A professor at MIT's Sloan School of Management and two colleagues have developed a mathematical model they say will predict movie revenues -- including second-weekend B.O. -- based largely on first-weekend online reviews from Yahoo! Movies.
Chris Dellarocas, along with
Neveen Farag Awad of the U. of Michigan Business School and
Michael Zhang of MIT, waded through data from numerous online movie sites before settling on Yahoo!'s.
They experimented with several models that used combinations of variables such as online reviews, professional critics' reviews, first-weekend B.O., production budgets, per-theater revenues and the number of online postings.
Along the way, the researchers discovered online reviews by women correlated much more closely with how well a movie does than reviews by men. As a result, their model gives more weight to the women's reviews.
Results of the research so far have been presented at academic conferences, with the next outing scheduled for June 25 at the Intl. Marketing Science conference in Rotterdam.
The research is still a work in progress; the profs plan to improve the accuracy of their model by taking into account other movies being released at the same time and the genre of the movies involved.
The formula and all explanations are online, in pdf format, at
http://ccs.mit.edu/dell/papers/movieratings.pdf
Contact the Variety newsroom at
news@variety.com