This dataset explores trending streaming services such as Hulu, Amazon Prime, Netflix, and Disney+. Our goal is to conduct a comprehensive analysis by visualizing the four datasets to answer questions about trends in TV shows, movies, directors, cast, countries, genres, ratings, etc., like the ones listed below:
- Which TV shows and movies are available on all the four streaming platforms? Which of them are unique to only one platform?
- What is the ratio of movies to TV shows within the four platforms?
- Which directors have directed movies and TV shows that are available on all the four streaming platforms? Which directors appear only on one of the streaming platforms?
- How many of the TV shows and movies are available in each country? How many of them are available across multiple countries and how many are only available in one?
- In which month were the most TV shows added to each platform? What about movies?
- ...
Drawing such connections can lead us to generate information that can be used to create a user-facing dashboard for viewers, producers, and other stakeholders in the streaming industry. It can help them make informed decisions about content to watch, create and distribute.
The "Streaming Symphony" created from these datasets (Hulu, Amazon Prime, Netflix, and Disney+) provides insights into movie and TV content on streaming platforms. The datasets include:
- show ID
- type
- title
- director
- cast
- country
- date added
- release year
- rating
- duration
- genre
- description
Although the dataset's collection method is unclear, it was gathered in mid-2021 and provided by Shivam Bansal on Kaggle, a community platform for data scientists and machine learning enthusiasts.
The dataset provides a snapshot of available content on each platform up to mid-2021, which is why the datasets may not be 100% comprehensive or accurate with current trends. Nevertheless, the datasets allow for analysis of trends over time in genres across platforms and countries, useful for researchers studying the streaming industry. Information about directors and cast members can be valuable for analyzing industry patterns. Combining the datasets offers a comprehensive overview of streaming platform offerings for comparison.
As we can see from graph below, the shows and movies released on Netflix are mostly concentrated between 2010 and 2020. The number of shows and movies released on Netflix is increasing over the years since the mean and quarters are closer to the top of the chart.
From the bar chart below, we can see that the most shows on all platforms are rated for general audiences. The bars on the left side (indicating for less mature audiences) are much longer than the bars on the right side (indicating for more mature audiences).
From the plots below, we can see which directors have directed the most number of movies within each of the platforms. The Hulu dataset only mentions three directors throughout the entire dataset and they all appear exactly once, which is why the results look similar for all three of them.
I am interested in what genre each platform uploads the most, like what each genre is the most dominant on each platform and within each platform what's the distribution like. Also I am interested in what genre each director have participated in and which genre each director participated the most. The other point I want to dig into is what genre the actor participated the most.
What genre each director have participated in and which genre each director participated the most & what genre the actor participated the most
Regarding the directors and actors who have participated in various genres, it appears that each director and actor has their preferred genre. This preference could be why they are more likely to show up in shows of specific genres. It is also interesting to note that there are few directors or actors who have participated in more than 20 shows, except on Amazon Prime. For instance, Mark Knight directed 109 shows in the "special interest" genre, which is the highest number for any director in a single genre across all platforms. Similarly, Gene Autry participated in 31 shows in the "western" genre, which is the highest number for any actor in a single genre across all platforms.
**Question 2 (Alex): What is the feature distribution by year, and are there any trends for specific genres over the years.
So I started by creating a basic line graph to observe the general trend of how many features were released per year that made it onto a streaming service. My first observation was that the streaming services tended to favor newer content, either by popularity or exclusives (such as Netflix shows like Stranger Things). It makes sense that as the population has gone up since our oldest feature was released, that the number of features has increased exponentially, along with the population. The comparison of population and number of features vs time follow a very similar pattern. More people watch newer content, so these streaming services tend to favor newer content then classic movies and TV shows. This is very reflective in the following plot.
Next I wanted to analyze this further, by genre. My goal was to check for any trends regarding genre across this 100 year time period. Though the number of features per genre can vary greatly, we can still compare relative trends between similar genres. PG-13 movies and TV-14 TV shows compare at 3162 features and 3111 features respectively. Though they share a similar number of features, PG-13 follows a much cleaner trend compared to TV-14. The amount of PG 13 trends upward much sooner than TV-14.
Generally, if we ignore feature count by genre, and just analyze the trends of each genre, we can see there aren't many outliers. No specific genre has a noticeable change between 1920 and 2000. Most trend upward exponentially from around 2000 onward. The last thing I noticed was that the TV prefix tended to follow the same pattern. Minus TV-Y7, every TV genre trended downward in 2021, the end of our dataset. A quick google search fills in why. “Historically, the fall television schedule was created to help auto advertisers promote their new car models.” Which makes sense, given that our dataset was last updated at the end of September 2021. The dataset has not been updated with fall shows, leaving the plunge shown in figure 3.
