Mini Project 02: The Business of Show Business

Introduction

In this project, I will be utilizing data from the Internet Movie Database (IMDb) to discover what the next best movie is to produce. Before settling on my decision, I will dive into Hollywood’s history to identify key characteristics of successful movies, identify successful directors and actors, and examine some of Hollywood’s most famous flops. With all this information compiled together, I will pick a crew and a movie to remake, and then pitch the idea to the higher ups at my company.

Diving into the Data

Firstly, I must download the data of Hollywood’s history into my rstudio. Since there were a sizable amount of people recorded in the industry, I decided to restrict my attention to people with at least two “known for” credits. This all can be seen in the folded code below:

Click here to see how the data was downloaded 
library(tidyverse)
── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.4     ✔ readr     2.1.5
✔ forcats   1.0.0     ✔ stringr   1.5.1
✔ ggplot2   3.5.1     ✔ tibble    3.2.1
✔ lubridate 1.9.3     ✔ tidyr     1.3.1
✔ purrr     1.0.2     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
library(dplyr)


get_imdb_file <- function(fname){
  BASE_URL <- "https://datasets.imdbws.com/"
  fname_ext <- paste0(fname, ".tsv.gz")
  if(!file.exists(fname_ext)){
    FILE_URL <- paste0(BASE_URL, fname_ext)
    download.file(FILE_URL, 
                  destfile = fname_ext)
  }
  as.data.frame(readr::read_tsv(fname_ext, lazy=FALSE))
}


NAME_BASICS      <- get_imdb_file("name.basics")
Rows: 771476 Columns: 6
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (6): nconst, primaryName, birthYear, deathYear, primaryProfession, known...

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
TITLE_BASICS     <- get_imdb_file("title.basics") 
Warning: One or more parsing issues, call `problems()` on your data frame for details,
e.g.:
  dat <- vroom(...)
  problems(dat)
Rows: 4729628 Columns: 9
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (8): tconst, titleType, primaryTitle, originalTitle, startYear, endYear,...
dbl (1): isAdult

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
TITLE_EPISODES   <- get_imdb_file("title.episode") 
Rows: 8585773 Columns: 4
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (4): tconst, parentTconst, seasonNumber, episodeNumber

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
TITLE_RATINGS    <- get_imdb_file("title.ratings") 
Rows: 1490875 Columns: 3
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (1): tconst
dbl (2): averageRating, numVotes

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
TITLE_CREW       <- get_imdb_file("title.crew") 
Rows: 10521826 Columns: 3
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr (3): tconst, directors, writers

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.
NAME_BASICS <- NAME_BASICS |> 
    filter(str_count(knownForTitles, ",") > 1)

Now that I restricted the worker’s data set, I am curious to see how many obscure movies are in my data, as I want to avoid creating a low selling movie. Below is a chart that depicts the number of titles in comparison to their number of IMDb ratings.

TITLE_RATINGS |>
    ggplot(aes(x=numVotes)) + 
    geom_histogram(bins=30) +
    xlab("Number of IMDB Ratings") + 
    ylab("Number of Titles") + 
    ggtitle("Majority of IMDB Titles Have Less than 100 Ratings") + 
    theme_bw() + 
    scale_x_log10(label=scales::comma) + 
    scale_y_continuous(label=scales::comma)

Here, we can see the a majority of the movies in this data set have less than 100 ratings. To limit my data set further, I threw out any title that had less than 100 ratings. Looking at the quarterly ranges below, this shows that this drops about 75% of our data.

TITLE_RATINGS |>
    pull(numVotes) |>
    quantile()
     0%     25%     50%     75%    100% 
      5      11      26     100 2954275

So, I performed this drop by performing the following code:

TITLE_RATINGS <- TITLE_RATINGS |>
    filter(numVotes >= 100)

We now want to do the same for our “Title” tables, which can be seen below.

Click here to see how this was done 
TITLE_BASICS <- TITLE_BASICS |>
    semi_join(TITLE_RATINGS, 
              join_by(tconst == tconst))

TITLE_CREW <- TITLE_CREW |>
    semi_join(TITLE_RATINGS, 
              join_by(tconst == tconst))

TITLE_EPISODES_1 <- TITLE_EPISODES |>
    semi_join(TITLE_RATINGS, 
              join_by(tconst == tconst))
TITLE_EPISODES_2 <- TITLE_EPISODES |>
    semi_join(TITLE_RATINGS, 
              join_by(parentTconst == tconst))

TITLE_EPISODES <- bind_rows(TITLE_EPISODES_1,
                            TITLE_EPISODES_2) |>
    distinct()


rm(TITLE_EPISODES_1)
rm(TITLE_EPISODES_2)

#| cache: false

Now that our data has been significantly reduced, we can move on!

Task 1: Column Type Correction

After performing the glimpse function to examine each table, it was clear that some columns appeared as “character (string)” vectors, when they should be numeric. This is due to the null or N/A values that appears in the data sets. Since R cannot read these files as is, we must mutate the tables in a way that R can read them numerically. This is done with the following code:

Click here to see how this was done 
library(dplyr)

