---
Back in 2011, a very popular TV show called [Friday Night Lights](https://en.wikipedia.org/wiki/Friday_Night_Lights_(TV_series) "Friday Night Lights") was wrapping up it's final season. As with all shows that rally cult followings, fans cried out for a followup movie; but this one was different. What this meant is that in it's lifetime as a story "Friday Night Lights" was a real life situation that inspired a book, which was turned into a movie, which became a TV show which could potentially become a movie again. This is when I personally became interested in the "cross section" of pop culture. I found myself thinking about questions …show more content…
So which ones should we choose to adapt? Is it possible that we can predict which TV shows will make succesfully movies? Let's give it a shot.
#### How Can We Do This?
For our first pass, we can approach this problem using **linear regression**, a [supervised learning](http://garretthoffman.github.io/hipster_game/ "The Hipster Game, or, a Very Serious Introduction to Core Concepts in Supervised Learning") technique, where our predictive relationship is defined using a function of the form **y** = **B0** + **B1** * f(X1) + **B2** x f(X2) + ... + **BN** * f(XN). f(X*i*) is tradiitionally equal to X*i*, however, in some cases other function may be a better fit (e.g. log(X*i*) or X*i*2).
#### What Do We Know Already? (The Data)
Approximately 65 TV show adaptations were sampled. This is a smaller sample size then we would normally like, however, it is not uncommon to come across problems with limited information available for analysis. Data on the movies and and corresponding TV shows was collected from Box Office Mojo, IMDB and Wikipedia using python webscraping tools such as **requests** and **beautiful