2.1 Data Input

Import data csv. Make sure our data placed in the same folder with our R project data.

imdb <- read.csv("data_input/imdb_top_1000.csv")
imdb <- subset(imdb, select = -c(Poster_Link,Overview)) #drop "Poster Link" & "Overview" columns
head(imdb)

2.2 Data Inspection

str(imdb) #chek data types of each columns

#> 'data.frame':    1000 obs. of  14 variables:
#>  $ Series_Title : chr  "The Shawshank Redemption" "The Godfather" "The Dark Knight" "The Godfather: Part II" ...
#>  $ Released_Year: chr  "1994" "1972" "2008" "1974" ...
#>  $ Certificate  : chr  "A" "A" "UA" "A" ...
#>  $ Runtime      : chr  "142 min" "175 min" "152 min" "202 min" ...
#>  $ Genre        : chr  "Drama" "Crime, Drama" "Action, Crime, Drama" "Crime, Drama" ...
#>  $ IMDB_Rating  : num  9.3 9.2 9 9 9 8.9 8.9 8.9 8.8 8.8 ...
#>  $ Meta_score   : int  80 100 84 90 96 94 94 94 74 66 ...
#>  $ Director     : chr  "Frank Darabont" "Francis Ford Coppola" "Christopher Nolan" "Francis Ford Coppola" ...
#>  $ Star1        : chr  "Tim Robbins" "Marlon Brando" "Christian Bale" "Al Pacino" ...
#>  $ Star2        : chr  "Morgan Freeman" "Al Pacino" "Heath Ledger" "Robert De Niro" ...
#>  $ Star3        : chr  "Bob Gunton" "James Caan" "Aaron Eckhart" "Robert Duvall" ...
#>  $ Star4        : chr  "William Sadler" "Diane Keaton" "Michael Caine" "Diane Keaton" ...
#>  $ No_of_Votes  : int  2343110 1620367 2303232 1129952 689845 1642758 1826188 1213505 2067042 1854740 ...
#>  $ Gross        : chr  "28,341,469" "134,966,411" "534,858,444" "57,300,000" ...

colSums(is.na(imdb)) #chek number of missing value of each columns

#>  Series_Title Released_Year   Certificate       Runtime         Genre 
#>             0             0             0             0             0 
#>   IMDB_Rating    Meta_score      Director         Star1         Star2 
#>             0           157             0             0             0 
#>         Star3         Star4   No_of_Votes         Gross 
#>             0             0             0             0

From our inspection we can conclude:

1. imdb data contain 1,000 rows and 14 columns after we drop 2 columns (“Poster Link” & “Overview”).
2. In column 15, Gross should have numeric data type, so it must be changed before the data processing.
3. Metascore is an average score given by the film critic and there are 157 movies that do not have a Metascore.

2.3 Data Cleansing & Coercions

From Data Inspection, we need to change Gross data type not into numeric by using as.numeric(). But before we can do that, we need to drop the comma symbol by using gsub()

imdb$Gross <- as.numeric(gsub(",","",imdb$Gross))

str(imdb)

#> 'data.frame':    1000 obs. of  14 variables:
#>  $ Series_Title : chr  "The Shawshank Redemption" "The Godfather" "The Dark Knight" "The Godfather: Part II" ...
#>  $ Released_Year: chr  "1994" "1972" "2008" "1974" ...
#>  $ Certificate  : chr  "A" "A" "UA" "A" ...
#>  $ Runtime      : chr  "142 min" "175 min" "152 min" "202 min" ...
#>  $ Genre        : chr  "Drama" "Crime, Drama" "Action, Crime, Drama" "Crime, Drama" ...
#>  $ IMDB_Rating  : num  9.3 9.2 9 9 9 8.9 8.9 8.9 8.8 8.8 ...
#>  $ Meta_score   : int  80 100 84 90 96 94 94 94 74 66 ...
#>  $ Director     : chr  "Frank Darabont" "Francis Ford Coppola" "Christopher Nolan" "Francis Ford Coppola" ...
#>  $ Star1        : chr  "Tim Robbins" "Marlon Brando" "Christian Bale" "Al Pacino" ...
#>  $ Star2        : chr  "Morgan Freeman" "Al Pacino" "Heath Ledger" "Robert De Niro" ...
#>  $ Star3        : chr  "Bob Gunton" "James Caan" "Aaron Eckhart" "Robert Duvall" ...
#>  $ Star4        : chr  "William Sadler" "Diane Keaton" "Michael Caine" "Diane Keaton" ...
#>  $ No_of_Votes  : int  2343110 1620367 2303232 1129952 689845 1642758 1826188 1213505 2067042 1854740 ...
#>  $ Gross        : num  2.83e+07 1.35e+08 5.35e+08 5.73e+07 4.36e+06 ...

Missing value check:

colSums(is.na(imdb))

#>  Series_Title Released_Year   Certificate       Runtime         Genre 
#>             0             0             0             0             0 
#>   IMDB_Rating    Meta_score      Director         Star1         Star2 
#>             0           157             0             0             0 
#>         Star3         Star4   No_of_Votes         Gross 
#>             0             0             0           169

Missing value appears after changing the data type in Gross. This happens because the null object was previously counted as a character.

Because we cannot analyze films without Gross, we need to find the missing value and remove it from the dataframe using -which().

imdb <- imdb[-which(is.na(imdb$Gross)),]

colSums(is.na(imdb)) #check if there's still any missing value

#>  Series_Title Released_Year   Certificate       Runtime         Genre 
#>             0             0             0             0             0 
#>   IMDB_Rating    Meta_score      Director         Star1         Star2 
#>             0            81             0             0             0 
#>         Star3         Star4   No_of_Votes         Gross 
#>             0             0             0             0

Check the data shape:

dim(imdb)

#> [1] 831  14

Finally, we have imdb data that contain 831 rows and 14 columns which ready to be processed and analyzed

3.1 Top 5 High Gross Movies

• What 5 top movies are getting highest gross amount?

top5_gross <- imdb[order(imdb$Gross,decreasing = T),][1:5,]
top5_gross

summary(top5_gross)

#>  Series_Title       Released_Year      Certificate          Runtime         
#>  Length:5           Length:5           Length:5           Length:5          
#>  Class :character   Class :character   Class :character   Class :character  
#>  Mode  :character   Mode  :character   Mode  :character   Mode  :character  
#>                                                                             
#>                                                                             
#>                                                                             
#>     Genre            IMDB_Rating     Meta_score     Director        
#>  Length:5           Min.   :7.80   Min.   :68.0   Length:5          
#>  Class :character   1st Qu.:7.80   1st Qu.:75.0   Class :character  
#>  Mode  :character   Median :7.90   Median :78.0   Mode  :character  
#>                     Mean   :8.06   Mean   :76.8                     
#>                     3rd Qu.:8.40   3rd Qu.:80.0                     
#>                     Max.   :8.40   Max.   :83.0                     
#>     Star1              Star2              Star3              Star4          
#>  Length:5           Length:5           Length:5           Length:5          
#>  Class :character   Class :character   Class :character   Class :character  
#>  Mode  :character   Mode  :character   Mode  :character   Mode  :character  
#>                                                                             
#>                                                                             
#>                                                                             
#>   No_of_Votes          Gross          
#>  Min.   : 809955   Min.   :659325379  
#>  1st Qu.: 834477   1st Qu.:678815482  
#>  Median : 860823   Median :760507625  
#>  Mean   : 934068   Mean   :778736742  
#>  3rd Qu.:1046089   3rd Qu.:858373000  
#>  Max.   :1118998   Max.   :936662225

ggplot(top5_gross, aes(x = Gross,y = reorder(Series_Title, Gross))) +
  geom_col(aes(fill = Gross), show.legend = F) +
  labs(title = "Top 5 Movies based on Gross Revenue",
       x = "Gross Revenue",
       y = NULL) +
  geom_label(aes(label = comma(Gross)), hjust = 1.05) +
  scale_fill_gradient(low = "red", high = "black") +
  theme_minimal()

From the data visualization above, we can conclude:

1. Top 5 High Gross Movies have an average gross of $778,736,742, with an average rating of 8.06.

2. Star Wars has the highest gross at $936,662,225.

3. The success of Star Wars is well explained by the investopedia:

   When Star Wars: Episode VII—The Force Awakens was released in 2015, after a gap of 10 years, Walmart ran an advertisement campaign titled “A New Generation of Fans is Born.” The retail behemoth had it right. The longevity and endurance of the Star Wars franchise mean that it spans a broad age range from a 4-year-old kid, newly introduced to the franchise through its latest movie installment, to the 70-year-old Grandpa reliving its thrills with the original. Targeting multiple demographics is also a smart strategy. It evokes a range of emotions—nostalgia, happiness, excitement—in fans.

   source: https://www.investopedia.com/articles/investing/102215/why-star-wars-franchise-so-valuable.asp

4. 3 out of 5 films are franchise or series films (Star Wars, and Marvel Cinematic Universe - Avengers). In general, franchises film have a large number of communities and fan base, so that can be a high contribution to the gross.

5. Almost all films were released after 2000, where the number of cinemas and access to films is increasing every year. This is also a factor in the high gross amount.

6. How about Avatar? The key to Avatar’s success lies in the year it was released in 2009, where in that year CGI was an extraordinary thing when compared to 2018 and above.

   Avatar came out in 2009 and took the world by surprise at just how far technology had come. It showed everyone (whether they were in the industry or audience) what levels CGI could be taken to. In 2021, it’s no big deal when a movie is mostly computer-generated, but many would argue that Avatar is the film that gave way for every film that followed it to push technology further than before. James Cameron said that this was a movie that he had been planning, creating, and waiting to make for years. What held him up for so long?

   source: https://gamerant.com/avatar-regain-spot-highest-grossing-film/

3.2 Top 5 High Rated Movies

• What 5 top movies are getting highest rating?

top5_rating <- imdb[order(imdb$IMDB_Rating,decreasing = T),][1:5,]
top5_rating

summary(top5_rating)

#>  Series_Title       Released_Year      Certificate          Runtime         
#>  Length:5           Length:5           Length:5           Length:5          
#>  Class :character   Class :character   Class :character   Class :character  
#>  Mode  :character   Mode  :character   Mode  :character   Mode  :character  
#>                                                                             
#>                                                                             
#>                                                                             
#>     Genre            IMDB_Rating    Meta_score    Director        
#>  Length:5           Min.   :9.0   Min.   : 80   Length:5          
#>  Class :character   1st Qu.:9.0   1st Qu.: 84   Class :character  
#>  Mode  :character   Median :9.0   Median : 90   Mode  :character  
#>                     Mean   :9.1   Mean   : 90                     
#>                     3rd Qu.:9.2   3rd Qu.: 96                     
#>                     Max.   :9.3   Max.   :100                     
#>     Star1              Star2              Star3              Star4          
#>  Length:5           Length:5           Length:5           Length:5          
#>  Class :character   Class :character   Class :character   Class :character  
#>  Mode  :character   Mode  :character   Mode  :character   Mode  :character  
#>                                                                             
#>                                                                             
#>                                                                             
#>   No_of_Votes          Gross          
#>  Min.   : 689845   Min.   :  4360000  
#>  1st Qu.:1129952   1st Qu.: 28341469  
#>  Median :1620367   Median : 57300000  
#>  Mean   :1617301   Mean   :151965265  
#>  3rd Qu.:2303232   3rd Qu.:134966411  
#>  Max.   :2343110   Max.   :534858444

ggplot(top5_rating, aes(x = IMDB_Rating,y = reorder(Series_Title, IMDB_Rating))) +
  geom_col(aes(fill = IMDB_Rating), show.legend = F) +
  labs(title = "Top 5 Movies based on IMDB Rating",
       x = "IMDB Rating",
       y = NULL) +
  geom_label(aes(label = IMDB_Rating), hjust = 1.05) +
  scale_fill_gradient(low = "red", high = "black") +
  theme_minimal()

From the data visualization above, we can conclude:

1. Top 5 High Rated Movies have an average gross of $151,965,265, with an average rating of 9.1.
2. The Shawshank Redemption has the highest Rating at 9.3.
3. As in Top 5 High Gross Film, the franchise film also has a high rating, such as The God Father and Batman (The Dark Knight). So the community and fan base becomes one of the important factors in contributing to film rating and revenue.
4. Unlike Top 5 High Gross, almost all films were released in the year under 2000. Why are old movies rated so high? This can happen because of the lack of a spectrum of reviews on the film. As we know, film review sites such as IMDB were only launched in 1993, and were popular in 1996 and above, so before that the spectrum of reviews for films was very limited.