NAME_BASICS <- NAME_BASICS |>
  mutate(birthYear = as.numeric(birthYear),
         deathYear = as.numeric(deathYear))
Warning: There were 2 warnings in `mutate()`.
The first warning was:
ℹ In argument: `birthYear = as.numeric(birthYear)`.
Caused by warning:
! NAs introduced by coercion
ℹ Run `dplyr::last_dplyr_warnings()` to see the 1 remaining warning.
TITLE_BASICS <- TITLE_BASICS |>
  mutate(isAdult = as.logical(isAdult),
         endYear = as.numeric(endYear),
         startYear = as.numeric(startYear),
         runtimeMinutes = as.numeric(runtimeMinutes))
Warning: There were 3 warnings in `mutate()`.
The first warning was:
ℹ In argument: `endYear = as.numeric(endYear)`.
Caused by warning:
! NAs introduced by coercion
ℹ Run `dplyr::last_dplyr_warnings()` to see the 2 remaining warnings.
TITLE_EPISODES <- TITLE_EPISODES |>
  mutate(seasonNumber = as.numeric(seasonNumber),
         episodeNumber = as.numeric(episodeNumber))
Warning: There were 2 warnings in `mutate()`.
The first warning was:
ℹ In argument: `seasonNumber = as.numeric(seasonNumber)`.
Caused by warning:
! NAs introduced by coercion
ℹ Run `dplyr::last_dplyr_warnings()` to see the 1 remaining warning.

Now that the data tables are cleaned up and ready to be used, I can begin to explore the data.

Task 2: Instructor-Provided Questions

Firstly, I want to find out how many movies are in our data set? How many TV series? And how many TV episodes?

Click here to see the code 
Number_of_Movies <- TITLE_BASICS |>
  filter(titleType == "movie") |>
  summarize(Number = n())

print(Number_of_Movies$Number)
[1] 89305
Number_of_TVshows <- TITLE_BASICS |>
  filter(titleType == "tvSeries") |>
  summarize(Number = n())

print(Number_of_TVshows$Number)
[1] 16742
Number_of_Episodes <- TITLE_EPISODES |>
  summarize(Number = n())

print(Number_of_Episodes$Number)
[1] 3026526

After running the code above, I found that there were 89,305 movies, 16,742 TV shows, and 3,026,526 TV Episodes.

Secondly, I want to know who the oldest person is in our data set?

To do this, I performed the following code:

oldest_person <- NAME_BASICS |>
  filter(is.na(deathYear)) |>
  arrange(birthYear) |>
  slice(1)

print(oldest_person)
     nconst       primaryName birthYear deathYear primaryProfession
1 nm0393833 William S. Hooser      1836        NA             actor
                 knownForTitles
1 tt0157755,tt0015357,tt0408145

I quickly saw that the birth year of this person was 1836. So unless our oldest person is 188 years old, we need to put a restraint on how far back their birth year is, in case their death was not recorded. Since the oldest living person in the world is 116, the earliest possible birth year can be 1908. Many individuals came up for this birth year, however once I Google searched the ten oldest people in the world, none of them were on that list. I even saw that a few of these names did have deaths recorded on Google, but not on the database. Since there are only about 500,000 living individuals over the age of 100, which is only about .007% of the population, I decided to set my birth range to 100 years ago. Though someone may be slightly older than this and still alive, I thought this would be a safe bet to assume.

oldest_person <- NAME_BASICS |>
  filter(is.na(deathYear),
         birthYear >= 1924) |>
  arrange(birthYear) |>
  slice(1)

print(oldest_person)
     nconst     primaryName birthYear deathYear           primaryProfession
1 nm0001693 Eva Marie Saint      1924        NA actress,producer,soundtrack
                           knownForTitles
1 tt0047296,tt0053125,tt0348150,tt1837709

Thus, I found that Eva Marie Saint is the oldest and still alive person in our database. Her career spanned over eighty years and she won numerous awards for her works. She was even born close to New York, in Newark, NJ!

Thirdly, there is one TV Episode in this data set with a perfect 10/10 rating and at least 200,000 IMDb ratings. I want to discover what episode this is and what show it belongs to.

To do this, I restricted my data to only show me these restraints. This can be seen in the following code:

Click here to see how this was done 
Best_Episode <- TITLE_RATINGS |> 
  filter(averageRating == 10,
         numVotes >= 200000)
Best_E_series <- TITLE_EPISODES |>
  filter(tconst == "tt2301451")

print(Best_E_series)
     tconst parentTconst seasonNumber episodeNumber
1 tt2301451    tt0903747            5            14
Best_Episode_name <- TITLE_BASICS |>
  filter(tconst == "tt0903747")

print(Best_Episode_name)
     tconst titleType primaryTitle originalTitle isAdult startYear endYear
1 tt0903747  tvSeries Breaking Bad  Breaking Bad   FALSE      2008    2013
  runtimeMinutes               genres
1             45 Crime,Drama,Thriller

From doing the commands above, I found that the only episode that reached these ratings was episode 14 of season 5 in Breaking Bad. Looking this up on Google, I found that the title of the episode was “Ozymandias.” This episode is rated so highly due to the brilliance of how show runners depicted the main protagonist getting everything taken away from him. Pivotal events happen, including a main character’s demise which became a significant turning point for the narrative. Over a decade later, this episode is still raved and talked about by fans.

Fourthly, what four projects is the actor Mark Hamill most known for?

