The Complete Sight and Sound Greatest Films of All Time Database

*48/*6&23*480'&7 02, *48/*6&23*480'&7 02, 

 $"%&.*$02,3 $"%&.*$02,3

*33&24"4*0/3)&3&3"/%"1340/&20+&$43  2"%5"4&&/4&2



)&0.1-&4&*()4"/%05/%2&"4&34*-.30'--*.&)&0.1-&4&*()4"/%05/%2&"4&34*-.30'--*.&

"4"#"3&"4"#"3&

"4*&0/*"

)&2"%5"4&&/4&2*48/*6&23*480'&7 02,

07%0&3"$$&33404)*3702,#&/&94805&453,/07

02&*/'02."4*0/"#0544)*3702,"4)4413"$"%&.*$702,3$5/8&%5($!&4%3

*3$06&2"%%*4*0/"-702,3"4)4413"$"%&.*$702,3$5/8&%5

)*3702,*3."%&15#-*$-8"6"*-"#-&#84)&*48/*6&23*480'&7 02, 

0/4"$4$"%&.*$02,3$5/8&%5

THE COMPLETE SIGHT AND SOUND GREATEST

FILMS OF ALL TIME DATABASE

KATIE DONIA

A capstone project submitted to the Graduate Faculty in Data Analysis and Visualization in

partial fulfillment of the requirements for the degree of Master of Science, The City University

of New York

2024

KATIE DONIA

iii

The Complete Sight and Sound Greatest Films of All Time Database

Katie Donia

This manuscript has been read and accepted by the Graduate Faculty in Data Analysis

and Visualization in satisfaction of the capstone project requirement for the degree of

Master of Science.

_________________________ __________________________________________

Date Ellie Frymire

Thesis Advisor

_________________________

Date

_________________________________________

Matthew K. Gold

Program Director

THE CITY UNIVERSITY OF NEW YORK

ABSTRACT

The Complete Sight and Sound Greatest Films of All Time Database

Katie Donia

Advisor: Ellie Frymire

Once every ten years, Sight and Sound, a film magazine produced by the British Film

Institute (BFI), conducts an international poll of critics and film journalists to determine the

“Greatest Films of All Time.” The first poll was conducted in 1952, and ten years later the BFI

decided to make their poll decennial, allowing for the list to change and thus to demonstrate how

time, changing opinion, and the continuous creation of new films change the film

canon. However, despite the opportunity for long-term analysis, there is not an easy way online

to see how a film’s ranking has changed between polls outside of opening lists in multiple

tabs. To combat this, the goal of my project is to increase the accessibility of data pertaining to

all iterations of the Sight and Sound Greatest Films of All Time Poll by creating a centralized

database. This project includes the transcription of every ten-film ballot published by Sight and

Sound for all 8 iterations of the poll, as well as the calculated number of votes and rank for each

film, organized into a workbook. This project also includes an interactive website that displays

each film’s biographical information and ranking in every poll and allows for filtering and

sorting. Together, these deliverables allow for a previously unprecedented ability to analyze

changes over time to the Sight and Sound poll and thus the film canon.

TABLE OF CONTENTS

List of Figures……………………………………………………………………………vi

Digital Manifest…………………………………………………………………………vii

Technical Specifications…………………………………………………...…………..viii

Narrative………………………..…………………………………………….…………..1

Background and Objective……………………………………...……………..…………1

Data Collection………………………………………………………………….…………3

Website Development………………………………………………………….…………..7

Evaluation and Future Steps………………………………………………..…..………..9

Appendix A: List of Variables………………………………………………………….12

Appendix B: Glossary of Functions……………………………….…………….……..14

Appendix C: Figures……………………………………………………………..……..16

Reference List………………………………….....……………………………………..19

LIST OF FIGURES

Figure 1: Early sketch of website design and functionality………………………….16

Figure 2: Early sketch of proposed expanded film slide…………………………….17

Figure 3: Finished look at website functionality……………………………………...18

vii

DIGITAL MANIFEST

1. Capstone White Paper (PDF)

Code

Files for the webapp are stored inside a GitHub repository:

https://github.com/bondie00/parcel-react

viii

TECHNICAL SPECIFICATIONS

The code for this project was written using the JavaScript framework React. No

additional libraries were used. The code is stored in a public GitHub repository and the website

is hosted on GitHub Pages and Vercel. The site depends on data resources stored in a JSON file

in the code repository. The data resources contained in the JSON file were copied from the main

data tab of the Google Sheets workbook which stores all the data transcribed and calculated for

this project.

NARRATIVE

Background and Objective

Once every ten years, Sight and Sound, a film magazine produced by the British Film

Institute (BFI), conducts an international poll of critics and film journalists to determine the

“Greatest Films of All Time” [1]-[8]. Each selected voter submits a ballot of ten films they think

are the greatest of all time. The ballots are aggregated, and a list is created by ranking the films

by number of votes received. The first poll was conducted in 1952, and ten years later the BFI

decided to make their poll decennial, allowing for the list to change and thus to demonstrate how

time, changing opinion, and the continuous creation of new films change the film canon.

From 1952 to 1992, the results of the poll were published in Sight and Sound as a top ten

list plus runners-up and accompanied by a large selection of the received ballots [1]-

[5]. Although the internet did not exist or barely existed for these polls, the top ten plus runners-

up lists can currently be found on the current Sight and Sound website [9]-[13]. However, these

lists of around 12 to 25 of the most voted for films only make up a small portion of the complete

set of every film that received at least one vote that year. The only way to find the films that

were voted for but did not make the official list is to go directly to the published ballots. This

requires looking through the original magazine publications, which can be digitally accessed via

Sight and Sound’s digital archive for the cost of a yearly subscription [14]. At this level,

however, it is not clear how many votes each film received. There are some hobbyist websites

that attempt to calculate the number of votes per film based on published ballots [15][16], but

those calculations have never been done in a transparent way with data to back it up before.

As the BFI moved into the digital age, it became easier to reach out to more and more

voters, and so the number of ballots, and accordingly the number of films that received votes,

increased dramatically throughout the 2002, 2012, and 2022 polls. Sight and Sound expanded

their lists out to the top 100 films in 2012 and the top 250 in 2022. Additionally, results were

published physically in Sight and Sound magazine but also digitally on the Sight and Sound

website. Suddenly, every ballot could be published online without the worry of taking up

physical page space. The 2002, 2012, and 2022 websites were fairly robust [18]-[20]. On the

2002 and 2012 websites, users could access an alphabetical list of every film voted for and click

on the titles to see who voted for them [21][22]. This was the first website that could display

how many votes each and every film received without having to manually tally it up. However,

the robust 2002 and 2012 websites no longer exist; the 2012 poll replaced the 2002 poll’s place

on the website, as did the 2022 poll to the 2012 poll. All that exists online for 2002 and 2012

now are a top ten list for 2002 and a top 100 list for 2012 that mirror how the earlier polls are

presented [17][23]. The data for the 2022 poll that is currently online does include all ballots,

but the only films you can click on to see how many total votes they received are the ones in the

top 250 [24]. Still, there is no way to see how many votes each and every film received.

Along with these barriers to data accessibility within each poll, there is absolutely no way

to quickly answer questions about how the ranked lists have changed over time, even though that

is a stated function of the poll. For example, if someone is looking at the 2022 top 250 list and

wants to know where the current first place film ranked in 2012 or any past poll, their only

option is to open another tab, search for the past poll, and then scroll through the list until they

find the film they are looking for and can compare rankings. There also has never been a way to

apply filters to the lists that would allow for a closer analysis of the film canon in terms of

release years, countries of origin, and more, for example, if someone is interested specifically in

how French films rank versus American films.

These barriers illustrate that before being able to answer questions about how the Sight

and Sound poll has changed over time, there needs to be better access and better organization to

the underlying data. As such, this project became focused on creating a centralized database

with the most complete look at all Sight and Sound data possible. The project’s deliverables

include a Google Sheets workbook that contains a transcription of every ten-film ballot

submitted to and published by Sight and Sound for all 8 iterations of the poll, as well as a main

data sheet that lists every film that has ever been voted for along with its release year, director,

country of origin, and number of votes and rank in each poll iteration. This main data sheet

feeds the other half of my project, an interactive website that displays the films along with their

biographical information and their rankings in one centralized location, and includes a robust

sorting and filtering feature.

