Foucault pendulum in the Oregon Convention Center. Photo by Scott Dexter, CC BY-SA 2.0 https://creativecommons.org/licenses/by-sa/2.0, via Wikimedia Commons
Introduction
In these two 2024 preseason articles, I will be discussing how the risk profile the Earthquakes play with this season will determine their chances of moving up to a home-advantaged playoff spot or even back to last in the West.
Riskless Business
When it came out, I touted Tiotal Football’s post called Football Tactics and Rolling Dice on his Absolute Unit blog as the introduction to my 2024 preseason series. If you haven’t read it, I recommend you stop reading this right now and go read that. I don’t have the level of relationship with Tiotal that former ASA colleague-turned-Athletic-writer John Muller does, but suffice it to say that Tiotal and I have seen soccer similarly over the past few years. Tiotal recently followed-up that first post with another called On Randomness and Parity.
Without just brazenly repeating here what Tiotal’s articles are talking about, let me attempt to summarize: in soccer, as in any team sport, the amount of risk you decide to take on when you try to score or when you defend, along with the level of talent that you have/superior tactics you are able to play, determines how many chances you will get but can also determine how much you leave yourself open to your opponent scoring, too.
More often than not, the team with more chances to score, given equal talent, will win. Better talent and/or tactics give your team an advantage given an equal number chances, because, presumably, those advantages will give you better chances or reduce the quality of your opponent’s chances. But in soccer a team can often look better and be better, but they will take a draw or a loss on the day.
The question is: How much risk and what risks should you take to give yourself the best opportunity for winning?
At the beginning of the 2023 season, two things were equally true:
- The Quakes attack took more risks
- The new defense and goalkeeper hadn’t yet gelled
This led to scores at the beginning of the season like: 2-1, 3-1, 3-0, and 2-2. The Quakes only played in two draws until May 27th, a home 0-0 against wooden spoon-winning Toronto and an infamous 1-1 draw against the Red Bulls in the longest stoppage-time regulation game in MLS history.
The Quakes were figuring out how to play within Luchi’s philosophy of rest defense. (Something I was begging for under Almeyda and Covelo.)
Then two things became equally true:
- The Quakes attack took fewer risks
- The new defense and goalkeeper gelled pretty well
After that point, the draws started coming. The Quakes drew 5 of their next 9 and ended the season with 6 draws in their last 8, for a total of 14 draws — tied with NYCFC for the most in Major League Soccer in 2023. Their scale was evenly balanced with 10 wins and losses on either side of the draws.
Thanos would be proud.
Results Don’t Equal Risk
The problem wasn’t the rest defense structure. The problem was how often the Quakes took risks in getting forward, despite Luchi preaching breaking lines and playing with “verticality”. Either it wasn’t often enough or they didn’t have the right players to do it effectively.
The Shots per Attacking Third Entry metric below shows the Quakes were toward the bottom of the league at getting shots from their trips into the final third. Their biggest attacking issue wasn’t getting into the final third, it was doing it in a way where they could unsettle the defense and goalkeeper to get off a shot. Plodding your way slowly into the final third, even if you played between the lines along the way to get there, is not taking any risks. Playing slower also allows the defense and keeper to remain set, relying on some super-human golazo or colossal defensive mistake to get a goal.
It’s important to keep in mind, however, that more risk doesn’t necessarily equate to better results. Almeyda’s Quakes teams well-illustrate this point: they took on a lot of risk, but with inferior tactics and, most would say, inferior talent, the risk profile was much too big. As Trevor Wojcik and I demonstrated last preseason, Almeyda’s most likely outcome, using his teams’ 2021 and 2022 data in a 10,000 season simulation, was a 35-point season — last in the West. His most realistic upside didn’t even approach the playoff line, set at 47 in 2022.
The biggest question Luchi Gonzalez had coming into the 2023 season was how much to swing the risk pendulum while improving the defensive talent. As Trevor showed in his 2024 Season Preview article for American Soccer Analysis, the roster and tactical changes improved the Quakes by five expected points, from 35 to 40. Fortunately, in reality they overperformed that by four more points to get to the Western Conference play-in game. (Thank you, Daniel.)
However, in the process, the attack took a significant step back going from 52 goals (455 shots at 0.097 xG per shot for 44.16 xG) to 38 goals (412 shots at 0.097 xG per shot for 39.9 xG), including penalties. The league average in 2023 was 422 shots for 45.3 goals and 44.2 xG or 0.105 xG per shot, putting the Quakes firmly below average in shot quality and very slightly below on shot quantity.
This shows up when we look at a Where Goals Come From boxplot (above). It shows San Jose attempted fewer shots from progressive passes (left-most plot) than the league average (black line through the green box). These shots from progressive passes come via through balls, cutbacks, half-space passes into the box, and dangerous crosses. They are where the highest quality shots come from (outside of penalties), resulting in ~40% of all goals in MLS. 42% of the Quakes’ goals in 2023 came on these shots — they just needed more of them.
The typical MLS team since TAM-era 2017 (not counting the weird 2020) shoots 12 times a game (408 shots) for 0.11 xG per shot. Just being average with your shots and xG per shot gives you an upside of around +20 goal differential and 65 points and a downside of around -20 goal differential and 30 points. This wide range of outcomes is going to be important in my next article.
The Quakes defense still surrendered a monstrous 471 shots for 48.7 xG (0.103 xG per shot) and 43 goals against (yes, that 5.7 goals against fewer than expected goals against is pretty much all Daniel). In 2022, they gave up 481 shots, but for 58.12 xG (0.12 xG per shot) and 69 goals. So the defense improved almost exclusively by giving up less dangerous shots than reducing total shots against. Plus, for much of the season, Daniel was saving shots normal keepers don’t.
No Playoff Home Cooking
There are three key issues that prevented the Quakes from playing at least one playoff game at home last season after looking like they could get a home-advantaged spot in the first half of the season:
1. The attack had less-and-less bite by the eye test and the key metrics as the season went on.
Just check out the downward trend line below for proof. The same trend also shows up looking at Goals Added (g+) per game.
2. Way too many shots were given up on the road giving the Quakes very thin margins to get any away wins.
As I mentioned already, the defense greatly improved on the quality of shots conceded, but San Jose was outshot almost 2-1 on the road, creating a massive disadvantage for them.
3. Not creating and taking advantage of mistakes near the goal
Talk to Luchi Gonzalez for 15 minutes any day of the week, and you’ll understand how much he likes to press. The main purpose of a press should be to create high turnovers or at least force the defenders with the ball to boot it long over the press to get out of trouble. Luchi also talks about transition goals being the most common and important goal in soccer. All good: I agree with him.
The problem is the Quakes were not good at forcing mistakes that turned into goals. In fact, they were second-to-last in the league by a key measurement: “Individual Play” goals.
Individual Play goals and expected goals are a measurement of how opportunistic a team is at taking advantage of defensive mistakes both in the attacking and defending phases of play. Examples of Individual Play shots are: stealing a ball and shooting without needing a pass, rebounds, shooting a loose or deflected ball in the box, and other situations that don’t require a pass before a shot. If it’s your team scoring, you’ll probably call them “hustle goals”, and if you are getting scored on, they are “garbage goals”. Regardless, the Quakes scored a total of four of these goals last season. At the top? St. Louis and New England with 17 of these. Yes, I realize teams just passed the ball to St. Louis around their own box sometimes, but getting 17 of these goals is a good indication of just how badly a team wants to get to the ball in-and-around the box.
For pressing, one of the common metrics is Passes Per Defensive Action (PPDA). Usually, this measurement is taken in the attacking 60% to capture mid-block actions. However, to measure activity near the penalty box, I’ve adjusted this measure to the attacking 40% of the pitch and compared it to the Expected Goals from Individual Play shots. What we are trying to measure is pressure higher up the pitch and potential resulting shots from winning the ball high. While this needs more research, the starting correlation is pretty good as seen in the upper-right corner blow. The Quakes position in this graph is not, as you can see by the teams around the Quakes in 2023 (most missed the playoffs or finished in a low playoff spot).
Conclusion (for now)
The Earthquakes were ineffective risk takers in 2023. They took fewer good risks as the season went on until it was practically too late. As a result, they barely survived an unlucky result to end the season, thanks to Portland’s game 34 loss.
To improve on 2023, they will need to take more risks and get better outcomes from the risks they take. In effect, they’ll need to take more risks to score more goals without being affected by the potential downside of a defensive breakdown and being counter-attacked.
Fortune favors the bold. The bold also courts disaster. In Part 2 of this short series, we will examine what risks are worth taking in MLS and how the Quakes are targeting to be more effective at risk-taking in 2024.
