Drumming Upstream Season One: WRAPPED
Let's crunch some numbers
This week, instead of sharing my thoughts on, and performance of, a song from my Spotify Liked list, I thought I’d get in the spirit of the season and summarize the project’s first full-ish year.
If you spent time on social media last week you probably saw about as much Spotify branding as a human being can stand. The app, as they have done each of the last six years, served each of its users a personalized summary of their listening habits from the last 10 months. Spotify Wrapped, as this feature is called, is designed to make Spotify users feel good, or at least unique. The app congratulates the user on their dedication to their favorite artists, the range of genres they listen to, and the raw minutes of audio that they ingested. For example, my Spotify Wrapped claims that I listened to more music than 86% of US Spotify users. I believe it, but without any data about the rest of the world I’m not sure exactly how proud of that stat I should be.
Accompanied as they are by bright and colorful graphics, these statistics are designed to be shared on social media. And boy howdy do they get shared! My Instagram feed was thick with friends and acquaintances patting themselves on the back and/or cringing at their own listening habits. Meta-commentary on the trend arrived in my feed near simultaneously. Curmudgeons chided users for expressing their supposed “unique” taste (itself a pick-me way of expressing individuality, but hey, go off), others cracked jokes at Apple Music and Tidal users for not being able to join in the “fun” (karmic retribution for Apple users’ performative disgust for the green text bubbles of Android owners), while musicians struggled to preach to anyone but the choir about how paltry streaming payouts are.
All of this, including meta-meta commentary like this blog post, serves as an extended marketing campaign for Spotify. Given how far Spotify’s stock has plunged in the last year I’m sure Ek and co. are grateful for all the hard work we’ve put in to keep them top of mind. Look, I wear band shirts, clearly I’m not in principle too good for paying to advertise someone else’s product. As per usual my issues with Spotify are more obtuse and ungainly.
First, as I tried to articulate last year, frequency does not imply favoritism. Spotify tells me that my favorite song this year was “Virginal II” by Tim Hecker, a song I streamed on-repeat while writing about it for this very newsletter in order to prove a point about how even harsher styles of ambient music can be turned into background music. Anyone who’s read Drumming Upstream this year knows that track isn’t even in my top 10. I may be a statistic outlier as far as Spotify is concerned, but I saw similar versions of this miscalculation play out all over my feed. If you had an unobtrusive playlist that you put on in the background while working, guess what? All of those songs are headed straight to the top of your most played list regardless of whether you could match a name and face to the tune.
That is of course, if you forgot to turn on private mode. Users concerned about how their Wrapped might look at the end of the year can prevent the app from tracking their listening habits by going “private”. I’m sure that this data is still tracked somewhere in Spotify’s servers, but for the sake of your pride they’ll exclude any incognito listening from your end of year summary. Without knowing who is hiding what it is impossible to take any of Spotify Wrapped’s conclusions at face value. And speaking of incomplete data sets, what about all of the music that people listen to in December? Are Spotify waging a war on Christmas music? Do they want to spare my results from the Stevie Wonder marathon I’m in the midst of?
Finally, Spotify only gives its users the most basic of counting stats and papers over the more complex data with vagueness. I am certain that Spotify keeps track of the tempo of the music that you listen to, and this latest Wrapped admits to tracking your listening habits against the time of day. Why not show what tempos I prefer when, instead of reducing those habits to lifestyle-branding-soup? Why not tell me where and when my favorite songs are from? All I’m asking is for Spotify to tell me something that I don’t already know.
I have never been less surprised by Spotify’s raw numbers than I was this year. All five of my top songs have entries in Drumming Upstream, as do each of all but one of my Top 5 artists (the fifth was the subject of an early 2022 episode of Lamniformes Radio). The only data point that raised my eyebrows was that the genre Broadway, standing in for the two musicals I had to learn in 2022, hung on at the 5th spot of my Top 5 Genres with only two months worth of plays.
Lucky for me, I have my own data to draw from. In the interest of filling in the gaps where Spotify cannot, and to commemorate the first year of many undertaking this ridiculous project, I present to you: Drumming Upstream Wrapped
Fair warning: I am not a statistician or a data analyst. I just think numbers are neat and I like grouping and regrouping pieces of music to see if any connections emerge. Most of the time I walk a with little more than a Westbrookian “that’s pretty interesting” but I always have fun digging around. After the number crunching, I’ll offer a few subjective superlatives and reflect on the project as a whole.
Total Number of Songs: 21
Total Number of Distinct Acts: 21
Total Drumming Upstream Word Count: 62,939
Songs Learned on Drums: 16
Songs with Live Drummers: 9
Songs with Produced Drums: 7
Drumless Songs: 5
Average Song Length: 6:54
Longest song: 16:32
Shortest song: 3:04
Average BPM: 96 bpm1
Average Word Count: 2,997
Countries Represented (# of Artists):
The United States of America (12)
United Kingdom (3)2
France (2)
Canada (1), Germany (1), Norway (1), Sweden (1)
Genres Represented (# of Songs):
Heavy Metal (4)
Ambient (4)
Rap (2)
Rock (1), Dance (1), Classical (1), R&B (1), Folk (1)
Years of Release (# of Songs): 1912, 1982, 1984, 1995, 2001, 2002 (2), 2004, 2008, 2010, 2011, 2013 (4), 2014 (3), 2015 (3)
Best Written Entry: “Écailles De Lune (Part 1)” by Alcest
Although the song itself didn’t crack the top 10, I think this entry represents everything I want to accomplish in Drumming Upstream. The balance between the personal, historical context, lyrical analysis, and music theory is spot on. The flow from paragraph to paragraph is the smoothest I was able to accomplish this year. And it’s got some solid one liners! If you want my recommendation for which entry to send someone who might be interested in the project, this is the one.
