Nate Silver was correct when he predicted the recent election state primaries that cost the Republics the White House, (see High-Risk Primaries Could Cost Republicans in 2014). Now, Nate is going back to his sporting roots. His initial success was in the statistical analysis of baseball with his PECOTA system for projecting baseball player performance and careers. However, now he is making a different bet on a different sport. It is football. In his recent prediction (Nate Silver Picks the Super Bowl!) after analyzing both the 49ers and the Ravens, Nate predicts that the 49ers will be the Super Bowl winners in today’s upcoming battle in New Orleans.
Can statistics predict a super bowl winning team? There are a number of unknown variables and we cannot quantify which team has the most resolve and determination to win. Both team have fought hard to reach the super bowl, but how much do they want win. In a recent twitter post, Rakesh Nair, @mynameisnair, created a firestorm of debate arguing that “Unless analytics products incorporate sentiment & heart, they cannot accurately predict outcomes.” It is unknown for which team Mr. Nair is rooting…however, it, perhaps, is the Raven given his recent slightly acerbic and highly polarizing tweets or he may be simply engaging in an intellectual argument about the maturity of today’s statical models to integrate emotion and sentiment as variables.
Can big data analytics predict super bowl winners? Do we have enough data on each team and the players to gain insight into the challenge. Perhaps. However, right or wrong…one data point doesn’t make a solid track record (reference: The Black Swan & Fooled by Randomness Blog by Nassim Teleb). We will see if Nate is right or wrong tonight.