Analysis

NWSL quarter-season report card: Grading with advanced statistics

ISI Photos-Daniel Bartel

Who’s underachieving? Who’s overachieving? And who is just plain achieving? Chad Murphy breaks down each NWSL team with advanced statistics.

We are part of The Trust Project What is it?

With the NWSL season already a quarter over (yes, seriously), it’s a good time to take stock of what has been going on and how teams are doing. Rather than doing the eyeball test, I look at the underlying statistics behind each team’s performances to talk about how teams have done and make some predictions on what we’ll see for the rest of the season.

For most of my analysis, I use a measure known in the soccer analytics community as “expected goals” or “xG.” The idea is that goal-scoring is subject to a fair amount of variance, or luck, and in a small sample like we have so far variance can make a big difference in the table. Using expected goals is one way of looking at the things underlying a team’s performance and gives us a better picture of who is overachieving, who is underachieving, and what we can expect in the remaining fixtures.

Expected goals calculations involve some fairly complicated math, but the concept behind them is surprisingly simple. The basic idea is that each time a player takes a shot it has a certain probability of scoring. This probability can be calculated by knowing some simple facts like where the shot was taken, how far it was from goal, the angle between the ball and the center of the goal, whether it was kicked or headed, whether it was assisted from a cross, the speed of the attack, and where the nearest defender was when the shot was taken.

When I plug these numbers into my formula, I get an expected goal (xG) value between 0 and 1. This number represents the likelihood of a shot scoring, with 0 meaning that players will never score from that position, 0.5 meaning players have a 50 percent chance of scoring, and 1 meaning there is a 100 percent chance of scoring. I have collected data on every shot taken in 2016, calculated this number, and added it up for each player and then for each team. I use these numbers as the basis of my analysis of the NWSL season so far.

NWSL’s 10 teams are broken down into three groups: overachievers, underachievers, and regular achievers below based on the underlying statistic of expected goals created by each team and then expected goals allowed for each team, all of which have played five games.

Overachievers

Let’s start with the overachievers. To be an overachiever, a team has to be performing well above its underlying statistics. This means they’re either scoring significantly more goals than expected, or the same underlying statistics are higher than one would have predicted at the beginning of the season. The Spirit and Red Stars are both buoyed by their defense, conceding significantly fewer goals than one would expect, while I include the Orlando Pride here because Orlando is dramatically exceeding preseason expectations for an expansion team.

Washington Spirit

The bad news: They Spirit have conceded one goal despite opponents’ xG values of 3.5. This overperformance ultimately means we might expect them to start conceding more in the future, but for now their defense and opponents’ ability to find the back of the net has put them as the surprise leaders through the first quarter of the season.

The good news: they’re a close 4th in xG scored, and are 1st in xG allowed. This looks like a recipe for continued success and if they can maintain their form they look to be a shoe-in for a playoff spot. They’ve been a bit lucky to only concede one goal, and I would expect them to start giving up a few more goals soon, but even with that scoring at such a high rate while both conceding few chances and actual goals puts them in a good position for the remainder of the season. Crystal Dunn has been a particular bright spot, leading the team with 1.34 xG, putting her in the top 10 league wide despite not actually finding net.

Chicago Red Stars

The bad news: The Red Stars are in the middle of the pack both in terms of xG created and xG allowed. That may not be enough to sustain them in 2nd place, especially given how well the Orlando Pride are performing in their inaugural season.

The good news: Captain Christen Press is leading the league in xG creation, and has been consistently among the top players every week so far. She has a combined xG score a little over 2, and has scored a league-leading three goals. She’s living up to her prominent role in the team, and if she can maintain this pace, the Red Stars will continue to hang around the top of the table.

Goalkeeper Alyssa Naeher has also been a tremendous signing for Chicago (sorry Breakers fans), and looks to be a big reason why the Red Stars have conceded fewer goals (three) than the underlying statistics would predict. Her presence in goal has been a huge boost to the Red Stars, and looks to be one of the best acquisitions of the year. If these two players keep performing at a high level, the Red Stars look to be ready to mount a serious challenge this year. Both, however, will miss time to U.S. national team commitments.

Orlando Pride

The Orlando Pride’s success has possibly been the surprise of this season. Many people didn’t expect much from the Pride as an expansion franchise, but Orlando is currently fourth in the table and fans could easily even be disappointed with the club’s position there. The Pride lead the league in xG created and sit 2nd in the league in xG allowed. Alex Morgan is unsurprisingly doing well for the newcomers, but Lianne Sanderson has performed admirably as well. The English forward is currently 2nd in the league in xG creation, only behind Christen Press. As the only team with two players in the top 5 of expected goals, Orlando is in a strong position to maintain their spot or even improve for the rest of the season.

Continue reading: Which teams are performing just as expected?