Chasing the narrative: What the numbers tell us (if anything) about Michael Bradley’s World Cup

19 Comments

After the U.S.’s third group game, the counter-narrative hit full steam, with Major League Soccer’s official website putting its weight behind the rosy view of Michael Bradley’s play. According to FIFA, the United States central midfielder covered more ground in group stage than anybody in the tournament, a factoid that got a full writeup as the league lauded one of its biggest stars. It’d be nice if Bradley’s other numbers got the same attention, but it’s understandable why MLS is trying to promote Bradley’s cause.

So why don’t we do that? Why don’t we give the full statistical record some attention? If the “Bradley’s been awful” narrative is already out there and the counter-narrative’s gaining steam, why don’t we try to take a broader look at Bradley’s production? Let’s collect the numbers, provide some context, and see if there really is anything that confirms the general assessment. Is there anything in the broader statistical record that says Bradley’s been bad?

The subtext of that assessment is more complicated, though. It’s not that Bradley has been bad in the absolute sense (some people are saying this, though). It’s that he hasn’t come close to meeting expectations. This is a player around whom head coach Jurgen Klinsmann has built his formation. It’s not only fair to judge him against a normal player’s expectations but against the performance his coach wants from his most important player. Klinsmann and the U.S. have incurred a type of opportunity cost by setting up as they do.

For his part, Klinsmann has weighed in on the issue, but that’s not the point, here. The point is to see if the statistics, as flawed and limited as they are, offer any support to either claim. What numbers say Bradley’s failed to meet expectations? Or, what statistics support the opposing view, that Bradley’s been fine and has become a scapegoat?

Some numbers we won’t discuss here: Goals (Bradley doesn’t have any, though he should); assists (again, none); distance covered (it speaks to effort, not efficacy). Those numbers are pretty self-evident. We don’t need to dig deeper.

Instead, we’ll look at some of the secondary statistics, though these numbers don’t mean much independent of their context. So we’ve done our best to give that to you. Not only do you get the raw number, but you see where Bradley ranked within the team and within the game.

Of course, that’s only part of the context. The team’s style of play has also been a big factor in Bradley’s numbers. It’s hard to lead a game in passing when your team’s willing to play without the ball. For the most part, that’s what the U.S. has done.

There’s also Bradley’s role within the team to consider, one which establishes some expectations for his performance. As the highest man in the U.S.’s central midfield, we should expect Bradley to be among the team’s leaders in key passes, touches, and passes every game.  As a focal point for the U.S. in transition, Bradley should see more contested time on the ball than his teammates, meaning he’ll likely lead the U.S. in times dispossessed and turnovers, even on good days.

What we’re really looking for here are outlying numbers – something that supports the notion Bradley has been particularly poor:

(All numbers are available via WhoScored.com.)

Opponent Key Passes Touches Dispossessed Passes Turnovers
Ghana (W, 2-1)  0  52  3  43  2
Ranks (Team/Game)  T6/T16  4th/11th  1st/T1  1st/T6  2nd/T4
Portugal (D, 2-2)  1  82  1  69  0
Ranks (Team/Game)  T4/T8  1st/3rd  T1/T3  1st/2nd  T7/T11
Germany (L, 0-1)  1  61  2  49  2
Ranks (Team/Game)  T1/T3  1st/9th  T2/T3  1st/9th  T3/T3

 

There are other numbers we can look at, like passes completed, through balls, and dribbles. I chose these because they require the fewest caveats, whether those caveats be based on the nature of the data, context dependence, or other, more philosophical issues.

The numbers that stand out: The lack of chances creates (key passes – passes that lead to shots), and two games where the overall pass numbers are low. That Bradley had three ‘dispossessions’ and two turnovers against Ghana (when his passes and touches were both low) supports the idea he was far below his standards in the opener. The larger body of data only supports one other claim: Bradley hasn’t been creating changes. The degree to wish you can separate that from the U.S.’s tactics is up to you.

The defensive numbers, where, given the U.S.’s deficits in possession, you’d expect Bradley to be among his team’s as well as the game’s leaders:

Opponent Tackles Interceptions Fouls
Ghana (W, 2-1)  1  1  2
Ranks (Team/Game)  T7/T13  T3/T5  T3/T4
Portugal (D, 2-2)  1  2  2
Ranks (Team/Game)  T4/T9  T4/T8  T1/T3
Germany (L, 0-1)  4  0  3
Ranks (Team/Game)  1st/1st  T8/T13  2nd/2nd

 

Again, what’s our goal here: To find something that clearly highlights Bradley’s struggles. Among the defensive numbers, there’s nothing convincing in either direction. Bradley had a good night in tackles against Germany, but he only generated one more turnover than he did against Portugal.

All of the in-team and in-game ranks look decent enough, considering players like Kyle Beckerman are better positioned to rack up tackles, while Matt Besler would be the best bet to lead the team in interceptions.

Even in the fouls number, there’s nothing eye-catchingly bad. While there’s little here that tells us how Bradley actually played, the description the numbers offer doesn’t support many conclusions.

source: Getty Images
Michael Bradley speaks to the media during training at Sao Paulo FC in Brazil. Through three games at the World Cup, the U.S. central midfielder has no goals and no assists while leading the tournament in distance covered. (Source: Getty Images)

So we’re left back where we started: Relying on our eyes, our analysis, and the more obvious numbers. The extent to which any of that’s reliable is unclear. He missed an easy goal against Portugal and has failed to generate many chances for his teammates, but it’s reasonable to assume the absence of Jozy Altidore and the U.S.’s tactics are heavily skewing those results (at least, the chance creation). People may be seeing what they want, judging Bradley against what they hope will happen instead of a more equitable benchmark.

Me? I tend to agree with the broader opinion. This is as bad as I’ve seen Bradley play in a long time. I think Bradley’s been below average compared to other midfielders in the tournament, and I think it’s completely fair to judge him against a higher standard, given his coach has made decisions to put him in advantageous positions. Though I think, given his track record, it’s highly unlikely Bradley’s poor performances will continue, I think his first 270 minutes in Brazil speak for themselves.

How do I back that up, though? Persuasive language, mostly. I can also highlight particularly poor touches, appeal to my own self-inflated authority, or rely on the wisdom of crowds, however “wise” that may be. Sometimes, I actually manage a complete, cogent argument, though it’s pretty rare.

What I can’t do is rely on the numbers. I can’t lean on distance covered. I can’t build a case on chances created. In context, there’s little in the statistical record that tells us anything about Bradley’s World Cup, whether I’m trying to trumpet the counter-narrative or pile on.