Germany’s historic victory over Brazil is the kind of game that makes ordinary analysis virtually pointless. It is so out of line with any sort of sensible expectations that it short circuits our conventional methods of processing a game. Rather than acknowledging the unlikeliness of the outcome, its improbability, we are even more prone than usual to fall back on narratives that suggest the inevitability of the outcome.
Nate Silver, reflecting on fivethirtyeight’s prediction of a Brazil win, used the game as an opportunity to think more generally about the statistics of truly unlikely events (though critics will suggest, not entirely without cause, that Silver was also constructing an apologia for his model’s shortcomings). Though I don’t have anything to say about the statistics of the game, I’m similarly interested in using the game as an example of a more general situation: of how we narratively transform contingency into inevitability.
In his typically sharp analysis of the game, Michael Cox began by suggesting that “this ridiculous scoreline was an entirely fair reflection of Germany’s dominance.” To be clear, Germany’s dominance is not in question, let alone that Germany was always the game’s likely winner. But I am interested in thinking about what it means to call 7-1 a fair reflection of the two teams. I doubt Cox believes that if Brazil and Germany were given a replay this weekend, the result would again be a six-goal victory (yesterday’s third place game may show how poor this Brazil team is, but there’s still a gulf between 3-0 and 7-1). If not, then what does it mean to think of this outcome as a fair reflection of anything?
A critical aspect of thinking about the loss is the fact that the game went from 0-1 to 0-5 in six minutes. That’s six minutes from what between relatively even teams would look a tight game that anyone could still win to an unwinnable embarrassment under any circumstances. While I’m not claiming that this was a contest between relatively even teams, it was certainly was a game between historically comparable nations. Whatever the case, six minutes is a very short time in which to alter a game plan or to recognize and respond to a fatal flaw.
To say that the team or Scolari should have reacted within that time frame is natural, but it’s probably an unreasonable expectation. Teams oftentimes have rough going in the five minutes after conceding a goal. In most games, the opportunities created in this period aren’t realized–as is the case for most chances at any point in any game–and some new level of normality for the game is established going forward. But here, Germany’s chances all came good in that brief period, putting the game out of reach and effectively ending the contest. Is that evidence of a superiority that should clearly have been recognized and countered, or just a statistical anomaly? How could one even distinguish between those hypotheses?
The first 22′ of the game, of course, are a different matter. That opening period could reveal signs of a fatal flaw that should have been recognized and reacted to. Was that the case here? Michael Cox is very clear–and clearly right–that Brazil’s greatest weakness was down their left (Germany’s right) flank, in behind Marcelo. He lists six instances of Germany getting in behind Marcelo before Germany’s second goal, though on two of those instances, Marcelo actually did make recovering runs. Even so, four lapses in 20 minutes is enough to ask why Marcelo and Brazil didn’t alter their approach.
The reason, pretty clearly, is that Marcelo’s attacking posture was an integral part of Brazil’s own game plan, providing support for Hulk in trying to breach Germany’s high line down the flanks and particularly on the left. As had been the case in Germany’s two previous games, against Algeria and France, Luiz Felipe Scolari identified the Germans’ high defensive line as a point of attack. Moreover, he seems to have taken the lesson that competing in the midfield with this German side is a losing battle, and instead sought to circumvent the middle, both via direct balls from the back four and from attacks down the flank, especially down the left. Presumably, the focus on the left had less to do with targeting Philipp Lahm (still a great individual defender) than with trying to exploit Lahm’s more attacking posture (relative to Höwedes). But the decision to focus on the left, to the extent that it would encourage Marcelo’s natural attacking style, would almost necessarily involve leaving openings in behind.
Now, there’s a lot to critique in Scolari’s specific implementation of this game plan. As Cox notes, Fred’s lack of pace made him even less likely to contribute in this game than in earlier ones (where he was already of questionable value). While placing Hulk at the key point of attack makes sense in theory, he might nonetheless have proven more effective starting in his more natural right position, or (as I suggested) up top in place of Fred. I actually think that Bernard’s pace makes good sense given Scolari’s game plan, but others could differ, especially if the decision was made at the expense of a start for Willian (who must have not yet reached full fitness–even Scolari wouldn’t have omitted his most obvious replacement for Neymar otherwise). If fullbacks were expected to play a major role in the attack, placing a good two-way player (like either Oscar or Ramires) ahead of them might have allowed for better balance between attack and defense.
Having said all that, it’s difficult to argue with the objectives of Brazil’s game plan. Accepting that Scolari might have developed a better specific approach, the real question for those opening 22′ is whether or not Brazil’s attack was showing opportunities that would justify the threats to their backline that Cox identified. Knowing the result in advance, it’s difficult not to focus on Germany’s strengths and Brazil’s weaknesses; but let’s try to turn that around. Here’s a timeline of the lead up to Germany’s 2nd goal focusing on Brazil’s opportunities rather than their shortcomings (to make sense of the following, it might help to review Cox’s timeline of the opening 25′ of the game by way of comparison–I’ll be referring to it at various points in my counter-narrative):
At 0:15, Marcelo came forward and fed Fred, who sent long ball over the top for Bernard that challenged Höwedes. This was the start of 2′ of Brazilian possession in Germany’s final third. At 1:28, Marcelo crossed a ball back into the box in the aftermath of a Brazilian corner kick, but no one was there. At 2:16, he recovered an attempted clearance and took the first shot of the game.
At 3:07, Marcelo wasn’t really involved as David Luiz hit a long ball to Hulk down the left flank and he crossed dangerously in for Bernard, a good chance, though Manuel Neuer intercepted. This led to Germany’s first break in behind Marcelo at 3:17, and while that was clearly Marcelo’s error, the overall play looked like an exchange of chances for both sides. On ESPN’s feed, it was Brazil’s chance rather than Germany’s that was deemed worthy of a replay.
At 8:28, not long after Germany’s first real chance (at 6:50, noted by Cox), Marcelo stripped Müller of the ball and forced the striker into a professional foul in order to halt a counter-attack.
At 9:35, as Cox noted, Marcelo conceded in attack, again from a long ball to Hulk down the left flank. Though a clear individual error, it was decidedly not an example of Marcelo taking up a tactically unwise position, as his effective recovery (also noted by Cox) demonstrated.
Although that concession of a corner kick did set up the opportunity for Germany’s first goal, blaming Marcelo’s defensive lapse seems extreme, particularly given the corner kicks conceded as a result of Brazil’s attacking strategy at the other end of the pitch. The real explanation of that goal was some clever and effective obstruction of David Luiz by Miroslav Klose, allowing Müller to reach the back post unmarked and thus able to easily finish the opportunity.
If anyone was at fault, you’d probably have to blame Júlio César or Luiz, so this too looks like an individual error rather than a tactical shortcoming in need of change.
At 11:32, immediately after the restart, Marcelo hit a long ball up the left flank for Hulk, but he couldn’t get ahold of the ball.
At 13:00, Marcelo won a long ball from Müller, sparking an offensive run up the middle for David Luiz that fed Hulk, who couldn’t get his pass through for Fred in the middle. This led to a German counter-attack up the same flank that was stopped by a foul from a very well-positioned Marcelo at 13:22 before Müller entered Brazil’s half. It seems harsh to include this incident, as Cox has, on the list of German opportunities to get in behind Marcelo.
At 16:37, Hulk fed Marcelo breaking dangerously into Germany’s box–Cox acknowledges this as “one of Brazil’s best moments”–and only a perfectly-timed tackle by Lahm, one that could have been misread as a penalty opportunity on a different day, kept the play from developing. This was probably as good a chance as Khedira’s opportunity at 6:50.
At 17:42, Marcelo’s advanced position allowed him to play a cross to Oscar on the right side of the box to create a half-chance for Brazil. These two chances underscore Brazil’s attacking threat in the period after conceding that first set-piece goal.
Cox notes another opportunity in behind Marcelo at 18:35, and it’s certainly the case that, having joined the attack, Marcelo chose not to track back on the German counter-attack. But in reality that attack was hardly threatening at all, suggesting–at least in this instance–that Dante, Luiz Gustavo and Oscar had the situation well in hand, and that Marcelo’s attacking position was sustainable.
So, before conceding their second goal, Brazil’s strategy of using Marcelo and Hulk to attack down the left had ceded 3-6 chances (before Germany’s second and third goals) and created about the same number. All chances are not equal, of course, and it is certainly true that Brazil’s chances were all created in traffic, while Germany was finding open space in their attacking moves. Nonetheless, prior to Germany’s six-minute explosion, you would probably describe them as having the better of an open game being contested by both teams, each creating opportunities down the same flank. Had Marcelo managed a slightly greater purchase on his 3′ long shot or Hulk done a better job on his 4′ cut back for Bernard, had Marcelo gotten off his 17′ shot or Lahm’s tackle been a fraction off (or less accurately observed by the referee), does it seem likely that the end result would have been a 7-2, 7-3, or 7-4 Germany win?
Germany was clearly the better side, probably the best of the tournament heading into today’s final; while Brazil have looked underwhelming throughout. Germany’s victory may, indeed, have been something like inevitable. But expecting Brazil to have recognized a 7-1 drubbing was on the way and to have taken precautions to avoid it, especially in the opening 22′ of the game, is to ask them to predict a high improbable future. That’s a lot harder than predicting a highly improbable past. Anyone can do the latter, but that doesn’t make it any less improbable.