Skip to content

ChristianFMadsen/spotifyPlaylistAnalysis

Repository files navigation

Analysis of my playlist

I spend a lot of time every day listening to music and so I thought why not try to combine my love for data and music and see where that leads me. I have a number of different metrics at my disposal that I believe can give an insight into what kind of musical taste I have as well as more behavioural statistics. The playlist doesn't have any particular mood it is just a collection of songs that I like and currently has >3000 songs on it. From my most common artists and genre analysis it is clear that I mostly listen to hip hop / rap / rock / metal / EDM. The average duration of a track is a little over 4 minutes with some tracks particularly by TOOL pulling the average up. I started using spotify in late 2011 however I added most of my songs from 2015-2017 and it also looks like I have added a good amount of songs so far this year. Mid 2015 is the year that i started studying physics at Aarhus University and as a natural by product I met a lot of new friends in that time period and went to a lot of social gatherings which exposed me to all kinds of new music. As for 2023 I've found a lot of new music by watching streams and hearing other peoples playlists and I have also really gotten into drum'n'bass (dnb) this year. Most of the songs that are on my playlist has been released between 1995 and 2023 which is essentially my lifespan. My guess is that this is a pretty common pattern for most people as we're naturally exposed to more music from the time period that we live through. I also have a fair amount of songs from the 60's 70's and 80's but those are almost exclusively from very famous artists (eg. Queen / Rolling stones / AC/DC) whereas the songs from '95-2023 are more of a blend.

Distribution of various song metrics

I have also tried to look at the distribution of some metrics that spotify provides for each song based on their own analysis, these are: Danceability, Energy, Loudness, Tempo, Valence and Popularity.

  • Danceability is a number between 0 and 1 that tries to quantify how suitable a track is for dancing based on for example tempo, rhytm stability etc..

  • Energy is also a number between 0 and 1 representing the perceived intensity and activity in a song; a typical energetic track is fast, loud and noisy.

  • Loudness is the overall loudness of a track in dB with typical values in the range -60 to 0dB.

  • Tempo is an estimation of a tracks beats per minute (BPM).

  • Valence is a number between 0 and 1 which describes the mood conveyed by a track in terms of positivity, so high valence tracks sound happy/cheerful/euphoric and low valence tracks sound sad/depressed/angry.

  • Popularity is a number between 0 and 100 where 100 is most popular. It is mostly based on total number of plays and how recent those plays are thus songs that are being played a lot now will generally have a higher popularity compared to songs that were played a lot in the past.

The majority of the songs on my playlist have a danceability ~0.4 to ~0.8 so pretty high which is mostly due to my fondness of EDM where that is the express purpose of most songs, however despite this I am a terrible dancer. In terms of energy the majority is above 0.5 with a lot of songs in the 0.8-1 range. This isn't really a huge surprise either considering the genres that I listen to a lot are metal and EDM which is usually high energy. I specifically remember being blown away by the energy in most of Rage Against the Machine's songs which is one of the main reasons that they are one of my favorite artists. The whole spectrum of valence is represented in the playlist with the majority being somewhere around the middle i.e. most songs are somewhat neutral in terms of being considered positive/negative. The loudness for most of my songs seems to be pretty located at the upper end of the loudness range between -10 to 0dB. Again, not really surprising as most metal and EDM (in particular dnb) tracks are very "active" and therefore perceived as being loud. Most tracks tempo is between ~90 to ~140 with a noticeable spike around 170. The spike around 170 is due to my new-found love for dnb where tracks are usually at or around 174 BPM. It should be noted that whatever their method is for estimating BPM it is not infallible as all of the songs in the top 5 for tempo are reported to have over 200 BPM but none of them are nowhere near that. Finally in terms of popularity a lot of songs are at 0 with the rest distributed relatively evenly up to around 90. A lot of the EDM and some hip hop that I listen to aren't really very popular in the grand scheme of things so I am not surprised that there is a big spike at 0.

So to sum it all up, most of the songs that comprise my playlist are high energy, loud, danceable, neutral in terms of valence and not too popular according to the metrics provided by spotify. Considering I mostly listen to hip hop / metal / EDM that seems reasonable.

Releases

No releases published

Packages

No packages published

Languages