Skip to content

Extensive pre-modeling EDA to derive insights for Microsoft's new movie studio

Notifications You must be signed in to change notification settings

klsalcedo/microsoft_movie_analysis

Repository files navigation

Microsoft Movie Analysis

Group Members: Joanne de Leon and Katarina Salcedo

Motivation

Provide Microsoft with an understanding of how to create a successful movie studio. We will give insights on how to produce high grossing and top-rated movies thus providing the tools to launch a well known and highly regarded studio.

Data

  1. To get information on movie ratings and characteristics we used two Rotten Tomatoes datasets that contained a total of 33517 movie ratings from 20 different genres.
  2. For information on movie preformance in theaters, we looked at a dataset from The Numbers which provided information on release dates along with domestic and worldwide gross. This set had 4387 movies ranging from 2000 to 2019.

Methodology

  1. The Rotten Tomatoes datasets were merged togeter using the ID column to create a new dataset with information on movie ratings and their genres. We decided to examine their movie ratings to see which genres were the most rated - no matter the rating - to ascertain which types of movies were most frequently watched as we hypothesize that the two are related. To closely examine the distribution of ratings along with the genres asscociated, we subdivided the ratings into high, average, and low.
  2. A subset of The Numbers data containg years ranging from 2000 to 2019 was taken to account for the rate of inflation. This data was then cleaned and release dates were seperated into two columns containing the release month and release year. We were then able to group movie titles into their release months and years then plot this data. From this, we were able to track box office trends.

Findings

  1. Since Drama and Comedy were by far the most rated and hence the most watched films, concentrating on producing films in the these genres will ensure viewer turnout as well as introduce the public to the studio’s brand.

  2. Significant numbers of viewers watch movies in the Action & Adventure, Mystery & Suspense, Romance, and Science Fiction & Fantasy genres, no matter the rank of rating, implying that the quality of the films in these genres is irrelevant as to whether a viewer will watch. Therefore creating films in these genres will yield a consistent viewership as well as increase exposure of the studio.

For all three subsets of ratings, high, average and low, the same genres are repeatedly represented as the genres with most ratings. Namely, Drama, Comedy, Action & Adventure, Mystery & Suspense, Romance, Science Fiction & Fantasy, Art House & International, Kids & Family, and Horror. With Classics showing up in the high ratings but Animation being represented in the average and low ratings.

Screen Shot 2021-06-10 at 7 38 59 PM

Screen Shot 2021-06-10 at 7 39 22 PM

Screen Shot 2021-06-10 at 7 39 39 PM

  1. We narrowed down the months which have consistently brought in the highest box office sales to June, July and November with February and May coming second to these. Microsoft can increase movie profits by aiming to release new movies during these peak months.

Screen Shot 2021-06-10 at 4 18 09 PM

  1. Based on the yearly trend, we advise Microsoft movie studio to consider streaming their movies in addition to releasing them in theaters to reach a greater audience and ensure success.

Screen Shot 2021-06-10 at 4 39 20 PM

About

Extensive pre-modeling EDA to derive insights for Microsoft's new movie studio

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •