Early Season Data Check-in

Jeremy Ebobisse, Rodrigues, Jonathan Mensah, and Cade Cowell form a wall in front of a free kick against Los Angeles FC at Levi’s Stadium on May 6, 2023. These four players will likely be key to how successful the 2023 San Jose Earthquakes will be. Photo credit: Sara Nicole Mindful Photography.

Just 11 games may seem to be a fairly small sample size, but it is about a third of the way through the season, and metrics like expected goals (xG) begin to have predictive power around then. So I wanted to check in on a few major topics of interest and see if advanced metrics from our friends at American Soccer Analysis, FBRef, and Statsbomb had anything to say about them, written from the perspective of someone (like me) who is data-literate but certainly no Jamon Moore.

Luchiball vs. the Almeyda’s “Bushido“

The early returns are unambiguous: the team is massively better this year, at least some of which is attributable to the coaching change. Their current xG differential is +0.18/game, which would be the best figure that San Jose has managed since ASA began tracking xG in MLS about 10 years ago. The turnaround from the Almeyda era (his last three years were -0.28, -0.39, and -0.41 respectively) is stark. If you isolate to even game states, which can be a more accurate measure of team quality given that tactics tend to change when you’re in front or behind, the Quakes are a remarkable +0.39/game, which is 2nd place in the entire league, and would of course also be far and away the best figure in San Jose since xG was tracked.

In terms of actual goal differential, the team is underperforming expectations somewhat, which paradoxically is a good thing: it means that they have been slightly unlucky, and therefore their results stand on a very solid foundation. Fivethirtyeight‘s power rating currently assesses the squad’s chances of making the playoffs at 76%, pretty good for a team that has qualified just twice in the last decade.

Essentially, there should be no doubt at this point that firing Almeyda and hiring Luchi was a good decision.

The Remade Defense

The club completely overhauled the defensive personnel of the team in the last 12 months, with new players at each of the four defensive positions and major acquisitions in goal and defensive midfield. In addition, one of the key reasons Almeyda was let go was to put in place a coach who had more traditional defensive tactics than the man-marking system.

The good news is that all those changes appear to be working. Their current xG allowed is 1.33/game, which is the best number since those Dominic Kinnear teams in 2015 and 2016 which prized defensive solidity first and foremost. Unlike those systems, however, it hasn’t come at the expense of offense, with the 2023 team registering 1.54 xG/game, the second-highest figure in the last ten years of Quakes soccer. However, I’d note that while 1.33xGA is good for San Jose in historical terms, it is still middle-of-the-pack for MLS.

In terms of individual players, Rodrigues is, believe it or not, 9th amongst all players in MLS in G+. Not center backs, all players. Now this is partially due to quirks in that metric, and some of it has to do with the unorthodox playing style that sends him well up the pitch, but I would say that it accurately suggests he’s having a strong season, one poor error against Monterey Bay notwithstanding. His volume of defensive actions is simply monstrous, he regularly breaks lines as a dribbler, and he is 5th on the team in xG, which is unusual for a defender. Despite his maverick playing style, he (mostly) hasn’t been caught out as regularly as the similar Francisco Calvo, minimizing defensive damage.

I personally have been very impressed with Jonathan Mensah’s performances, and xPass indicates he’s a well-above-average passer, in addition to the noticeable calm he brings to the back. He has a shockingly low volume of defensive actions (Cade Cowell is literally the only outfield player with fewer), but I think that speaks more to why defensive actions aren’t the best way to measure a defender: Mensah prevents actions from being necessary in the first place.

For the fullbacks, Paul Marie has continued his remarkable career progression, recording the second-best G+ on the team and earning one of the starting spots. Miguel Trauco, whom he beat out for it, actually has decent numbers too (particularly as a passer). Carlos Akapo doesn’t stand out quite as much since he’s less of a creator in possession, but he’s been defensively solid.

All in all, the coaching change and massive influx of new personnel have clearly improved the defensive performance, so credit to the Front Office for both. But there’s still some way to go before it’s a championship-level unit.

Goalkeepers and the Daniel Signing

Without a doubt, the most surprising and controversial move the club made in the offseason was the decision to spend a transfer fee, international slot, and sizable wages to acquire a keeper (Daniel) in an attempt to upgrade on homegrown incumbent JT Marcinkowski. The case for the move was that Daniel’s advanced shot-stopping metrics were spectacular in his native Brazil and that the net value of those goals was undervalued in the global transfer market compared to the scorers on the other end of the equation.

Unfortunately, Daniel was injured in the second match and we have essentially no sample size on which to evaluate him. There were some shaky moments, and there was a really good save in the cup match, but with keepers, you really can’t go off of so little.

