clock menu more-arrow no yes

Filed under:

Immeasurables change the state of games

New, 17 comments

The impact of game state on the Sounders' statistical performance should let the team feel better about Sunday's shaky win, but not by much.

"But, have you seen our stats?!"
"But, have you seen our stats?!"
Mike Russell

If a tree falls in Seattle with 64,000 fans shouting around it, does it make a sound?

I've previously lamented the sparsity of meaningful soccer statistics. For every action leading to a goal - or even a good chance at goal - there are 100 measured events that don't appear to impact the game's result in any meaningful way. For those few events that matter, myriad factors (often best described as "luck") make the difference between a pass and a key pass, a blocked shot and a chance at goal, a keeper in position or off the line, and a ball off the post or in the net.

Courtesy the chalkboard app on mlssoccer.com, here is the "buildup" to the Seattle Sounders' key opening goal in the 6th minute of Sunday's game against the Portland Timbers.

martins150830

Missing from this schematic is a recovery and successful pass from Jack Jewsbury which makes the difference between the end of Nelson Haedo Valdez's unsuccessful pass and his "unsuccessful" tackle on Nat Borchers in the box.We also don't see (and didn't even record) Brad Evans' challenge for the loose ball once Borchers lost possession - a key factor in the ball reaching Obafemi Martins for the recovery and shot. Three Portland defenders - and at least as many attempted clearances - don't show up at all, because they failed.

Valdez had a good run, and kept with the play even when his service to Martins went awry. Marco Pappa had a good entry pass to start the play, and Stefan Frei distributed well with a throw following an easy save against Lucas Melano on the other end, but the main factors that made the first goal were the balance and elite athleticism of Obafemi Martins, coupled with a handful of loose ball actions we explicitly do not measure. From there, the game changed.

unmeasure

The plot above depicts several performance indices in 10-minute windows sliding in 5-minute intervals across game time (each data point represents 10 minutes of play, and each data point shares 5 minutes with its neighbor on either side along the line), representing changes in game state with key events. Seattle struggled with passing accuracy throughout the first half, including the period before the goal (the aforementioned save on Melano resulted from nice service from Darlington Nagbe, who had received the ball with plenty of open space in front of him from Lamar Neagle). But after the goal, Portland pressed forward strongly, unleashing a flurry of shots not limited to their dangerous set pieces, and forcing Seattle into desperate clearances.

The "clearance bias" concept (here formulated as clearances divided by the sub of attempted tackles, interceptions, and defensive blocks) is something I've discussed before as a measure of whether a defense is attempting to transition into attacking play. A bunkering team should be expected to clear more often, relative to its total defensive actions, given the danger of opposing players flooding the defensive half and the absence of distribution targets upfield.

The Sounders were fairly lucky not to concede in the 1-0 game state, but they didn't. After the game went 2-0, and after the half, Seattle continued to look disjointed, but settled down on the ball enough to reduce Portland's chances - even after the Timbers drew within 1. The decline in clearance bias for Seattle once Cristian Roldan, Gonzalo Pineda, and Osvaldo Alonso subbed into the midfield reflects the team's improved passing accuracy and possession in the middle third.

But..

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

Those who clicked on the link above (and may other Sounders fans besides) will certainly recognize Caleb Porter's line from Portland's 2-1 playoff victory in 2013. Then, he referred to a large volume of attacking numbers Seattle put up trying to come from behind. Now, he expresses more or less the opposite sentiment about a very similar 2-1 result, having the opposite beneficiary...

We wanted to come here and not just sit back and defend but attack and I think we showed that throughout the whole game. It was maybe 20 shots to six at their place so we are happy with the performance...

Caleb Porter will say whatever he wants to say about statistics, provided they put himself and his team in a favorable light.

Still, what I said about Portland's defensive posture not fully containing Seattle 2 years ago applies twice as strongly to Sunday's game. Certainly playing with a lead changed the Sounders' play - it unfortunately made them still more likely to concede. Seattle didn't play well, but played just well enough to win. When the factors we can't or won't measure swing the other way, the team will have to play better.

Raw stats for this work were gathered from OPTA via whoscored.com.