Figure 2 is showing the difference between a popular genre like G rated movies, vs a less prominent genre like PG. The top graph is more fluid, and less impacted by outlying years compared to the varying number of PG features we see in the bottom plot. Less features results in an increase in volatility. Figure 3 follows this as well, less data means more uncertainty in trends.
I was able to analyze the trend of features by year, and look into various hypotheses for why streaming services favor newer content. As the population increases, the number of features does too. Most likely because younger audiences prefer newer, more modern content. Every genre we analyzed followed the same pattern of increasing exponentially, while TV shows specifically fell off in the first half of 2021, indicating that more TV shows get released in the fall. A quick google search aligned with this theory.
My Reserch Questions revolved around Directors and Casts and tying them together with the other componenets of the datasets. On a broader perspective, I was trying to look into how the presence of certain directors and actors on each streaming platform impact user engagement and satisfaction, what the most common director and actor pairings are across platforms, and how these pairings vary by genre. Additionally, I tied these aspects to other compnenets of the datasets such as duration and type. Specifically, my research questions are:
- What is the distribution of Movies and TV Shows for each of the Streaming Platforms?
- Which director(s) appears to be directing the most number of the movies/TV shows within the four platforms?
- Which directors have directed movies and TV shows that are available on all the four streaming platforms? Which directors appear only on one of the streaming platforms?
Unfortunately, there are no directors that appear on all four of the platforms.
- Are there any TV shows and movies across all the platforms which have the same cast as other TV show(s) or movie(s)?
Unfortunately, there are casts mentioned in the Hulu dataset, which is why there exists no plot for the casts of the Hulu Dataset.
- For each director, what are the different genres they have directed movies and TV shows of? For each cast member, what are the different genres they have appeared in?
For these questions, the results are in tabular form but I do have general statistics for reference:
Netflix
Number of Directors: 8235
Number of Casts: 44367
Number of Genres: 42
Amazon Prime
Number of Directors: 8235
Number of Casts: 44367
Number of Genres: 31
Disney Plus
Number of Directors: 1121
Number of Casts: 5899
Number of Genres: 44
Hulu
Number of Directors: 3
Number of Casts: 0
Number of Genres: 36
Ethan:
For question 1, The complete analysis and visualizations can be found here. In this project, I analyzed various genres across different streaming platforms, as well as the directors and actors who participated in those genres.
My findings show that drama is a popular genre across all platforms except for Disney+, where family and animation are the most popular. Additionally, the least popular genres differ across platforms. I also found that each director and actor has their preferred genre, which could explain why they are more likely to appear in shows of specific genres. Interestingly, few directors or actors have participated in more than 20 shows, except on Amazon Prime.
Overall, this project provides insight into the dominant genres on various streaming platforms, as well as the preferences of directors and actors when it comes to genres. These findings could be useful for me and others interested in producing content for these platforms or for individuals interested in exploring specific genres.
Alex:
Here is a link to the rest of my python code. This has some extra visualiztions, along with the opther questions and their respective graphics.
Overall, streaming services favor newer features versus older ones, while TV shows tend to be released later in the fall, compared to movies. There aren't any noticeable trends for specific genres over the years this dataset was recorded. This project allowed me to approach data analysis in my own way. I was able to test various visuals and coding strategies on my own, compared to the forced nature of the labs.
Anitej:
In conclusion, I was able to find the correlation between Directors/Casts and other components of the datasets, specifically genres and number of appearances. Here is a link to the complete analysis of the datasets regarding this topic, done by me. Overall, We find that:
- Netflix, Amazon Prime, and Disney Plus are streaming services to go to when looking for movies and Hulu is the one for TV shows.
- Except for Hulu, there are at least 4 directors in each dataset who've directed more than 10 movies in that dataset.
- Bear Grylls is the only cast that appears on all of the datasets.
- The most number of times a cast has appeared in a dataset is above 50, on Amazon Prime.
- The most number of Genres present in a streaming platform is 44, on Disney Plus.
- The most number of Directors present on a platform is 8235, on Amazon Prime.
- The most number of Casts present on a platform is 64126, on Netflix. This detailed analysis of the datasets in relation to the directors and Casts allowed me to successfully answer my research questions and come to a conclusion.