Chad Marshall's first year in Seattle garnered a well-deserved Defensive Player of the Year award; the veteran CB was the key component of a shield-winning defense despite heavy rotation and spotty performances from the second CB and a regular fullback pairing on different-but-mutually-undesirable ends of the aging curve. On the numbers, whoscored.com rated Marshall 4th in the league in overall performance, ranking 13th in clearances per game and 1st in interceptions. So, when Jeremiah asks me whether long travel and cramped seating have impacted Marshall's performance, answering should be straightforward, right? See, for yourself, above.
The accumulation of individual defensive statistics is primarily determined by role rather than player quality, additionally modified by game pace. These facts make individual defensive performance extremely difficult to measure - particularly from game to game. Nevertheless, if Marshall is systematically susceptible to struggles on the road in contrast to his teammates, we might expect such changes to show him picking up fewer of the overall team actions. The header plot depicts Marshall's share of team defense - that proportion of tackle attempts, interceptions, clearances and blocks for which he is individually responsible - for each of the 40 games he's started as a Sounder. Red triangles mark away games, green squares denote home games, and the overall picture shows little difference. The largest multi-game perturbation in Marshall's responsibility with Seattle appears associated with the early-2014 experiment with the 4-3-3 formation and early efforts to develop effective tactics. Median defensive share is somewhat greater for home games, but mean shares for home (~18%) and away(~17%) illustrate a fairly insignificant shift in performance (so little that it could also be attributed to tactical differences):
Unique performance in draws is likely attributable to small sample size, and the situational splits otherwise show little specific change in Marshall's performance - the middle 50% of the data represented by the shaded boxes covers a very similar range. If he is uniquely affected by travel, it leaves no detectable impact on basic defensive numbers.
If defensive action rates are inextricably tied to role, is there a better means of assessing performance?
Thus far in 2015 the Sounders have faced quite a few single-forward setups (allegedly) - including all of Marshall's starts. We can broadly assign the opposing forward as being uniquely the responsibility of Seattle's centerbacks (with the unfortunate side-effect that Marshall's performance cannot be separated from that of Brad Evans). The figure above depicts how often those forwards were used (by passing rate) and how dangerous they were (by chances - shots + key passes + assists) against Seattle in comparison to their overall performance on the season (negative values mean the player passed or created more often against Seattle, positive numbers indicate relatively weaker performance). Bearing in mind that shot frequency is a poor measure of quality, each of Gabriel Torres, Kei Kamara, and Patrick Mullins appear to have been more dangerous matched up with Marshall than at other times in 2015, and form a tentative trend of weaker performance with travel distance. The sample size of this final plot is far too small to be considered meaningful, but a similar approach may be worth a look in the future.
Raw stats are gathered from OPTA via whoscored.com.