# Wolfram Alpha Does Its Best to Analyze the Super Bowl

Can an analytics engine predict the Super Bowl?

Wolfram Alpha isn’t making any guarantees, but the self-billed “computational knowledge engine” is willing to crunch the numbers and throw out a few data points about this Sunday’s Super Bowl XLVII, in which the Baltimore Ravens will face off against the San Francisco 49ers in New Orleans.

Now approaching its fourth anniversary, Wolfram Alpha is an altogether different search engine from either Google or Bing. Unlike the latter two, which return pages of blue hyperlinks (or thumbnail previews of visual media) in response to queries, Wolfram Alpha produces a definitive, usually numerical answer. Type in the name of a celebrity, for example, and the output includes dates of birth and death, height, a chart of notable accomplishments, and even a graph of Wikipedia page hits over the years.

Developed by Stephen Wolfram and based his Wolfram Research’s Mathematica analytical platform, Wolfram Alpha contains more than 10 trillion pieces of data cultivated from primary sources, along with tens of thousands of algorithms and equations. Solving complex math problems is a particular strength: simply type one into the search bar, and the system produces an answer. (An extensive rundown of Wolfram Alpha’s capabilities is available on its “About” page.)

But how does the system feel about Sunday’s game? “There can be years between games with the Ravens and 49ers,” read a Feb. 1 posting on Wolfram Alpha’s blog. “And when they play, the Ravens objectively get more done. They score more touchdowns and they move across the field more frequently.”

This year, however, the two teams’ stats are “annoying similar,” the posting continued. “They have the same number of touchdowns, nearly the same number of points, yards, and field goals, and similar third down conversion percentages.” All that being said, the 49ers have done all that in fewer plays, a potential sign that the team plays “more efficiently.”

The 49ers have played a game at the Mercedes-Benz Superdome this season and won; the Ravens played at the stadium once in recent times, and won as well. So the venue is sort of a draw, at least as far as historical data is concerned.

On the player front, Wolfram Alpha also ran a graph comparing the sacking history of Ray Lewis, a Ravens’ star linebacker, with that of 49ers’ linebacker Aldon Smith, and found their data radically different:

It also compared the stats of Colin Kaepernick, the replacement for injured 49ers quarterback Alex Smith, and found some data points that could help San Francisco, at least with regard to rushing yards:

But there’s always an element of unpredictability to these sorts of things: no matter how accurate the analytics engine backing Wolfram Alpha, there’s always the chance that something—a freak injury, a stray gust of wind at a crucial moment—could alter the dynamics on the ground in ways unpredictable by historical data.

This isn’t the first time that Wolfram Alpha’s done something designed to boost its profile in the eyes of your average Web surfer. In January, for example, it upgraded its Personal Analytics for Facebook platform, giving users the ability to mine their own social-networking data in new ways.

Image: Wolfram Alpha