About Billie Charts

Learn more about my personal music charts, their origins, and how year-ends are tallied!

What is this?

The Billie Charts are a personal initiative I started in April 2013, to create an archive of my music listening habits throughout the years. This was inspired by Billboard's Hot 100, a weekly tally of the most listened songs in America. Though charts were made occasionally between April and August, it wasn't until August 11th that I officially began doing consistent updates to my charts.

From August 2013 to August 2017, I curated a daily chart, which mostly updated every evening. I would sit down, take a look at the songs in my playlist, and determine which ones felt like the most relevant songs that day. I first started by doing a Top 10, and then graduated to a Top 20 in 2015.

For the first 2 years of doing charts, I had no easy way to gather statistics on my listening habits, so I had to rely on my own memory to determine which songs did best. However, in January 2015 I would find out about Last.FM. Although it wasn't directly used for my charts at the time, I would lean on it to help bridge any lapse in memory.

In August 2017, the charts saw their biggest formatting change to date, becoming the Weekly Top 50. Since memory wasn't going to cut it for a weekly chart, I began fully relying on Last.FM data for creating the charts, and leaning on my own judgment to break ties (of which there were lots. And large ones too) The weekly top 50 format has stuck around to this day.

Why do this for 10+ years?

I will admit, it's quite a shock that I've kept at this for as long as I have. Since August 2013, I have been very consistent about updating these charts. The only blackout was in December 2015, when I was worried about the charts ruining my music listening habits. In hindsight, I regret stopping my charts at all.

I think the reason I regret it in the first place is that I just enjoy having that convenient archive of my music habits. Being able to go back and have a general idea of what I was into at any given time in my life is very fun! It also helps with my memory, given how much of my memory appears to function based on the songs I associate to certain moments of my life. A lot of my life pre-charts feels much more... vague in my mind as a result.

The other part of why I like having that archive is for a similar reason that chart-watching actual music charts is fun. Seeing trends come and go, the stories that are told through records being broken, the breakthroughs and come-backs, those are just so much fun to follow. Sure, you might think that there's not as much value in being invested in something I'm fully in control of. But I disagree! Mainly because I can't predict what future me will feel. Maybe a song I want to push as a big hit will wither on me in a few days. Maybe another song will take over out of nowhere. There is still a degree of uncertainty to where things will go that makes chart-watching my own music habits quite a lot of fun!

Where can I see these charts?

At the moment, I upload all my charts on CrownNote every week. I am pondering on a way to start uploading new charts to this website going forward, but this may not happen for a while. However, if you wish to get a cursory glance at which songs I was most into in a given year, I'm currently working to upload my official Year-End Top 100 lists to this website. You can consult them by going to the Year-End portion of the website! In the future, I may also add Spotify, CrownNote and Last.FM year-end lists to compare and contrast how each platform reports on my charting/listening habits.

Why are you making this part of your site look like Billboard?

Oh that's just me parodying the Billboard website. Specifically, the layout of the website is heavily inspired by how the site looked like around 2010-12, with the grays and gradient background. I find it fun to pretend like this part of my website is a fake magazine reporting on my music listening habits with the same level of professionalism as an actual music magazine would. Is it incredibly self-absorbed? I won't say no. But is it fun? Absolutely. And that's all that matters!

I also toyed with the idea of creating little editorials analyzing my charts' history, retrospectives, rankings, weekly chart updates, and more to really sell the "fake publication" angle of this sub-site. I'm not planning to do that right now, but it's something I'm considering for further down the line!

How do you calculate your Year-Ends?

Alright, this is gonna get very theoretical and a bit silly. Put on your math hats, folks.

The formula for calculating year-ends is typically very straightforward. Depending on the number of positions in the current chart year, I tally up points as such: #1 is the most points, and the lowest number is equal to 1 point. For mathematical folks, here's a handy formula to calculate the points a song earned on a given chart:

Points = Number of positions - Song position + 1

So for example, in a Top 10 chart, #5 would be worth 6 points.

This format is pretty straightforward in most cases. It will typically favor songs that have charted for longer periods of time over songs with high peaks, as it would only take 2 weeks at #26 to get the points from being at #1. Meanwhile, other methods (like pure play count and using the song position as a denominator in a fraction over a static number) might give too much weight to a song that performs amazingly one day but falters afterward. Since I'm interested in prioritizing songs with longevity, this doesn't really fit our needs.

However, there is an edge case that causes problems: If we have to calculate charts from 2 distinct chart formats, then there needs to be a way to bridge this gap. Since the frequency and number of positions can shift, trying to use this method as-is will weigh strongly in favor of the chart formats that have either more positions, or occur more frequently. There are only 2 years so far where this is an issue for calculating year-ends: 2013 and 2017.

2013 is a year that saw multiple format changes: The first 2 charts were a weekly top 25, then the next few charts from May to August 10th are scattered throughout, never closer than a week apart however. Then, on August 11th, the charts shift to a daily format where the above formula can easily be pasted in. Historically, I have only included the charts from August 11th to the end of the year. However, since a lot of songs get left out of the tally, I came up with a patchwork solution as follows:

2017 is a trickier case. The year is very evenly split between a Daily Top 20 from January to July, then a Weekly Top 50 from August to December. This makes calculating the top 100 songs of the entire year very awkward. My initial instinct was to go back and create the top 50s for the first half of the year. However, this method caused many songs that were genuine hits to be underrepresented in the Year-End list, and some less notable songs to chart surprisingly high. So my solution to balance things out was to create a list that combined both lists without weighing one too much over another. However, if we try to directly translate the score of a daily song into a weekly context, things will get skewed real fast. Main issue being that in the daily charts, song simply didn't chart nearly as long as they would in the weekly charts. The longest chart run in the 2017 dailies was 37 days, or just a little over 5 weeks. In that year alone, the longest charting song hit 21 weeks. On average, songs would typically not chart for more than 2-ish weeks in the daily charts, versus 4-6 in the Weekly era. Therefore, another method had to be used.

Thankfully, the number of points songs would get in the daily era were not all that far off from the numbers we get in the Weekly point tallies. So only by applying a slight modifier in favor of dailies, we get a chart that nicely combines both the weekly and the daily point tallies. I settled on 1.1 as the multiplier, since higher/lower values proved too lopsided. From there, I can simply add the weekly and daily scores together to create a total that we can compare.

Another pro of combining both scores, instead of trying to shove everything into a weekly chart lens, is that we don't get bogus stats for the song's peak and chart length. Since the daily charts are the ones that were used from January to July, the stats from the weekly readaptations feel arbitrary, even if they are technically based upon a more objectively correct data source. It's weird. Billie Chart science is weird.

This may have been a bit of a convoluted ramble, but hopefully it sheds some light on how I calculate my year-end lists!