Away, Bitter Away

There are two basic truths embodied in the graph above.

Firstly, teams perform better at home. From 2010 to 2014, there were 99 full season performances from MLS teams. Only 7 exhibited better records away than at home, and the largest margin in favor of road play was 0.353 ppg (2012's Chivas USA). On average, since the beginning of 2010, MLS teams have earned 1.016 ppg on the road and 1.674 at home. To put it another way, the average MLS squad at home matches the overall regular season performance of the 2013 Portland Timbers, or marginally superior to 2012's Seattle Sounders.

The second truth is more of an admonition not to trust small sample sizes. The extreme performances of Toronto and Orlando (the two lowest points) won't be sustained over a full season, and Columbus (the highest data point) will similarly lose its bizarre Home/Away disparity given half a chance. Seattle might have anticipated a weaker opponent given the Crew's overall performance so far in the young season, but Columbus has been dominant at home.

I believe that a significant portion of the difficulty teams face playing on the road is often self-inflicted. Defensive tactics applied too early in the  game are often counterproductive. After Saturday's game, I was curious to see whether tactical changes might explain some of the difference for these two teams.

These plots represent the changes in player share of basic team passes and defensive actions between home and away performances in 2015 (as was more thoroughly discussed here and here). Several players exhibit very consistent usage patterns home and away- particularly forwards (red) and goalkeepers (purple). Seattle and Columbus both show increased passing share away for a defensive or holding midfielder (in blue; Pineda for Seattle and Tchani for Columbus). The usage of Seattle's fullbacks (defenders are marked in pink) away also increases significantly. Mohammed Saeid's increased defensive share in holding midfield is the most dramatic change for Columbus (if anything, this would be the potentially conservative aspect of play on the road the Crew should look to change- the attack is the greatest strength of their team, and cautious play may do it more damage than it benefits the defense). Overall, the differences for Columbus are impressively slight; to explain their relative difficulty on the road, the original explanation of "fluke/sample size" may have to suffice.

The Sounders went on the road and faced a very good team. The result must be disappointing, but must not be unexpected. Seattle's current 1.4ppg on the road has been exceeded by only 10 of those 99 full season performances since 2010. If there are to be significant lessons in Saturday's 3-2 loss, I have a few suggestions:

-Columbus' attacking movement pulled the Sounders centerbacks out of their best positions to defend crosses back in front of goal. The team needs to deal with those circumstances more effectively.

-Seattle struggled to force scoring chances after pulling within 1 (in fact, their only shots of the last 15 minutes came from defenders Tyrone Mears and Chad Marshall).

-Osvaldo Alonso's defensive share has been erratic (see the precipitous blue drop in away play in the graph above). To the extent this can be attributed to tactics, it likely warrants a change. Seattle needs the honey badger in defense more than it needs him upfield.