Big data at the Super Bowl: Unlocking sports insights and football predictions

Post Comment
Social Media Manager, Internet of Things, IBM

Data and the information that comes from it has value across several industries. For automotive, electronic and industrial manufacturers, the data is put to use in manufacturing and operations. But what about niche areas such as sports? Plenty of people attend sporting events glued to their smartphones and churning data. From them, we can identify valuable sports insights, which can be useful for both sports franchises and the fans themselves.

Approximately 111.9 million people—not counting those who were watching the game through a streaming service—tuned in to watch Super Bowl 50. More than 70,000 attended in person, and of them, 50,000 likely had smartphones.

Beyond the tweets: Seeing trends in sentiment

Consider a few ways how all that data can be mined and used for action. Back in 2012, the University of Southern California Annenberg Innovation Lab (AIL) conducted some real-time analysis of Super Bowl quarterbacks Tom Brady and Eli Manning. They looked at thousands of public tweets and analyzed them for tone: sarcasm, sincerity, snarkyness, positivity and so on. And they discovered a few things. Public sentiment, for example, is notoriously fickle. Consumer insights shift rapidly, and they also discovered that, overnight, Manning beat Brady when it came to positive sentiment: 66 percent to 61 percent.

An element of predicting performance based on player stats was also discovered. By plugging the data into a predictive analytics tool, you can get a good idea of which player is going to perform best or which team is going to win. This information is, of course, of interest particularly to fantasy football players or gamblers. This technology may seem a little too much like science fiction to imagine a world in which it is possible, but envision the world of Rodney Ruxin. What would he do if he had these tools at his disposal? I can pretty much guarantee he’d mop the floor with Pete, Jenny, Taco and Andre.

Fan-based insights and actions

Four years later, we’ve come a long way. Now, we don’t just collect, analyze and interpret the data, we can use it to take real actions. Recently, IBM paired up with the Ottawa Senators to incorporate behavior-based, predictive analytics to enhance the in-stadium experience.

This market is valuable; according to PricewatershouseCooper (PwC), global sports market revenues are projected to grow annually at 3.7 percent from $121.4 billion in 2010 to $145.3 billion in 2015. And now we’ve got the tools to help sport franchises activate their monetization efforts. Fan insights and clustering can help identify attitudes and behavior trends to align with personal preferences. And visualized preferences, lead scoring, a key performance indicator (KPI) dashboard and data governance are all part and parcel of the toolset.

Fans have the choice to turn to their devices to watch whatever they want, and sports venues are missing out on that attention. A new way to activate customers is needed as an option that enhances fan engagement. To learn more, see what Peter O’Leary, chief marketing officer (CMO) and vice president of ticketing at the Ottawa Senators Hockey Club, has to say in this video.