Around The League // Week 10
Just four games this weekend, but no shortage of things to break down to start February. Welcome back to Around The League, your home for numbers and nuance from the past week of action in the Major Arena Soccer League. We’ll take advantage of the light week to discuss a new way for us to break down the game, but we’ll start in Baltimore, where the Blast earned a comfortable 8-2 win over Empire.
Baltimore Gets Back to Winning Ways
One could argue that no team needed a win this weekend as badly as Baltimore. Last season’s Ron Newman Cup Semifinalists sat 7th in the table going into their game against Empire and knew a win would see them jump into the 6th and final playoff spot with half of the season left to play.
The Blast endured a poor run of results in January, dropping four of their five games since ringing in the new year. With a host of injuries affecting their lineups during that time, this weekend was an opportunity to see those faces again and get back in the win column.
Head coach David Bascome’s side enjoys playing in transition, working the ball up the field with pace, finding and exploiting their opponents’ defensive lapses. Key to this approach is a goalkeeper capable of turning defense into attack at the blink of an eye. Julian Rodriguez demonstrates exceptional aptitude in this area, and against Empire, he was at his very best.
The 27-year-old recorded two assists, pushing his current points streak to four straight games, and he currently leads all goalkeepers in points with six assists. If that wasn’t enough, he’s been one of the league’s best between the pipes, conceding 4.94 goals fewer than expected.
Another product of coach Bascome’s coaching philosophy is his team’s attacking defenders. Chad Poarch and Oumar Sylla led the way with two goals and an assist each as the Blast’s defenders continued their dominance this campaign. When looking at Shots taken per game among all defenders, Poarch, Sylla, and Patrick Thompson make up three of the top four spots.
If we look at their expected goal statistics, the Blast are among the only teams to excel on both sides of the ball, with Kansas City being their lone companion. They’ve scored 5.187 more goals than expected, and they’ve conceded 5.482 fewer goals than expected, indicating an above-average ability to both create dangerous goal-scoring chances and prevent them.
After a taxing front-half of the season, the Blast will welcome some rest with no games scheduled this weekend, but they’ll quickly turn their focus to their next game against the Ambush at home on Saturday, Feb. 14, at 4:00 PM EST. Last season, St. Louis was the only team to beat the Blast in Baltimore, so as with most games this season, nothing is guaranteed.
The Importance of Kelvin
Given that this weekend was relatively light, I’d like to take this opportunity to advance our collective understanding of the game. We’re going to explore a statistic I use to contextualize the game called Kelvin, named after Utica’s star forward, Kelvin Oliveira, due to his relative over-performance in front of goal last season.
The stat takes a team’s Goal Differential (GD) and subtracts it from their Expected Goal Differential (xGD). The resulting difference, positive or negative, helps us identify which teams are over- or underperforming in front of goal. In a single-game sample, Kelvin can be misleading, but now that seven of the eight teams have played at least 12 games, the figure becomes a brighter light in the dark as we attempt to isolate skill from the stats.
The key to this data is tracking and establishing a standardized xGD to measure against GD. For this, I take our Hoxie statistic, which is shot differential for the entire season, and multiply both Shots For and Shots Against by the league’s average shot conversion rate (0.223).
This gives us an idea of how many goals a team would be expected to concede if the average player were taking the average shot. Teams with a positive Kelvin are overperforming, while teams with a negative Kelvin are underperforming.
Given that Kelvin includes measurements for both offensive and defensive play, we have additional ways to contextualize a team’s performance. Naturally, it’s possible for two teams to have the same Kelvin stat, but arrive at that figure in different ways. Enter, GMxG and xGAMGA; two seemingly random amalgamations of letters that represent offensive and defensive performance.
GMxG stands for “Goals Minus Expected Goals” and shows us relative performance in attack. A team with a high GMxG stat scores more goals than expected, and vice versa. Conversely, xGAMGA stands for “Expected Goals Against Minus Goals Against” and indicates relative defensive performance. A team with a high xGAMGA concedes fewer goals than expected, which initially sounds counterintuitive, but the goal is for positive stats to be good and negative stats to be bad. As such, a team with a low xGAMGA would concede goals at a higher rate compared to the league average.
So which teams fall into these categories? We’ll start on offense with the Milwaukee Wave, who have scored an extra 11.072 goals in their first 13 games. This overperformance indicates an offense that generates more dangerous attacking opportunities than the rest of the MASL and, in turn, conversion rates significantly higher than average.
On the other side of the ball, the Kansas City Comets are notorious for breaking this basic xG model. They give up many shots, which inflates the expected goals figure, but the shots they concede are from long range, tight angles, or with several defenders behind the ball. This explains why the Comets traditionally lead the league in blocked shots. They protect the dangerous areas of the turf and don’t give up a swath of high-value shots. This season, they’ve conceded 14.012 fewer goals than expected.
You may be asking yourself, “Why are we not attributing a team’s shooting overperformance to their finishing ability”? This is a fair question. The answer lies in years of data tracking from outdoor soccer, which shows that shot conversion rates fluctuate significantly and are highly unpredictable.
This is also true for the MASL. Some of the league’s very best attackers go through periods of drastic underperformance, while players not normally known for their finishing may experience the exact opposite. Generally speaking, teams and players do not develop a supernatural ability to finish. Regardless, even outdoor soccer’s xG model quantifies a team’s ability to create chances by estimating how many goals the average player should be expected to score. Here, we arrive at a similar answer by using a team’s relative performance to quantify the difference between expected and actual outcomes.
As a result, chance creation becomes the inference. Teams that overperform their xG are creating more dangerous goal-scoring opportunities. They use the boards to find unmarked teammates, make runs to the back post, take shots in and around the penalty area, and capitalize on transitional moments with fewer defenders between them and the goal. This is where we need to watch the games to develop a full understanding of what we’re seeing.
Remember, data is not a replacement for our prior understanding of the game; it’s a tool to help us determine what to look for and where to find it. Coaches who embrace data are better for it, as are fans who use it to gain an enhanced understanding of what they’re seeing.