When Gotham FC general manager Yael Averbuch West was an active player, with a career that spanned nine years, she had no use for analytics.
“I definitely, as a player, never had access to analytics that I felt were helpful,” she said.
“I think it’s something that we’ve struggled with on the women’s side, codifying the use of data and analytics to drive decision making. And that’s for a multitude of reasons,” said Orlando Pride general manager Haley Carter, also a former player.
Advertisement
Carter, too, has experienced a dearth of data in her career until now, alongside teams lacking both the expertise to gain insight from those numbers and then the ability to present it to coaches who usually aren’t data experts themselves.
Now, NWSL general managers like Carter and Averbuch West find themselves in a bold new world where analytics aren’t just becoming more common for NWSL teams, but are increasingly considered a necessity. Not every team lists an analytics position yet, but multiple teams in the league — like Gotham, Pride, Washington Spirit, Bay FC and Angel City — consider analytics to be an important part of gaining a competitive advantage in several areas.
In speaking to various data analysts and general managers, it’s clear that any club with serious aspirations will have to make this a staffed-out, full-time department, whether it’s to gain in-game advantages or to find the best players. Those who can’t or won’t fill such a position will lag behind.
Then and now: analytics in women’s soccer
There seems to have been a rapid evolution happening in the last six years. Companies like StatsBomb, SkillCorner, Analytics FC, and Wyscout now offer decent data platforms across multiple women’s leagues and national teams — although not nearly as comprehensive as men’s soccer yet.
Averbuch West said that when she was a player, at best, she might have gotten some performance data such as the number of sprints in a game. “That’s not helpful,” she said, “because maybe we had the ball the whole time and I didn’t need to sprint, or maybe I’m not as fast as other people so my sprint wasn’t clocking as a sprint.”
That was also the experience for Carter, who was part of the early years of NWSL where she said that most coaches tended to rely on the eye test more than data, for good or for ill. She’s still a believer in trusting her coaches’ experience and ability to evaluate players but backed up now by data used smartly in the right context. She trusts her technical staff to be able to evaluate players with relative precision and consistency, but now wants to use data to support their decision making.
![](https://cdn.theathletic.com/app/uploads/2024/02/26145630/USATSI_21176288-scaled.jpg)
“That’s like the science of football, versus the art of football,” she said. “And I think the answer is probably somewhere in between.
For Bay FC GM Lucy Rushton, who comes from a deep background in data analysis, including five years as head of technical recruitment and analysis at Atlanta United and seven years as a senior performance analyst at Reading, the data that clubs need to start gaining advantages is here already, at least for a select number of leagues.
Advertisement
“We probably don’t have the historical data,” she agreed, “but my argument was that you don’t need that. Very rarely do you go back any more than two seasons really because then so much has gone on in those last two years that how relevant is performance data from two years, three years ago anyway?”
How data is being used
Different teams have different approaches to implementing an analytics team, whether it’s in how much they want to lean on data to help drive decision-making or the areas in which they want to apply it.
At Bay FC, director of data analytics Arielle Dror said they’re working to establish a department with her on pure data analytics, a performance analyst who’s more video-focused, and a joint data/video analyst who can do a little bit of both.
At the Spirit, head of data analytics James Hocken listed three main pillars of use: team and player analysis, opposition analysis and recruitment. The last part, said Hocken, is where he thinks data can provide the biggest advantage.
At the Pride, Carter is also aiming to dig out any gains possible in recruitment, pairing her newly-hired director of scouting and analytics Kat Conner with talent ID scout Aaron Hegenberger. Together, they’ve built out position-specific profiles based on how they want to play. It helps them find good fits for the team and gives them somewhere to start in finding apples-to-apples comparisons, particularly when considering players in different leagues.
Rushton, though, warned of the difficulty in translating results between leagues when scouting potential international players, although this isn’t limited to the women’s game.