Click here to see how this was done 
Mark_Hamill <- NAME_BASICS |>
  filter(primaryName == "Mark Hamill")

print(Mark_Hamill)
     nconst primaryName birthYear deathYear       primaryProfession
1 nm0000434 Mark Hamill        NA        NA actor,producer,director
                           knownForTitles
1 tt0076759,tt2527336,tt0080684,tt0086190
Mark_Hamill_Projects <- TITLE_BASICS |>
  filter(tconst == "tt0076759") 

print(Mark_Hamill_Projects)
     tconst titleType                       primaryTitle originalTitle isAdult
1 tt0076759     movie Star Wars: Episode IV - A New Hope     Star Wars   FALSE
  startYear endYear runtimeMinutes                   genres
1      1977      NA            121 Action,Adventure,Fantasy
Mark_Hamill_Projects <- TITLE_BASICS |>
  filter(tconst == "tt2527336") 

print(Mark_Hamill_Projects)
[1] tconst         titleType      primaryTitle   originalTitle  isAdult       
[6] startYear      endYear        runtimeMinutes genres        
<0 rows> (or 0-length row.names)
Mark_Hamill_Projects <- TITLE_BASICS |>
  filter(tconst == "tt0086190") 

print(Mark_Hamill_Projects)
     tconst titleType                               primaryTitle
1 tt0086190     movie Star Wars: Episode VI - Return of the Jedi
                               originalTitle isAdult startYear endYear
1 Star Wars: Episode VI - Return of the Jedi   FALSE      1983      NA
  runtimeMinutes                   genres
1            131 Action,Adventure,Fantasy
Mark_Hamill_Projects <- TITLE_BASICS |>
  filter(tconst == "tt0080684") 

print(Mark_Hamill_Projects)
     tconst titleType                                   primaryTitle
1 tt0080684     movie Star Wars: Episode V - The Empire Strikes Back
                                   originalTitle isAdult startYear endYear
1 Star Wars: Episode V - The Empire Strikes Back   FALSE      1980      NA
  runtimeMinutes                   genres
1            124 Action,Adventure,Fantasy

Using the command above, I found that Mark Hamill was known for four Star Wars titles, including: Star Wars: Episode IV - A New Hope, Star Wars: Episode VIII - The Last Jedi, Star Wars: Episode V - The Empire Strikes Back, and Star Wards: Episode VI - Return of the Jedi.

Fifthly, I want to know what TV series, with more than 12 episodes, has the highest average rating?

To compute this, I did the following:

Click here to see how this was done 
episode_amount <- TITLE_EPISODES |>
    group_by(parentTconst) |>
    summarise(total_episodes = n(), .groups = 'drop')

tv_ratings <- TITLE_BASICS |>
    filter(titleType == "tvSeries") |>
    inner_join(TITLE_RATINGS, by = "tconst") |>
    inner_join(episode_amount, by = c("tconst" = "parentTconst"))


highest_rating_series <- tv_ratings |>
    filter(total_episodes > 12) |>
    arrange(desc(averageRating)) |>
    slice(1)

highest_rating_series |>
    select(primaryTitle, total_episodes, averageRating)
  primaryTitle total_episodes averageRating
1  Craft Games            318           9.7
print(highest_rating_series)
      tconst titleType primaryTitle originalTitle isAdult startYear endYear
1 tt15613780  tvSeries  Craft Games   Craft Games   FALSE      2014      NA
  runtimeMinutes                  genres averageRating numVotes total_episodes
1             NA Adventure,Comedy,Family           9.7      151            318

Thus, I found out that Craft Games was the highest rated series with over 12 episodes, with an average rating of 9.7.

Lastly, I wanted to know more about the TV series Happy Days. The TV series Happy Days (1974-1984) gives us the common idiom “jump the shark”. The phrase comes from a controversial fifth season episode (aired in 1977) in which a lead character literally jumped over a shark on water skis. Idiomatically, it is used to refer to the moment when a once-great show becomes ridiculous and rapidly looses quality.

I want to know if it is true that episodes from later seasons of Happy Days have lower average ratings than the early seasons?

To do this, I must restrict my data set to only show me information about Happy Days. At first, I only restricted this by the name, but I quickly realized that there were numerous projects with the same name. Because of this, I also added the restrictions of the start year and end year of the series. By doing the following code, I find the necessary tconst that identifies the series. Then, I ran the following code to get the average rating of each season and compare them.

Click here to see how this was done 
TITLE_BASICS |>
  filter(primaryTitle == "Happy Days",
         startYear == "1974",
         endYear == "1984") 
     tconst titleType primaryTitle originalTitle isAdult startYear endYear
1 tt0070992  tvSeries   Happy Days    Happy Days   FALSE      1974    1984
  runtimeMinutes              genres
1             30 Comedy,Family,Music
Happy_Days <- TITLE_EPISODES |>
  filter(parentTconst == "tt0070992")

Happy_Days_Scores <- inner_join(Happy_Days, TITLE_RATINGS, join_by(tconst))
Happy_Days_Seasons <- Happy_Days_Scores |>
  group_by(seasonNumber) |>
  summarize(averagescore = mean(averageRating, na.rm = TRUE)) |>
  arrange(seasonNumber)
library(DT)
datatable(Happy_Days_Seasons)

By looking at the table above, we can see that the scores did in fact drop towards the later seasons, with Season 8 being the lowest scoring season. For the first half of the show, they average at a high 7 rating, but a downward decline began after season 3. Season 11, which was their final season, inevitably scored back in the 7 range, but the score still did not beat any from the first four seasons.

Task 3: Custom Success Metric

For my success metric, I decided to create a scale that assesses the total popularity of a work. To do this, I decided to multiply the average rating that it got by the number of votes. My logic is that a work cannot be successful without each of these metrics: a low scoring average with a lot of votes means that the movie flopped and a high scoring average with no votes means that the movie is too obscure. However, if a movie is averaged highly and has a lot of votes, this means that it is widely known and highly regarded. Thus, I believe that this is an accurate way to measure success.

To test my theory, I will run some tests to see if it holds up against actually highly regarded movies and some of the more obscure ones in Hollywood.

Firstly, I will choose the top ten movies on my metric and confirm that they were box office successes.

