Comparing Categories Using Gestalt Principles in Music Data
Rolling Stone’s 500 Greatest Albums of All Time attempts to rank the world’s best music albums based on the votes of music critics, artists, and industry figures. Originally published in 2003, the list has undergone several revisions, with the most recent update occurring in 2023. A look at the dataset summary reveals that this is the 2012 revised edition, as the most recent album release year is 2011.
Number Year Album Artist
Min. : 1.0 Min. :1955 Length:500 Length:500
1st Qu.:125.8 1st Qu.:1970 Class :character Class :character
Median :250.5 Median :1976 Mode :character Mode :character
Mean :250.5 Mean :1979
3rd Qu.:375.2 3rd Qu.:1988
Max. :500.0 Max. :2011
Genre Subgenre
Length:500 Length:500
Class :character Class :character
Mode :character Mode :character
I chose to analyze two aspects of the dataset: count of artists with albums classified as Hip-Hop, and the rankings trend of The Rolling Stones albums over time.
Hip-Hop Artists with the Most Albums in the Top 500
Filtering the Genre column for albums classified as Hip-Hop returned 29 rows.
# A tibble: 29 × 6
Number Year Album Artist Genre Subgenre
<dbl> <dbl> <chr> <chr> <chr> <chr>
1 48 1988 It Takes a Nation of Millions to Hold Us … Publi… Hip … Conscio…
2 118 2005 Late Registration Kanye… Hip … None
3 123 1986 Raising Hell Run D… Hip … None
4 134 1994 Ready to Die The N… Hip … Thug Rap
5 138 1992 The Chronic Dr. D… Hip … Gangsta
6 144 1988 Straight Outta Compton N.W.A Hip … Gangsta
7 153 1991 The Low End Theory A Tri… Hip … Conscio…
8 219 1986 Licensed to Ill Beast… Hip … None
9 228 1987 Paid in Full Eric … Hip … None
10 242 1984 Run D.M.C. Run D… Hip … None
# ℹ 19 more rows
Sorting by Artist count revealed that Kanye West and Jay-Z were tied for the most Hip-Hop albums in the Top 500, with 3 each. I chose is a standard horizontal bar chart to visualize the data, with artist name labels on the Y axis for better legibility. I also added a default fill aesthetic for the 21 discrete artists in the set.
Genre labels for each album in the dataset can be combined (eg, the Beastie Boys’ Paul’s Boutique is labeled as Hip Hop, Rock, Funk/Soul, therefore not included in the table), so an improved visualization could include any combined genres in the dataset that include Hip-Hop.
My visualization includes minor gridlines and a light grey background, but it isn’t possible for artists to have half of an album on the Top 500. A better visualization could keep the grey background, but remove the minor gridlines for less visual noise.
The Rolling Stones Album Rankings in the Top 500
Filtering the Artist column for albums by The Rolling Stones, (a partial namesake of Rolling Stone itself) returned 10 rows. The Rolling Stones are tied with with the Beatles for the second most albums in the Top 500.
# A tibble: 10 × 6
Number Year Album Artist Genre Subgenre
<dbl> <dbl> <chr> <chr> <chr> <chr>
1 7 1972 Exile on Main St. The Rolling Stones Rock Blues R…
2 32 1969 Let It Bleed The Rolling Stones Rock Blues R…
3 58 1968 Beggars Banquet The Rolling Stones Rock, Funk… Blues R…
4 64 1971 Sticky Fingers The Rolling Stones Rock Classic…
5 109 1966 Aftermath The Rolling Stones Rock Blues R…
6 116 1965 Out of Our Heads The Rolling Stones Rock Blues R…
7 180 1965 The Rolling Stones, Now! The Rolling Stones Rock, Blue… Pop Roc…
8 213 1981 Tattoo You The Rolling Stones Rock Classic…
9 270 1978 Some Girls The Rolling Stones Rock Blues R…
10 357 1967 Between the Buttons The Rolling Stones Rock Blues R…
My goal for this visualization is to see if The Rolling Stones’ album rankings on the Top 500 increased or decreased over time.By adding a trendline over a scatter plot of the table data, I can see that there is a slight decline overall from their earliest ranked album, 1965’s The Rolling Stones, Now!, to their latest ranked album, 1981’s Tattoo You. This might indicate a creative decline in the band’s output since 1972’s Exile on Main St., their highest ranked album. As the Y axis limit is large (0 to 500), I labeled the points with the numeric album ranking for better legibility. I also reversed the Y scale, so the higher ranked albums would appear closer to the top of the scatterplot.
Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
ℹ Please use `linewidth` instead.
`geom_smooth()` using formula = 'y ~ x'
As noted in Better Data Visualizations, scatterplots can be difficult for readers to fully understand, even with a trendline showing the general direction of the relationship between the points.
The visualization is also missing album titles; I had a difficult time labeling the points with the entire album names in a way that was readable, so I left the ranking labels only.
Conclusion
I chose two aspects of the Top 500 that are unrelated to each other for the most part - a variety of hip-hop artists, and a single rock band with ranked albums in three different decades. However, the wealth of information in the dataset could show how the waxing and waning popularity of the two genres over the decades affects rankings. Does the Rock genre, which dominates the list with 249 entries, start to disappear by the time hip-hop’s popularity starts increasing in the 1980s? Comparing the differences between the original 2003 list and its subsequent revisions could also reveal trends in music criticism over time.
