By Jeremy Stierwalt
Service Line Director, Analytics
Optimal Solutions an NTT Data Company
Two weeks ago, the Seattle Seahawks clinched their Super Bowl spot with an exciting last minute interception by Richard Sherman. It was an impressive, dramatic play to end the game; the kind of nail-biting denouement that football fans live for. But it was Sherman’s post-game interview that set the social networks ablaze. Tweets with the name ‘Richard Sherman’ peaked at 1.3 million that day, and a deluge of Richard Sherman oriented posts and comments dominated Facebook, Reddit and YouTube. All of these comments, which were once only shared between fans at the local bar or water cooler, are now captured and measured via social media tracking. Social media platforms have become one of the largest data banks from which to learn and draw useful insights, and this ‘social sentiment’ data might actually be able to predict athletes like Sherman’s performance and the outcome of games. Sound far-fetched? I think not… and companies are investing billions of dollars on it… (including our parent company NTT)
It’s a similar phenomenon to the ‘Home Team Advantage’, in which teams playing on their own field are said to have a greater chance of winning with fan support (among other factors). There may also be a ‘social sentiment advantage’ at play. SAP InfiniteInsight and the SAP Predective Analtyics Platform can draw on the wealth of data provided by social media and attempt to make predictions about the big game.
First off, will Richard Sherman be able to back up his boastful claims with his performance on Sunday? Manning throws “Ducks”? What do the fans say? The self-proclaimed ‘best Cornerback in the League’ received 27.5% positive sentiment, 16.3% negative, and about 56% neutral sentiment mentions in the social sphere. It seems a few weeks time has mellowed the overall popular opinion of the player, which was overwhelmingly negative the days following his infamous interview.
Comparatively, Broncos QB Peyton Manning garnered 44.5% positive sentiment, 11% negative, and 44% neutral within the last week, making him the clear fan favorite between the two rival players.
As for the overall teams, social chatter is higher for Seahawks, but the Broncos received slightly higher positive sentiment at 37%, versus the Seahawks 33% positive sentiment. If the social networks are any indication of performance, Peyton and his Broncos seem to have a very slight edge for Sunday’s game. Go BRONCOS!
Other, more ‘scientific’ factors can also be measured using SAP’s Player Comparison Tool. This tool utilizes the SAP Predective Analytics platform and measuaables include items like:
- How jet lagged the players might be
- How good a team is on rushing/passing
- How the weather is predicted to be on game day (we query a national weather forecast service to get up to date weather data)
- How well teams perform on different types of turf
- A myriad of of other intangibles
As the technology continues to evolve, the big data industry will learn just how accurate social sentiment can be in forecasting outcomes of sporting events. I’m hoping my numbers will prove favorable for the Broncos today, but if there’s one thing I’ve learned from this unpredictable season, it’s that in life (and football), anything can happen.