Game State, Revisited

I hadn't planned another game breakdown so soon, but - as Dave recently discussed - Caleb Porter went and said something silly (again):

"You can put your hand up the puppet of stats and make them say whatever you want to say"

I suspect many readers of Sounder At Heart caught my earlier reference to a saying attributed to Benjamin Disraeli and popularized by Mark Twain "There are three kinds of lies: lies, damn lies, and statistics." The reference was tongue-in-cheek... statistics cited selectively may be extremely misleading - like giving only the prosecution's side of a case at a jury trial. The numbers are dispassionate - when you select a non-representative subset of the data, they mislead, and this is perhaps the most common transgression in public discourse. With soccer statistics, we're still trying to describe what the roughly objective numbers "mean" regarding overall performance, and whether the signal of on-field play exceeds the noise of the game. "Is the sample representative" and "are the numbers meaningful" - these are legitimate questions... but I hate the saying "stats can say anything" - they can't, so long you as you tell the whole story.

So, aside from press conference sound bites, what was Porter trying to say in the comment above? The Seattle Sounders dominated several statistical categories on Saturday, but still found themselves down 1-2 heading into the second leg. Porter claimed the statistical disparity was all part of the plan. Was it?

As before, the plots above represent a 10-minute sliding window of stats collected at 5 minute intervals. "Clearance bias," again, is Clearances/(Tackles Won + Interceptions + Recoveries). Portland numbers are in red/diamonds and Seattle in blue/squares. The vertical dashed lines represent the timing of the two Portland goals.

Portland came out into the first leg of the semifinal with a predictably conservative game plan: sitting back, playing direct, and clearing instead of directing. That approach is clear in each of the three phases of the game delineated above, and Seattle's numerical advantage would likely have persisted even in a scoreless draw. However, the severity of the statistical gulf between the teams may plainly be attributed to events as the game progressed.

After a brief interval following Johnson's goal, where Portland pressed and Seattle reloaded, the Timbers doubled down on a defensive posture. They tried to limit the scoring Seattle's opportunities by clogging their own defensive third. For roughly the middle 30 minutes of the game, the strategy wasn't necessarily perfect, but it certainly made an impact. The Sounders could pass accurately, but they couldn't generate shots at a high rate. Around midway through the second half, that began to change (also partially coinciding with Rosales subbing on for Scott in the 63rd minute). A defensive gaffe from Traore rewarded Portland's direct play once again in the 67th minute, and from that point on - even after Alonso pulled one back - the Timbers bunkered and the Sounders mostly failed to finish. Still, breaking down the phases of the game, it's clear that the defensive approach never wholly contained Seattle's attack.

Is Porter correct, or is he rationalizing? Both. Portland clearly adjusted to circumstance, and if the disparity in finishing persists a similar approach to Thursday's game would be promising... but the Sounders were clearly their own worst enemy in the first leg (woodwork's antics notwithstanding). With better finishing or better fortune, the phase 1 or 2 hold some hope for the Sounders. Phase 3 is a posture Portland should avoid. If the Timbers and Porter find a different form at home and generate opportunities equal or above those of Seattle, the Sounders will have a hard time playing themselves back into the tournament.