“What does a 65% pass success rate in the WSL look like in NWSL? How does it convert? That doesn’t happen in the men’s game. It doesn’t happen in any data at all, because we’re just not complex enough to have those models yet,” she said.
But realistically, analytics are what’s available to evaluate international players. “We can only watch the international games on TV. Going abroad and scouting is expensive. So obviously data helps us massively on those as well,” said Rushton.
![](https://cdn.theathletic.com/app/uploads/2024/02/26144248/USATSI_22621357-scaled.jpg)
“You have to have a very solid marriage between your scouting team and your data analytics department so that they’re on the same page as the players who move through the process,” said Carter. “You only need a couple of people to run the analytics, but you need a lot of talent ID scouts.”
But with the data that’s available on NWSL teams, at least, there’s enough of it to go beyond the basics now. Dror said that she and her team were putting infrastructure into place ahead of the 2024 season to do things like analyze tactical trends and correlate them with success to tailor tactical approaches. They’re taking the raw data and writing code that puts that data into more usable formats, and to some extent automates the process to make analysis go faster down the road.
Advertisement
Hocken is doing something similar at the Spirit, helping build custom metrics that they use to evaluate their players. Neither he nor Dror was specific about the kinds of things they were measuring, citing proprietary information and competitive advantage. But they both said that they have enough granular data now to be more flexible in how they’re able to look at the game, as opposed to being beholden to whatever a company decided to collect.
One specific example that did get cited was from a 2023 game that Gotham narrowly lost.
“Based on the statistics of the chances we created, core basic metrics that usually dictate a team’s success, based on the statistics, we would win 98% of the time,” said Averbuch West. “What that tells us is that it’s not a time to panic and change everything you’re doing for the next game. That’s the time to say, “OK, well, if we keep doing that, we’re going to win 98% of the time, so we should actually play exactly the same way every single game.’ And those types of reassurances on a team level are really helpful.
“For players, it’s about showing patterns when you do these things, you will probably be successful, so do more of those things.”
![](https://cdn.theathletic.com/app/uploads/2024/01/25092257/GettyImages-1792135031-scaled.jpg)
This wasn’t just using analytics to augment a coaching decision; it was an important educational piece too, creating something actionable for the players and the coaches.
“It’s just a deeper level of information, basically, like things that your eyes may miss,” she said. “Our coaches are really good, intelligent, a lot of times they can pull out probably what went wrong. But there are other things that data can show. For example, on all of our set pieces, the ball landed in these zones and the highest percentage of scoring is when the ball lands in these zones. … There are things that nobody notices or trends that you can’t observe until they’re quantified for you later.”
“I don’t think you can Moneyball the NWSL”
NWSL teams are still finding the balance between using analytics to enhance, but not necessarily to fully drive decisions.
“Player decisions are very nuanced,” said Carter. Just because the numbers have identified a good player for their system doesn’t mean there aren’t other make-or-break factors.
Advertisement
“A big thing we look at with scouting is also character, professionalism, work ethic, all the things that data doesn’t get into,” said Averbuch West. “But you need to have the references and have the conversations and the knowledge of the context in which a player is performing.”
Dror said scouting internationals is hard enough with data not always translating easily between leagues, but scouting young players can be next to impossible given the lack of historical data. Trades of established players might not need to go back too far, to Rushton’s earlier point about not needing years and years of data, but digging up talented young players is still a giant challenge because there isn’t a large amount of standardized high-quality data from youth pools. When looking at potential young players, there’s also a risk assessment to be made — will this teen phenom mature into a star player, or will they hit their ceiling early?
“If you’re looking at a player who’s 18 and you want to project who they’re going to be when they’re 24, I mean, it’s hard even with the amount of data we have on the men’s side, but it’s certainly not possible with the data that we have on the women’s side,” Dror said. “I think the best we can help is evaluating international signings and evaluating domestic players. College is hard, but not impossible.”
