Data Analysis: Movie Industry

  1. Context and Background

    Studio executives want to understand which film attributes contribute to higher audience scores. Using film genres and production costs as a starting point, this report addresses the core question: Do higher production budgets translate to better audience ratings?

    Specifically, we examine two main research questions: How are production budgets associated with IMDb audience ratings, and does the length of the movie affect viewer scores?

  2. Data

    This analysis uses the Movie Industry data set combining film records from 1980–2020 with metadata from The Movie Database (IMDB).

    Key Variables:

    tmdb_rating: Audience rating score on a 1–10 scale.

    budget_millions: Production budget scaled into millions of USD for readability.

    runtime_minutes: Total film duration in minutes.
    genre: Primary film category.

  3. Data Notice

    To ensure reliable cross-genre comparisons, the sample was restricted to the eight major genres (new_genres) containing genre categories with 300 or more films defined.

Distribution of Audience Ratings

Distribution of Production Budgets

Fitted Model for Budget and Audience Ratings

OLS & Interactions

                                        budget_model    interaction_model
                                        Budget Model   Runtime Interation
Dependent Var.:                          tmdb_rating          tmdb_rating
                                                                         
Constant                           6.136*** (0.0160)    3.928*** (0.0910)
budget_millions                   0.0010*** (0.0003)      0.0023 (0.0015)
runtime_minutes                                        0.0213*** (0.0008)
budget_millions x runtime_minutes                    -3.46e-5** (1.24e-5)
_________________________________ __________________ ____________________
S.E. type                                        IID                  IID
Observations                                   4,955                4,954
R2                                           0.00224              0.16246
Adj. R2                                      0.00203              0.16195
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Budget Model: When holding the rating constant, for every 1 million increase in a movie’s production budget we observed a 0.0009 point increase in its IMDB audience rating. However, while this coefficient is statistically significant at a 1% level (p < 0.01), this model alone does not explain well enough what is affecting audience ratings (R_2 = 0.00196). This tells us that financial investment alone will not heavily influence audience scores.

The Interaction Model: When holding our runtime constant, the effect of the production budget is estimated to be 0.0023, but this is statistically insignificant (p > 0.05). But, when we hold our production budget constant for every additional minute of runtime we will see an 0.0213 increase in audience rating, and we can see that this is statistically significant at a 1% level. The interaction term (budget_millions x runtime_minutes), will allow us to evaluate how the budget effect changes as movies get longer, we see a decrease of -0.0000346 in the rating effect per minute of runtime, and we can see that this is also statistically significant at the 1% level. This tells us that there is a positive association of production budget falling as film lengths increase.