The puppet speaks!

There are a nearly endless number of stories to tell through soccer statistics, even if one can’t choose the plot. Below, I’ve chosen three key points about Saturday’s final to discuss in the numbers.

Play The Best Team Available

In the 33rd minute of the Columbus Crew’s match in Seattle, a [seeming] lifetime ago on March 7, Lucas Zelarayan cut inside from the left wing on a dribble around Sounders starting winger Miguel Ibarra, deftly avoiding a challenge from João Paulo before dropping the ball off to Artur in space at the top of the box. Xavier Arreaga closed down the central midfielder, snuffing out the chance at a long shot, but leaving the back line undermanned. Artur’s quick pass out to the right found Luis Diaz with ample space from starting left back Joevin Jones, who failed to block a low cross into the six-yard box. Yeimar Gomez Andrade stepped toward the near post with the rushing Zelarayan (with the still-tracking Ibarra left in the dust). The cross found the unmarked Gyasi Zardes at the far post, for a 1-0 lead to Columbus. Despite Nico Lodeiro’s absence (with João Paulo and Cristian Roldan essentially trading off CAM duty) Seattle had a number of good chances to equalize as the game progressed, including an Ibarra goal called back for offside by VAR. Nouhou entered as a sub in the 59th minute and earned a penalty off Harrison Afful’s arm in minute 79 that Raul Ruidiaz converted (after his first attempt was saved).

Seattle’s lineup that day has some distinct differences from the team now: Lodeiro’s hamstring holding him out, Jones starting at LB, and Ibarra receiving his second of three starts on the season. The late-game Seattle subs were Justin Dhillon and Harry Shipp (who could have earned a game-winning assist, had Jordan Morris converted an open header). In 2020, such lineup churn has been constant. With Monday’s win against Minnesota, Brian Schmetzer has managed to repeat a starting 11 from game-to-game on four occasions, and Seattle has managed to [eventually] win each of those games. Such evidence fuels Schmetzer’s acknowledged preference for playing the “hot hand” and not changing a working scheme. But there are other things all four of those lineup repeats and all three playoff wins share: Ruidiaz/Lodeiro/Morris starting; any two of João Paulo/Cristian Roldan as the CM pairing; Yeimar Gomez Andrade starting at CB. Take a look at the many lineup permutations and results of the 2020 season below.

Assuming a first-choice lineup (a lineup that has only started 2 of 25 games), note the decline in performance with more changes.

The Sounders have been extremely successful when rotating within the first choice ~13 outfield players who fit similar roles within the 4-2-3-1. Although Shane O’Neill brings a different skillset than Arreaga, the team can play in a similar manner around both players. The same can be said for Jones on the wing, Alex Roldan at RB, and Cristian Roldan at CM. Changes that prompt significant tactical adjustments (Bruin at forward, or Delem at CM) tend to cause more problems. The “hot hand” is not the team that started Monday night’s game and fell to a two-goal deficit. It’s not even the team that launched the improbable comeback to win in stoppage time (though Gustav Svensson and Kelvin Leerdam surely demonstrated their value). The hot hand is the best team Brian Schmetzer can put on the field.

Team style by xPass

Take a look at the passing numbers for Seattle and Columbus, courtesy of American Soccer Analysis, focusing on the difference between pass completion rate (Pass %) and expected pass completion (xPass %) based on pass type and field position. At its best, Seattle is a possession team. Every outfield player distributes the ball with average or better success. Only Lodeiro, Morris, and Ruidiaz are expected to attempt a high proportion of risky pass attempts. Only Ruidiaz and Morris exhibit a substantially reduced role in simple possession. In contrast, each of Columbus’ four attackers has a distribution load at about 40 passes per game or less. This holds true even with the absence of Pedro Santos, who will most likely be replaced by Derrick Etienne Jr. or Youness Mokhtar. The absence of both Santos and Nagbe increases the distribution pressure on Zelarayan, and that may, in turn, inhibit his ability to crash the box himself. The relatively light passing load of the attacking band gives the backline a proportionally higher role in cycling basic possession than is the case for Seattle, particularly for the fullbacks. This, along with Darlington Nagbe’s well-below-average defensive activity for his position, puts more pressure on the fullbacks to stay at home than is the case for Seattle. Columbus plays with good width through mobile, aggressive attacking play from the front four, generating a high proportion of chances up close to goal.

With Nagbe, note also the difference in the “vertical” stat between him and João Paulo and Gustav Svensson. This value is an approximation of the average upfield distance of completed passes. In 2020, Nagbe’s 1.67 is the lowest among CMs/DMs recording at least 500 minutes on the field. Svensson and João Paulo rank second and ninth, respectively. Nagbe’s passes are extremely lateral for his role (which is part of the reason he is so effective in holding possession). Much of the work he does to progress upfield is by receiving and advancing on the dribble. With Nagbe now ruled out of the final, Columbus lacks a replacement to match his unusual play style. The most likely stand-in seems to be MLS journeyman Fatai Alashe, who could possibly provide a little bit more traditional DM defensive engagement, but not enough to justify the added pressure on Artur and others to move the ball between the defense and attack.

Defensively, Seattle will want to keep strong positional connectivity on the backline and be prepared for quick, short passing and skillful 1-v-1 play. These factors suggest that Leerdam and Arreaga should be preferred, if available, for the final. On the other end, Seattle needs to avoid excessive reliance on overlapping runs and hopeful crosses. Columbus’ CBs are reasonably effective in the air, and the flanks will have decent defensive support. In the middle, Artur will try to disrupt Lodeiro. Seattle’s three CMs, on their game, will be too much for the Columbus midfield to effectively close down. There will be chances for Morris and Ruidiaz to receive the ball in the box.

Arguably, the absences of Nagbe and Santos are the most disruptive to Columbus’ play that could occur, short of excluding Zelarayan. Although Caleb Porter may not choose an ultraconservative posture as the hosting team, he still may wish to modify the team’s usual approach, perhaps by emphasizing direct, counterattacking play to take advantage of their speed on the wings and attempt to bypass Seattle’s strong defensive midfield.

xG and the Columbus Defense

Columbus has allowed roughly 10 fewer goals in 2020 than expected by shots conceded, the most extreme defensive difference in the league. Tim Foss offered a partial potential explanation on Thursday — the Crew defenders put in a high number of shot blocks — but this doesn’t adequately explain the disparity. Often differences in GA vs. xGA may be attributed to goalkeeper success or to unusual game situations (e.g., a team that regularly defends a real lead sometimes will see xGA build up as the team sits back). Neither seems to apply here — Eloy Room’s GK-specific xG numbers are good but not spectacular, and Columbus’ xGA disparity seems to be distributed over a large number of games (both victories and losses). The Crew have likely done some things to maintain better defensive results than expected, but they’ve also likely been a bit lucky. The Sounders will get chances at goal, but remembering March 7 (or, for that matter, Monday), shouldn’t underestimate the challenge of converting those chances.