(Together, There’s a Method to the) March Madness
How Sports Data Viz Leverages the Power of Community
One of the primary reasons I’m as active in sports analytics as I am today is the constant learning fostered by an interactive community on Twitter. It started with a simple tweet three years ago where I tagged a couple people asking where I could find a certain data set. I got responses almost immediately. The rest, as they say (they don’t), is history (it isn’t).
So naturally, when the Data Visualization Society (DVS) spun up, I was excited to see another community dedicated to a niche within sports analytics — specifically sports visualization. Any guesses as to what the first ever post in that channel was?
It was, hand to my heart, a link to Sports-Reference websites. Because, as Jay-Z would remind us, “men lie, women lie, numbers don’t.” It felt slightly poetic, as most great sports analysis and visualization journeys often start with an intriguing question or an inviting slice of data, not entirely unlike how I had personally gotten started. For those that don’t know about Sports-Reference, you ought to read this delightful NYT feature about the website by James Wagner, but to paraphrase his own words: it’s a platform for an incredibly wide range of data across popular sports (hockey, baseball, basketball, etc.) that effectively serves as the most popular digital almanac of sports data.
Cross-pollination across sports is one of the most effective ways to cultivate new concepts and enhance analytics research within individual sports
In the DVS slack, the website has been cited at least 3 times already, just by my rough count, and all for various sports too! That versatility has been fun to see, as the discussions and data viz presented within the channel have spanned basketball, soccer, hockey, baseball, tennis, and even e-sports. Cross-pollination across sports is one of the most effective ways to cultivate new concepts and enhance analytics research within individual sports. As the sports-viz channel continues to grow, the foundation of sporting diversity that it’s built off of ought to serve us well.
For example, here’s a PassSonar Map that was shared in the channel by Ben8t, a data scientist who works with soccer (football?) analytics. It makes perfect use of a radial form factor to show directionality and distance of passes from various players within an 11-man formation. In this way, it reminds us of the classic wind rose.
Now, let’s step back for a second. Soccer isn’t the only sport where a player passes a ball in fluid gameplay. For starters, basketball players do it too! Just a month ago, the inimitable Todd Whitehead (@CrumpledJumper on Twitter), who’s long been a stalwart of data visualization in basketball, published a post with a very similar concept for charting player passing lanes in basketball. He also used a radial form factor to study players’ passing tendencies.
In the non-radial passing visualizations category, how about when within hours of each other, both Nic Wispinski and Nicholas of Canova Analytics shared their analytics dashboards for hockey and basketball, respectively? Nicholas’ website (bigleaguegraphs.com) features various tools for visualizing shot charts, how players are passing to each other, and even visualizations for player and team comparisons. Wispinski’s dashboard also leveraged the graceful efficiency of mapping onto a field-of-play diagram, plotting shot attempts, goals, and turnovers from the Rangers-Devils game earlier in January. However, mapping isn’t just for sports played on a physical field. Sven Charleer shared his team’s impressive work on enhancing the spectator experience by augmenting League of Legends streams with real time dashboards, including features such as how players had moved their characters across the map.
Cross-pollination isn’t just about sharing your own work, though. It’s also about an entire community coming together to share resources. Neil Richards identified Sports Viz Sunday, which hosts monthly data challenges for sports-themed data visualizations. Cameron Yick resurfaced a perhaps unforeseen marriage between Bloomberg and Major League Baseball from several years ago (or maybe let’s set that relationship status as “been dating for a while”) where Bloomberg did comprehensive analysis of pitch tracking data by partnering with Oculus. Heck, we even got treated to an entire thread of resources for tennis data, from Tennis-Data.co.uk to – courtesy of John Burn-Murdoch – Jeff Sackman’s GitHub repo and Ultimate Tennis Statistics. Sharing data and code is a major component of advancement in the public domain for analytics and visualization, so in this vein, Ryo Nakagawara also ought to be commended for opening up the floodgates and letting us in on where he keeps all his code for creating soccer visualizations.
Learning from other sports is a great way to enhance our own understanding and internalize new modes of analysis. It’s been encouraging to see the foundations for that kind of discussion organically cropping up in our channel from the jump, just even through the sheer presence of folks sharing similar work as they spot visualizations of interest. With all that said, to neatly wrap this up and tie it with a bow as you will, March is after all for the hoops. So I’ll leave y’all with Kyle Yetter’s March Madness Bracket Assistant. In the words of hip-hop luminary Future, the sports-viz channel is “balling like it’s March Madness.”