![](https://cdn.theathletic.com/app/uploads/2024/02/26144825/USATSI_22617432-scaled.jpg)
That’s where hands-on evaluation is still a much stronger tool.
“The thing about young domestic players is that you can actually bring them into your environment,” said Averbuch West. “There’s no better way to understand how a player fits in until you see them in their environment. Nobody can do a better job than that. So I think that’s a really important component if that’s possible.”
Case in point: in mid-February, 13-year-old McKenna Whitham joined preseason training with Gotham FC and the Spirit and KC Current.
Another set of players that current data misses is players outside of the biggest professional leagues — those who aren’t in the U.S., England, France, or even Sweden or Germany. Rushton said development is definitely lagging in several women’s leagues covered, which makes it even harder to find a hidden gem, the type of player in a smaller or semi-pro league who could make the step up.
Advertisement
“The perfect examples, the two Fijians, Cema (Nasau) and Sofi (Diyolowai) that we brought in on trial during the World Cup last year,” said Carter. “Those are athletes who have never had access to a professional performance environment. The women’s league in Fiji, nobody is tracking statistics. Statistics aren’t available on Opta or Statsbomb. Film is not available on Wyscout. But those athletes are out there, and I’d say that’s true in Oceania, that’s true in South America, it’s true in Africa.”
Dror agreed that the analytics were only as good as the data and that the data could carry its own biases. “Data isn’t neutral because of how it’s collected,” she said, a warning oft-emphasized in Intro to Statistics classes. The way a question is asked, who is asking it, and who gets asked can all influence a data set. Both Dror and Carter emphasized the need to have a varied analytics team to check each other’s work and to continually think outside of the numbers.
“I don’t think you can Moneyball the NWSL in the way that you can baseball,” said Carter. “It’s important that you codify it and you have a process that’s repeatable and scalable, but you also have to be cognizant of biases that are inherently built in by codifying and formalizing a process. You have to be cognizant of athletes that you might miss or attributes that you might miss because you are standardizing an approach to players.”
The immediate future
Everyone The Athletic spoke to was optimistic about the development and integration of analytics into every club over the next three to five years.
While Gotham is still working to hire a full-time analyst, Averbuch West said, “We see it as an absolutely crucial part of all of our decisions going forward in terms of how we prioritize our time, energy and everything we do.”
Carter said she sees every team eventually staffing a full-time analyst.
“For me having this position is about professionalizing the women’s game,” she said. “The future is continued investment in it because as we move forward, as the league evolves, we start getting into the business of women’s football and that is buying talented athletes before they become the next Sam Kerr and then moving them along for a fee.”
Advertisement
That could mean bringing in more players like Jun Endo and Kerolin, identifying and signing hugely talented players at the beginning of their careers instead of more towards the end.
Carter said she’d eventually like to see analytics complemented by a bevy of talent ID scouts with “every continent, and every regional confederation, working on a contractual basis to help you identify talent because that is how you avoid only relying on the established talent.”
Hocken, who spent two years as a performance analyst for Florida State University, is acutely aware of progress in coverage, going from collecting all his own college data to being able to buy it from a company like Wyscout.
Data says: she’s really, really good https://t.co/PzJTkeqBUa
— Arielle Dror (@arielle_dror) February 13, 2024
“Statsbomb for example, have doubled their coverage in one year of leagues,” he said. “It’s still not the same as the men’s game, but that’s a huge improvement in such quick time. And it’s not like they’re going to stop there and say all right, we’re done. They’re going to continue and grow.”
“I feel like we’re just about to reach that point where maybe a majority of the clubs have someone full-time on staff,” said Dror.
For her, the next step is building out the staff, and bringing in several people with complementary skill sets to be able to plan both short and long-term analysis.
“I would love this to be the most technologically advanced league,” said Dror, “where we’re just recruiting all these really talented people that wouldn’t ordinarily choose to work in sports but find this to be an interesting opportunity.”
(Top photo:Getty Images; Design: Eammon Dalton)