Data Collection

The first task was to transcribe a very large number of ballots. From 1952 to 1992, the

number of ballots received climbed steadily from around 60 to around 150, but shot up to almost

850 in 2012 and over 1600 in 2022. I explored using a web scraper to do the transcription work

for me, but ultimately chose not to; I decided the time needed to learn how to build or use a web

scraper would be better spent learning web development while manually transcribing. In

addition, as previously stated, the only way to access the individual ballots from 1952 to 1992

are via scanned and digitized versions of Sight and Sound magazine, from which it would be

very difficult to web scrape. I did reach out to the BFI via a research form submission on their

website to ask about how the official Sight and Sound poll data is stored and if it is available for

use, but did not hear back.

The first step was to enter by hand each ballot, including the voter’s name, country, and

the ten films they voted for, into a row on a Google Sheets spreadsheet, with a separate sheet for

each poll iteration. I quickly noticed that from poll to poll, decade to decade, Sight and Sound

was changing the way they formatted the ballots and film titles. For example, in the earliest

polls, Sight and Sound would typically list titles for European films in their original language,

but over time it became more common to refer to them by their English titles. Titles would need

to be standardized, but I decided that should be done after the transcription phase to ensure

standardization across all polls.

For the 1952 to 1992 polls, the Sight and Sound digital archives provided my source of

ballot data [1]-[5]. Because it was stated in the magazine how many ballots were received for

each poll, it became evident that my data for 1952, 1962, and 1972 was missing between 9% and

35% of ballots per poll. This is an issue with my data that cannot be fixed given my current

access to Sight and Sound data. For these polls, I have used the published number of votes for

the top ten plus runners-ups, because those are formally published numbers, although they do not

exactly match the counts calculated in my data workbook due to the missing ballots. I also found

that, even for the polls with all ballots accounted for, the counts calculated in my data workbook

did not always match what was published by Sight and Sound. I can only speculate why that

might be, because Sight and Sound’s behind-the-scenes data staging and tallying processes are

mostly a black box. For the sake of trusting my own processes and having transparent data, I

have used the number of votes and ranked order that my spreadsheets produce for the 1982

through 2022 polls, even if they differ from the official Sight and Sound lists.

To access digital versions of the 2002 and 2012 ballots, I used the wayback machine to

retrieve past versions of the Sight and Sound website [19][20]. The only poll I did not transcribe

by hand was the enormous 2022 poll. I was fortunate to find that other cinephiles have already

done the work of web scraping the most recent 2022 poll. A quick search on Reddit revealed

several spreadsheets that had been posted for public use [26]. I chose the one that matched my

existing format closest [25], but still had to do some cleaning to ensure all the ballots were

transcribed.

Once every poll had been transcribed onto its own sheet, total votes for each film for each

poll were calculated. To do this, a combination of the FILTER and FLATTEN functions were

used to transform the data into a usable shape, the UNIQUE function was used to list every film

title once, and the COUNTIF function was used to count the number of times each title occurs in

the transcribed ballots. After applying this to each sheet, I had complete versions of every poll;

not just top tens or top 100s, but exhaustive lists of every film voted for and how many votes it

received in each poll.

These lists were combined into one sheet such that every film voted for in each poll has

its own row. For example, Citizen Kane, which has appeared on every single Sight and Sound

poll, has eight rows on the all votes tab, each containing a poll year and the number of votes it

received that year. Sorting this sheet by film title starts to achieve this project’s goal of

conveying a film’s change in rank in one glance. This sheet was then used to fill out the main

data tab by hand. On the main data tab, unlike the all votes tab, each film has only one row, and

the ranking and number of votes in each poll make up sixteen different columns. This structure,

in which each film is essentially an object with many different characteristics, allows the data to

be transformed to JSON and for my website to use the JSON data to display a slide for each film

containing all of its information.

The transition from the all votes tab to the main data tab is where all of the data cleaning

and standardization took place, particularly regarding titles. The task here was to go through the

all votes tab alphabetically, with each new film getting a new row in the main data tab, making

sure to standardize the title in the process. In some unique cases, foreign films do not have

English titles, and are most commonly referred to by their original titles even in America. In

most cases, however, foreign films have English titles that are sometimes used while others elect

to use the foreign titles, as was the case with Sight and Sound. The problem is that deciding

which title to use is mostly a preference choice, and will be different even among different

publications. The same non-consensus can be true of a film’s release year and country of origin,

details that were also being standardized onto the main data tab. I decided to pick a source of

film information and abide by the decisions they made. The streaming service Mubi turned out

to be the best source of information about these films [27]. Nearly every film listed, even the

obscure art shorts, of which there are many in the database, have official Mubi pages that list the

English title, an alternate original language title if applicable, director, release year, country of

origin, and even genres. Mubi was used as the reference for film title, alternate title, director,

and year in my database.

The only information I reliably wanted another source for was country of origin. Out of

all the biographical information used by my database, country of origin is the least

standardized. There are many ways one might determine a film’s country of origin: where it was

shot, where it is set, the director’s nationality, the location of the production company that

funded it. All of these play a role in a film’s perceived country of origin. For example, spaghetti

westerns of the 1960s were shot in Italy by Italian directors, but set in the American West

starring American stars speaking English, and were probably financed by both the United States

and Italy, if not other European countries as well. Some films such as these are clearly split in

nationality, but as often as possible I wanted to narrow down the country of origin down to one

country so that analysis would be clearer. Mubi is prone to listing two or more countries or

origin, especially for independent films or films from poorer nations that often need to find

financing from multiple countries. When Mubi listed more than one country of origin, I would

consult Wikipedia to get an idea of the most perceived country or origin [28]. For example, if in

the opening lines of the film’s wikipedia article, it was called a “blank nationality” film, as in

“L’Atalante, an Italian film directed by…” then I chose to use that country as the country of

origin even if Mubi listed more. These are the data cleaning steps that were performed for each

and every film in my workbook in order to prepare the data for my website.

Website Development

From the first capstone brainstorm session, I wanted to do a project that incorporated all

of the Sight and Sound polls back to 1952. I had done a previous class project about the 2022

Sight and Sound poll [29]. That project, made in Tableau, used interactive charts and graphs to

tell a story about how the Sight and Sound poll looked in 2022 and how it had changed since

2012 (using the number of ranks a film increased or decreased between 2012 and 2022 as a

variable that I calculated by hand for each of the top 100 films). I was interested in how that

project could be expanded to tell a story about not only the changes from 2012 to 2022, but

across the history of the poll and thus the history of the film canon. The first thing I did was

gather the ranked lists from each earlier poll, which, as stated previously, was only around 20

films for each poll between 1952 and 2002. I was not happy with the mismatch in the amount of

data I had access to. Although it is true that there is an inherent dimensionality mismatch as

more and more ballots are submitted to each successive poll, I was not satisfied with having a

pool of 250 films from 2022 to compare to only 20 films from 1952, especially when the

comparison of ranks over time was my driving intention. I quickly realized I would need to

transcribe the data and calculate ranks myself, starting from the ballot level.

Figures 1 and 2 show that by the time I submitted my proposal, I had already decided the

website would be an exploratory database rather than a narrative story. During a class on play in

the digital humanities, I was reassured of the power of an exploratory database by Ian Bogost’s

Play Anything [30]. Bogost posits that anything can be played or had fun with if approached

with the right mindset, and that, as such, there are playgrounds all around us if we are willing to

engage with them. An interactive database is a digital playground waiting to be played with. It

may not look like fun on the surface, but if, as Bogost suggests, the database’s structures are

thoughtfully engaged with, it can contain hours of entertainment and answers to countless

different questions. An exploratory database can be just as much fun as a narrative story and

encourages deeper interaction with the data through personal manipulation.

Once I had the general idea for my website, I was ready to start coding my project in

JavaScript using React as the framework [31]. Components are a defining feature of

React. Components allow a section of html and associated coding to be reused as a template in

order to repeat structures with potentially different data. As every film would be getting its own

slide on my website, the slides were an obvious candidate for a component. As a component,

every film and its associated ranks can be displayed by looping through the main data array. The

film slides were also an introduction to working with divs. To display the ranks in a row on each

slide, I needed to use flexbox to force divs into rows rather than columns and to align the text

exactly how I wanted it.

