Stats at the Movies

This article, written by Brendan Bettinger, shows Bettinger attempting to calculate a movie's gross income using a variety of variables. From his calculations, which utilized Minitab, Bettinger was able to approximate an equation for a movie's gross based on reviews, budget, number of theaters, if it is a sequel or not, and if it is PG-13. He got the equation Gross = –80 + 0.6×RT + 0.5×Budget + 0.025×Theaters + 50×Sequel + 20×PG13, which has a decent R2 value of .65 (65%). I think this is interesting because I would never have thought it would be possible to so arbitrarily pick out a movie's income. I also found it cool that he made a regression with multiple variables.

This article relates to our AP Stats class because as of now we are learning to create regressions for a set of data. This article begins by creating a linear regression based on movie reviews, then goes more in depth. A couple of new things I learned from this is about regressions using multiple variables, and binary variables.

To see how well this model works, I am going to test it on "The Dark Knight Rises".

RT: 87
Budget: $230,000,000
Theaters: 15,000
Sequel: 1 (yes)
PG-13: 1 (yes)

Gross= -80 + .6(87) + .5(230) + .025(15,000) + 50 + 20 = $532,200,000
Actual: $1,076,169,641

So, for the TDKR, this doesn't quite work. However this movie was wildly successful and popular, so it was likely an outlier.

 

Finally, I will attempt to predict the revenue for Skyfall using this equation. Since there are no reviews or theater numbers yet, I will predict using Quantum of Solace's values.

 

RT: 65

Budget: $130 million

Theaters: 3500

Sequel: 1

PG-13: 1

 

Gross= -80 + .6(65) + .5(130) + .025(3500) + 50 + 20 = 181.5

With this, I predict Skyfall will gross $181,500,000.

 

And that is the statistics of movies!