One Time or Recurring Payments
Become a Patron of the Retro Arts
How the chart is objectively compiled
Notes from previous iterations of chart, circa Early 2021 to August 2021
— The previous version of the chart reflected everything “Synthwave/Retrowave/Outrun” released since Jan 1st 2021
– While comprehensive, the chart never really changed from week to week. No one wants to see what is essentially the same chart being released every week when all the big artists stay on top all year long.
– One of the reasons why this happens is that there is a huge disparity between the top/popular artists/tracks and pretty much everyone else who all trend on average so much lower than the
– Cannons and The Midnight are touring bands. Cannons just played Lollapalooza. That’s not something that most everyone else is doing, and the numbers for the whole year reflect this.
Notes detailing methodology, newest iteration – August 2021
— I changed the Methodology of the chart to reflect the last 4 weeks of New Releases
– New Releases meaning the tracks/releases must have been released from the time the data was scraped and 4 weeks ago, specifically starting at August 1st for this first iteration of this new chart
– I trialed and compiled versions of this chart going back 1 week, 2 week, 3 weeks, and 4 weeks, each time being manually processed, which is a ton of time, but I felt like I needed to do that because I really want to stop having to massage this “product.” One final methodology to rule them off, at least for now, because I’m not an expert programmer and can’t currently hire anyone else to do this very specific methodology, though I think this is a fantastic case study to continue to learn programming
– 4 weeks was the timeframe I’m going with because I want the chart to have a certain amount of turnover while also being a Top 100 chart. 4 Weeks of data ensures that I have at least 100 tracks plus the opportunity to add extra for the Supercharged version
– As for the Supercharged version, I should now be able to deliver an extra 100 tracks on top of the core 100 due to the self-imposed limit of 4 weeks. More weeks, more music on the Supercharged list. The number of extra Supercharged tracks can fluctuate depending on how much new music is released each week.
– As of 9.15.21 there are more than enough tracks being released to support a full Top 200 for the Supercharged version
Data Sources / I’m kind of typing this as a step by step process if
you follow along
– I scraped every single playlist that has Synthwave or Outrun or Retrowave in either its playlist title or
within its description. For example, this might come up “Unlimited Synthwave Drive List” or “The best retrowave music to drive around to at midnight”
– Number of playlists with those keywords: 12,220+ (the + is 3, it’s 12,223). For the most part part, this means ALL of the Syn/Ret/Out playlists that currently exist on Spotify
— From those 12,220 playlists, I culled it down to all lists that have 50 followers or more. The point of this is because the core theory of the chart is that we are using the metric of “added to influential playlists” to help guide us to find the most played/popular. 50 is a decent number because more than that I might cut off the chance of using a list that even a few other people follow, less than 50 and I have too many lists and just way too much information featuring outside genres
– After culling down the12,220, we’re left with 883 playlists with over 50 followers
— From this list, I sat down for about 20 hours total and opened each and every individual playlist to make sure that each playlist was 1 relevant to our genres and 2 update frequently enough
— Pure data scraping cannot help this part of the process. I was sitting there and checking out whether each list was primarily Synthwave/Outrun/Retrowave (some of them were like “Songs from
Stranger Things Soundtrack” “Songs from Drive Soundtrack” that aren’t new synthwave)
— The other thing was looking for, and this is an absolute necessity to verify by eyeballing each list by hand, is whether the owner of the list updates it frequently/often enough to even matter. So many lists are like “Best Synthwave of 2016” and feature all tracks from 2016, many even started to be an updated thing but then stopped updating October 2020
– There’s no point using data from a list that doesn’t update frequently enough. And I keep saying
enough because some don’t update every week but update every 2 weeks, or update in batches every 9 days or every month, etc. The mindset of this task is to keep a more open mind and intuition when determining whether the list is updated frequently enough to be usable. I’m pretty certain I have more data than I need, but that’s better than having less data, I’m trying to be super
accurate and more usable relevant data helps.
– Now I have a core list of lists using the methods above that both fall into the categories of Has Followers and is Somewhat Influential and Updates Frequently Enough to supply usable data
— The number of tracks that my core list of lists supplies me with is 39,773 (as of the last data scrape on 8.21.21 when I wrote this)
– From the 39,773 tracks, I then sort by date and lop off any releases older than Jan 1st 2021, which gave me 4,707 tracks to work with for this year
– Those 4,707 tracks can then be sorted by date timeframes, recent plays/popularity, numbers of plays, bpm, key, etc (this data comes along when scraped)
— The past 4 weeks of data includes 315 tracks, which I manually sit and sort through (to make sure irrelevant tracks aren’t included, i.e. The Weeknd or Elton John have had tracks and remixes posted to playlists in the genres, but are not necessarily an entity in our community
– I’m not a programmer so each time I make a chart, I have to sit there and just make sure that each track is or isn’t Synthwave/Retro. Data doesn’t really cover this anyway, this part of the process is also where creative experience and an open mind comes into play
– This is also part of the process that can be opened up to peer review. Now that we have these data defined tracks that were recently released in our Synthwave community, a group of us can go through that list and figure out whether maybe that particular The Weekend or Elton John remix should remain on the list
– I never want anyone to think I’m an arbiter of what is or isn’t anything, so to verify that the track belongs on this Synthwave-specific chart, I googled each and every artist and track along with the search strings including Synthwave or Retrowave or Outrun; looking for whether or not they pop up in our synthly communities and youtube channels and playlists, or if their soundcloud/spotify says synthwave or 80s or has deloreans and retro things of that nature
– I do a very deep dive to make sure that whatever ends up on the list is relevant because I’m trying to be as accurate and correct as possible, because this chart affects all of us, which is why I feel that it should be open to anyone who wants to work on it
— Lots of disco house tracks ended up not on the list. Some did, because I found out that the artist has a specifically retro/synthwavey feed on ig or twitter. Some darker harder tracks didn’t make it, some of them, some of the artists had one decent outrun electro hard dark track but you listen to everything else they have and the outrun track is clearly the outlier and the rest of hard rock or something.
— I can see how this comes off as curation, but this is actually a part of the process I want to open this up to other humans in our community to verify. This lady’s track on the list? Yay, Yay, Nay, etc. It’s now possible to allow other people to help make the chart more accurate.
– I’m able to use data to get us this far, but then the human element should be more inclusive to stave off naysayers or anyone who questions the project.
– Once it’s sorted, then it’s a matter of marking the relevant tracks 1-100 on the spreadsheet as I go down the list. Everything else, non-relevant tracks, duplicates, are deleted and what I have left is the Retro Power Chart
– Last Step: Export/Import the chart to Spotify for publication