Click here to see how I filtered my data to show this 
TITLE_RATINGS <- TITLE_RATINGS |>
  mutate(popularity = averageRating*numVotes) 

movie_ratings <- inner_join(TITLE_BASICS, TITLE_RATINGS, join_by(tconst))

movie_ratings <- movie_ratings |> select(tconst, titleType, primaryTitle, genres, startYear, averageRating, numVotes, popularity)

movie_ratings <- movie_ratings |> 
  filter(titleType == "movie") |>
  arrange(desc(popularity))


Top_ten <- movie_ratings |>
  slice(1:10)

print(Top_ten)
      tconst titleType                                      primaryTitle
1  tt0111161     movie                          The Shawshank Redemption
2  tt0468569     movie                                   The Dark Knight
3  tt1375666     movie                                         Inception
4  tt0137523     movie                                        Fight Club
5  tt0109830     movie                                      Forrest Gump
6  tt0110912     movie                                      Pulp Fiction
7  tt0816692     movie                                      Interstellar
8  tt0068646     movie                                     The Godfather
9  tt0120737     movie The Lord of the Rings: The Fellowship of the Ring
10 tt0133093     movie                                        The Matrix
                    genres startYear averageRating numVotes popularity
1                    Drama      1994           9.3  2954275   27474758
2       Action,Crime,Drama      2008           9.0  2935734   26421606
3  Action,Adventure,Sci-Fi      2010           8.8  2605397   22927494
4                    Drama      1999           8.8  2386403   21000346
5            Drama,Romance      1994           8.8  2311754   20343435
6              Crime,Drama      1994           8.9  2268830   20192587
7   Adventure,Drama,Sci-Fi      2014           8.7  2179321   18960093
8              Crime,Drama      1972           9.2  2059556   18947915
9   Action,Adventure,Drama      2001           8.9  2052301   18265479
10           Action,Sci-Fi      1999           8.7  2097695   18249947

The following are the top ten successes on my list:

  1. The Shawshank Redemption
  2. The Dark Knight
  3. Fight Club
  4. Forrest Gump
  5. Pulp Fiction
  6. The Godfather
  7. Interstellar
  8. The Lord of the Rings: The Fellowship of the Ring
  9. The Matrix
  10. The Lord of the Rings: The Return of the King

All of which were MAJOR successes in the box office. So far, this success metric is looking good!

Secondly, I will do the same as before, but I will choose the five lowest movies on my list and ensure they were of low quality.

Click here to see how I filtered my data to show this 
movie_ratings <- movie_ratings |> 
  filter(titleType == "movie") |>
  arrange(popularity)


bottom_five <- movie_ratings |>
  slice(1:5)

print(bottom_five)
     tconst titleType                             primaryTitle
1 tt0498734     movie                 Zwischen Glück und Krone
2 tt0386134     movie                        Yubiley prokurora
3 tt0309422     movie               The Case He Couldn't Crack
4 tt1773029     movie                                 Rockland
5 tt1666348     movie Vixen Highway 2006: It Came from Uranus!
                  genres startYear averageRating numVotes popularity
1            Documentary      1959           1.2      103      123.6
2                 Comedy      2003           1.1      128      140.8
3         Crime,Thriller      1981           1.0      153      153.0
4          Comedy,Family      2010           1.5      102      153.0
5 Action,Horror,Thriller      2010           1.1      145      159.5

The following are the 5 lowest rated on my list:

  1. Zwischen Glück und Krone
  2. Yubiley prokurora
  3. The Case He Couldn’t Crack
  4. 18’ler Takimi
  5. Parentesi tonde

These all proved to be more obscure and poorly-rated movies. Therefore, the metric does work both ways.

Thirdly, now I will choose a prestige actor and confirm that they have many projects that score highly on my metric. For this, I decided to look into Tom Hanks and Leonardo DiCaprio, as I believe they are two major actors in the industry currently.

Click here to see how I filtered my data to show this 
library(tidyverse)

Leo <- NAME_BASICS |>
  filter(primaryName == "Leonardo DiCaprio") |>
  separate_longer_delim(knownForTitles, ",")


colnames(Leo)[colnames(Leo) == "knownForTitles"] <- "tconst"

Leo <- inner_join(Leo, movie_ratings, join_by(tconst))

Hanks <- NAME_BASICS |>
  filter(primaryName == "Tom Hanks") |>
  separate_longer_delim(knownForTitles, ",")


colnames(Hanks)[colnames(Hanks) == "knownForTitles"] <- "tconst"

Hanks <- inner_join(Hanks, movie_ratings, join_by(tconst)) 

Leo_Hanks <- bind_rows(Leo, Hanks)

Leo_Hanks2 <- Leo_Hanks |> select(primaryName, primaryTitle, averageRating, numVotes, popularity)
datatable(Leo_Hanks2)

Here, we can see that the movies given back by my success metric are in fact box office successes. Movies such as Titanic, The Wolf of Wall Street, Cast Away, and Forrest Gump are all well-known movies that were incredibly well-rated. Thus, the metric still holds.

Fourthly, for my final test, I wanted to compare my metric with the highest-grossing movies of all time. After searching this data on Google, I found that some of the highest grossing movies of all time is Titanic with $2,264,750,694, Avatar: The Way of Water with $2,320,250,281, and Avatar with $2,923,706,026.

