Box-office Soothsayer

Coming soon is a methodology that uses software and the Web to forecast the
financial viability of a movie

If
'Hollywood is the land of hunch and the wild guess', is Bollywood any
different? Sleazy stuff and big bang action may no longer sell. Foretelling
box-office success is, therefore, a more potent task than guessing the number of
hot-scenes in Mallika Sherawat's next flick.  

Producers
who till now turned to astrologers for financial performance predictions of
their films, may now have a more logical tool to beat the unpredictability that
the movie industry is famous for. The method uses IT and statistics to predict
success before a film's theatrical release and promises to de-risk the
business for Hollywood investors like never before. It can work for Indian
films, but with customization.

In
an Oklahoma State University supported research project, information scientists
Ramesh Sharda and Dursun Delen have come up with a system to use neural networks
for predicting box-office success using seven key parameters: the value of a
star or a superstar of competition; genre; sequel; Motion Picture Association of
America ratings that assess the degree of sexual content, violence and adult
language; technical effects in the film; and the number of screens on which the
film is to be shown during its initial launch.

Right
now, this is more of a methodology than pre-packaged software. But Ramesh says
he will put the trained network and other comparable techniques into a web
system to let others enter inputs and generate a forecast. This system should be
operational sometime next year.

Neural(ogy) Networks mso-fareast-font-family:"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language: EN-US;mso-bidi-language:AR-SA">Neural(ogy) Networks An Artificial Neural Network is an information processing pattern that is inspired by the way biological nervous systems, such as the brain, process information. The key element is the structure of the information processing system, composed of a large number of highly interconnected processing elements working in unison to solve specific problems. It is capable of modeling extremely complex non-linear functions. “For many years, linear modeling has been the commonly used technique in capturing and representing functional relationships between dependent and independent variables, largely because of its well-known statistically explainable optimization strategies. In the problem scenarios where the linear approximation of a function was not valid (which was frequently the case), the models suffered accordingly. Now, such cases can easily be modeled with neural networks,” writes Ramesh in a study called 'Predicting box-office success of motion pictures with neural networks'. Applications of neural networks have been reported in many diverse fields addressing problems in areas such as prediction, classification, and clustering.

Prof Ramesh Sharda is reportedly working on the movie prediction software with major Hollywood studio. He is a graduate from Udaipur University

The
accuracy of the neural network model, Ramesh says, can be improved by adding
some other determinant variables such as production budget and advertising
budget, which are known to be industry trade secrets and are not publicly
released. The customization for the Indian film industry would mean analyzing
the historical data of films with a different set of parameters. Songs, for
example, play a big role in popular Indian films; movie ratings probably less
so.

Ramesh
and Prof Delen have been working on this model for the last seven years, mostly
collecting data from Hollywood and testing the model each year. The forecasting
problem here is converted into a classification problem, that is rather than
forecasting the point of estimate of box-office receipts, the duo classifies a
movie in nine categories, ranging from 'flop' (less than $1 mn) to
'blockbuster' (over $200 mn).

Ramesh
claims the neural network model, developed using a commercial software product
called NeuroSolutions, will predict the financial success of a motion picture
before its theatrical release with pinpoint accuracy 37% of the time and within
one category of performance, with 75% accuracy.

If
that really happens, studios, distributors and exhibitors will have more relaxed
times ahead. Movies will cease to be risky ventures.

"Times New Roman";mso-ansi-language:EN-US;mso-fareast-language:EN-US;
mso-bidi-language:AR-SA">-Goutam Das 

goutamd@cybermedia.co.in