JT Marcinkowski has a bit longer of a sample, although of course still quite small. His xPass is second on the team with +2.98/100, indicating he’s kept up his known strength playing the ball. But the primary GK advanced metric here, which evaluates number of shots stopped beyond expected, tells a bit of a different story depending on which provider you ask. Of the roughly 30 goalkeepers who have seen significant time, ASA has JT as the 3rd best keeper in MLS, FBRef has him tied for 11th, and StatsBomb has him in 17th. Unfortunately, data godfather Jamon Moore rates the accuracy of those three service’s GK models in the opposite order of JT’s standing. That means it’s probably most fair to consider him middle of the pack.

Last year, Marcinkowski’s G-xG was amongst the worst in the league, so his performances thus far have represented a significant improvement. He will always be limited by his height and reach compared to other keepers, but if he can consistently maintain this level, it somewhat undermines the case for spending big resources on an import, although we won’t know for sure until Daniel gets a meaningful run in the team. Then again, his injury has underscored just how important it is to have a high-quality backup, which the Quakes didn’t last year. That might be justification enough.

Are the highest-paid players justifying their wages?

You don’t need me, or “advanced” metrics, to tell you that Cristian Espinoza has been spectacular this season, but we might as well go there anyway. He’s top of MLS (literally 1st place) in xG+xA, and it’s not even particularly close. He’s tied for the league lead in actual goals scored. He leads the Quakes in key passes by a country mile. Oh, and all this while he’s maintained his stellar defensive work for a winger. He’s a god.

Espinoza’s fellow DP Jamiro Monteiro hasn’t been as good. While Monteiro has retained his strength as a passer (third in xPass) and still covers a huge amount of ground, his defensive actions remain in decline relative to his peak Philadelphia years. What our own Jamon Moore identified preseason was that for the Quakes to have an explosive offense, Monteiro would need to add more final-third productivity rather than just content himself with being a strong presser and possession-retainer. That unfortunately hasn’t happened: he’s bottom of any of the Quakes attacking players in xG+xA per game. He only has 8 key passes (to Espinoza’s 31). His role is different than those of other attackers, so this isn’t a dramatic red flag or anything, but he has definitely yet to contribute in attack enough to make this team dangerous or take the load off of Espinoza. His brutal miss against LAFC didn’t help.

Carlos Gruezo is a difficult one to evaluate, given that he was injured in his 3rd game and has only a few games back fully fit. I also am of the opinion that defensive midfield is the position that advanced metrics can tell us the least about, in isolation, since their roles are so intrinsically interconnected. That being said, Gruezo is towards the bottom of the team in G+ and xPass. His defensive actions are also about equivalent to Jackson Yueill, which indicates he’s not a pure enforcer. I’m not personally concerned, and the overall improvement of the team in general (and the midfield in particular) is dramatic, but I would say that Gruezo has yet to justify the big transfer fee. Then again, Matt Doyle says it’s still one of his favorite acquisitions from the offseason, so perhaps the numbers don’t tell the full story.

Ebobisse (2nd in goals and xG) and Yueill (tops in xPass and progressive passes) are both having strong seasons and are excellent TAM players. No deep discussion is needed there.

The Cade Cowell Conundrum

The eye test hasn’t been great with the Quakes’ young homegrown, and the data tells a somewhat interesting story underneath that. He’s actually 3rd on the team in xG+xA (at 2.99), although is extremely “unlucky” in that he’s only recorded 1 assist and no goals. He leads the team in goal-creating actions, primarily via dribbling, and is 2nd (behind Espinoza) in progressive receptions and carries. The optimistic view is that his luck is bound to turn around and that the creation is there; the pessimistic view is that his underperformance reflects an unusually poor ability to finish or find the final ball. However, I will note that his xPass is dead last in the team, which aligns with the impression that he can be wasteful. And his G+ is at the very bottom of the regulars, too.

All of this is interconnected with my concern that the primary remaining weakness of this team was a lack of attacking depth. If Espinoza’s supernova season cools off at all, and Monteiro stays at the same middling level, then who creates the goals? Ebobisse is a solid finisher but mostly doesn’t self-create in isolation. Cowell has done very little to convert his own work and can be wasteful.

Ousseni Bouda and Benji Kikanović both show up quite well in the advanced metrics, out-doing Monteiro and Cowell on a per-minute basis in several attacking categories, but neither has been trusted with heavy minutes yet, and neither has converted those goals and assists either. Niko Tsakiris hasn’t played yet this year, and even if he did, he’s still just a teenager.

That’s why I still view the attacking depth as the most important place for potential reinforcements in the immediate term (and, my sources indicate, the Front Office does too). It will take a lot of pressure off to have more options, especially as Espinoza inevitably cools off and other attackers miss time through international duty and injury. In the long run, it will depend more on whether Benji and Cade get offered from Europe, and whether Monteiro gets extended beyond this year.