Interesting write up! I was waiting for the movie tie in though. How much of the longevity of these songs is due to the fact they were highlighted in a movie? Maybe a future statistical analysis?
From a DJ's perspective: it used to be that being featured in a movie made a song a huge hit at parties. "Son of a Preacher Man" went from being an obscure song that would clear a dance floor to a high-energy floor-packing banger after being featured in Pulp Fiction, and there are countless other such songs in the '90s. "Soul Bossa Nova" by Quincy Jones comes to mind. In the digital age, people no longer watch a movie at the same time, so songs from films no longer have that effect. Streaming has exacerbated this.
Two pieces of information are necessary to give this analysis any credibility.
First is the geographic catchment. If the majority of Spotify users in the data set are from the USA, that gives an enormous bias. For example, Spotify (and for that matter Apple Music) will regularly 'force' rap and hip hop into any playlist – even if I have repeatedly notified them to "play less". In other words, the algorithm is constantly trying to load music it thinks is 'popular' into playlists that I have no interest in, for example, rap and hip hop. Is the geographic catchment known?
Second is the actual listening time of any piece of music. If 30 seconds is defined as being "listened to" then every rap and hip hop song that has been forced into my playlists could be in the data, especially when it can take me 30 seconds to reach the music-playing device and hit the "please, no more of this stuff" button, especially if, as usual, the iPad is the turntable and the iMac is the drawing board.
These are great points, and great next steps for an analysis like this. I think the work done here is still credible, for what it is. Like most good statistical analyses, this one leads to more questions. Good questions. Like your questions. Besides, it takes so much work to do what he already did! Why wait to share what he's found?
The chart of music acts that didn’t stand the test of time is interesting. Ricky Nelson and the Ames Brothers are favorites of mine, and George Strait is a favorite of seemingly everyone in Texas. Whether it’s a wedding or a high school dance, it’s rare that I don’t play at least a couple George Strait songs, and Brooks & Dunn’s “Neon Moon” is a must-play at just about any party here.
Blink-182 still packs *some* dance floors at weddings, night clubs, even teen parties, but I need to read the crowd, as it will flop at others. However, the song that does it is "All the Small Things." That will transform the right dance floor into a screaming, hugging, hands-in-the-air sing-a-long.
The one song that gets everyone onto every dance floor among every group of every age, and has nearly always been the highest energy moment of every event I've DJ'ed for the past 35 years is "Dancing Queen" by ABBA. Other songs have come and gone, but somehow that song has remained an absolute peak energy-building, floor-packing banger the entire time.
This is awesome! Thank you for it. I think classical music could make for a nice case study to compare with this. The question you ask about what would happen if music froze in time could be addressed there since new classical music (in the popular, traditional sense) stopped around the 1940s. But you could also do an analysis of the genres *within* classical. In 19th century Europe classical was pop! Classical stars like Liszt, Clara Schumann and Wagner generated wild frenzies at their concerts and were feted as superstars. And new classical dances were considered risque, hyper-sexualized and people dancing them were called basically all the things parents say about kids in every decade. Yet it evolved quickly, and splintered, across the 150 years from Mozart to, say, Rachmaninoff. Would be an interesting comparison.
Prog rock. Kept alive by passionate fans bit completely ignored by mainstream media. I’m finishing an article about narrative and literature in prog that mentions this.
My only pick with this is, that disco lasted considerably less than a decade, arising as a large cultural phenomenon around 1975-6, and being chucked out in about 1980 or so.
Looks like listening is random for a few years, as people's taste changes, so exponential, then decays linearly presumably as the diehard fans actually die. So we could expect the linear trend to continue to zero.
Love this. Dying to know why some songs that don't make the Billboard charts when they're released become popular years, even decades later. So many obscure 80s post-punk songs that never got radio play back in the day, but now you can hear them being played at the grocery store, or the mall. Why those songs? Why not others? So curious to me....
Interesting write up! I was waiting for the movie tie in though. How much of the longevity of these songs is due to the fact they were highlighted in a movie? Maybe a future statistical analysis?
From a DJ's perspective: it used to be that being featured in a movie made a song a huge hit at parties. "Son of a Preacher Man" went from being an obscure song that would clear a dance floor to a high-energy floor-packing banger after being featured in Pulp Fiction, and there are countless other such songs in the '90s. "Soul Bossa Nova" by Quincy Jones comes to mind. In the digital age, people no longer watch a movie at the same time, so songs from films no longer have that effect. Streaming has exacerbated this.