As I continued to design, it was not always clear which parts of the code should or should

not be in components. It can come down to personal choice, but I discovered that it was often

necessary to use a component even when I initially thought otherwise. For example, the filters

work by setting and using useStates which are applied when the apply filters button is

clicked. However, the array that contains the film results is not in a useState, and instead is

calculated and populated each time any useState is set. This meant that anytime any other

useState was triggered it would refresh the list of films. This was a big problem, because I was

using a lot of useState triggers, even to display and hide the filter choices, so every time a filter

category was expanded or closed, it would apply the filters before the apply button was

clicked. I realized that any part of the website that needed to be able to change without changing

the film results would need to be in a component to avoid competing useStates. As a result, the

code contains eight components, even though not many of the components show up multiple

times on the website.

Evaluation and Future Steps

This project was successful on multiple fronts. Although my deliverables do not

obviously spell out an answer to how the film canon has changed over time, they do provide the

raw numbers that can inform other researchers' writings in the field of film and the development

of the film canon. My data workbook is, to my knowledge, the only source of calculated vote

counts for every film on every Sight and Sound poll, particularly with data to back it up, and I

am proud to offer it publicly for anyone to use in future projects. Furthermore, my website is a

compelling artifact that displays a film’s change in rank over the course of the poll in one

centralized location, which was the driving intention of the project. And, although not a stated

intention of the project, I took my React and JavaScript skills from nearly nothing to a level of

competency with an intense desire to continue improving.

A shortcoming of this project is not achieving all of the visual goals for the website. My

original sketches include a line chart of rank over time for each film, which would render a

film’s change in rank visual and trends easier to identify than in raw numbers alone. I also

wanted to include a movie poster on each slide, and to display each film’s title in the font used

on the movie poster or on the film’s title card. These visual elements were meant to reveal

something about the films themselves, and without any visual elements, the database lacks

personality, even though the films it contains are brimming with personality. The largest hurdle

to implementing visual elements is the time it would take to collect images for every single film

in the database, but there may be a better way to implement images.

Another shortcoming of this project is that the website does not do a good job of

displaying the shape of the data. By using the filters and seeing how the number of films

changes, some of the shape regarding which decades or countries produce the most films with

the most votes can be gleaned. But, the shape of the data would be better served by histograms

and other charts that visualize these aspects of the data. As such, the ultimate goal for my

website would be to combine aspects of both exploratory and narrative data visualization by

incorporating charts onto the film slides and histograms into the filters.

I have a strong desire to continue working on this project, and I likely will, as I have

already laid out improvements I want to make. Eventually, I would like to launch future projects

that use the Sight and Sound polls to tell stories about how the film canon has changed, but I am

very satisfied with the result that I have laid the groundwork necessary to do so, and in the

process created a resource that anyone can use to tell these stories, too. My desire to continue

this project is entirely due to the success I feel I have already found.

APPENDIX A: LIST OF VARIABLES

App.js

currentPage - useState containing current page number

moviesPerPage - useState that sets number of films per page

poll - useState containing current poll year selected

sort - useState containing current sort method selected

countryFilters - useState containing currently applied country filters

yearFilters - useState containing currently applied year filters

directorFilters - useState containing currently applied director filters

titleFilters - useState containing currently applied title filters

prevStartYear - useState containing previously selected start year filter, to populate search bar

prevEndYear - useState containing previously selected end year filter, to populate search bar

nextState - object that collects filter selections before clicking the sort and filter button

yearData - list of results after poll year selection is applied

filteredData - list of results after all filters have been applied

indexLast - number of last result shown on the page

indexFirst - number of first result shown on the page

indexCurrent - list of results to display within the range of the page

CountryFilters.jsx

show - useState that shows or hides country filters

selected - useState containing currently selected countries, for use by checkboxes

Countries - list of country names to display as filter choices, organized by continent

Filters.jsx

clickedCOI - useState for expanding country of origin filters

clickedYears - useState for expanding release year filters

clickedDirector - useState for expanding director filters

clickedTitle - useState for expanding title filters

state - useState to force refresh, used on clear click

sortSel - useState containing current sort selection, to populate the dropdown menu