Click here to see how I filtered my data to show this 
Highest_Grossing <- data.frame(
  tconst = c("tt0499549", "tt4154796", "tt1630029", "tt0120338", "tt2488496"),
  LifetimeGross = c("$2,923,706,026", "$2,799,439,100", "$2,320,250,281", "$2,264,750,694", "$2,071,310,218")
)


High_Grossing <- inner_join(movie_ratings, Highest_Grossing, join_by(tconst))

The chart below depicts three of the above movies. It is clear to see that the popularity score correlates with their gross, as both are high. This solidifies my success metric even further.

datatable(High_Grossing)

Lastly, now that my success metric has been solidified, I need to restrict my data set to only show me “successes.” To do this, I must distinguish a number that once a movie surpasses it in its “popularity” scale, then it is deemed a success. The number I decided to pick is 5,000,000, since I feel as though every great movie should be able to surpass that. Then, I reduced my data set to fit that benchmark using the following code.

solid_or_better <- movie_ratings |>
  filter(popularity >= 5000000)

Task 5: Key Personnel

Now that I have my target genre, I want to identify a few actors and one director to really anchor my project. To create a strong team, I want one prestigious actor who has many successes in many genres, and preferably at least one major success in the action genre. With this major name, I hope to bring in their major following and their broad skill set to perfect the movie. For my second actor, I want to pick someone who has been apart of a highly rated movie, but is still not widely recognized yet. I want someone who is up-and-coming, that can build a strong profile around being in this movie. Then, the connection with these two main leads should bring good publicity.

For my prestige actor, I want someone who has been in the industry for a good amount of time and that is older with some experience. Thus, I decided that I wanted my actor to be at least 40, but no older than 75.

Click here to see how I filtered my data to show this 
alive_actors <- NAME_BASICS |>
  filter(birthYear >= 1949 & birthYear <= 1984) |>
  filter(is.na(deathYear)) |>
  separate_longer_delim(primaryProfession, ",") |>
  filter(primaryProfession == "actor") |>
  separate_longer_delim(knownForTitles, ",") 

solid_or_better2 <- movie_ratings |>
  filter(popularity >= 5000000)

successful_movies <- solid_or_better2 |>
  select(tconst, genres, popularity) 

colnames(successful_movies)[colnames(successful_movies) == "tconst"] <- "knownForTitles"


successful_actors <- full_join(alive_actors, successful_movies, join_by(knownForTitles))

successful_actors <- successful_actors |>
  arrange(desc(popularity))

successful_actors_average <-successful_actors |>
  select(primaryName, popularity) |>
  group_by(primaryName) |>
  summarize(totalpopularity = sum(popularity, na.rm = TRUE)) |>
  arrange(desc(totalpopularity))
datatable(successful_actors_average)
Warning in instance$preRenderHook(instance): It seems your data is too big for
client-side DataTables. You may consider server-side processing:
https://rstudio.github.io/DT/server.html

One by one, I decided to go down the list of people to see what their top works were. This can be seen in the code below.

Click here to see to see the code 
Orlando_Bloom <- successful_actors |>
  filter(primaryName == "Orlando Bloom")

Hans_Zimmer <- successful_actors |>
  filter(primaryName == "Orlando Bloom")

Hugo_Weaving <- successful_actors |>
  filter(primaryName == "Hugo Weaving")

Gary_Rizzo <- successful_actors |>
  filter(primaryName == "Gary A. Rizzo")

Leonardo_DiCaprio <- successful_actors |>
  filter(primaryName == "Leonardo DiCaprio")

I decided against Orlando Bloom, since he had 4 successful movies, three of which were apart of the Lord of the Rings franchise, and one Pirates of the Caribbean. I want a more diverse actor, rather than someone who mainly dominates the same universe.

Hugo Weaving was similar, with three Lord of the Rings movies and one movie titled “V for Vendetta.” Rizzo would have been perfect… until I researched and found out he was the music producer on these projects, and not an actor. Same with Zimmer.

Finally, I got to my seventh person on the list, Leonardo DiCaprio. Not only did his top movies dominate different genres, including Crime, Drama, Thriller, Romance, Comedy, etc., his top movie was an Action/Adventure mixed genre. Also scoring highest on his list on the popularity scale, Inception became the fourth highest-grossing movie of 2010, with DiCaprio as the star of the movie. Thus, it was clear that he was my first pick actor.

Next, I wanted to find someone younger and more up-and-coming. Thus, I chose to restrict the age gap to someone less than 30 years old. I did this with the following code:

Click here to see to see the code 
alive_actors_young <- NAME_BASICS |>
  filter(birthYear >= 1995 & birthYear <= 2024) |>
  filter(is.na(deathYear)) |>
  separate_longer_delim(primaryProfession, ",") |>
  filter(primaryProfession == "actor") |>
  separate_longer_delim(knownForTitles, ",") 

solid_or_better2 <- movie_ratings |>
  filter(popularity >= 5000000)

successful_movies <- solid_or_better2 |>
  select(tconst, genres, popularity) 

colnames(successful_movies)[colnames(successful_movies) == "tconst"] <- "knownForTitles"


successful_actors_young <- full_join(alive_actors_young, successful_movies, join_by(knownForTitles))

successful_actors_young <- successful_actors_young |>
  arrange(desc(popularity))

successful_actors_average_y <-successful_actors_young |>
  select(primaryName, popularity) |>
  group_by(primaryName) |>
  summarize(totalpopularity = sum(popularity, na.rm = TRUE)) |>
  arrange(desc(totalpopularity))


Tony_Revolori <- successful_actors_young |>
  filter(primaryName == "Tony Revolori")

I quickly decided upon Tony Revolori, as he was high on the list and was in a Spiderman movie recently that did fairly well. When I looked him up, he had some breakout awards and is expected to become a lot bigger in the upcoming years, especially as the projects he is currently working on begin to roll out.

At this point of the project, I realized that I had a slight oversight and did not include any female actresses into the mix. So, I decided to pick another younger actor to add to my team.

Click here to see to see the code 
alive_actress_young <- NAME_BASICS |>
  filter(birthYear >= 1995 & birthYear <= 2024) |>
  filter(is.na(deathYear)) |>
  separate_longer_delim(primaryProfession, ",") |>
  filter(primaryProfession == "actress") |>
  separate_longer_delim(knownForTitles, ",") 


successful_actress_young <- full_join(alive_actress_young, successful_movies, join_by(knownForTitles))

successful_actress_average_y <-successful_actress_young |>
  select(primaryName, popularity) |>
  group_by(primaryName) |>
  summarize(totalpopularity = sum(popularity, na.rm = TRUE)) |>
  arrange(desc(totalpopularity))



Taylor_Geare <- successful_actress_young |>
  filter(primaryName == "Taylor Geare")

By running this previous code, I found that Taylor Geare would be a perfect pick for the movie. Not only has she won one award and been in a few works recently, I discovered that she was also in Inception with Leonardo DiCaprio! At the time, she was five years old and she played Philipa, who was the daughter of DiCaprio’s character. I believe that this pairing can bring major publicity to our work, as fans of Inception will get to see this duo back together, especially in a way that Geare will be able to display her skills as an adult.

Now, I must find a director for my movie. Using the same technique as above, I began my search.

Click here to see to see the code 
alive_directors <- NAME_BASICS |>
  filter(is.na(deathYear)) |>
  separate_longer_delim(primaryProfession, ",") |>
  filter(primaryProfession == "director") |>
  separate_longer_delim(knownForTitles, ",") 

successful_directors <- full_join(alive_directors, successful_movies, join_by(knownForTitles))


successful_directors2 <- successful_directors |>
  group_by(primaryName) |>
  summarize(totalpopularity = sum(popularity, na.rm = TRUE)) |>
  arrange(desc(totalpopularity))

Wally_Pfister <- successful_directors |>
  filter(primaryName == "Wally Pfister")
datatable(successful_directors2)
Warning in instance$preRenderHook(instance): It seems your data is too big for
client-side DataTables. You may consider server-side processing:
https://rstudio.github.io/DT/server.html

Coming in second on the list, Wally Pfister easily proved to be the strongest candidate for this movie. Not only did he score incredibly highly on this list, he also worked on major movies such as The Dark Knight, The Prestige, Transcendence, and much more. However, most notably, he also worked on Inception with Leonardo DiCaprio and Taylor Geare. A picture of DiCaprio and Pfister can be seen below.

Thus, not only have they all worked together in the past so they know each other/have good chemistry, they were also able to create a masterpiece of a movie before in the same genre we are aspiring to. This could stir a lot of good media, as it not seen very often that three people from a very successful movie team up again to work on a remake of another successful movie. Thus, I feel very confident about the potential team I have put together.

Task 6: Finding a Classic Movie to Remake

Now, I must find a classic movie in this genre to remake. I want to find a movie that has not been remade in the past 25 years, has a large number of IMDb ratings, has a high average rating, and has a fan base that wants a remake to happen. To do this, I must first filter my data sets to give me a movie that fits these first three restraints.

Click here to see to see the code 
remakemovie <- movie_ratings |>
  filter(startYear < 1999) |>
  filter(averageRating >  8) |>
  filter(numVotes > 20000) |>
  arrange(desc(popularity))
datatable(remakemovie)

I want to look at what the top movies are in the action genre, however, I also want to stay away from remaking any Star Wars movies. Thus, coming in at number 14, I debating on remaking Léon: The Professional, however, when I looked up if fans wanted a remake, they seemed entirely opposed. Many fans believe the film is perfect as is and it would not do any justice for it to be remade today. Thus, I decided to keep looking.

At number 31, I found Die Hard, a movie made in 1988 about an NYPD cop that has to take matters into his own hands when a group of robbers take control of the building he is in, holding everyone hostage except for him.

When I looked up if fans would want a remake of Die Hard, numerous articles came up showing great approval for this. Some of these are linked below:

Thus, a Die Hard remake is the best bet for a successful action movie.

However, before we move further, our legal department must reach out to certain members of the original movie production to secure the rights for the project. So, let’s find out who is still alive from the original.

Click here to see to see the code 
Die_Hard <- NAME_BASICS |>
  separate_longer_delim(knownForTitles, ",") |>
  filter(knownForTitles == "tt0083658") |>
  filter(is.na(deathYear))

