Cannibal Corpse. Again.

Table showing a list of musical artists and their statistics from Spotify and Last FM. Cannibal Corpse is in third place when sorted by listener to playcount ratio.
Cannibal Corpse fans can’t get enough. Good thing there aren’t any all-you-can-eat corpse buffets.

I added a new “ratio” column to my web app because I think it can be misleading to look only at Listeners, Followers, and Playcounts when comparing artists (or albums &tracks). Many people will buy an album or follow an artist because they’re popular at the moment then never listen again. Even just plays can be misleading so I wanted to know how often followers/listeners actually played that artist’s music.

I expected Bowie and Zeppelin to be at the top when sorting by ratio, but Cannibal Corpse?!

Even when comparing CC to artists with similar stats in other columns (below), they beat the snot out of … Journey, Motley Crue, Ozzy, Clapton, and Tom-Freakin-Petty & the Heartbreakers! Tom Petty solo (think “Free Fallin”) is even lower at 1:13.

Table comparing Cannibal Corpse with artists having similar statistics for Last FM playcount.

The brand-new next item on my Data Visualization to-do list is now to graph which genres have the highest listener-to-playcount ratio. I hypothesize it’s metal. According to my local record store owners, used metal on vinyl is more expensive and yet still sells faster than other genres and they absolutely can’t keep Christian Metal from flying off their shelves.


Volume 4’s illusion of unpopularity

Scatterplot showing studio album playcounts fetched from the API. X-Axis is Black Sabbath's career over time. Y-Axis is playcounts from 460k to 15 million.
Album playcounts fetched from the API. X-Axis is Black Sabbath’s career over time. Y-Axis is playcounts from 460k to 15 million.

In a recent Reddit post, I showed a scatterplot (above) showing each Black Sabbath studio album’s playcount from LastFM. Volume 4 was the absolute lowest — even beneath every Ian Gillan and Tony Martin album (no disrespect to either, I’m just talking data and conventional perception). I and most commenters were surprised so here are some more screenshots explaining how that happened.

Table showing the same data from June 19. The only albums lower than Vol. 4 are obscure bootlegs.

Looking at my column chart for their Spotify popularity scores (as of today), I noticed one release of Vol 4 has a popularity score of 2 while the others are more respectable and what we’d expect.

Spotify contains several releases of each album. Some countries get even more than USA users.

When I get my data from MusicBrainz‘s API, it looks like this:

Children of the Grave is a bootleg of Vol 4 so it is the same “release group” from MusicBrainz.

MusicBrainz is where I get info about artists and all their albums and tracks. I’ve collapsed the data for each album so you can see there are 12 total versions of Vol 4. I push that list to LastFM to pull the number of listeners and playcounts for each album and track.

A “release group” is an album. A “release” is each edition or version of that album — Remastered, collectors’s edition, comes with a poster, whatever.

LastFM doesn’t have data for all releases but sends what it does have. For reasons I won’t bore you with, I only grab data for the first release in each release group which, in this case, is that f***ing Children of the Grave bootleg. As I write this, I have a solution in mind that should also fix another oddity in my app.

I hope you’re nerdy enough to have enjoyed this.