Header.jsx

show - useState for expanding header information

selection - useState containing information subheading selection

Pagination.jsx

pageNumbers - array containing all page numbers listed out

displayNumbers - array containing page numbers that will be displayed as buttons

SearchBar.jsx

clickedSearch - useState for displaying dropdown selection list under search bar

searchTerm - useState containing current search term, to populate search bar

Selection.jsx

show - useState that determines whether a director or title selection is displayed above the search

bar

YearFilters.jsx

searchTerm - useState containing the inputted start year filter, to populate search bar

searchTerm2 - useState containing the inputted end year filter, to populate search bar

APPENDIX B: GLOSSARY OF FUNCTIONS

App.js

applyClick() - on sort and filter button click, applies filters by setting filter-related useStates

according to contents of nextState

applyPoll(year) - on poll selection click, displays only the films from the selected poll year by

setting the poll useState

paginate(pageNumber) - on page button click, navigate to that page by setting currentPage

useState to clicked page number

countryFilters.jsx

countryHandler(event) - add selected countries to and remove unselected countries from list of

filters for nextState

click() - on continent click, show associated country filters

selectAll(event) - on continent checkbox, select all associated country checkboxes

Filters.jsx

toggleCountries() - on country of origin title click, expand area by setting clickedCOI useState

toggleYears() - on release year title click, expand area by setting clickedYears useState

toggleDirector() - on director title click, expand area by setting clickedDirector useState

toggleTitle() - on title title click, expand area by setting clickedTitle useState

handleSort(e) - add sort type selection to list of filters for nextState

clearFilters(code) - remove filters from list of filters for nextState depending on which clear

button was clicked

Header.jsx

toggleShow() - on more information click, expand are by setting show useState

SearchBar.jsx

clickHandler(event) - update search bar text by setting current search term useState

searchBarClick() - on search bar click, display dropdown selection list by setting clickedSearch

useState

useOutsideClick(event) - track a click that is outside the dropdown selection list to collapse the

dropdown selection list

select(event) - on dropdown selection list content click, add director or title to list of filters for

nextState

Selection.jsx

unselect(event) - on click of previously selected director or title, remove director or title from list

of filters for nextState and stop displaying above search bar by setting show useState

YearFilters.jsx

handleStart(event) - add inputted start year to list of filters for nextState

handleEnd(event) - add inputted end year to list of filters for nextState

APPENDIX C: FIGURES

Figure 1: Early sketch of website design and functionality

Figure 2: Early sketch of proposed expanded film slide

Figure 3: Finished look at website functionality

REFERENCE LIST

[1] British Film Institute, “As the Critics Like It,” Sight and Sound, vol. 22, no. 2, pp. 58-60, 1952.

[Online]. Available: Sight and Sound Digital, sightandsounddigital.bfi.org.uk.

[2] British Film Institute, “Top Ten,” Sight and Sound, vol. 31, no. 1, pp. 10-14, 1962. [Online].

Available: Sight and Sound Digital, sightandsounddigital.bfi.org.uk.

[3] British Film Institute, “Top Ten 1972,” Sight and Sound, vol. 41, no. 1, pp. 12-16, 1972. [Online].

Available: Sight and Sound Digital, sightandsounddigital.bfi.org.uk.

[4] British Film Institute, “The Top Ten 1982,” Sight and Sound, vol. 51, no. 4, pp. 243-246, 1982.

[Online]. Available: Sight and Sound Digital, sightandsounddigital.bfi.org.uk.

[5] British Film Institute, “Top Ten: 250 Verdicts,” Sight and Sound, vol. 2, no. 8, pp. 18-30, 1992. Sight

and Sound Digital.

[6] British Film Institute, “The Ten Best Movies of All Time,” Sight and Sound, vol. 12, no. 9, pp. 24-

50, 2002. [Online]. Available: Sight and Sound Digital, sightandsounddigital.bfi.org.uk.

[7] British Film Institute, “The Greatest Films of All Time,” Sight and Sound, vol. 22, no. 9, pp. 39-71,

2012. [Online]. Available: Sight and Sound Digital, sightandsounddigital.bfi.org.uk.

