Suppose it's the end of 2013, and you want to develop a forecast in preparation for the March 2014 home entertainment release of the Disney movieFrozen.

The 2013 model does not yet have at least 52 weeks of home video unit sales data for all films released in 2013 so the 2013 model will not be ready to use until the end of 2014. Therefore, at the end of 2013, the 2012 regression model is the most recent model available.

What are some of the shortcomings and difficulties associated with the regression model Disney made for its forecast forFrozen? If you worked for Disney Studios, how might you improve the forecasting model? What other factors, besides gross box office, might you take into consideration when developing a forecast for home video units?