Two pieces of information are necessary to give this analysis any credibility.
First is the geographic catchment. If the majority of Spotify users in the data set are from the USA, that gives an enormous bias. For example, Spotify (and for that matter Apple Music) will regularly 'force' rap and hip hop into any playlist – even if I have repeatedly notified them to "play less". In other words, the algorithm is constantly trying to load music it thinks is 'popular' into playlists that I have no interest in, for example, rap and hip hop. Is the geographic catchment known?
Second is the actual listening time of any piece of music. If 30 seconds is defined as being "listened to" then every rap and hip hop song that has been forced into my playlists could be in the data, especially when it can take me 30 seconds to reach the music-playing device and hit the "please, no more of this stuff" button, especially if, as usual, the iPad is the turntable and the iMac is the drawing board.
These are great points, and great next steps for an analysis like this. I think the work done here is still credible, for what it is. Like most good statistical analyses, this one leads to more questions. Good questions. Like your questions. Besides, it takes so much work to do what he already did! Why wait to share what he's found?
The chart of music acts that didn’t stand the test of time is interesting. Ricky Nelson and the Ames Brothers are favorites of mine, and George Strait is a favorite of seemingly everyone in Texas. Whether it’s a wedding or a high school dance, it’s rare that I don’t play at least a couple George Strait songs, and Brooks & Dunn’s “Neon Moon” is a must-play at just about any party here.
Blink-182 still packs *some* dance floors at weddings, night clubs, even teen parties, but I need to read the crowd, as it will flop at others. However, the song that does it is "All the Small Things." That will transform the right dance floor into a screaming, hugging, hands-in-the-air sing-a-long.
The one song that gets everyone onto every dance floor among every group of every age, and has nearly always been the highest energy moment of every event I've DJ'ed for the past 35 years is "Dancing Queen" by ABBA. Other songs have come and gone, but somehow that song has remained an absolute peak energy-building, floor-packing banger the entire time.
This is awesome! Thank you for it. I think classical music could make for a nice case study to compare with this. The question you ask about what would happen if music froze in time could be addressed there since new classical music (in the popular, traditional sense) stopped around the 1940s. But you could also do an analysis of the genres *within* classical. In 19th century Europe classical was pop! Classical stars like Liszt, Clara Schumann and Wagner generated wild frenzies at their concerts and were feted as superstars. And new classical dances were considered risque, hyper-sexualized and people dancing them were called basically all the things parents say about kids in every decade. Yet it evolved quickly, and splintered, across the 150 years from Mozart to, say, Rachmaninoff. Would be an interesting comparison.
Doing my part to keep the Paul Davis and .38 Special legacies alive.
"ABBA's "Money Money Money" generated ~2.8 million plays 28 years after its debut"
Am I misunderstanding something? That song is 48 years old, not 28.
"2 Legit 2 Quit" also looks to be about 20 years off.
I am 100% surprised that “Mr. Jones” was not mentioned in this article. That song will not go away. But maybe that’s a radio vs Spotify thing.
Prog rock. Kept alive by passionate fans bit completely ignored by mainstream media. I’m finishing an article about narrative and literature in prog that mentions this.
My only pick with this is, that disco lasted considerably less than a decade, arising as a large cultural phenomenon around 1975-6, and being chucked out in about 1980 or so.
Macarena is outdated in the US? You learn something new everyday.
Last year, my class (Poland) made essentially the entire school dance. HUNDREDS of teens/very young adults!
One day Spotify created a playlist called "Pop Punk's Not Dead," and it was an instant winner for me.
Looks like listening is random for a few years, as people's taste changes, so exponential, then decays linearly presumably as the diehard fans actually die. So we could expect the linear trend to continue to zero.
👍 🎸🎶🍻
Love this. Dying to know why some songs that don't make the Billboard charts when they're released become popular years, even decades later. So many obscure 80s post-punk songs that never got radio play back in the day, but now you can hear them being played at the grocery store, or the mall. Why those songs? Why not others? So curious to me....
I found this very interesting but how can this be written without mentioning The Beatles?
This dataset predates The Beatles being on Spotify. Which is a bummer.
This was fun. Can you do poetry by decade next? :P