Worst Written Entry: “Covered In Writing” by Eluvium
One lesson I’ve learned from this project is that I do not have much to say about ambient music. Or at least, I have yet to develop the skill of writing about ambient music. I think I made some strong observations about the genre in my Ehlers, Hecker, and Barwick pieces, but by the time I got to this Eluvium track I was running on empty. I’m honestly not sure what the hell I was talking about in this one. This is the only entry that felt like an obligation to put together. If you got something out of this piece, congratulations, I didn’t.
Best Performed Entry: “Keep You On My Side” by Chvrches
In general my drumming got better as the year went on. That’s good! It means this project is working. This song had the added benefit of being recorded to a click, which meant that it was way easier to lock in with the recorded part. So even though I think “Hallowed Be Thy Name” is my favorite performance, “Keep You On My Side” gets the slight technical edge.
Worst Performed Entry: “Bolivia ‘95” by Scott Walker
This song was the first real challenge of project both practically and conceptually. Grappling with that challenge made me a better drummer, and a better appreciator of Walker’s work, but the fruits of that labor didn’t show up until after I moved on from this song. I view this performance as a necessary failure, one that showed me what I needed to improve on going forward, but it is not a recording I enjoy revisiting.
Since this is only the first year of the project, I’m not sure how proud to be of turning in 21 entries. A full year of 52 entries was never realistic. There were tours to go on, two musicals to learn, an unexpectedly broken snare, and a case of COVID to contend with. Even under ideal circumstances, it took a while for my production process to get into the groove. When I started this project in January I had no experience with video editing or heck, even video shooting. A year later, the video is usually faster to produce than the written portion.
Speaking of the written portion, 63k words!! Wow! That’s two The Stranger’s worth of words. While I never had a target length for these letters going in, an average of 3k feels right. Still, at this rate it’ll take me another 1,377,000 words to finish the remaining 459 songs. That’s as many words as Infinite Jest, Gravity’s Rainbow, and The Recognitions combined. Yikes! While many of the songs left to cover will take 3k or more words to tackle, I hope that I’ll learn the discipline to keep the pieces that don’t need to be that long from ballooning.
I’m glad to see that the drumless songs only make up a small chunk of the overall material. I tried to hold myself to an unspoken rule of keeping the percentage of drumless songs beneath 50% for the project. That doesn’t look like it’ll be a problem any time soon.
I’m going to keep my eye on that live vs produced drummer split, although if I had guess live drummers will maintain a solid lead for the whole project.
I’m tickled pink by the composite song suggested by the average length and tempo. A near 7-minute song that never cracks 100 bpm? Are you sure it’s not a song from Sisyphean? I expect this average length to drop over time, but not by much. There are some Big Boys coming up next season that’ll keep us away from radio-friendliness for a long time.
No surprise that the USA and the 2010s are overrepresented in the country and year fields, considering that this data set was built by an American during the 2010s. It is worth noting that the songs from the 2000s make up the bulk of the top 10 on the Leaderboard. Also not a surprise, since that decade covered my teen years. Seeing Ravel on the far end of the chronological spectrum makes me chuckle. Incredibly, there’s an even older piece of music coming down the pipeline, although I might not get to it for a few years.
To better incentivize my continued progress through my Liked list, I’ve decided to allow myself one act of re-organization on the Leaderboard for every ten songs that I complete. That means this year I have two chances to move a song up or down on the Leaderboard.
I’m going to use my first re-organize token to drop “Plays John Cassavetes 2” by Ekkehard Ehlers down to number 8, below “Delorean Dynamite” by Todd Terje and above “Covered In Writing” by Eluvium. While I’m still a big fan of the piece, I think I let my fondness for the memory of hearing the track for the first time inflate my appreciation for it in the present. If I’m being honest with myself, the less cerebral tracks that sat below Ehlers for most of the year do more for me these days.
For my second act of re-organization, I’m going to bump “Bolivia ‘95” by Scott Walker up exactly one spot. It’s a minor change, sure, but it’s one that I think will make future ranking easier.
Speaking of future entries, I’m already knee deep in research and rehearsal for the next batch of songs. My plan is to stock up some easier tunes in December so that I can hit the ground running in January with a backlog of videos to draw from in case any unexpected illnesses or gear malfunctions come my way. Working chronologically the next batch of songs is heavy on, well, heavy stuff. There are both personal and professional reasons for the spike in metal and hardcore in my Liked list, which I’ll explain in due time. However, in an effort to keep us all from burning out on high volume guitars I’m going to attempt to front load some non-metal tracks. Think of it as a way to ease into the new year.
As for the remainder of this old year, I’ll have a few more round-up style pieces coming your way before the ball drops. Thanks again for reading. More soon.
Only songs learned on drums were used for this data set. Not worth it to go back and trace out the tempos for the songs without drums. For songs with multiple or varying tempos I found a median and excluded extreme outliers that didn’t represent a majority of the song’s tempo.
Though he was born in the United States I’ve included Scott Walker in the UK for this census. Walker’s music career took place while he lived in England and to my knowledge he never moved back to the States after splitting town. Walker made British music, is what I’m trying to say. Chvrches, from Scotland, also fall into the UK… FOR NOW.
For the purposes of this newsletter I’ve decided to classify Have A Nice Life and Black Wing as punk acts on the basis of Dan Barrett’s history in emo/post-hardcore, HANL’s obvious debt to post-punk, and for the deliberately lo-fi/lo-budget production origins of the music.
Once again, Scott Walker makes classification a headache. As weird and un-radio friendly as “Bolivia ‘95” is, it still makes the most sense to think of Walker as a pop artist, writing and singing his version of pop songs. That they are not as popular as other pop does not change that context.