Welcome back to Sports Analytics 101, a series of blog posts outlining the core concepts behind sports analytics in non-technical terms. You can find all available installments in the series here.
Whether by reading sportswriters who frequently cite advanced metrics or by watching a broadcast that flashes key data points across the screen, many fans are first introduced to sports analytics concepts through the media. Sports media’s job is to tell the story of sports, and sports analytics can be a powerful tool in that storytelling process.
For example, let’s say a journalist, Kevin, wants to write about the career arc of a particular player, Megan. In his piece, Kevin illustrates the pivotal moments in Megan’s career, including the day she was drafted, the day she hit the game-winning shot, and the injury that sidelined her for the season.
Kevin also wants to provide data to support certain claims about Megan. For example, Kevin wants to show that Megan had her most productive season after returning from injury. To do this, Kevin looks at a variety of advanced stats related to Megan’s position and plots them over time to illustrate Megan peaking in the season immediately following the injury. Kevin might use this to build the narrative of a “comeback season.” Kevin also wants to compare Megan to the player drafted right before her, and uses advanced stats to show which aspects of the game each player excelled at relative to the other.
Sports journalists don’t have to use analytics to tell stories in this way, but using analytics can, if used well, put additional weight behind the narrative. Kevin could have just claimed that Megan had her best season following the injury by highlighting a few impressive moments, or by simply citing Megan’s average points in the seasons before the injury and season after the injury. This may be just as convincing to some readers, but the added support of good analytical evidence strengthens the case. Of course, analytical evidence can also be used incorrectly or without important context, so just because a journalist uses analytics doesn’t necessarily mean it’s used well (and that doesn’t just apply to journalists: team employees can certainly use analytics incorrectly or without context too).
Beyond citing existing metrics, media members might themselves develop new metrics to use in their work. As it turns out, significant strides have been made in analytics by media members developing their own metrics for public consumption. To be clear, I’m using the term “media member” quite liberally here to include everything from full-time reporters for national outlets to unpaid bloggers with an entirely different day job. There have been significant contributions made to analytics by media members across the spectrum. In fact, many amateur, hobbyist sports analytics bloggers have been hired by professional teams. It’s a tried and true career path to go from blogging about sports analytics on the side to working in the field full-time.
What might lead a media member to develop a new metric? For one, they may simply be interested in contributing to the collective sports analytics knowledge base. They may also be interested in using the new metric to make a point or tell a story.
Let’s say a blogger, Simone, is interested in which soccer players are best at taking corner kicks. Simone could try to watch a lot of video of players taking corner kicks, but that would be incredibly time consuming (remember resource constraints that prevent teams from watching and digesting vast quantities of film?). Simone could alternatively pull some very basic stats on corner kicks, like the number goals that have resulted from a particular player taking corner kicks. This might be marginally helpful, but Simone feels that using those metrics is too simplistic, and that a lot of factors contribute to a goal being scored from a corner kick, some of them very much out of the player’s control. With this in mind, Simone sets out to develop a more comprehensive metric of corner kick-taking skill.
Simone acquires a dataset of corner kicks, including the positioning of all players on the field when the kick is taken, and uses it to develop a metric that quantifies how well a corner kick taker places the ball in high-value areas. With this metric, Simone ranks all corner kick takers in a particular league, supplementing her rankings with qualitative analysis of why a certain player may or may not be good at taking corner kicks, or why any contextual gaps in the metric might be causing a few odd results. Of course, while this metric might contribute to a good piece of content, it could also theoretically be useful to an actual team.
Below is a list of additional questions that members of the media might answer with the aid of analytics. Note, however, that the scope of topics in sports analytics that the media could theoretically cover is vast, and encompasses all of the potential questions from previous posts that teams and leagues might ask of analytics. Fans naturally care about the same things that front offices do: Is it worth signing this free agent pitcher at the price his agent is asking? If we traded our first-round draft pick, what is fair compensation?
If fans are interested in knowing it, media members who cater to those fans are almost certainly interested in covering it.
Therefore, a limited set of additional media analytics questions:
This post concludes the "Why" portion of Sports Analytics 101, which has covered the constraints that make analytics useful and analytics use cases for three key stakeholders: teams, leagues, and the media. There are certainly other stakeholders not covered in this series that can find value in analytics (sports bettors and amateur athletes, for example), but many of these other stakeholders' use cases overlap with the use cases for the three stakeholders we've examined.
The next post will begin the "How" portion of this series. We'll dive into how the different types of metrics work and explore they ways analytics is implemented in practice. Stay tuned.