Continuum: John Mayer’s Career Visualized (Python, Tableau)
“One, two, one, two, three…”
I remember unwrapping the plastic wrap from the CD that I had just bought at the local mall, the whir of the car’s CD player as it loaded the disc and then hearing the opening words of “Waiting On The World to Change” as we pulled out of the mall parking lot. The album was John Mayer’s Continuum and it played a formative role in my musical evolution by fusing pop, soul and blues into an easily accessible format that would lead me to explore the lineage of those styles through other artists like Stevie Ray Vaughan and Sam Cooke.
However, the album differed considerably from Mayer’s previous work such as his debut studio album, Room for Squares (2001), which was largely acoustic and his later work such as the country/folk inspired Paradise Valley (2013). The degree of Mayer’s expansive exploration of different musical styles seems particularly unique so I was curious about visualizing his musical evolution through each of his studio albums. I used the Python library Spotipy to retrieve the catalog data (e.g. track duration, popularity) and audio features (e.g. energy, liveness) as determined by the Spotify API. You can explore the full interactive version here!
Some interesting insights I found:
Popularity is defined as how many plays each track on the album has received and how recent those plays are. Interestingly, Continuum is rated as more popular than his most recent album, The Search for Everything. It would be interesting to explore the weighting of the algorithm because it appears that the cumulative plays of Continuum are a greater influence on the metric than the more recent plays of TSFE.
The audio feature metrics are intended to quantify an artist but they also reveal how much the influence of a producer can affect those same metrics. For example, Heavier Things ranked first in both loudness and energy which I suspect is due to the influence of producer Jack Joseph Puig and the use of horns and loops on the album.
I was surprised by the results of some audio metrics and would be interested in learning more about what features most influence those metrics. For instance, I’ve always considered Room for Squares an acoustic album because of songs like “No Such Thing” (written about my actual high school!) and “Why Georgia” but it was actually ranked as Mayer’s least acoustic album.