print(Die_Hard)
       nconst           primaryName birthYear deathYear
1   nm0000435          Daryl Hannah      1960        NA
2   nm0000631          Ridley Scott      1937        NA
3   nm0000707            Sean Young      1959        NA
4   nm0001026        Joanna Cassidy        NA        NA
5   nm0001579    Edward James Olmos      1947        NA
6   nm0003256          Joshua Pines        NA        NA
7   nm0003666         Debbie Denise        NA        NA
8   nm0007072       Katherine Haber      1944        NA
9   nm0022370          Vickie Alper        NA        NA
10  nm0029959            Tim Angulo        NA        NA
11  nm0045630    Michael Backauskas        NA        NA
12  nm0047460      Robert D. Bailey        NA        NA
13  nm0048193            Beau Baker      1953        NA
14  nm0048396             Don Baker        NA        NA
15  nm0049792         Peter Baldock      1949        NA
16  nm0081795   William Biggerstaff        NA        NA
17  nm0099156     Christian Boudman        NA        NA
18  nm0114926       Winnie D. Brown        NA        NA
19  nm0123746            Tom Burton        NA        NA
20  nm0137698      Elizabeth Carlon        NA        NA
21  nm0164425          Robert Clark        NA        NA
22  nm0172782           Alan Collis        NA        NA
23  nm0173645            Gary Combs      1935        NA
24  nm0180108        Peter Cornberg        NA        NA
25  nm0190038           Eugene Crum        NA        NA
26  nm0193345           Greg Curtis        NA        NA
27  nm0199344          Stephen Dane        NA        NA
28  nm0203343 Howard Brady Davidson        NA        NA
29  nm0213057           Lisa Deaner        NA        NA
30  nm0214303        Michael Deeley      1932        NA
31  nm0218540     Carolyn DeMirjian        NA        NA
32  nm0220984        Linda DeScenna      1949        NA
33  nm0238734           David Dryer      1943        NA
34  nm0244956            Syd Dutton      1944        NA
35  nm0252914              Bud Elam        NA        NA
36  nm0266684       Hampton Fancher      1938        NA
37  nm0270670         Jane Feinberg        NA        NA
38  nm0274928      Michael Ferriter        NA        NA
39  nm0280029      William E. Fitch        NA        NA
40  nm0281551        Linda Fleisher        NA        NA
41  nm0283845           Stephen Fog        NA        NA
42  nm0285893          Terence Ford      1945        NA
43  nm0290519          Chris Franco        NA        NA
44  nm0292476       Terry D. Frazee        NA        NA
45  nm0302091          Steve Galich        NA        NA
46  nm0302406         Joe Gallagher      1953        NA
47  nm0303421        Kurt P. Galvao        NA        NA
48  nm0312920         Michael Genne        NA        NA
49  nm0325242        Joyce Goldberg        NA        NA
50  nm0333849         David Grafton        NA        NA
51  nm0341454         Cary Griffith        NA        NA
52  nm0356040           Robert Hall        NA        NA
53  nm0362046   David R. Hardberger      1948        NA
54  nm0362257          Alan Harding        NA        NA
55  nm0367242   Graham V. Hartstone      1944        NA
56  nm0369220          Donald Hauer        NA        NA
57  nm0372264            Les Healey      1950        NA
58  nm0381500            Linda Hess        NA        NA
59  nm0384603       Richard L. Hill        NA        NA
60  nm0390897  Richard E. Hollander        NA        NA
61  nm0394803           Bob E. Horn        NA        NA
62  nm0404527   Gillian L. Hutshing        NA        NA
63  nm0438325        Michael Kaplan        NA        NA
64  nm0452191       Madeleine Klein      1946        NA
65  nm0453139           Crit Killen        NA        NA
66  nm0461802       Michael Knutsen        NA        NA
67  nm0487514         James Lapidus        NA        NA
68  nm0507809        Terry E. Lewis        NA        NA
69  nm0513924          Marci Liroff        NA        NA
70  nm0518430        Basil Lombardo        NA        NA
71  nm0519436          Ronald Longo        NA        NA
72  nm0523370       Stephanie Lowry        NA        NA
73  nm0560094        Linda Matthews        NA        NA
74  nm0570488            Tim McHugh        NA        NA
75  nm0573488    Gregory L. McMurry      1952        NA
76  nm0590114         Michael Mills        NA        NA
77  nm0616694          Donald Myers        NA        NA
78  nm0624443          Tony Negrete        NA        NA
79  nm0640384         John O'Connor        NA        NA
80  nm0649687    James F. Orendorff        NA        NA
81  nm0655669       Shirley Padgett        NA        NA
82  nm0672065         Peter Pennell      1939        NA
83  nm0672459    David Webb Peoples      1940        NA
84  nm0689279     Thomas R. Polizzi        NA        NA
85  nm0689332   George Polkinghorne        NA        NA
86  nm0689503        Tama Takahashi        NA        NA
87  nm0694138           Ivor Powell        NA        NA
88  nm0703436        David Q. Quick        NA        NA
89  nm0709588          Gary Randall        NA        NA
90  nm0711047           Karen Rasch        NA        NA
91  nm0722541    Victoria E. Rhodes        NA        NA
92  nm0728110        Richard Rippel        NA        NA
93  nm0745076     Jonathan Rothbart        NA        NA
94  nm0747438     Thomas L. Roysden      1944        NA
95  nm0761836     William Sanderson      1944        NA
96  nm0768102          Fumi Mashimo        NA        NA
97  nm0769731       Steve Schaeffer        NA        NA
98  nm0770435          John Scheele        NA        NA
99  nm0774039       Suzie Schneider        NA        NA
100 nm0775562 Richard Peter Schroer        NA        NA
101 nm0778771     John A. Scott III        NA        NA
102 nm0779860             Tom Scott      1948        NA
103 nm0791107    Victor A. Shelehov        NA        NA
104 nm0794135        Arthur Shippee      1948        NA
105 nm0811435       David L. Snyder        NA        NA
106 nm0816170         Tom Southwell        NA        NA
                                          primaryProfession knownForTitles
1                                 actress,producer,director      tt0083658
2                     producer,director,production_designer      tt0083658
3                            actress,miscellaneous,director      tt0083658
4                            actress,producer,miscellaneous      tt0083658
5                                   actor,producer,director      tt0083658
6              visual_effects,editorial_department,director      tt0083658
7                     visual_effects,producer,miscellaneous      tt0083658
8                            miscellaneous,producer,actress      tt0083658
9                                             miscellaneous      tt0083658
10         visual_effects,cinematographer,camera_department      tt0083658
11               visual_effects,editorial_department,editor      tt0083658
12                           visual_effects,director,writer      tt0083658
13                     sound_department,actor,miscellaneous      tt0083658
14                visual_effects,camera_department,director      tt0083658
15              sound_department,editorial_department,actor      tt0083658
16                             art_department,miscellaneous      tt0083658
17        visual_effects,miscellaneous,editorial_department      tt0083658
18                      costume_department,costume_designer      tt0083658
19                   director,producer,animation_department      tt0083658
20                                           visual_effects      tt0083658
21        art_department,special_effects,make_up_department      tt0083658
22                producer,production_manager,miscellaneous      tt0083658
23                          stunts,assistant_director,actor      tt0083658
24           production_manager,assistant_director,producer      tt0083658
25                                          special_effects      tt0083658
26                           special_effects,art_department      tt0083658
27          art_department,art_director,production_designer      tt0083658
28                                transportation_department      tt0083658
29                                           visual_effects      tt0083658
30                  sound_department,producer,miscellaneous      tt0083658
31                                                  actress      tt0083658
32                set_decorator,production_designer,actress      tt0083658
33                         visual_effects,camera_department      tt0083658
34                    visual_effects,special_effects,writer      tt0083658
35                         visual_effects,camera_department      tt0083658
36                                    actor,writer,director      tt0083658
37             casting_director,casting_department,producer      tt0083658
38      visual_effects,special_effects,editorial_department      tt0083658
39                                        camera_department      tt0083658
40                                           visual_effects      tt0083658
41                                           visual_effects      tt0083658
42               actor,assistant_director,camera_department      tt0083658
43                                        camera_department      tt0083658
44                                          special_effects      tt0083658
45            special_effects,visual_effects,art_department      tt0083658
46   sound_department,editorial_department,music_department      tt0083658
47              producer,editorial_department,miscellaneous      tt0083658
48                  camera_department,cinematographer,actor      tt0083658
49                             visual_effects,miscellaneous      tt0083658
50                                           visual_effects      tt0083658
51                                        camera_department      tt0083658
52                                           visual_effects      tt0083658
53         visual_effects,camera_department,cinematographer      tt0083658
54                         visual_effects,camera_department      tt0083658
55                    sound_department,editorial_department      tt0083658
56           assistant_director,production_manager,producer      tt0083658
57             editor,editorial_department,sound_department      tt0083658
58                miscellaneous,producer,production_manager      tt0083658
59                           special_effects,art_department      tt0083658
60        visual_effects,miscellaneous,animation_department      tt0083658
61                      costume_department,costume_designer      tt0083658
62                   editor,music_department,visual_effects      tt0083658
63                costume_designer,costume_department,actor      tt0083658
64                             miscellaneous,actress,stunts      tt0083658
65            special_effects,art_department,visual_effects      tt0083658
66                                            miscellaneous      tt0083658
67        costume_designer,costume_department,miscellaneous      tt0083658
68                                           art_department      tt0083658
69                  casting_director,producer,miscellaneous      tt0083658
70                                           art_department      tt0083658
71                                           visual_effects      tt0083658
72   music_department,sound_department,editorial_department      tt0083658
73                      costume_department,costume_designer      tt0083658
74                         visual_effects,producer,director      tt0083658
75        visual_effects,miscellaneous,animation_department      tt0083658
76                 make_up_department,actor,special_effects      tt0083658
77                                          special_effects      tt0083658
78                                         sound_department      tt0083658
79     camera_department,cinematographer,assistant_director      tt0083658
80                                           art_department      tt0083658
81                                       make_up_department      tt0083658
82                                         sound_department      tt0083658
83                                   writer,editor,director      tt0083658
84           producer,assistant_director,production_manager      tt0083658
85                                           visual_effects      tt0083658
86                         camera_department,visual_effects      tt0083658
87                       producer,writer,assistant_director      tt0083658
88                                           art_department      tt0083658
89          production_designer,art_director,art_department      tt0083658
90  editorial_department,script_department,sound_department      tt0083658
91                              assistant_director,producer      tt0083658
92                                           visual_effects      tt0083658
93                  visual_effects,producer,special_effects      tt0083658
94         set_decorator,art_department,production_designer      tt0083658
95                                    actor,archive_footage      tt0083658
96        visual_effects,animation_department,miscellaneous      tt0083658
97                                music_department,composer      tt0083658
98             visual_effects,producer,animation_department      tt0083658
99                           visual_effects,special_effects      tt0083658
100               assistant_director,producer,miscellaneous      tt0083658
101                                          art_department      tt0083658
102                         music_department,composer,actor      tt0083658
103                                       camera_department      tt0083658
104                                          art_department      tt0083658
105         production_designer,art_department,art_director      tt0083658
106         art_department,production_designer,art_director      tt0083658

Thus, 135 people from the original are still alive. Some actor and actresses are still alive as well, so we may want to consider adding them in the plot as a “fan service,” such as the main actor Bruce Willis.

Task 7: Write and Deliver Your Pitch

For our next production, a remake of the 1988 Die Hard would be a strong competitor in the industry, starring Leonardo DiCaprio, Taylor Geare, and Tony Revolori. Directed by Academy Award winning Wally Pfister, the movie is bound to have an artistic expression like no other, similar to his eight other success. A guest appearance from Bruce Willis will have the fan base, who is practically begging for a remake, running to theaters.

From the 1990s to the 2000s, the Action genre experienced over a 350% growth in successes, becoming one of the most wanted and rewarding genre in the industry. With beloved actor Leonardo Dicaprio combining forces with innovative director Pfister and up-and-coming Geare, we will give Inception fans a wanted reunion in this remake, growing our anticipated fan base even further. Lastly, including young actor Revolori with resonate with the younger generation, especially those tied to Marvel, as he is a fan favorite in that sphere. Thus, this combination would be an absolute recipe for success, especially since the action genre has the most successes in the 2020s thus far! If we move forward with this plan, we can easily be the next success!