3.3 Top 5 Highest Gross & Rating Movies

Based on data on Average Rating, the average value of film reviews fell each year, and start to level out a bit more in 1996, 1998, and 1999. Those are the respective launch years for IMDb, Rotten Tomatoes, and Metacritic. To get a more precise analysis result, the imdb data will be generated for 1996 and above.

imdb_slice <- imdb[imdb$Released_Year > 1996,]
head(imdb_slice)

• What is the IMDB rating range on the data?

unique(imdb_slice$IMDB_Rating)

#>  [1] 9.0 8.9 8.8 8.7 8.6 8.5 8.4 8.3 8.2 8.1 8.0 7.9 7.8 7.7 7.6

Based on the results above, the IMDB rating is in the range of 7.6 to 9.0. If it is assumed that the high rating is 8.5, then the top 5 films with the highest gross can be identified through the following command:

imdb_rate <- imdb_slice[imdb_slice$IMDB_Rating > 8.5,]
imdb_top5 <- imdb_rate[order(imdb_rate$Gross,decreasing = T),][1:5,]
imdb_top5

summary(imdb_top5)

#>  Series_Title       Released_Year      Certificate          Runtime         
#>  Length:5           Length:5           Length:5           Length:5          
#>  Class :character   Class :character   Class :character   Class :character  
#>  Mode  :character   Mode  :character   Mode  :character   Mode  :character  
#>                                                                             
#>                                                                             
#>                                                                             
#>     Genre            IMDB_Rating     Meta_score     Director        
#>  Length:5           Min.   :8.70   Min.   :74.0   Length:5          
#>  Class :character   1st Qu.:8.80   1st Qu.:84.0   Class :character  
#>  Mode  :character   Median :8.80   Median :87.0   Mode  :character  
#>                     Mean   :8.84   Mean   :86.2                     
#>                     3rd Qu.:8.90   3rd Qu.:92.0                     
#>                     Max.   :9.00   Max.   :94.0                     
#>     Star1              Star2              Star3              Star4          
#>  Length:5           Length:5           Length:5           Length:5          
#>  Class :character   Class :character   Class :character   Class :character  
#>  Mode  :character   Mode  :character   Mode  :character   Mode  :character  
#>                                                                             
#>                                                                             
#>                                                                             
#>   No_of_Votes          Gross          
#>  Min.   :1485555   Min.   :292576195  
#>  1st Qu.:1642758   1st Qu.:315544750  
#>  Median :1661481   Median :342551365  
#>  Mean   :1832014   Mean   :372675332  
#>  3rd Qu.:2067042   3rd Qu.:377845905  
#>  Max.   :2303232   Max.   :534858444

ggplot(imdb_top5, aes(x = Gross,y = reorder(Series_Title, Gross))) +
  geom_col(aes(fill = Gross), show.legend = F) +
  labs(title = "Top 5 Movies based on Rating & Gross",
       x = "Gross Revenue",
       y = NULL) +
  geom_label(aes(label = comma(Gross)), hjust = 1.05) +
  scale_fill_gradient(low = "red", high = "black") +
  theme_minimal()

From the data visualization above, we can conclude:

1. 5 top movies with highest rating and gross amount have an average gross of $372,675,332, with an average rating of 8.84.

2. The Dark Knight (Batman Franchise) has the highest rating and gross. The success of this film cannot be separated from the role of one of the actors, Heath Ledger who has been voted as the best Joker portrayal of the character of all time at LADbible.

3. The genre of top 5 films is Action.

4. Both The Lord of the Rings and Inception were directed by Christopher Nolan.

5. It’s surprising that the Star Wars and Avengers franchises don’t make the top 5. Although it has a high gross, it turns out that the film did not get good reviews.

6. Animation and CGI don’t seem to be enough to get high ratings, a film needs a rich plot and a lot of character development to attract audiences like The Lord of the Rings has done.

   The Lord of the Rings movies gave some characters more development and others a greater role in the story. The movies, too, therefore, deserve a lot of credit for bringing the story to a new audience.

   sources: https://gamerant.com/lotr-why-still-popular/