My final project is a visualization of the Sight and Sound Poll of the Greatest Films of All Time.  Sight and Sound is a monthly film magazine published by the British Film Institute. Once every ten years, critics are asked to submit a ballot of ten films they think are the greatest of all time. A list is then compiled that ranks approximately one hundred films based on how many times they appeared on critics’ ballots.  The poll was first conducted in 1952 and has been redone every ten years since.  When I first looked over the 2022 list, I was disappointed in it, especially in the representation (or lack thereof) of more modern films.  But, I knew that the list would be a great source of data by which a closer inspection could reveal trends I was missing at first glance.  Particularly, I was interested in tracking how both country of origin and production year affect a film’s position on the list.  To do this, I looked at the 2022 list in comparison to the 2012 list, filtering by geographic location and decade of release.  My driving research questions included what kinds of films are maintaining dominance in the film canon, what kinds of films are losing dominance, and what kinds of films get to be added to the canon.

My project begins on a landing page that explains the premise of the Sight and Sound poll and the goals of my visualization, as I have just done in this paper.  Down the side of the landing page is a very basic visualization of the entire list of greatest films in order.  I wanted a way for my audience to explore the list in the same way someone first seeing the list published on Sight and Sound’s website would, but without having to leave my visualization.  I think this is especially helpful to people who have never seen the 2022 list and who may want to first explore it without any filters or any conclusions being drawn.  The films are represented by dots and are arranged in order of ranking.  The tooltip tells you the name of the film, the year it was made, its director, its country of origin, and its ranking on the 2022 list.

From the landing page, my project consists of five different pages of visualizations.  The first page is called “2022 list overall”, and goes about summarizing the makeup of the list by both country of origin and decade produced.  The top section of this page focuses on country of origin.  The first visualization is a world map that uses color density to represent the number of films from each country that are on the list.  While the tooltips do tell you how many films from each country are on the list, I think the usefulness of the world map lies in showing trends on a continental scale.  For example, the world map illustrates just how dominant Europe is on the list, with almost every European country being represented, while Africa has only one country represented, and South America none.  Because frontrunners USA and France have so many more films on the list than any other country, I initially found that the color density array was not giving me what I wanted in terms of color. USA and France were showing up very dark, but every other country was a very similar shade of light red.  As such, I adjusted the scale of the color legend so that the darkest red represented about half of America’s total.  This way, countries like Italy, Japan, and the USSR, who have far less films than USA and France, but still a relatively high amount compared to other countries, would be colored dark enough on the map.

The graph underneath the world map uses the same data and illustrates a similar idea, but it breaks films out into individual dots which gives a better idea of the number of films on the list from each geographic area than the less accurate color density.  This graph also organizes the films by ranking on the y-axis, so you can start to get an idea of where films from each area are more likely to appear on the list.  For example, this graph allows you to see that USA’s films are concentrated at the top half of the list, with many films in the top 20, whereas France, although highly represented, is far less represented in the top 20.  I gave USA, France, Italy, USSR, and Japan their own columns, because they are the countries with at least five films on the list, but I could not give each country its own column or else the graph would become too spread out.  So, I grouped the rest of Europe, excluding France, Italy, and the USSR into one column, and the rest of Asia, Africa, and Oceania, excluding Japan, into one column.  

Continuing the breakdown by country of origin, the next graphs visualize the changes between the 2022 and the 2012 list.  This section consists of two stacked bar charts.  The top one illustrates films that rose in rank between the 2012 and 2022 lists, shown in green, and films that dropped in rank, shown in red.  The bottom one illustrates films that were on the the 2022 list but not on the 2012 list, shown in blue, and films that were on the 2012 list but not on the 2022 list, shown in dark red, as well as films on the 2022 list that were released between 2012 and 2022 and thus could not have been on the 2012 list, shown in light blue.  Changes between the 2012 and 2022 lists are an important part of my driving questions to establish how the film canon is changing over time.

The bottom half of this page mostly repeats the graphs from the country of origin section, but focuses instead on the decade each film was made.  The world map is replaced with a bar chart, obviously because decades cannot be represented on a map, but the idea remains that the bar chart shows the overall trend of which decades are most represented, while the graph beneath it adds the dimension of ranking, illustrating that newer films are more likely to appear on the bottom half of the list.  I chose to break the changes in rank into two different bar charts, because having six different colored sections in each stacked bar was too much to handle when looking at every film on the list and made it harder to identify the trends. 

Back to the landing page, I also created a page of visualizations for four different geographic areas that I could identify distinct trends within (the United States, France, Europe, and Asia, Africa, and Oceania). The visualizations on these pages only include films from the determined country/continent(s).  The first graph on each of these pages is similar to the dots by ranking graphs on the overall page, using a dot to represent each movie and placing rank on the y-axis and year produced on the x-axis.  I added text that identifies trends in ranking and also brings in director, identifying certain directors who are particularly dominant in certain decades.  Next is a bar chart illustrating the number of films from that country/continent by decade they were released.  Again, the bar chart quickly illustrates the overall trend in decades while the dots by ranking elaborates into where specific films fall on the list.  

Like the overall page, the bottom half of each geographic page focuses on changes between 2012 and 2022.  First, I used a pie chart to identify how many films from each area rose in rank, dropped in rank, were new to the list, and were dropped from the list.  The pie chart allows you to see generally if more films rose or dropped in rank, which helps determine how the area is faring in the film canon.  Underneath the pie chart is a stacked bar chart that uses the same colors and illustrates the same idea, but also allows a look into specific films, as each film receives its own tooltip.  I thought it was more appropriate in these pages to combine all six colors into one stacked bar chart, because less films are being looked at, and if I broke it into two bar charts, some bars would be very small or nonexistent.  The stacked bar chart also breaks the change in rank down by decade, so you can see changes by decade within the country/continent.  Originally, I wanted to make separate pages for geographic areas and for decades, but I decided that since I was able to tell a story about decades within the country/continent pages, that I did not need separate pages for decades.  

As I was creating this visualization, I found I was able to make very pointed observations that answer my driving questions. As such, I focused on telling my audience exactly what I see in the data, rather than creating a more exploratory visualization where the audience draws their own conclusions.  That is why I also used quite a lot of text to summarize the takeaways from each graph.  Overall, I illustrated a cohesive story about how the countries I termed the powerhouse countries, the countries with the most representation on the list, are actually the countries that are losing ground in the film canon to make way for emerging countries that are beginning to establish their filmmaking presence.  Similarly, in terms of production year, older films are losing ground in the film canon to make way for more recent films.  This is something I was not able to identify from just looking over the list.  I needed to do an in depth analysis of the changes between 2012 and 2022 to illuminate that the list is actually, in my opinion, moving in the right direction toward representation of less traditionally represented countries and representation of more modern film.  I think these takeaways would be of great interest to people who, like me, have a large amount of film knowledge, but I also think that I presented the information in a digestible enough way to be of interest to people with less film knowledge, as well.