stifling defense overall.

On the other hand, Klinsmann defenses were worse in one-on-one play, yielding 11.6 attempts at a 36.2% success rate compared to 7.5 attempts at a 28.9% success rate.

*(For the final third: here’s how I’ve done it. In the MLS chalkboard feature, you’ll notice there are alternating strips of color. I’ve drawn the rectangle inclusive of the last four bars.)

Conclusion

Many of these statistics are suggestive rather than definitive, owing to some of the caveats discussed above. Nevertheless, if you clear the mind of some of the distractions of both eras (Jose Torres at left back! Johnny Bornstein at left back!) it would appear Klinsmann is succeeding at many of the goals he set out. Certainly, however, some of Klinsmann’s tactical decisions are lacking—he didn’t coach any of the games in the five-game tournament as brilliantly as Bradley’s master class against Jamaica in the first knockout round of the Gold Cup.

Evolution is too often a slow and painful process.

About these ads

9 responses to this post.

  1. Posted by 2tone on 2012/07/16 at 7:33 AM

    I didn’t know Eddie Gaven refused a Gold Cup call-up. What a shmuck.

    Reply

    • Posted by dth on 2012/07/16 at 8:59 AM

      Yeah, not sure it was a huge loss at all.

      I think he, like Brad Guzan, was getting married?

      Reply

  2. Posted by mbw on 2012/07/16 at 8:49 PM

    Minor quibble: I think this year’s games were considerably tougher than last year’s, and that performance this year matched expectations while last year it fell short.

    First claim: If you use ESPN’s power index (which provides anticipated goals-for and goals-against figures*) and accept Nate Silver’s premise that home-country advantage is worth about half a goal per game, and if you treat last year’s Gold Cup final as a neutral-site game rather than a US home game (as seems reasonable), the anticipated goal difference from last year’s games was about +4.8 (or .8 per), from this year, about +2.9 (or .58 per).

    Second claim: Results-wise, the ESPN-Silver model would have anticipated five wins, one loss, and no draws last year (against an actual four wins, two losses, and no draws) and two wins, one loss, and two draws this year (which is exactly what happened). Goal difference last year was .3 worse than anticipated, this year, .02 better.

    Enjoyed your piece, as always.

    * Data-quality caveat: I couldn’t find mid-2011 SPI ratings and had to use present ratings for the 2011 results. This presumably distorts the calculations a little bit — though not much. (I’m not sure in which direction.)

    Reply

    • Posted by Jared on 2012/07/17 at 5:48 AM

      Expectations were definitely not met last year, statistically or otherwise especially with the loss to Panama.

      I think using this year’s data for last year’s might lower the rankings of the teams involved at least for Mexico who has dropped 6 spots in the ELO. Not that there would be a change in the anticipated US loss. I also think you could use the Gold Cup final as an away game for the US based on my recollection of the crowd makeup.

      I think one thing that makes the stat based analysis difficult is that it was very tough to tell how much Scotland cared about that game. Last year we know that the teams went out to compete at their best.

      Reply

      • Posted by schmutzdeck on 2012/07/17 at 6:14 PM

        Jared,

        “I think one thing that makes the stat based analysis difficult is that it was very tough to tell how much Scotland cared about that game. Last year we know that the teams went out to compete at their best. “

        You are 100% correct. The Gold Cup was a tournament that BB and most of the players knew could well determine whether BB kept his job.
        Unless they lost every game 10-0 there was no way JK was getting fired over the simulated tournament.
        That sort of intangible stressor was impossible to replicate in JK’s version of baseball’s “simulated games”. So the comparisons have to be made on a curve so to speak.

        Reply

  3. Another good article from TSG. However, when comparing these two coaches… it should be noted that Bradley was the coach for the last 6+ years while Klinsmann is just now finishing year 1. Then look at the numbers and Klinsmann already has the team playing at a level similar to Bradley’s, if not better in certain areas.

    Also, consider Bradley had Donovan/Dempsey throughout his 6+ years (predominantly healthy as well) while not until the recent 5 games “tournament” did Klinsmann have both on the same pitch. Just wanted to make those obvious comments while reviewing the results. I am happy with the change and by WC2018… I expect to see a much better team on the pitch. That is right… I’m looking that far ahead.

    Reply

    • AG,
      You are not entirely fair.

      I’m happy to have JK as well but it should be noted that this team is still basically Bradley’s team and if they are playing at a level similar to Bradley’s or better, it should be no surprise.

      The core is still there and the surrounding players are now better (Jones, Fabian, etc.). Some of the holdovers, MB90, Torres, Gomez, Jozy, etc. are better now than they were a year or two ago. And MB90, Torres and Jozy got better because they got better at their clubs not with the USMNT. I also think Gomez is a better forward now than he showed when he was at the World Cup, though I thought he did well there.

      And there appears to be a better group of young and upcoming players (Shea, Boyd, Corona, etc.) to push the older ones than what Bradley had.

      The point is this team was likely going to be better no matter who BB’s replacement was. JK has put his stamp on it but it is still a long way from truly being his team. That will come after 2014 Cup, if he stays around that long.

      Reply

Leave a Reply

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

WordPress.com Logo

You are commenting using your WordPress.com 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

Follow

Get every new post delivered to your Inbox.

Join 251 other followers

%d bloggers like this: