To my mind, predictive analytics represents an interesting intersection between statistics, economics and other disciplines, all for the purpose of making decisions about the likelihood of certain events. One of the most impressive models of predictive analytics has been the model built by the firm called Epagogix. This algorithm uses only the script of a movie to make a forecast of the expected receipts at the box office. More recently, a number of financial houses converted their models and used them to predict the outcome of the soccer World Cup that ended last month.
And now two professors from the academy have published a paper here in which they discuss the value of an algorithm that would guide movie houses on the potential box office receipts by application of their mathematical formula. From Sarah MCBrides, blog post here, it looks like they have identified some useful nuggets for making the most out a script and studio bosses may be calling on them shortly. My concern though is not about the ability of a cold model to predict fairly complex business outcomes but that the sample of movies (200) appears to be small and much less as compared to the number of movies produced worldwide.
As I undertake to read the full paper, I have no doubt that this algorithm is definitely more useful than the gut instincts of seasoned movie making experts. As I had noted a while back, experts so called, are not as sophisticated thinkers as they claim. There's competition for Epagogix now and that can only be good for the movie industry.
HT: Freakonomics Blog
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment