How Big Data is helping sports teams find the winning edge

During the 2015-16 English Premier League soccer season, underdog Leicester City surprised many by capturing the title for the first time ever. But going into the next season, the team struggled. Coaches and managers were challenged to understand why the team was not dominating as it had the year prior.

Soccer is a complex and strategic game, and common statistics rarely tell the full story of how a team succeeds and fails. But working with STATS Ltd., a sports data company, and employing machine learning tools and a plethora of data, Leicester City coaches were able to gain more insight into the team’s performance.

“The assumption was they were great offensively, but the data actually showed otherwise. It was actually their defensive prowess that was crucial to their victory,” said Patrick Lucey, chief data scientist at STATS. “Their organized defensive structure allowed them to reduce the quality of their opponents’ chances across the different scoring methods. The data highlighted their disruptive game that made them one of the most difficult teams to attack against.”

Armed with data, Leicester City coaches and managers honed those strengths and refocused the team. 

Data is not just for sports teams. Many refer to data as the new oil and there is much talk about how to apply Big Data insights to the enterprise. But many organizations struggle to implement Big Data programs. Looking at how sports teams are using Big Data could provide enterprises insight into how Big Data can help them find their own winning formulas.

Lucey believes that because data has already disrupted the status quo in sports, businesses can learn a lot from the sports industry when it comes to determining how to leverage Big Data.

For example, by structuring the millions of data points within video footage, STATS can model how athletes react to others on the field by looking at thousands of prior instances as models — giving them the ability to test new plays through machine learning. 

“In sports, the data is big, but it’s granular, so that allows us to model specific applications in context,” said Lucey. “It allows teams to answer very specific questions.”

For example, in terms of soccer, given a specific shot, what’s the likelihood the team will score?

“The data can tease it out, then we can come to understand the data we have and find the context that leads to those very specific answers,” said Lucey. “It’s a tremendous time saving tool that also helps teams and coaches do their jobs better.”

Another sports data company, Sportradar, is helping National Football League teams better analyze large data sets with an analytics tool it calls radar360. The web-based, statistical analysis application is currently used by all the NFL teams and by NFL media properties to analyze a combination of in-game statistics, historical stats dating back to 1920 and post-game subjective statistics.

“It provides NFL teams the capability to slice and dice data using endless combinations of filters and scenarios to very quickly gain deep analysis and insight on their team or on their opponents,” said Ashok Balakrishnan, senior vice president of product management and technology.