Kevin Nosbusch is an IBM senior technology consultant based in Milwaukee, Wisconsin. In 1973 he played for the Fighting Irish during Notre Dame’s National Championship season, and went on to play for the San Diego Chargers.
When I played football at the University of Notre Dame and for the San Diego Chargers, broadcast television and radio were the primary ways fans enjoyed the game. There was no ESPN, no sports talk radio, the Internet was only known by DARPA scientists and social media didn’t exist.
Gosh, I sound pretty old. But in just 30 years the media and sports industries have been completely transformed by technology. Today, fans are not only Tweeting about their favorite players and teams, but just last week at the Pro Bowl athletes were participating in the virtual conversation on the field at Twitter stations.
This week, IBM and the University of Southern California Annenberg Innovation Lab (AIL) are conducting an analysis of social media trends related to Super Bowl Quarterbacks Tom Brady and Eli Manning. By analyzing hundreds of thousands of public tweets they’ll determine the fans’ sentimental favorite – the people’s champion if you will.
Like previous analyses on movies, retailing and baseball, analytics makes it possible to understand positive, negative and neutral sentiments of social media commentary, distinguish irony, and even apply machine learning to figure out which tweets are just background noise and those that are truly important. So we can now break down the Big Show’s compelling story lines in a way that traditional media never could.
Take the eternal New York/Boston sports rivalry. Is this a grudge match for the Patriots who had their perfect season ruined by the Giants four years ago? Will Tom Brady be the first quarterback to win four Super Bowls since Joe Montana? With his second Super Bowl in four years, has the Giants’ Eli Manning earned elite status and taken up his brother Peyton’s mantle as arch-rival of Tom Brady? How will the family dynamic affect Eli Manning who will play on his brother’s turf in Indianapolis during Super Bowl XLVI?
Talk radio and sports columns are full of pundits and prognosticators discussing these topics ad nauseam. Coaches and players are focused on playbooks and game strategy. Organizations running commercials are waiting anxiously for consumer reaction. In this context, fans today have an opportunity to both share and learn from others instantly; and they’re providing researchers with an unfiltered voice that is ripe for analysis.
But this is about so much more than analyzing which QB fans are rooting for in Super Bowl XLVI. Uncovering hidden insights from public Twitter comments can help a slew of people, from marketers to the NFL franchise owners, better understand opinions toward players, teams and products. These opinions can no longer go unnoticed, which is why the NFL has wisely established a Super Bowl Social Media Command Center.
We can see from the work that IBM and USC are doing that this notion of shared community engagement continues to expand exponentially thanks to other social media platforms like Facebook, Google+, LinkedIn and YouTube. In this context, sentiment analysis can be applied to learning how consumers perceive brands, reputations of companies and organizations, and new television shows and movies – just to name a few.
So which quarterback will be the social MVP of Super Bowl XLVI? We’ll find out later this week.
In the meantime, the data that IBM and USC are gathering affirms what I’ve known since my playing days — statistics are an important part of sports. And today gathering social media data can be a powerful tool that impacts the bottom lines for businesses of all sizes in all industries.
- Learn more about IBM and USC AIL social media analysis projects.
3:26 pm
[...] my colleague, and former NFL player Kevin Nosbusch posted on Wednesday, IBM and the University of Southern California Annenberg Innovation Lab are conducting the first [...]
Posted by: Super Bowl Analysis Takes Us beyond the Tweets « A Smarter Planet Blog

3:41 pm
Hello, Kevin. Looking good! Great article.
Denise E.
Posted by: Denise Evans
12:25 pm
So, just to understand, IBM and AIL use specific software in order not only to search and find posts in Tweeter about these two Quarterbacks, but also to identify the sentiment of the tweets. This is like the software identifies specific words or signs that determine sentiment. For example, when a post has many exclamation marks, then obviously the tweet shows enthusiasm.
This is extremely interesting, but it has to face a big challenge which is credibility and people’s differentiation. Many people express their sentiment about someone, like a Quarterback, with different ways. Thus, some words may sometimes show positivity, while the same words may in other cases show irony. Hence, it would be interesting if the software can identify sentiment not only based on words and signs, but also on their sequential use in tweets.
Posted by: Lefty
1:03 am
Facebook, Google+, LinkedIn and YouTube. In this context, sentiment analysis can be applied to learning how consumers perceive brands, reputations of companies and organizations, and new tel
Posted by: エドハーディー
1:03 am
s ended for the football season now, but The Tennessean newspaper was running a prediction market (crowd sourcing) on Pro and College football games, statistics, etc.
It acted like options on the price of
Posted by: エドハーディー
8:33 pm
Fear the beard!
Posted by: Grits
5:49 pm
Pizza Pete, we’re not yet sure who the winner is, but we’ll announce findings tomorrow. in the meantime, agree… Dan was a prolific passer and decent guy… but my beard was better
Posted by: Kevin Nosbusch
3:53 pm
The “game” has ended for the football season now, but The Tennessean newspaper was running a prediction market (crowd sourcing) on Pro and College football games, statistics, etc.
It acted like options on the price of stocks on the stock market. If you bought “The NY Giants will defeat the Green Bay Packers in the playoffs when the probability was nearly 50/50 the value of your investment would grow if the sentiment of the crowd turned against your pick. They could change drastically in value during the game if it looked like an upset was likely.
See Football Futures at https://secure1.foresightondemand.com/FOOTBALLFUTURES/trader/login/index
Posted by: Phil Hartman
9:36 am
So who’s the “Sentimental favorite” and most popular QB this year?
Apparently the guy whose younger brother is playing Sunday:
http://townhall.com/political-cartoons/2012/01/31/95872
PS. And the guy that Kevin N. played with at QB in San Diego was also a long time sentimental favorite of mine: Mr. Fouts.
Posted by: Pizza Pete