[8] British Film Institute, “The Greatest Films of All Time,” Sight and Sound, vol. 33, no. 1, pp. 27-112,

2022. [Online]. Available: Sight and Sound Digital, sightandsounddigital.bfi.org.uk.

[9] British Film Institute, “The Greatest Films of All Time… in 1952,” BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/polls/greatest-films-all-time/1952

[10] British Film Institute, “The Greatest Films of All Time… in 1962,” BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/polls/greatest-films-all-time/1962

[11] British Film Institute, “The Greatest Films of All Time… in 1972,” BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/polls/greatest-films-all-time/1972

[12] British Film Institute, “The Greatest Films of All Time… in 1982,” BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/polls/greatest-films-all-time/1982

[13] British Film Institute, “The Greatest Films of All Time… in 1992,” BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/polls/greatest-films-all-time/1992

[14] British Film Institute, “Digital Addition and Archive,” Sight and Sound. [Online]. Available:

https://sightandsounddigital.bfi.org.uk

[15] “Why Only a Top 10? 2002 Sight & Sound Top Films Survey of 253 International Critics & Film

Directors,” Cinemacon. [Online]. Available: https://www.cinemacom.com/2002-sight-

sound.html

[16] “Sight & Sound Top Films Critics Surveys 1952 - 1992,” Cinemacon. [Online]. Available:

https://www.cinemacom.com/sight-and-sound.html

[17] British Film Institute, “The 100 Greatest Films of All Time 2012,” BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/polls/greatest-films-all-time-2012

[18] British Film Institute, “The Greatest Films of All Time”, BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/greatest-films-all-time

[19] British Film Institute, “Sight and Sound Top Ten Poll 2002,” BFI. [Online]. Available:

https://web.archive.org/web/20030819195634/http://www.bfi.org.uk/sightandsound/topten/poll/.

[20] British Film Institute, “Analysis: The Greatest Films of All Time 2012.” BFI. [Online]. Available:

https://web.archive.org/web/20191223235153/https://www.bfi.org.uk/films-tv-

people/sightandsoundpoll2012.

[21] British Film Institute, “Sight and Sound Top Ten Poll 2002: Who Voted?,” BFI. [Online].

Available:

https://web.archive.org/web/20031003172036/http://www.bfi.org.uk/sightandsound/topten/poll/li

st.php?list=films

[22] British Film Institute, “Analysis: The Greatest Films of All Time 2012: All Films.” BFI. [Online].

Available: https://web.archive.org/web/20190704184435/https://www.bfi.org.uk/films-tv-

people/sightandsoundpoll2012/films

[23] British Film Institute, “The Greatest Films of All Time… in 2002,” BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/polls/greatest-films-all-time/2002

[24] British Film Institute, “The Greatest Films of All Time: All voters”, BFI. [Online]. Available:

https://www.bfi.org.uk/sight-and-sound/greatest-films-all-time/all-voters

[25] Project Parallax, “S&S 2022 Analysis,” datasheet. [Online]. Available:

https://docs.google.com/spreadsheets/d/1F6HMntP2scCHhAvFc77iHONhf4U0cKkXQY3zrfVZ

nXE/edit

[26] “Detailed Information on the Sight & Sound 2022 Results,” Reddit. [Online]. Available:

https://www.reddit.com/r/TrueFilm/comments/11k7po7/detailed_information_on_the_sight_sou

nd_2022/

[27] “MUBI: Watch and Discover Movies,” Mubi. [Online]. Available: https://mubi.com/en/us

[28] “Wikipedia,” Wikipedia. [Online]. Available: https://www.wikipedia.org/

[29] K. Donia, “The 2022 Sight and Sound Poll, Visualized,” Tableau Public. [Online]. Available:

https://public.tableau.com/views/SightandSound-PublicVersion/HomePage?:language=en-

US&:display_count=n&:origin=viz_share_link

[30] I. Bogost, “Playgrounds, Everywhere” in Play Anything: The Pleasure of Limits, the Uses of

Boredom, and the Secret of Games. Basic Books, 2016, pp. 1-27.

[31] React. “API Reference.” React. [Online]. Available: https://react.dev/reference/react.