Making feature films is a high-risk venture. Studios invest millions of dollars creating the next blockbuster with no way of knowing up-front whether their huge investments will turn a profit. When deciding whether to fund a particular movie, what are some pieces of information that would help ensure that the film would be successful? It would help to know that you are investing in a good story, with memorable characters. Something that will connect with the audience. Our project explores the relationship between books and movies to answer the question: do book adaptations perform as well as movies? What does the data tell us about how book adaptations perform after being released as movies, and are they reliably better than the average movie?
To analyze the success of book adaptations, we performed EDA on over 45,000 films from March 1910 to July 2017 using a dataset originating from the TMDB Open API. In this dataset, there were 829 films based on novels. To extract insights on the books themselves we combined this data with Goodreads data such as book publisher, page count, and book rating. Matching the two datasets on ‘title’ revealed 105 matching movies and books that were used in this analysis.
Hypotheses we evaluated, along with results, are below:
- Are ratings of Book Adaptations statistically different from most films? Y
- Are Book ratings positively correlated to Film ratings? Y
- Are Horror Adaptation ratings statistically different from ratings of other film genres? N
- Is there a relationship / correlation between the number of reviewers of a book, and the movie rating? N
- Is there a relationship / correlation between the number of Goodreads book reviews and the movie rating? N
Raw Data: Link
- Ashley Howell Gates (PM)
- Matthew Bishop
- Lauren To