Facebook Stats Show Who’s Got Game in NBA Playoffs

Dan Grady's picture
 By | May 14, 2014
in WebFOCUS, Analytic Dashboards, Big Data, Business Analytics, business analytics, dashboard, Dashboards, Data Discovery, Data Visualization, Facebook analytics, feedback analytics, Sentiment Analysis, sentiment analysis, share of voice, Social Analytics, Social Media, text analytics, WebFOCUS, Marketing
May 14, 2014

The NBA playoffs are in full swing and many fans are already calling 2014 the best year ever. A number of playoff records have already been set and we haven't yet finished the second round.  A record 5 of the 8 first-round series went to a deciding Game 7, which has never happened before. In fact, the most Game 7’s ever played in the playoffs is 5 and we’ve reached that already this year with 3 rounds to go! In addition, a record 8 overtime games have been played, with Oklahoma City and Memphis playing 4 straight overtime games in their series. Another playoff first.

The games alone would be enough to have NBA fans buzzing and lighting up social media.  But the NBA also had a major PR controversy to deal with when TMZ posted audiotapes featuring Los Angeles Clippers owner Donald Sterling making offensive remarks. The league had to act quickly and on Tuesday, April 29, the NBA commissioner announced that, effectively immediately, he was banning Sterling for life from having any association with the Clippers organization or the NBA. Another NBA first.

With all this activity, I thought it would be interesting to analyze some social media data, so I pulled commentary from the Facebook pages of each of 16 teams that made the NBA playoffs for the week of April 28th through May 5th.  I used Information Builders WebFOCUS Business Intelligence & Analytics platform to access the data, apply various analytical techniques like sentiment and word frequency, and then visualize the data in various formats to look for trends, patterns and outliers. 

What’s especially cool about social media is how it makes new data available for comparative or competitive analysis. For example, we can access and analyze what the public is saying about our own brand, and we can do the same for our peers. That is essentially what I’ve done here for the 16 teams in the playoffs: 

  • I captured 117,847 total posts and comments across the 16 Facebook pages for the week.
  • Those comments were created by 75,602 unique individuals (or accounts).
  • From a share of voice perspective, the San Antonio Spurs were the winners with 19.87% of the conversation. The Golden State Warriors and Indiana Pacers followed with 12.24% and 12.09% respectively.
  • The Atlanta Hawks surprisingly drew the lowest amount of social support with only 1.16% of the conversation.  I say surprisingly because as an 8 seed they pushed the first seeded Pacers to Game 7. I would have expected a little more engagement there.
  • From a sentiment analysis standpoint, the team that generated the highest percentage of positive posts/comments was the Toronto Raptors with 38.7% of their 2,544 posts/comments scoring positively. The Memphis Grizzlies came in second at 31.9%.
  • The Miami Heat and the San Antonio Spurs, both participants in last year’s final games, had the lowest negativity on their pages at 3.79% and 4.76%. 
  • The Atlanta Hawks were the losers in this analysis with a league-leading 13% negativity on their page.

Perhaps the biggest winner of the week was Adam Silver, the first-year commissioner of the NBA.  When I use a search-based application to analyze the data (also WebFOCUS), I discovered that 45% of the posts/comments that mention “Silver” scored positive to only 13% negative.  This favorable feedback is likely a result of how quickly and decisively he responded to the controversy and was able to return the attention to the games themselves.

Which most NBA fans would agree has been – “Fantastic!”

If you want to learn more about this technique and low to leverage it, I am hosting an upcoming webinar on the topic of “Feedback Analytics” which will go more in depth into the use cases. You can sign up for it here: