Nov 22, 2020; Kansas City, Kansas, USA; San Jose Earthquakes goalkeeper JT Marcinkowski (18) punches out a shot on goal during the second half against Sporting Kansas City at Children’s Mercy Park. Mandatory Credit: Denny Medley-USA TODAY Sports
At his postgame interview after the 4-0 romp in Carson over the Los Angeles Galaxy on October 14, Matías Almeyda told the media that “this team has an identity.”
San Jose had just won their fourth in five games, their best run in the shortened, bonkers, 2020 season. Almeyda was reacting to that, of course, but also reacting to where this team is in his tenure and as an organization. As Chris Wondolowski rides off into the sunset (or up 680, or whatever), Jessie Fioranelli and Almeyda are remaking the team in their image. This will be a three-part series looking into what this identity is:
- What it means defensively
- What it means on the attack
- Whether this identity is even any good or not, or what a “good” identity is
I should start by saying: I’m optimistic about the club’s direction. This might be because I’m a born optimist, or because there isn’t much else to be optimistic about right now, or because the winter transfers really haven’t gotten underway.
But it’s also because this is the first offseason in which San Jose has been making progress on an identity rather than starting from scratch since, what, 2013? There is momentum here, but not perfection. Which is fine! As someone who is helping raise a kid during a pandemic, I’ve really come to embrace that there is no “perfection,” that no individual in a society ever reaches their final form. Life is not an RPG, and neither is a sports team. There is only constant reiterations and tinkering, and making improvements daily, weekly, monthly.
The Earthquakes have to do all that, and oh boy, as we discuss the defense we’ll talk well past tinkering. But first, let’s set a baseline for discussing what an MLS team is and does:
Key Passes as Identity Formation
There are a ton of statistics out there to talk about how soccer is played. When discussing “identity,” I think it’s common to focus on tempo, pressing, formation: how fast to play, how confrontational to play, and where to play from. These are your basic toggles in Total Football Manager, after all.
I’d like to look at something different: key passes. Not because I think the things listed above are not important, but because they are player-dependent. If you have three great centerbacks, play three-at-the-back. If you don’t have good passers, just blast the opposing midfielders in the heels. Key passes get to something more direct: where are you trying to hurt your opponents? Where are your opponents trying to hurt you? To make a gruesome metaphor: it’s not about choosing a sword, it’s about aiming for the heart. You’d think more teams would remember to do that…or you might not, if you’ve watched a lot of MLS over the years.
Speaking of MLS over the years, key passes are also nice because my colleague Jamon Moore (in his work for American Soccer Analysis) made a great dataset showing what types of key passes are made, by quantity:
The “xG” numbers show the relative value of those key passes: i.e., the average cutback is worth .32 expected goals (32% chance of scoring a goal) while the average longball forward is worth .10 expected goals. A team that tries to hit cutbacks will likely have more success than a team that settles for longballs.
A quick word about these pass types: they’re fairly straightforward, with the two less-defined categories of “None” and “Regular” included. “None” is used where there is no key pass. Rebounds, shots from a poor defensive action, and individual shot creation all fit into this category. “Regular” can be defined as “simple passes, such as square passes, or passes not reaching a progressive pass length.” To help illustrate it, here are all of Thomas Muller’s key passes in two recent Bundesliga seasons:
The relative value of these passes can also be seen by comparing key passes that lead to shots with key passes that lead to goals:
The left-hand table here is the same as the one above. Here, you can see for example that progressive passes make up 15% of all shots but 20% of all goals. Or, progressive passes are weighted towards being threatening, while corner kicks (7% vs 5%) are not.
A good team will focus on getting key passes that are on the left of these tables and allowing key passes to the right of these tables. As a one-line instruction to players: make decisions for your teammates easy, make decisions for your opponents hard.”
Defending Antagonistically
In a preseason interview, Jesse Fioranelli proudly said “we’re one of the most annoying teams in MLS,
“not because we’re rambunctious, but the way Matías has trained the players. We can be competitive if we stick to our plan. That’s why we were exciting to watch. It might be part of the reason we fell short [last year]. Maybe we weren’t used to that pressure and expectations.”
This annoying-ness is the Earthquakes defense at their best: buzz around, make field players receive the ball facing their own goal and force longballs from the opposing keeper. I’m not going to repeat Joseph Lowery, who did a great job explaining it in a piece for The Athletic in July 2019. But it’s fun watching a frustrated pair of centerbacks force passes into a constantly-rotating San Jose front four:
For all of the talk on Almeyda’s 1v1 system, the Earthquakes at their best defend as a team and with their forwards. San Jose can be picked apart in transition but when they keep the game in front of them and have the wingers and Chris Wondolowski press centerbacks into the touchlines, the opponents make bad passes into whomever Judson is defending, and concede possession.
Defending by Attacking
This goes into another truism about Almeyda’s defensive tactics: they’re exhausting for both teams. San Jose actually toned down their press by the end of 2020 (likely because of fixture congestion as much as any tactical nous), but they were able to maximize their pressing by maximizing their own time on the ball. If you have 30 minutes of “press” in your team’s legs, that’s only 55% of the opponent’s time on the ball if you’re conceding 60% (or 54 minutes) of possession. But if you can hold on to the ball for 60% of the time, that 30 minutes of press can make 83% of the opponent’s 36 minutes of on-the-ball time a living hell.
This can be tough to see on any given matchday, because some teams (Minnesota, Portland) are comfortable on the back foot and happy to concede possession to San Jose. But in the whole, the pattern is clear:
Red signifies less possession (and a losing position), blue signifies more possession (and a winning position) for San Jose.
San Jose had a ton of possession in the Magnus Eriksson era of 2020, and again in that October-November run. What’s interesting is that even though the Earthquakes scored more in the second half of games, they often didn’t hold as much of the ball as they did in the first half. Keeping the opponents chasing the spread-out attack (which we’ll get to in the next installment) tired them out and made them easier to defend against in the second half.
Holding on to possession is this San Jose’s key way to keep the opponent off the board. As we’ll see in the next section, however, there were times this season that it felt like it was the only way the Quakes wouldn’t get scored on.
Over, Under, Around and Through: Key Passes Against
Because he has access to the data and is brilliant about these things, Jamon Moore plotted each team’s key passes conceded on goals against (what sorts of passes led to each goal):
Some thoughts:
- Only the Galaxy gave up more cutback assists than San Jose
- Only Houston gave up more progressive pass assists than San Jose
- Nobody gave up more set piece assists than San Jose
- Nobody gave up more corner kick assists than San Jose
Because I am very curious about these things, and have no soccer to watch currently, I noted how each of these goals against San Jose occurred. The following table doesn’t precisely corroborate with the one above. I noted the “transition” moment of what broke down the Earthquakes defense, what the assist was (if relevant), and how the goal was scored:
Not a very fun thing to relive, in my opinion! There’s some pretty gross games in there, particularly that September 2 LAFC match where the Angels just won turnovers in the midfield and passed it down the Earthquakes’ throat. Gross!
That said, from September 27th through November 4, you can see a whole lot more of that crossing…the Earthquakes still conceded a ton, but they at least were conceding from lower-quality passes.
Set Piece Nightmares
More stable defending, getting men behind the ball…none of this matters if you’re still letting dudes dunk on you in every set piece.
The truth is, I have no clue how San Jose was so bad at this, or whether this badness is repeatable. It’s an old saw in soccer that set piece prowess isn’t repeatable year after year. I’m not sure if same goes for set piece defense.
A lot of this occurred after Kashia was replaced by Jungwirth in the starting XI, and I suppose there’s something to not having a single outfield player over 6’ in these situations. Fixing set pieces alone probably nets San Jose a handful more points next season, and if you’re frustrated to have this boil down to “Somebody Should Do Something About All The Problems,” 2,000 words in, well, imagine how I feel.
Improving in 2021
If that September-November run is the way the Quakes want to defend, you can see the efficacy: use Judson to tamp down transition moments, have active forwards frustrate the back line into bad passes, and 1v1 defending to keep opposing attackers from receiving the ball with their faces to goal. It’s not a bad plan!
But, as I wrote back in August, San Jose still doesn’t really have a Plan B if everything doesn’t work out perfectly. If Judson needs to stop running at some point, if a new forward misses a rotation, if a defender gets caught between two attackers, it goes to pot.
Coming into 2021, the Earthquakes need someone who can do a halfway-decent Judson impersonation. They also need to be careful that any high-priced attacking piece they bring on is unafraid of doing the drudgery of shuttling between centerbacks in the 65th minute of a Wednesday fixture.
And they also have to do something about the set pieces! There is probably some fanbase-wide consensus that the team really hasn’t gotten over Clarence Goodson’s retirement. It’s been years since San Jose’s had a player who can organize a backline, pass through the midfield, and dunk a basketball. I would say something about replacing Florian Jungwirth in the back, but San Jose tried to do that the past few offseasons and always ends up starting Flo in games where the season’s on the line.
The fact is, the Earthquakes have had a historically-bad defense in three of the past four years. Almeyda’s new San Jose identity involves defending aggressively, annoyingly, and obnoxiously. It has shown glimpses of success and wide gawps of failure. Depending on what happens at right centerback next year — if a Flo replacement comes in as many expect, or not — there will be a complete personnel overhaul from the shambolic 2018 defense headed into 2021. With a defense to fit his style, there’s no excuse for the defense to go weeks surrendering goals by the handful again in 2021. And yes, that includes set pieces.