Since February this year, I’ve been creating a playlist for every month of the year with the music that I especially enjoy that month and it’s been great.
In college more than ever before, songs will attach themselves to memories. Every month or two, make a mix cd, mp3 folder, whatever – just make sure you keep copies of these songs. Ten years out, they’ll be as effective as a journal in taking you back to your favorite moments. 
The music that is in these playlists is unquestionably music that I like. But why do I like this music? I find myself unable to answer this question. How do I know I like a particular song enough to add it to this month’s collection?
I currently add a song to this playlist when it evokes a ‘this should be on my playlist’ feeling. I call this Feeling™. I pause whatever I’m doing to open Spotify and add it to my playlist.
Sometimes it occurs that the song I’m trying to add is already in my playlist so I get a - “This song is already on your playlist” warning. I love this warning. This implies that not only did I completely forget that I listened to this song before, it also managed to evoke the same “this should be on my playlist” feeling for a second time.
Here’s what I wanted to know about this particular feeling™
- Can I trust it? How reproducible is it?
- If it’s not reproducible, is it mood-based? This goes directly against reproducibility because if it’s mood-based then it can’t be reliably reproduced. Unless I reproduce the mood too.
I came across an opportunity to test this.
A friend with similar music tastes shared her Spotify-Top-100-songs-in-2019 playlist with me.
Normally, when someone shares their playlist with me, I listen through it once and add the songs that evoke Feeling™ to a separate playlist.
This time, I decided to do the above twice. Listen through it twice and add the songs that evoke the Feeling™ to different playlists. And then compare the two.
I was interested in finding the percentage of common songs. That is, which songs reliably reproduced the Feeling™ twice. If the number was high then it would mean a high rate of reproducibility and that I can trust this feeling to be somewhat reliable and continue using it.
Some notes on avoiding bias
One strategy would’ve been to simply add ALL or have a lower/more-liberal criteria for adding songs in Playlist 2. The relative count of the songs in each of the playlist also matters. I should strive to be just as selective.
Avoid looking at the name of the song before you make the decision to add it or not. The recognizability of the name might tip you off whether it was already in the earlier playlist.
I only got the idea for this towards the end of Playlist-1. I may not have been paying full attention to all the songs and end up paying extra attention during the second listen. I also may have accidentally skipped some songs over due to sleep and missed making the judgment of whether they needed to be added or not.
Original Playlist contained 100 Songs
Playlist 1: 21 Songs
Playlist 2: 58 Songs
The results suggest that I appear to have fallen into the first trap of subconsciously ending up with a more liberal-criteria when adding songs the second time. The extra attention definitely impacted this too.
19/21 songs are common in both of them. That is 90%.
I listened to some of the other 37 songs that weren’t in playlist 1 and I definitely enjoyed them. So either they didn’t evoke the same feeling the first time around or I was not paying attention or I had a higher criteria or they grew on me in the meantime.
I’m not too sure what to make of these results. One thing is certain, when I lower the bar, all the songs that previously evoked the feeling still evoked the feeling.
Other questions to ponder
- Why did those 19 songs in particular evoke this feeling reproducibly? Why do people like the music that they do? Are music tastes innate? Are music tastes heritable?
- How malleable are music tastes? Can you force yourself to like music that another person likes? If a person you like shares music, will you subconsciously be more inclined to alter your tastes? Based on an anecdote of one instance, I can say that moderate amount of like-ness for a person did NOT influence my music tastes to any appreciable degree. Still not a fan of the songs on her playlist.
1 - Here’s one other person who does this - https://frankchimero.com/listening/
2 - Earlier in the year, I similarly went through a playlist with 68 songs. It appears I added 23 of them to the monthly playlist. That is 34%.
3 - That’s it. That’s my justification. “Why do you like this song? It evoked a ‘this should be on my playlist’ feeling.” Is this odd? Don’t people like music they can relate to? I don’t relate to “hope is a dangerous thing for a woman to have” but I still enjoyed the song and it very much deserves its place on my monthly playlist. Why though? Isn’t this how this is for everybody?