Leveraging Statistics to Improve Performance
A Minnesota Lynx data scientist shares key factors to making it work
Thanks for reading the Her Hoop Stats Newsletter. If you like our work, be sure to check out our stats site, our podcast, and our social media accounts on Twitter, YouTube, Facebook, and Instagram. You can also buy Her Hoop Stats gear, such as laptop stickers, mugs, and shirts!
Haven’t subscribed to the Her Hoop Stats Newsletter yet?
Starting at a young age, Madison Schiller had two main passions: math and basketball. Thankfully after being recruited to play basketball for Cal State East Bay, majoring in computer science, and getting her master’s in business analytics, she found a profession that combined her two loves. As the recently-hired data scientist for the Minnesota Lynx, Schiller now spends her days surrounded by numbers and being an integral part of a sport she adores. On a regular basis, she mines basketball data, develops predictive models, creates visualizations, and interprets analyses to help her team find success on the court. Schiller shares a few ways other teams and players can ensure they’re maximizing the value of their data and discusses how the league can further support these efforts going forward.
A supportive organization
Luckily for Schiller, the Lynx have prioritized analytics for a long time—having an in-house statistician, Paul Swanson, for many years—and recognized the value of not only adding additional staff, but also expanding their expertise and skill sets in order to make the most of the data and analytical tools available to them.
One of Schiller’s initial tasks was to develop a more optimal data structure so the Lynx could leverage their college and WNBA data to identify and evaluate their free-agent and draft targets. Schiller also created new metrics and developed additional player evaluation models, which helped increase the team’s chances of selecting the best players available for the needed positions.
“Predictive analytical models are never going to be 100 percent accurate,” Schiller said. “Various models have different strengths and weaknesses and can vary in how they weight certain elements. But if a player scores well on multiple models, then it’s a good sign.”
Schiller also spent time automating some of the team’s scouting reports, ensuring the data are presented in the most actionable way and anticipating other analytic needs the Lynx may not have had at the forefront. The Lynx have been very appreciative of Schiller’s efforts.
“Cheryl [Reeve] and Clare [Duwelius] love analytics,” Schiller said. “I focus on giving them as much information as I can, so they can make the best decisions they can. They’re always wanting to progress and learn more about data and analytics. They know that’s a competitive advantage we want to have.”
Receptive players
Supportive leadership is important to capitalizing on data. Having players who are willing, or even eager, to use data to improve their performances is equally critical. Schiller believes the way the Lynx coaching staff has presented the data to the players has been key.
“Cheryl [Reeve] is really, really good about expressing what statistics—the ones we actually care about—to the players that she thinks is necessary at the time,” Schiller said. “We don't ever want to overwhelm any player or anything like that. She's pretty good at it, so they're really receptive.”
But it isn’t just showing the players their data, it’s getting them to use the information to make changes to their game. Although that can often be easier said than done, Schiller thinks the Lynx players’ ability to utilize the feedback will increasingly help them find more success.
“Habits are hard to change quickly,” Schiller said. “But the players are really receptive, and they're willing to work on it. And I think we're valuing the right things to win games. So I think, ultimately, in the long run, we will hit those metrics and improve.”
When asked about specific statistics players should focus on, Schiller recommended starting with the “Four Factors,” which are four possession-related statistics and include effective field-goal percentage, free-throw rate, turnover percentage, and offensive rebounding percentage. She also noted that players can do this relatively easily—and on their own—using publicly available basketball statistics sites. While Schiller was a college player, she routinely relied on CBB Analytics as well as Her Hoop Stats, both of which allow college players to quickly rank or compare themselves against their peers on a variety of metrics, including The Four Factors, across Divisions I-III.
A forward-thinking league
Although Schiller lights up when discussing the data and tools currently available to her, she also cautions that a pretty large data gap still exists when comparing the data and tools available to the WNBA versus what’s available to the NBA. One of the most notable differences is access to a player tracking tool called Second Spectrum, which collects multi-frame per-second metrics using cameras. Without this tool, WNBA teams have significantly less data available to them, must manually collect data for several key metrics, are much more limited in analytical and reporting capabilities, and as a result, lack the same level of intel for decision-making.
“I know we don't have as much data resources or data as the NBA, but we're striving for that,” Schiller said. “We have people on our staff collecting statistics because we don't have the player tracking statistics from the cameras. Second Spectrum collects everything. It tracks the ball, where the defenders are, and how close they are to the shooters. I know there’s so much you can do with it. Hopefully one day we'll be able to get that, and it'll expand so much more of what we can do with data.”
But Schiller is hopeful the WNBA, in particular the Lynx, will continue to increase the data, tools, and other analytical resources available to them.
“I definitely think we're trying to get where the NBA is,” Schiller said. “I know we have limited resources. But I think when we're given more resources, we'll utilize them. I know the Lynx always want to be on top of it. They’re always big on analytics. As soon as they had the resources, they really wanted to take advantage of it. And I'm super grateful. I love being in this role.”
Thanks for reading the Her Hoop Stats Newsletter. If you like our work, be sure to check out our stats site, our podcast, and our social media accounts on Twitter, YouTube, Facebook, and Instagram. You can also buy Her Hoop Stats gear, such as laptop stickers, mugs, and shirts!
Haven’t subscribed to the Her Hoop Stats Newsletter yet?
I would think +/- stats are important - as it relates to both individual and team stats while that player is in the game.