For this visualization, we graphed the length of movies according to their release date, and gave each data point a color on the continuous color scale based on its popularity. As the visualization shows, there has been a slight (but not statistically significant) decrease in the average length of a movie in the past 100 years or so. Furthermore, according to the color scale, there does not appear to be any significant correlation between a movie’s length and its popularity. While a few extra long movie outliers are some of the least popular, we don’t believe these have any significant impact on the overall correlation between a movie’s length and its popularity—audiences seem to equally enjoy short and long movies!
For this visualization, we graphed the release date of movies and compared them to their popularity at the time of data collection. As the visualization shows, there may in fact be a very slight negative correlation between these two factors. While it is certainly evident that our dataset contains many more movies from more recent decades as compared to more distant ones, it also does seem to show a very small decline in a movie’s popularity according to its release date. In short, audiences may tend to prefer older movies to newer ones!
For this visualization, we made a graph of the popularity of all of the data in our dataset, and we gave each data point a color according to its movie genre. As the visualization shows, the genre of a movie does not seem to have any significant effect on that movie’s popularity, as all the colors seem to be relatively equal in terms of their position on the y-axis (popularity). It is interesting to note, however, that the “Drama” genre, followed by “Comedy” and then “Action,” are by far the most prominent genres in our dataset. Given our dataset’s large size, this may suggest that most movies made in the past 100 years or so tend to be one of these genres, perhaps implying that these genres reach a wider audience than the others. Therefore, while it seems that moviegoers tend to equally favor all movie genres, it may also be the case that the genres of Drama, Comedy, and Action tend to draw in the largest number of audience members.
For this visualization, we used a categories diagram to analyze whether there is any correlation between a movie’s popularity and whether it has won any awards. As the visualization shows, there does seem to be a slight correlation between popularity and awards won—but only in one direction. While it seems to be the case that, as a movie’s popularity increases, so too does the chance that it has won at least one award, the visualization also shows a rather large number of movies which have not been very popular but have still won at least one award. Therefore, we have concluded that, while the most popular movies may be more likely to win an award, some of the least popular movies may also win at least one—perhaps suggesting a disconnect between the interests of audiences and those of film critics.
For this visualization (along with the corresponding visualizations 6 and 7), we analyzed whether the amount of experience possessed by an actor (or actress or director) had any effect the popularity of the movies they were in (or made). To do this, we plotted the number of movies starring each actor against the average popularity of the corresponding actor. As the visualization shows, there really seems to be no correlation between the number of movies an actor has starred in and the average popularity of movies starring that actor, suggesting that the level of experience of the starring actor in a movie does not affect that movie's popularity. It is worthing noting, however, that many of the actors in the dataset had only one movie listed for them, so this correlation may or may not have differed if there were more datapoints of actors starring in multiple movies.
As with the previous visualization, this visualization plots the number of movies of actresses in the dataset with the average popularity of the movies of these actresses. As with actors, there does not seem to be any significant correlation between experience and popularity—although it may perhaps be a slightly more positive correlation than for actors. Again, too, it is worth noting the relatively small number of dataset actresses who have starred in more than one or two movies, so this correlation may or may not have shifted had we had more extensive data.
As with the previous two datasets, this visualization plots the number of movies a director has made against the average popularity of movies made by that director. Similarly to the previous two, there does not seem to be any significant correlation between a director's experience and the popularity of his or her movies, suggesting again that the experience levels of those involved in a movie do not necessarily affect that movie's popularity.
This visualization, along with the following two in 9 and 10, attempt to assess whether inculding any particular actors, actresses, and/or directors has any effect on a movie's popularity. To do this, we took the top 250 most popular movies in our dataset and determined the actors who had starred in the most of them. As the pie charts were originally difficult to read due to the sheer volume of actors/actresses/directors who had been involved in just one of the top 250 movies, we reduced these visualizations to the top 30 of each of these groups of individuals, allowing us to view the top 30 actors, actresses, and directors of the 250 most popular movies in our dataset. As this visualization shows, actor John Wayne clearly leads the way for most movies making the top 250, starring in 12 out of those 250. He is followed by Paul Newman, with 6, and then Marlon Brando and Roger Moore with 4 each. Therefore, if moviegoers are searching for their next favorite movie, they would do well to choose one starring one of these actors! Filmmakers, however, are not so lucky. Since none of these actors are alive anymore, they will need to resort to actors lower down on the list if they are looking to improve their chances of maing a successful movie.
This visualization, like the previous one, attempts to find those 30 actresses starring in the greatest number of the 250 most popular movies in our dataset. As the visualization shows, Elizabeth Taylor comes in as the actress starring in the greatest number of most popular movies, with 6 popular credits to her name, followed closely by Diane Keaton, with 5. Sophia Loren and Greta Garbo are tied for third, starring in 4 top 250 movies each. Luckily for filmmakers, this time, Diane Keaton is still living and pursuing an acting career, so they would likely do well to select her for a lead role if they are looking to make a popular movie!
As with the previous two, this visualization is aimed at analyzing the 30 directors of the greatest number of the 250 most popular movies in our dataset. As the visualization shows, John Ford and Taylor Ray are tied as directors of the most top 250 movies, with 7 each. Sydney Pollack and Ingmar Berman both come in second, with 4 directorial successes, and Stanley Kubrick, Federico Fellini, Alfred Hitchcock, and Jud Taylor all follow closely behind, with 3 directorial movies making the top 250.
Overall, we have learned lots about what goes into making a good movie! Surprisingly, there do not seem to be any factors which correlate heavily to the popularity of a movie. While audiences may slightly prefer older movies to more recent ones, and while there may be a slight—and probably negligible—positive correlation in a starring actress's level of experience, it seems that the factor most affecting the popularity seems to lie in the people behind it. While not seemingly correalted to experience level in any way, there do seem to be some standout actors, actresses, and directors which audiences seem to prefer most. In the end, however, we have found that popular movies really can be made in a plethora of different ways, meaning that there really does seem to be an audience for all sorts of movies!